Salidroside protects RGC from pyroptosis in diabetes-induced retinopathy associated with NLRP3, NFEZL2 and NGKB1, revealed by network pharmacology analysis and experimental validation

Abstract

Objective

To investigate the effect of salidroside (SAL) in protecting retinal ganglion cell (RGC) from pyroptosis and explore associated molecular network mechanism in diabetic retinapathy (DR) rats.

Methods

HE, Nissl and immunofluorescence staining were used to observe the retinal morphological change, and the related target genes for salidroside, DR and pyroptosis were downloaded from GeneCard database. Then Venny, PPI, GO, KEGG analysis and molecular docking were used to reveal molecular network mechanism of SAL in inhibiting the pyroptosis of RGC. Lastly, all hub genes were confirmed by using qPCR.

Results

HE and Nissl staining showed that SAL could improve the pathological structure known as pyroptosis in diabetic retina, and the fluorescence detection of pyroptosis marker in DM group was the strongest, while they decreased in the SAL group(P < 0.05)). Network pharmacological analysis showed 6 intersecting genes were obtained by venny analysis. GO and KEGG analysis showed 9 biological process, 3 molecular function and 3 signaling pathways were involved. Importantly, molecular docking showed that NFE2L2, NFKB1, NLRP3, PARK2 and SIRT1 could combine with salidroside, and qPCR validates the convincible change of CASP3, NFE2L2, NFKB1, NLRP3, PARK2 and SIRT1.

Conclusion

Salidroside can significantly improve diabetes-inducedRGC pyrotosis in retina, in which, the underlying mechanism is associated with the NLRP3, NFEZL2 and NGKB1 regulation.

Introduction

Diabetic retinopathy (DR), mainly caused by neurovascular damage of retina, is the primary cause of visual impairment, which affects the vision of patients and even leads to blindness [19]. It was reported that there were approximately 463 million DM patients worldwide and this figure was expected to increase to 700 million patients in the following 25 years [21]. Some research proved that non-proliferative diabetic retinopathy(NPDR) was present in 25% of patients in 5 years after DM diagnosis, in 60% at 10 years and 80% at 15 years [10]. On the other hand, proliferative diabetic retinopathy(PDR) was found in 2% of patients with DM duration of less than 5 years and in 15.5% of patients who had DM for 15 years or more [13, 26]. In the early stages of diabetes, it is increasingly recognized that complex neuronal, glial, and microvascular abnormalities gradually disrupt retinal function. However, the concrete underlying cellular mechanism keeps to be known.

Pyroptosis, also known as cell inflammatory necrosis, is a programmed cell death that is involved in the development of a variety of microvascular complications of diabetes [4, 14]. Studies have found that the mechanism of pyroptosis in diabetic microvasculature is mediated by the activation of inflammasomes such as NLR Family Pyrin Domain Containing 3(NLRP3) and the activation of its downstream effector Caspase-1, resulting in the release of a large number of inflammatory factors [15]. Caspase-1 promotes interleukin-1β (IL-1β), Interleukin 18 (IL-18) forms inflammatory cells and increases the inflammatory response [17]. Under pyroptosis condition, Caspase-4/5/11 binds to bacterial lipopolysaccharide and becomes activated by oligomerization [25, 36]. Activation of Caspase-4/5/11 divides Gasdermin D (GSDMD) protein, and the generated active N-terminal domain of GSDMD protein can mediate the dissolution of the cell membrane, and finally activate the NLRP3 inflammasome to activate Caspase-1, produce IL-1β, and eventually lead to cellular hypertrophy [38]. Previously, the important role of pyroptosis in the occurrence and development of DR has been well known. Therefore, it is possible to develop the effective treatment methods on traditional Chinese medicine for DR prevention.

