Roflumilast Attenuates Microglial Senescence and Retinal Inflammatory Neurodegeneration Post Retinal Ischemia Reperfusion Injury Through Inhibiting NLRP3 Inflammasome

Abstract

Purpose

Retinal ischemia-reperfusion (RIR) injury is implicated in various retinal diseases, leading to retinal ganglion cells (RGCs) degeneration. Microglial senescence exacerbates inflammation, contributing to neurodegeneration. This study aimed to investigate the potential therapeutic role of Roflumilast (Roflu) in ameliorating microglial senescence and neuroinflammation following RIR injury.

Methods

C57BL/6J mice underwent RIR surgery, and Roflu treatment was administered intraperitoneally. BV2 microglial cells were subjected to oxygen-glucose deprivation and reoxygenation (OGD/R) to simulate ischemic conditions in vitro. SA-β-gal staining was used to detect cellular senescence. Quantitative PCR and ELISA were used to examine the levels of senescence-associated secretory phenotype (SASP) factors. Hematoxylin and eosin (H&E) staining was performed on retinal sections to assess retinal morphology and thickness. Surviving RGCs were labeled and quantified in retinal whole-mounts using immunofluorescence (IF). Furthermore, Western blot and IF staining were used to quantify the proteins associated with the cell cycle and NLRP3 inflammasomes.

Results

Roflu treatment reduced microglial senescence, ROS production, and secretion of pro-inflammatory cytokines in OGD/R-exposed BV2 cells. It also restored cell proliferation capacity and reversed OGD/R-induced cell cycle arrest. In vivo, Roflu alleviated retinal senescence, preserved retinal thickness, and protected against RGCs death in the RIR mouse model. Mechanistically, Roflu inhibited the NLRP3 inflammasome activation and suppressed DNA damage signaling pathway in microglia.

Conclusions

Roflu exerts neuroprotective effects by mitigating microglial senescence and inflammation via inhibition of the NLRP3 inflammasome in RIR injury. These findings suggest that Roflu may serve as a promising therapeutic strategy for retinal diseases associated with ischemic injury by targeting microglial senescence.

Retinal ischemia-reperfusion (RIR) injury is a pathological condition that occurs in various clinical scenarios, such as retinal artery occlusion, diabetic retinopathy, and glaucoma.¹ RIR injury is associated with various pathobiological events, including oxidative stress, inflammation, and death of retinal ganglion cells (RGCs), resulting in alterations in the retinal microenvironment.^2–4 The inflammatory cascade triggered by RIR injury plays a pivotal role in exacerbating neurodegeneration and vision loss,⁵ highlighting the importance of understanding the underlying mechanisms driving retinal inflammation.

Microglia, the resident immune cells of the retina, have emerged as key orchestrators of the inflammatory response during RIR injury.⁶ These highly dynamic cells respond rapidly to changes in the retinal microenvironment, adopting an activated phenotype characterized by the release of pro-inflammatory cytokines and chemokines.⁷ Our previous study has shown that RIR can induce microglial polarization and excessive release of inflammatory factors, ultimately resulting in retinal inflammation in the ischemic area and degeneration of RGCs.⁸ Recent evidence suggests that senescence-associated alterations in microglial function may contribute to dysregulated inflammatory responses in various neurological disorders, and reversing senescent microglia can effectively protect neuronal function, which prevents cognitive impairment and other neurodegenerative disorders in elderly individuals.^9,10

Microglial senescence is accompanied by the acquisition of a senescence-associated secretory phenotype (SASP), characterized by the secretion of pro-inflammatory cytokines, chemokines, and growth factors.¹¹ These SASP factors create a pro-inflammatory microenvironment, perpetuating chronic inflammation to neighboring cells, further amplifying the pro-senescence phenotype.⁶ Furthermore, senescent microglia exhibit impaired phagocytic clearance, mitochondrial dysfunction, and aberrant crosstalk with other immune cells, all of which contribute to the amplification of inflammatory responses and exacerbation of retinal pathology.¹² Specifically, inhibition of senescence-related inflammation has become a common mechanism for pharmacological interventions to exert beneficial effects in neurodegeneration. However, in current research on RIR injury, there is limited study on the relationship between inflammation and microglial senescence. Therefore, exploring how to reduce retinal neuroinflammation and microglial senescence after RIR holds promise for providing new therapeutic strategies in repairing neuronal damage after RIR.

Roflumilast (Roflu), a selective phosphodiesterase-4 (PDE4) inhibitor, has emerged as a promising therapeutic agent for modulating inflammatory responses in various disease states like chronic obstructive pulmonary disease and brain ischemia-reperfusion injury.¹³ By inhibiting the activity of PDE4, Roflu effectively elevates intracellular levels of cyclic adenosine monophosphate (cAMP), leading to the downregulation of pro-inflammatory mediators and the suppression of immune cell activation.¹⁴ Inhibiting PDE4 with Roflu can reduce neuroinflammatory responses of microglia, promote cognitive recovery, and support neuro-regeneration in cerebral ischemia mice.^15,16 Our previous research has shown that Roflu mediates the inhibition of pro-inflammatory cytokines production in the retina by microglia following RIR injury.⁸ This suggests that Roflu may be a promising strategy for treating retinal diseases, but its underlying mechanisms remain to be explored. Inspired by the findings of previous studies, we conducted the present study to explore the mechanism by which Roflu improves microglial senescence and mitigates retinal inflammation following RIR injury. We found that Roflu can regulate microglial senescence and exert neuroprotective effects by inhibiting the NLRP3 inflammasome, indicating its potential as a therapeutic agent for reversing retinal damage and neurodegeneration induced by RIR injury.

Methods

Animals

Eight-week-old C57BL/6J male mice were obtained from the Animal Center of Southern Medical University (Guangzhou, China). Mice were housed under a 12-hour light/dark cycle with ad libitum access to food and water. All experimental procedures were approved by the Animal Care and Use Committee of Southern Medical University (Guangzhou, China) and conducted following the Statement for the Use of Animals in Ophthalmic and Vision Research from the Association for Research in Vision and Ophthalmology. A total of 62 mice were recruited for this study. Fourteen mice were assigned to the sham surgery group. Among the 48 mice that underwent RIR surgery, 12 mice were excluded from analysis due to absence of neurological abnormalities (7 mice) or death (5 mice). Data from 50 mice were reported.

Drug Treatment

For in vivo experiments, Roflu (Selleck Chemicals, USA) was dissolved in 0.9% saline containing 5% Tween-80 and 0.5% dimethyl sulfoxide (DMSO) based on previous studies.⁸ For in vitro experiments, Roflu was dissolved in DMSO to a stock concentration of 100 mM. Prior to each experiment, Roflu was diluted to the desired final concentration, with a final DMSO concentration of 0.1%.

RIR Model Establishment

RIR surgery was conducted as previously described.¹⁷ Adult male mice were anesthetized by intraperitoneal injection of sodium pentobarbital at a dose of 50 mg/kg. Prior to surgery, 1% tropicamide eye drops (Santa, Japan) were used for pupil dilation, and 0.5% lidocaine hydrochloride (Santa, Japan) was used for corneal anesthesia. A 33-gauge needle connected to saline was used to maintain intraocular pressure within the range of 110 to 120 mm Hg for 60 minutes. The RIR model was induced in one eye of mice, whereas the contralateral eye remained untreated. Sham surgery was performed in another group of mice by inserting a needle into one eye without elevating intraocular pressure. Eye drops (Carbomer, Bausch, USA) were applied to moisten the eyes during ischemic injury. After 60 minutes of retinal ischemia, the needle was carefully withdrawn to restore normal intraocular pressure and reestablish retinal vascular perfusion. Ofloxacin ointment (Santa, Japan) was applied to the experimental eye to prevent bacterial infection. Mice in the DMSO group and Roflu group underwent RIR surgery. One hour after reperfusion, mice in the DMSO group and Roflu group were intraperitoneally injected with 1 mL of DMSO and Roflu solution, respectively.

Cell Culture

BV2 cell and human embryonic kidney cell line HEK293T were cultured in high-glucose DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. BV2 cells were maintained in a cell culture incubator at 37°C with 5% CO₂ and 95% air. Cells were divided into the normal control group, the oxygen-glucose deprivation and reoxygenation (OGD/R) model group, and the OGD/R + Roflu group. After overnight adherence, cells in the drug treatment group were pretreated with Roflu (10 µM) for 1 hour before establishing the OGD/R model. Briefly, cells in the control group were washed once with PBS and then continued to be cultured in high-glucose culture medium under aerobic conditions (95% air + 5% CO₂), whereas cells in the OGD/R group were cultured in sugar-free Dulbecco's modified Eagle's medium (DMEM) instead of culture medium and then transferred to an anaerobic culture box (95% N₂, 5% CO₂) for 4 hours. After OGD, cells were transferred back to aerobic conditions and cultured in high-glucose culture medium instead of sugar-free culture medium for 24 hours.

Senescence-Associated β-Galactosidase Staining

According to the instruction of senescence-associated β-galactosidase (SA-β-gal) Staining Kit (C0602; Beyotime, China), BV2 cells or retinal sections were washed with PBS, fixed with the special fixative for β-galactosidase staining in the kit at room temperature for 15 minutes, and then washed 3 times with PBS. After removing the PBS, 1000 µL of staining working solution (prepared immediately before use: 10 µL staining solution A + 10 µL staining solution B + 930 µL staining solution C + 50 µL X-gal) was added to each well to cover the cells. The cells were placed in a sealed, humidified container, incubated at 37°C in the dark overnight, and photographed under a microscope (IX73; Olympus, Japan). At least 3 different areas were randomly selected, with each area containing 100 cells, and the number of SA-β-gal-positive cells was counted.

Cellular Reactive Oxygen Species Detection

The BV2 cells were collected from 6-well plates and analyzed using the reactive oxygen species (ROS) Assay kit (S0033; Beyotime, China) according to the manufacturer's protocol. BV2 cells were treated with 10 µM dichloro-dihydro-fluorescein diacetate (DCFH-DA) at 37°C for 30 minutes, followed by washing with serum-free culture medium 2 times to remove excess DCFH-DA that did not enter the cells. Cells were observed under an inverted fluorescence microscope (Eclipse; Nikon, Japan).