Salidroside (SAL) an effective Chinese medicine that has anti-inflammatory [35], antioxidant [39], hypoglycemic [23] and other effects, as well as significantly hypoglycemic and neuroprotective effects, has been well known, so as that it can be used for the treatment of neurodegenerative diseases, cardiovascular diseases, diabetes, cancer and many other diseases [41]. Current treatment methods for DR, such as intravitreal injection of remizumab, retinal laser photocoagulation and vitrectomy, have no optimal therapeutic effect, with only t neuropathy to be delayed [13]. Comparatively, Salidroside can inhibit neuronal apoptosis and reduce the release of inflammatory factors, which also has a variety of pharmacological activities mild, safe and cheap, at the same time few adverse reactions, with long-term use advantage [5]. Also, Salidroside can inhibit the apoptosis of retinal pigment epithelial cells and retinal endothelial cells induced by hydrogen peroxide through the mechanism of anti-oxidative stress [27]. But the evidence on salidroside for retina glanglion cells (RGC) protection is completely unknown and the underlying pertinent gene mechanism is not clear and waiting to be elucidated.

In this study, we primarily explored the protective effect of salidroside for RGC in DR model and determine related gene changes, by network pharmacology and molecular docking [31, 40], combined with quantitative PCR validation. Our findings will further enrich the knowledge in preventing pathogenesis of DR, and provided a theoretical basis for the clinical usage of salidroside in DR treatment.

Methods

Preparation and administration of animal model

SD rats weighting 180 ± 200 g were purchased from the Department of Experimental Animal Science, Kunming Medical University,approved by the Animal Experiment Ethics Review Committee of Kunming Medical University, the approval number is KMMU20220894.The weight of the rats was 180–200 g. The animals were housed in a 12 h light/dark cycle at room temperature of 20–25 °C with 45% to 65% relative humidity and provided with standard food and water. After feeding for 3 days, the health status of the animals was observed and recorded. The rats were divided into 3 groups: normal control(CON) group, SAL treatment(SAL) group and diabetic model(DM) group.

After rats were weighed, STZ it was dissolved in sterile citric acid-sodium citrate buffer with a pH value of 4.5, and a concentration of 0.1 mol/L was prepared for later usage. In detail, before the model preparation, rats were fasted for 12 h and subjects to intraperitoneally injection with STZ at 65 mg/kg. After 2 h of modeling, they started to eat, then the blood of the tail vein tip was detected with a blood glucose meter after 3 days of administration, and the rats with fasting blood glucose greater than 16.7 mmol/L were used as the diabetes model, of which, the number of animals in DM group-up to 10 rats were designed as normal control.

After 6 weeks of feeding, the diabetic model group were divided into DM group (n = 9) and SAL administrated group (n = 8), except 3 rats died during the process The SAL solution was prepared with normal saline, and the rats in the treatment group were and treated by intragastric administration until 12 weeks, whereas, rats in the diabetic group were given the same amount of normal saline intragastric administration, and 2 rat died during the latter 4 weeks correspondingly, whereas,7 rats was continuously survived for 4 weeks in SAL administrated group except 1 rat died. All rats survived was carefully given nursing during all process till 3 months.

Sample harvest

Body weight and blood glucose levels were measured every 2 weeks during the 3 month treatment period. After 3 months of treatment, the rats were sacrificed by excessive anesthesia with 200 mg/kg sodium pentobarbital. After death was confirmed by monitoring for cessation of breathing and heartbeat, eyes from each rat in all group were isolated and fixed with 4% paraformaldehyde at room temperature.

Frozen section

After the eye tissue was taken, it was fixed in the same PFA fixative overnight, and the fixed eyeball tissue was taken and frozen in 10% (about 7 h), 20% (about 4 h) and 30% (about 12 h) sucrose solution respectively. Protection (4 °C) was kept until the tissue samples completely sink in each solution. Then we took out and absorbed the water; then cut off the excess flesh tissue around the eyeball, and injected a little sucrose into the cornea to make it more plump. Next, they were embedded with a plastic pipette Short fix with O.C.T., filled with O.C.T., at − 20 °C for 30 min, and the 3 eye tissues from each group were cut into 10 µm in coronal surface of the eyeball.

Hematoxylin–eosin staining (HE staining)

Tissues from DM group, CON group, and SAL group were put in an oven at 37 °C for 10 min, and washed three times with PBST for 1 min each, then added hematoxylin staining solution for about 4 min, and rinsed with tap water (purple). Subsequently, blue-returning solution was added to return to blue, with quickly wash with tap water and then add differentiation solution was added, and rinsed with tap water for 5 min. Lastly, eosin was added for about 2 min, and all sections were passed 85%, 95%, 100% I, 100% II in sequence (about 1 min), and TOI, TOII was transparent for 3 min each, air-dried, sealed with neutral gum, and examined by microscopy.

Nissl staining

Sections from DM group, CON group, and SAL group were put in an oven at 37 °C for 10 min, similar with above procedure like HE staining. Nissl staining solution (covering the tissue) was then used to stain tissue for about 3–5 min, then pour off the dyeing solution with distilled water, and quickly pass 70%, 100%, TOI and TOII transparent for 3 min each, then air dry and seal with neutral gum lastexamine under microscope.

Immunofluorescence detection of retinal expressions of NLRP3, GSDMD, Caspase-1, IL-1β, and IL-18

After eyeball embedding (DM group, CON group, SAL group) tissues were washed with PBST for 5 times/1 min; 3% goat Serum + 0. 3% Triton-100, were used to incubate at room temperature for 3 h, then primary antibody including IL-1β(R)(1:400); IL-18(R)(1:200); GSDMD(R)(1:200);NLRP3(R)(1:400);Caspase-1(R)(1:300), were used to incubate tissue overnight at 4 °C; washed with PBS for 5 times/1 min; Next, the secondary antibody known as Goat Anti-Rabbit 488 was used at room temperature for 3 h; washed with PBS for 5 time/1 min; and observed by fluorescence microscope after mounting. Lastly, DAPI (1:3000) and anti-fluorescence and anti-fade mountant were overlap the sections to avoid fluorescent detection.

Gene query of diabetic retinopathy, pyroptosis, salidroside

Related genes forpyroptosis, and salidroside-related were downloaded from GeneCards (GeneCards—Human Genes|Gene Database|Gene Search), after inputing their related keywords.

Venny intersection diagram

In venny2.0.1, the three genes among diabetic retinopathy, pyroptosis and salidroside are used as the intersection genes among them. The URL is https://bioinfogp.cnb.csic.es/tools/venny/index.html [31, 40].

GO and KEGG analysis

The intersection genes were imported into to Metascape database(https://metascape.org/gp/index.html#/main/step1) to perform GO and KEGG analysis, we selected custom analysis, then select "BP", "Enrichment" in turn CC", "MF", and "KE GG" for analysis. Export the image for analysis [31, 40].

Protein–protein interaction and screening of Hub genes

In STRING, the website known as: https://cn.string-db.org/was used. After the interaction between genes is analyzed through cross genes, the link of multiple proteins are selected, and we performed click analysis to export the protein interaction diagram and the table of interaction relationship. The Hub gene was then screened in Cytoscape using protein interaction tables derived from PPI.

Molecular docking verification

We used the screened core genes to query the 2D protein structure of the gene in PDB at: https://www.rcsb.org/, and searched for the chemical structure of the drug in Pubchem at: https://pubchem.ncbi.nlm.nih.gov/, for molecular docking with Autodock software.

Real-time quantitative polymerase chain reaction(qRT-PCR)

Total RNA from the retina of three groups was extracted via using the Trizol Reagent(TaKaRa) and subject to reverse transcription via using T100TM Thermal Cycler(BIO-RAD). The RT-PCR analysis was performed in C1000 Touch TM Thermal Cycler(BIO-RAD) by using the SYBR Premix Ex TaqTM Kit (TaKaRa). The GAPDH gene was used as an endogenous control for sample normalization. All primer sequences are shown in Table 1.