Immunofluorescence

Untreated or treated BV2 cells were seeded onto slides, and retinal sections or retinal flat mounts were permeabilized with 0.5% Triton X-100 for 30 minutes and blocked with 5% BSA for 1 hour. The samples were then incubated overnight at 4°C with primary antibodies against IBA-1 (1:100, ab89874; Abcam, UK), p-p53 (1:100, T40061; Abmart, China), p21 (1:250, 10355-1-AP; Proteintech, USA), p16 (1:800, #18769; Cell Signaling Technology, USA), pRB (1:500, YP0240; Immunoway, China), γH2A.X (1:100, T56572; Abmart, China), H3K9me3 (1:100, P37961-5; Abmart, China), LaminB1 (1:100, P60054; Abmart, China), and 53BP1 (1:100, T56831; Abmart, China). The next day, the samples were incubated with Alexa Fluor 488 (donkey anti-goat IgG or donkey anti-rabbit IgG, 1:500; EarthOx, USA) or Alexa Fluor 594 donkey anti-rabbit IgG (1:500; EarthOx, USA) at room temperature for 1 hour. DAPI was added to stain cell nuclei for 5 minutes, and images were captured under a microscope (Eclipse; Nikon, Japan).

H&E Staining

Mouse eyes were collected and fixed overnight in 4% paraformaldehyde at 4°C. After dehydration in graded sucrose solution, the eyes were embedded in paraffin, and 4 µm thick sections were obtained from each paraffin block. Retinal images were captured using a slide scanner (Teksqray, China), and the overall retinal thickness was measured at 500 µm from the center of the optic disc using Image J software (National Institutes of Health, USA). The average of three sections per eye was taken.

Cell Cycle Analysis

When the cell confluence reached 70% to 80%, cells were digested with 0.25% trypsin, terminated with culture medium, and all adherent cells were collected by pipetting. According to the instructions of Cell Cycle and Apoptosis Detection Kit (C1052; Beyotime, China), the cells were collected and washed with 1 mL of PBS, fixed overnight in pre-chilled 70% ethanol, washed again with pre-chilled PBS, and stained with propidium iodide staining solution (1 sample: 0.5 mL staining buffer + 25 µL propidium iodide staining solution [20×] + 10 µL RNase A [50×]) for 30 minutes at 37°C in the dark. Cell cycle distribution was analyzed using the FACScan system from Becton Dickinson, and Flowjo software was used to calculate the percentage of cells in each phase. The ratio of G0/G1 phase cells was calculated as G0/G1/(G0/G1 + S + G2/M), and the ratio of S phase cells was calculated as S/(G0/G1 + S + G2/M).

EdU Cell Proliferation Assay

BV2 cells were treated with 10 µM EdU for 8 hours according to the instructions of the kFlour555 Click-iT EdU imaging kit (Keygen Biotech, China). Cell proliferation was observed using a confocal microscope (TCS SP5; Leica, Germany).

Western Blot

Retinal tissue and BV2 cells were rapidly frozen at −80°C. Tissue and cells were lysed on ice in RIPA buffer containing 1% protease inhibitor and 1% phosphatase inhibitor. After sonication and centrifugation, the supernatant was collected, and protein concentration was quantified using the BCA protein assay kit. The supernatant was boiled at 95°C for 10 minutes, and proteins (15 µg/mL) were separated by 10% SDS-PAGE gel electrophoresis and transferred to a PVDF membrane (0.22 µm). The PVDF membrane was then blocked with blocking solution and incubated overnight at 4°C with primary antibodies against p-p53 (1:1000), p53 (1:1000, #2524; Cell Signaling Technology, USA), p21 (1:2000), p16 (1:1000), pRB (1:1000), γH2A.X (1:5000), H3K9me3 (1:1000), LaminB1 (1:5000), and GAPDH (1:8000, RM2002; Ray Antibody Biotech, China). The next day, the membrane was incubated with HRP-conjugated secondary antibodies at room temperature for 1 hour. Protein expression levels were detected using a chemiluminescence imager, and grayscale analysis was performed using Image J software.

Enzyme-Linked Immunosorbent Assay

To detect the levels of inflammatory factors TNF-α, IL-1β, and IL-6, cell culture supernatants were collected 24 hours after OGD/R. ELISA was performed according to the manufacturer’s instructions (Youpin Biotech, China). Absorbance at 450 nm was measured using a multi-mode microplate reader (Spectramax i3x MD5, USA).

Quantitative Polymerase Chain Reaction

Total RNA was extracted, and reverse transcription was performed using the PrimeScript RT reagent kit (Takara, China) according to the manufacturer’s instructions. Real-time PCR was performed using 2 × SYBR Green qPCR Master Mix (B21202; Bimake, USA). Quantitative analysis was performed in three steps using the Light Cycler480 Real-Time PCR system (Roche, Switzerland). GAPDH was used as an internal control, and target mRNA expression levels were calculated using the 2-ΔΔCt method. The primer sequences used in this study are listed in Supplementary Table S1.

Cell Transfection

A small interfering RNA (siRNA) target in NLRP3 (GYVB40082114; JiYuan Biotech, China) was used to silence the transcript-level expression of NLRP3 in BV2 cells, and off-target effects were assessed using a negative control (NC) sequence. Briefly, cells in antibiotic-free complete medium were seeded at a density of 1 to 4 × 10⁵/well and transfected for 24 hours until they reached approximately 50% to 70% confluency. The Lipofectamine 3000 kit (L3000015; Invitrogen, USA) was used for transfection, and siRNA diluted in Opti-MEM was added to a complete medium without FBS and added to cell-containing wells (>2 mL per well). After 6 hours of transfection, the medium was replaced with DMEM supplemented with 10% FBS and the transfection efficiency was verified using quantitative PCR (qPCR) and Western blot after 48 hours.