Table 1 Primer sequences

Statistical processing

SPSS and PS software were used for statistical analysis and graphing of the obtained data. Measurement data in each group were expressed as mean ± standard deviation (x ± s), and one-way analysis of variance was used. P ≤ 0.05 was considered statistically significant.

Results

Observation of structural changes of rat retina by HE staining and Nissl staining

HE staining reported that the retinal layers were structurally complete and neatly arranged, with normal cell morphology, and the inner limiting membrane was clearly visible. However, the number of cells in the inner and outer nuclear layers decreased and the arrangement was sparse, in the DM group, in which, the ganglion cells were partially edema, and the blood vessel-like structures that broke through the inner and outer plexiform layers were seen. Moreover, the number of ganglion cells was reduced, and the boundary between the inner and outer nuclear layers was unclear, with disordered arrangement and vacuolar degeneration. Compared with the CON group, the retinal thickness in the DM group were definitely decreased (P < 0.001), while it becomes thickness in the SAL group, when compared with in the DM group (P < 0.001) (Fig. 1A).

Fig. 1

HE and Nissl staining of retinal tissue in rats with SAL treated with diabetic retinopathy. A HE staining of retina tissue of rats in each group and quantitative histogram of retinal thickness in each group of rats. B Nissl staining of retina tissue of rats in each group and quantitative histogram of total neuron counts in each group of rats. (CON: normal control group; DM: diabetic model group. RPE: retinal pigment epithelium; ONL: outer nuclear layer; INL: inner nuclear layer; RGC: retinal ganglion cell. Bar = 50 μm)

Nissl staining showed that the retinal structure of the rats in the normal control group was clear, with neatly arranged. Whereas, in the DM group, the number of ganglion cells appeared obvious edema with disordered and sparse distribution. Moreover, in the SAL group, the edema of cells in each layer was significantly reduced, with relatively regular arrangement, when compared with in DM one (P < 0.001) (Fig. 1B).

Immunofluorescence staining

The results of immunofluorescence staining showed that the five marker genes known as GSDMD, Il-18, Il-1β, NLRP3, and Caspase-1 present the fluorescence intensity in the normal group with the weakest level, but the DM group was the strongest, while it decreased in the SAL group. In detail, compared with the CON group,, the fluorescence intensity of Il-1β in the DM group got an increase (P < 0.001), while the fluorescence intensity of in the SAL group for Il-1β decreased significantly (Fig. 2A); For Il-18 and Caspase-1, the fluorescence intensity in DM group are also increased compared with a CON group, and they decreased in the SAL group, when compared with the DM group (Figs. 2B and 3A). The change of GSDMD is similar with above marker, its fluorescence intensity increased in the DM group (P < 0.001), and decreased in SAL group (Fig. 3B); Lastly, NLRP3 increased in the DM group, and decreased in SAL treatment group, also (P < 0.001) (Fig. 3C).

Fig. 2

Fluorescence layout (50 µm) and statistics of CON group, DM group, and SAL group. A Retinal fluorescence intensity comparison of IL-1β in each group. B IL-18 retinal fluorescence intensity comparison in each group

Fig. 3

Fluorescence layout (50 µm) and statistics of CON group, DM group, and SAL group. A Fluorescence intensity comparison in each group for Caspase-1. B Retinal fluorescence intensity of GSDMD groups. C NLRP3 retinal fluorescence intensity in each group

Screening of diabetic retinopathy genes, pyroptosis genes, and salidroside genes

To detect the molecular network mechanism, we performed work pharmacology analysis. Firstly of all, 3832 diabetic retinopathy genes, 254 pyroptosis genes, and 20 salidroside genes were collected from Genecards. Then, they were formed cross analysis. Next, the three groups of genes were analyzed in venny2.0.1, and therewere 6 intersecting genes, namely: NFE2 Like BZIP Transcription Factor 2 (NFE2L2), Sirtuin 1 (SIRT1), Nuclear Factor Kappa B Subunit 1 (NFKB1), NLR Family Pyrin Domain Containing 3( NLRP3), Caspase 3( CASP3), Parkin RBR E3 Ubiquitin Protein Ligase(PRKN), and Fig. 4 Gene expression is shown at the end of references, also Table 2.