Lentiviral Infection

For lentiviral production, HEK293T cells were cultured in 10 cm dishes until 60% to 70% confluence. Subsequently, cells were co-transfected with lentiviral vectors PpLV3-CMV-NLRP3 (mouse)-EF1a-CopGFP-Puro plasmid (3 µg), pMD2.G vector (2 µg), and psPAX2 vector (3 µg) in 750 µL Opti-MEM. After 72 hours of transfection, viral particles were collected from the supernatant of cultured HEK293T cells through a Millex-GP filter device (Jiefei, China) with a pore size of 0.45 µm. For lentiviral infection, cells were incubated for 8 hours with 1 mL of packaged lentivirus supplemented with 10 µg/mL of polybrene (Biotek, USA). Cells were subsequently screened with 0.5 µg/mL puromycin for 8 days. Infection efficiency was observed daily by fluorescence microscopy and cells were infected with more than 90% efficiency, followed by OGD/R and Roflu treatment.

Quantification of RGCs

Mice were euthanized with an overdose of anesthesia, and the retinas were harvested and fixed in 4% paraformaldehyde. They were then incubated in goat serum containing 0.5% Triton X-100 for 1 hour, followed by incubation with mouse anti-Brn3a antibody (1:100, sc-8429; Santa Cruz Biotechnology, USA) overnight at 4°C. After 3 washes the next day, they were incubated with Alexa Fluor 488 (1:2000, E032210; EarthOx, USA) at room temperature for 1 hour. The retinas were then flat-mounted on slides, divided into four quadrants, and coverslipped. Three images of 0.3588 mm² each were captured at 1 mm intervals between each quadrant, representing the peripheral, inner, and central retina. The number of Brn3a-positive cells in each independent region of the retinal flat mounts was calculated using Image J software, averaged, to determine the quantity of RGCs.

Statistical Analysis

Statistical analysis was performed using IBM SPSS 20.0 software. All data were presented as mean ± standard deviations (mean ± SDs) and were first tested for normal distribution and homogeneity of variance. Data conformed to normal distribution and homogeneity of variance and were compared between two groups using the t-test, among multiple groups using 1-way analysis variance (ANOVA), followed by Tukey's test. In case of non-normal distribution, non-parametric tests such as the Kruskall-Wallis test and Mann-Whitney U test were used. A P value less than 0.05 was considered statistically significant.

Results

Roflu Regulates OGD/R-Induced Morphological and Metabolic Changes in BV2

In order to assess whether Roflu could ameliorate the senescence of microglia induced by the OGD/R model, Roflu was pretreated before OGD/R modeling. The dosage of Roflu used in this study was determined based on our previous research.⁸ Senescent cells exhibit an increase in the number and the size of lysosomes, and SA-β-gal serves as a biomarker for detecting cellular senescence. Under pH 6.0 conditions, blue signals indicate cellular senescence.¹⁸ In this study, we found that almost no β-gal staining cells were detected in the control groups. However, the level of positively stained cells increased after OGD/R stimulation and decreased after Roflu treatment (Fig. 1A). Additionally, metabolic disturbances in senescent cells lead to weakened mitochondrial function and increased production of reactive ROS. We measured the fluorescence intensity of DCFH-DA in BV2 exposed to OGD/R by using immunofluorescence. As shown in Figure 1B, compared to the control group, the level of ROS in BV2 treated with OGD/R significantly increased, whereas the fluorescence intensity of ROS in cells decreased significantly after Roflu treatment.

Figure 1.

Roflu regulates OGD/R-induced morphological and metabolic changes in BV2. (A) The cellular senescence status and the quantification of SA-β-gal positive cells were determined by SA-β-gal staining method (20×, scale bar = 100 µm). (B) The production of ROS and the quantitative analysis of ROS expression levels were measured by staining with DCFH-DA (20×, scale bar = 100 µm). (C) The expression levels of inflammatory factors TNF-α, IL-1β, and IL-6 secreted by microglia were determined by ELISA. (D) The mRNA expression levels of SASP-related inflammatory factors TNF-α, TGF-β, IL-1β, IL-6, PAI-1, MCP-1, and SIRT1 were analyzed (data are presented as mean ± SD from 3 or more independent experiments, conducted on different dates, with 3 parallel samples in each group, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).

Furthermore, supernatants of OGD/R-treated BV2 were collected, and the levels of inflammatory factors TNF-α, IL-1β, and IL-6 secreted by microglia were detected using ELISA (Fig. 1C). Compared to the control group, the levels of inflammatory factors TNF-α, IL-1β, and IL-6 were significantly upregulated. Similarly, we used qPCR to detect the mRNA levels of SASP-related inflammatory factors. As shown in Figure 1D, after OGD/R stimulation, the expression levels of TNF-α, TGF-β, IL-1β, IL-6, PAI-1, and MCP-1 increased, whereas the expression level of the anti-aging factor SIRT1 decreased. Interestingly, opposite changes were observed after Roflu treatment. These results indicate that OGD/R facilitates microglial senescence, whereas Roflu plays a role in delaying microglial senescence.

Roflu Mediates OGD/R-Induced Disturbances in Microglial Cell Cycle and DNA Damage Signaling Pathways

A decline in cellular proliferation capacity is a biomarker of aging, and changes in the cell cycle reflect the proliferative potential of cells.¹⁹ EdU cell proliferation assay showed that compared to the control group, the proliferation capacity of BV2 under OGD/R significantly decreased, whereas it improved after treatment with Roflu (Figs. 2A, 2B). Similarly, flow cytometry analysis results showed that compared to the control group, most cells in the OGD/R treatment group were arrested in the G0/G1 phase, with a reduction in the number of cells in the S phase. Following Roflu treatment, the proportion of cells in the G0/G1 phase decreased, whereas the proportion of cells in the S phase increased (Figs. 2C, 2D). These data indicate that OGD/R insult leads to cell growth arrest and significantly decreased proliferation, whereas Roflu reverses cell growth arrest and improves proliferation capacity.

Figure 2.

Roflu mediates OGD/R-induced disturbances in microglial cell cycle and DNA damage signaling pathways. (A, B) Cell proliferation capability was assessed using EdU incorporation assay and quantitatively analyzed (40×, scale bars = 50 µm). (C, D) Cell cycle analysis was performed by flow cytometry. The distribution of cells in G0/G1 phase, G2/M phase, and S phase was determined. (E) The results of Western blot showed that the expression of LaminB1 was decreased while H3K9me3 expression increased after OGD/R, these changes were reversed in the Roflu group. (F) Immunofluorescence staining was conducted to detect the expression levels of LaminB1 and H3K9me3 proteins, and fluorescence intensity was quantitatively analyzed. (G) Western blot results showed the protein expression levels of γH2A.X, p53, p-p53, p21, p16, and pRB. (H) Immunofluorescence staining was performed to detect the expression levels of γH2A.X, 53BP1, p-p53, p21, p16, and pRB proteins, and fluorescence intensity was quantitatively analyzed (data are presented as mean ± SD from 3 or more independent experiments, conducted on different dates, with 3 parallel samples in each group, *P < 0.05, **P < 0.01, and ***P < 0.001).

Figure 2.

Continued.

We further examined the signaling molecules associated with the cell cycle. Western blot was used to detect LaminB1 and H3K9me3 proteins. It was found that after OGD/R insult, LaminB1 expression decreased significantly, whereas H3K9me3 expression increased. Conversely, these changes were reversed in the Roflu group (Fig. 2E, Supplementary Fig. S1A). Due to the loss of LaminB1 levels induced by OGD/R, changes in nuclear membrane integrity in BV2 were observed through immunofluorescence, with a significant decrease in fluorescence intensity within the nucleus. Interestingly, H3K9me3 fluorescence intensity increased significantly, and these changes were reversed after treatment with Roflu (Fig. 2F, Supplementary Fig. S1B). Additionally, Western blot results showed that in the control group, the levels of γH2A.X, p-p53, p21, and p16 were very low, whereas pRB was highly expressed. In contrast, the expression of γH2A.X, p-p53, p21, and p16 proteins increased in the OGD/R group, whereas pRB expression decreased. Consistently, these changes were reversed after treatment with Roflu (Fig. 2G, Supplementary Fig. S1C). Moreover, immunofluorescence was used to further observe the changes in fluorescence intensity of γH2A.X, 53BP1, p-p53, p21, p16, and pRB proteins. Immunofluorescence results also showed similar trends (Fig. 2H, Supplementary Fig. S1D). These findings demonstrate that OGD/R induce DNA damage, further activate the CDKi signaling pathway, and induce stable cell cycle arrest, whereas treatment with Roflu can reduce DNA damage signaling transduction, attenuate microglial cell cycle arrest, and improve microglia proliferation capacity.

Roflu Improves Cell Senescence and RGCs Death Induced by RIR Injury

We performed whole-mount staining of retina tissues using SA-β-gal. As depicted in Figure 3A, minimal staining was observed in retina tissues of each control group, whereas a significant increase in positive staining was evident in the retina subjected to RIR injury. However, treatment with Roflu alleviated the extent of positive staining. Subsequently, we examined the expression levels of SASP inflammatory factors in the damaged retinas of RIR-insulted mice. Results revealed elevated expression levels of inflammatory factors, including TNF-α, TGF-β, IL-1β, IL-6, PAI-1, and MCP-1, along with decreased expression of SIRT1 compared to the control group. Interestingly, a reverse trend was observed following Roflu treatment (Fig. 3B).

Figure 3.

Roflu improves cell senescence and RGCs death induced by RIR injury. (A) The degree of retinal tissue senescence was evaluated using SA-β-gal staining. (B) The mRNA expression levels of SASP-related inflammatory factors TNF-α, TGF-β, IL-1β, IL-6, PAI-1, MCP-1, and SIRT1 in retinal tissue were measured. (C) Immunofluorescence staining and quantitative analysis of Brn3a⁺ ganglion cells in retinal flat-mounts were performed (20×, scale bar = 100 µm). (D) H&E staining of retinal sections and quantitative analysis of retinal layer thickness and IPL thickness were conducted (40×, scale bar = 50 µm). (E) Mice were treated with Csf1r antagonist PLX5622 to deplete microglia before modeling, followed by RIR injury and Roflu treatment. Neuroprotective effects were evaluated 7 days after RIR. Immunofluorescence staining of IBA-1 in retinal flat-mounts was performed to validate microglial depletion efficiency and quantify the number of IBA-1⁺ cells (20×, scale bar = 100 µm). (F) Immunofluorescence staining of Brn3a⁺ ganglion cells in retinal sections from different groups was conducted (40×, scale bar = 50 µm; data are presented as mean ± SD from 3 or more independent experiments, conducted on different dates, with 3 parallel samples in each group, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).

To validate the neuroprotective effect of Roflu, retinal whole-mounts were subjected to Brn3a immunostaining to label surviving neurons. As illustrated in Figure 3C, Roflu pretreatment preserved a considerable number of RGCs compared to the control group. Additionally, H&E staining of mouse retinal images showed a reduction in retinal thickness and inner plexiform layer (IPL) thickness after RIR injury, whereas Roflu preserved the overall retinal thickness, particularly the thickness of the inner IPL (Fig. 3D). To further elucidate whether the clearance of microglia affects the therapeutic effect of Roflu after RIR, we depleted microglia using the Csf1r antagonist PLX5622, followed by RIR injury and Roflu treatment in mice (Supplementary Fig. S2). Subsequent retinal whole-mount and IBA-1 immunofluorescence staining revealed a depletion of over 90% of microglia following PLX5622 administration (Fig. 3E), and Brn3a immunostaining of retinal sections indicated a considerable preservation of RGCs after Roflu pretreatment compared to RIR, whereas PLX5622 attenuated the therapeutic effect of Roflu (Fig. 3F). These data collectively suggest a potential association between the therapeutic effect of Roflu post-RIR injury and microglial senescence.

Roflu Inhibits the DNA Damage Signaling Pathway Induced by RIR

Immunofluorescence staining was used to detect γH2A.X, p21, and pRB, IBA-1, and DAPI. Compared to the control group, the co-staining of γH2A.X, p21, and IBA-1 increased, whereas the co-staining of pRB and IBA-1 decreased post-RIR injury, and these changes were reversed following Roflu treatment (Figs. 4A, 4B). Subsequently, Western blot was utilized to quantitatively analyze changes in the protein levels of LaminB1, H3K9me3, γH2A.X, p-p53, p53, p21, p16, and pRB in each group. Results indicated an increase in the protein expression levels of γH2A.X, p-p53, p21, and p16 following RIR, along with a significant decrease in pRB expression, which were reversed by Roflu treatment (Figs. 4C, 4D). These findings further demonstrate that RIR injury in vivo induces DNA damage, leading to activation of the CDKi signaling pathway, inducing cell cycle arrest, whereas Roflu treatment reduces DNA damage signal transduction.

Figure 4.

Roflu inhibits the DNA damage signaling pathway induced by RIR. (A, B) Retinal sections of mice from control group, RIR group, and Roflu group were stained with IBA-1 (green) and γH2A.X, p21, and pRB (red) as well as DAPI (blue) for immunofluorescence co-staining. The number of γH2A.X, p21, and pRB positive microglia was quantitatively analyzed (40×, scale bar = 50 µm) at 3 days post RIR injury. (C, D) Protein expression levels of LaminB1, H3K9me3, γH2A.X, p53, p-p53, p21, p16, and pRB were quantitatively analyzed (data are presented as mean ± SD from 3 or more independent experiments, conducted on different dates, with three parallel samples in each group, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).

Roflu Inhibits Microglial Senescence and Retinal Inflammation Via NLRP3 Inflammasomes

First, using siRNA transduction, we successfully knocked down NLRP3 in BV2 cells, as confirmed by qPCR analysis (Supplementary Fig. S3A). Additionally, we evaluated the protein levels of NLRP3, apoptosis-associated speck-like protein containing a CARD (ASC), IL-1β, and cleaved caspase-1 (Cl-caspase-1) via Western blot and measured the secretion of Cl-caspase-1 by BV2 cells using ELISA. As shown in Figures 5A and 5B and Supplementary Figure S3B, the expression of these proteins was significantly upregulated following OGD/R, whereas treatment with Roflu or NLRP3 knockdown substantially reduced their levels. We further investigated the effect of Roflu on NLRP3 inflammasome expression in the retinas of mice subjected to RIR injury, yielding consistent results (see Supplementary Figs. S3C and S3D). Collectively, these findings suggest that Roflu inhibits microglial senescence by suppressing the NLRP3 inflammasome in both in vitro and in vivo models.

Figure 5.

Roflu inhibits microglial senescence and retinal inflammation via NLRP3 inflammasomes. (A, B) Protein expression levels of NLRP3, ASC, IL-1β, and Cl-caspase-1 were detected using Western blot and quantitatively analyzed after knockdown of NLRP3. (C) The degree of SA-β-gal staining and the quantification of SA-β-gal positive cells were analyzed after knockdown of NLRP3 in microglia (20×, scale bar = 100 µm). (D) The mRNA levels of SASP-related inflammatory factors secreted by microglia were measured after knockdown of NLRP3. (E) Immunofluorescence staining was used to detect the expression levels of p-p53, p21, p16, and pRB proteins, and fluorescence intensity was quantitatively analyzed after knockdown of NLRP3 (40×, scale bar = 50 µm, data are presented as mean ± SD from 3 or more independent experiments, conducted on different dates, with 3 parallel samples in each group, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).

To elucidate the role of the NLRP3 inflammasome in microglial senescence under OGD/R conditions, we further assessed the phenotypic changes associated with microglial senescence and the mRNA levels of SASP-related inflammatory factors following NLRP3 knockdown. As shown in Figure 5C, SA-β-gal staining revealed a significant reduction in positively stained cells in the OGD/R + siNLRP3 group compared to the OGD/R + siCtrl group. Furthermore, the expression levels of TNF-α, TGF-β, IL-1β, IL-6, PAI-1, and MCP-1 were decreased (Fig. 5D). Subsequently, we detected the expression of p-p53, p21, p16, and pRB proteins through immunofluorescence. The results indicated that p-p53, p21, and p16 levels were elevated in the OGD/R + siCtrl group, whereas the pRB levels were low; in contrast, the OGD/R + siNLRP3 group exhibited the opposite pattern of changes (Fig. 5E and Supplementary Fig. S3E). These findings demonstrate that NLRP3 inhibition mitigates OGD/R-induced microglial senescence, mirroring the protective effects of Roflu.

NLRP3 Overexpression Attenuates the Protective Effect of Roflu Against OGD/R-Induced Microglial Senescence

To further validate whether Roflu ameliorates OGD/R-induced microglial senescence through NLRP3 inflammasomes, we overexpressed NLRP3 in BV2 cells using a lentivirus-mediated approach. Western blot and qPCR results confirmed successful overexpression (Supplementary Figs. S4A and S4B). Additionally, after NLRP3 overexpression, we assessed the expression levels of NLRP3, ASC, IL-1β, and Cl-caspase-1. As shown in Figures 6A and 6B, compared to the control group, NLRP3 inflammasome expression significantly increased following OGD/R induction. However, in the Roflu and Roflu + OE-NC groups, these protein levels were notably downregulated after OGD/R, an effect that was reversed by NLRP3 overexpression. Moreover, we included an additional control group (OGD/R + OE-NLRP3 without Roflu treatment) and examined the expression level of Cl-caspase-1 using immunofluorescence and Western blotting. As shown in Supplementary Figures S4C and S4D, the OGD/R + Roflu + OE-NLRP3 group exhibited a slight reduction in Cl-caspase-1 protein levels compared to the OGD/R + DMSO + OE-NLRP3 group. These results suggest that while Roflu reduces Cl-caspase-1 levels, there is a limit to its efficacy when NLRP3 is overexpressed.

Figure 6.

NLRP3 overexpression attenuates the protective effect of Roflu against OGD/R-induced microglial senescence. (A, B) Protein expression levels of NLRP3, ASC, IL-1β, and Cl-caspase-1 were detected using Western blot and quantitatively analyzed after overexpression of NLRP3. (C)The degree of SA-β-gal staining and the quantification of SA-β-gal positive cells were analyzed after overexpression of NLRP3 in microglia (20×, scale bar = 100 µm). (D) The mRNA levels of SASP-related inflammatory factors secreted by microglia were measured after overexpression of NLRP3. (E) Immunofluorescence staining was used to detect the expression levels of p-p53, p21, p16, and pRB proteins after overexpression of NLRP3 (40×, scale bar = 50 µm, data are presented as mean ± SD from 3 or more independent experiments, conducted on different dates, with 3 parallel samples in each group, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).

We further verified the phenotypic changes of microglial senescence and mRNA levels of SASP-related inflammatory factors after NLRP3 overexpression. As shown in Figure 6C, SA-β-gal staining revealed minimal β-gal staining in the OE-NC group, whereas positively stained cells increased in the OE-NLRP3 group. Additionally, compared to the OE-NC group, the expression levels of TNF-α, TGF-β, IL-1β, IL-6, PAI-1, and MCP-1 were elevated, whereas SIRT1 expression decreased in the OE-NLRP3 group (Fig. 6D). Subsequently, immunofluorescence was used to detect LaminB1, H3K9me3, γH2A.X, 53BP1, p-p53, p21, p16, and pRB proteins. The results showed that in the OE-NC group, H3K9me3, γH2A.X, 53BP1, p-p53, p21, and p16 levels were very low, whereas LaminB1 and pRB were highly expressed. In contrast, the OE-NLRP3 group exhibited the opposite pattern (Fig. 6E, Supplementary Fig. S5). Collectively, these findings indicate that Roflu mitigates OGD/R-induced microglial senescence through the modulation of NLRP3 inflammasomes.

Discussion

In this study, we investigated the therapeutic effects and mechanisms of short-term Roflu in ischemic retinal diseases. Our main findings are as follows: (1) Roflu effectively reduced neuroinflammation and RGC death following RIR injury, thereby restoring retinal structure and visual function; (2) Roflu exerted neuroprotective effects by inhibiting microglial senescence; (3) Roflu exerted its anti-microglial senescence effects through the regulation of NLRP3 inflammasome, and (4) NLRP3 overexpression largely reversed the protective effects of Roflu. These findings suggest promising clinical applications of Roflu in ischemic retinal diseases. Furthermore, PDE4 and its downstream molecules emerge as novel therapeutic targets for mitigating microglial senescence induced by RIR.

Substantial evidence indicates that RIR is a major mechanism underlying diseases, such as glaucoma and diabetic retinopathy, resulting in irreversible vision loss globally. During the reperfusion phase following ischemia, the retina produces a large amount of inflammatory and chemotactic factors due to oxidative stress, abnormal cell metabolism, and other factors.²⁰ These factors attract immune cells, such as microglia, to the ischemic area, triggering activation of microglia in the retina. Activated microglia release inflammatory mediators and pro-inflammatory factors, exacerbating cellular damage. Concurrently, neuronal cells undergo apoptosis or necrosis, leading to neuronal dysfunction and vision loss.²¹ Therefore, there is an urgent need to explore novel drugs for the treatment of RIR and inhibition of RIR-mediated inflammatory neurodegeneration.

Roflu, as a second-generation PDE4 selective inhibitor, has shown potent anti-inflammatory effects in respiratory and central nervous systems, but its role in retinal diseases has been rarely studied. Cerebral-related research suggests that abnormal cAMP signaling is associated with cognitive impairment in neurodegenerative diseases, such as Alzheimer's disease.²² PDE4 inhibitors lead to accumulation of cAMP, suggesting a potential role for Roflu in treating retinal neurodegenerative diseases. RIR injury damages various cell types, but RGC loss has been the main focus of models. Our current data indicate that after RIR injury, retinal IPL thickness decreases, along with a reduction in RGC numbers. Treatment with Roflu alleviates ischemic neuronal cell damage induced under RIR conditions, indicating its protective effects. Additionally, our previous research found a significant relationship between Roflu's neuroprotective effects after RIR injury and microglia activation.⁸ This study found that using PLX5622 to eliminate microglia reversed the pharmacological effects of Roflu in vivo, further confirming that microglia are the target cells of Roflu treatment.

Cellular senescence refers to irreversible cell growth arrest after reaching a certain number of cell divisions.²³ One of the most important aging pathways is stress-induced premature senescence (SIPS), which can be induced by cytokines, DNA damage, oxidative stress, and external stimuli.²⁴ Studies have shown that cellular senescence is significant in acute and chronic tissue damage/dysfunction induced by ischemia-reperfusion injury.²⁵ Senescent cells undergo irreversible cell cycle arrest, accompanied by changes in function, morphology, and gene expression.²⁶ It has been suggested that in neurodegenerative diseases and cerebral ischemic diseases, senescent microglia may be in a state of sustained activation, leading to excessive release of inflammatory factors. Moreover, senescent microglia exhibit a decreased ability to clear cellular debris and metabolic waste, which may lead to accumulation of toxic substances in the neural environment, exacerbating neuronal damage and showing weakened ability to support neuronal survival and maintain synaptic function.^27,28 However, there is limited research on the role of microglial senescence in RIR injury. This study explored the induction of microglial senescence after RIR injury and the role of Roflu in delaying microglial senescence. In the in vitro OGD/R model, stimulated microglia exhibited a typical cellular senescence pattern, including increased SA-β-gal-positive cells, G0/G1 cell cycle arrest phenotype, slowed cell proliferation, and decreased number of cells in the S phase. Furthermore, increased expression levels of senescence-related proteins H3K9me3, γH2A.X, 53BP1, p-p53, p21, and p16, and decreased expression of LaminB1 and pRB were observed, all indicating cellular senescence. Functional changes also occurred during cellular senescence, impairing normal cellular functions. Our study found that after OGD/R stimulation of microglia, secretion of inflammatory mediators including IL-1β, IL-6, TNF-α, TGF-β, MCP-1, PAI-1, and ROS increased, further confirming the pathological aging of microglia induced by OGD/R. Treatment with Roflu reversed these phenotypic changes. In the in vivo RIR-induced retinal aging model, we found that RIR injury induced retinal aging, manifested by increased SA-β-gal staining intensity and secretion of inflammatory mediators, such as IL-1β, IL-6, TNF-α, TGF-β, MCP-1, and PAI-1. Additionally, microglia co-stained with senescence-related proteins further indicating that RIR can induce retinal microglial senescence, and treatment with Roflu can reverse retinal microglial senescence.

It is crucial to elucidate the mechanisms by which Roflu delays microglial senescence and provides neuroprotection after RIR injury. In our previous study, we found that Roflu modulated microglia activation and retinal neuroinflammation through the Nrf2/STING pathway. It was demonstrated that STING directly interacted with NLRP3 and attenuated the polyubiquitination of NLRP3 to induce NLRP3 activation.²⁹ Studies have also shown that inhibiting NLRP3 inflammasome can alleviate endothelial cell senescence.²⁶ The NLRP3 protein forms a complex with ASC, which then binds to cysteine protease caspase-1 to form an inflammasome, further activating caspase-1, cleaving pro-IL-1β to its mature form IL-1β, thereby mediating inflammation.³⁰ IL-1β has been shown to have a strong influence on cellular senescence,³¹ and studies have also shown that NLRP3 overexpression and exogenous IL-1β supplementation induce senescence phenotypes in bone marrow stromal cells from patients with myelodysplastic syndrome³²; inhibiting NLRP3 can delay endothelial cell senescence in diabetic mice.³³ In our study, we also found similar phenomena, namely activation of NLRP3 inflammasome during RIR injury. Roflu treatment inhibited the formation of NLRP3 inflammasome, as well as the maturation and release of IL-1β, thereby preventing pathological senescence of microglia.

In conclusion, this study demonstrates that Roflu alleviates microglial senescence, thereby inhibiting RGCs apoptosis in RIR-injured retina. Our findings establish Roflu treatment as a potential therapeutic approach to alleviate retinal neurodegenerative diseases associated with microglial senescence. However, further research, including studies using human retinal cells, is needed to evaluate its clinical potential and support its broader application in future clinical settings.