Fig. 4

The intersection of diabetic retinopathy gene, pyroptosis gene and salidroside gene venny. Blue represents DR-related genes, yellow represents pyroptosis-related genes, and green represents salidroside-related genes

Table 2 Gene List

GO and KEGG analysis

The intersection genes were imported into to Metascape database to perform GO and KEGG analysis, we selected custom analysis, and selected "BP", "CC", "MF", "KEGG" in "Enrichment" for detail analysis. Only the biological process (BP) is enriched in the GO analysis. The first 9 pathways are: Signaling, response to stimulus, negative regulation of biological process, positive regulation of biological process, localization, metabolic process, biological regulation, regulation of biological process, biological process involved in interspecies interaction between organisms. In addition, two molecular functions (MF) were acquired which includes DNA-binding transcription factor binding, protein domain specific binding. At last, KEGG signaling pathways are pointed Lipid and atherosclerosis, Alcoholic liver disease, Parkinson disease (Fig. 5).

Fig. 5

GO and KEGG pathway analysis

Construction of PPI network and screening of Hub genes

In the STRING, the website is: https://cn.string-db.org/, and the interaction between genes is analyzed by using the intersection genes. It can be seen that an interaction relationship exists among 6 genes. According to the degree value, the Hub genes are screened and visualized, and sorted according to the degree value from large to small, in order: CASP3, SIRT1, NLRP3, NFE2L2, NFKB1, PARK2. (Fig. 6B, C).

Fig. 6

Construction of PPI network and screening of Hub genes. A Interaction network of intersecting targets. B Screening of hub genes. C The degree value of hub genes

Molecular docking verification

The screened core genes were imported into PDB (https://www.rcsb.org/) to query the 2D structure of the protein. Meawhile,the chemical structures of drugs were searched in the Pubchem database (https://pubchem.ncbi.nlm.nih.gov/). Then, molecular docking was performed with Autodock software. Consequently, SIRT1, NLRP3, NFE2L2, NFKB1, and PARK2 can form stable molecular structures with salidroside, while CASP3 cannot dock with salidroside. Through molecular docking verification, we speculated that NFE2L2, NFKB1, NLRP3, PARK2, and SIRT1 genes have regulatory effects on the pyroptosis of RGC, so as that they, could be considered as possible target for salidroside to prevent diabetic retinal damage (Fig. 7).

Fig. 7

Molecular docking verification. A Salidroside-NFE2L2. B Salidroside-NFKB1. C Salidroside-NLRP3. D Salidroside-PARK2. E Salidroside-SIRT1. F Salidroside-CASP3

qRT-PCR

Compared with the normal control group, the mRNA levels of NLRP3, NFE2L2, and NFKB1 in the diabetes model group were significantly increased, while they decreased in the SAL treatment group (P < 0.05), with the statistical difference. Differently, the mRNA level of SIRT1 in the diabetes model group was lower compared with the normal control group, and it got the lowest level with statistic significant, when compared with in DM group. Lastly,, the mRNA level of PARK2 in the diabetes model group increased, but there is no effect after SAL treatment (Fig. 8).

Fig. 8

The results of qRT-PCR. (From left to right, from top to bottom represent the relative expression of NLRP3, SIRT1, NFE2L2, NFKB1, PARK2 in CON, DM, SAL group)

Discussion

In this study, we found that the retinal layers in the CON group were the most complete and neatly arranged, with normal cell morphology, clearly visible internal limiting membrane, and ganglion cells arranged in a single layer. Comparatively, in the DM group, the number of ganglion cells gradually decreased with obvious edema increased, and the distribution was disordered and sparse. Whereas, in the SAL group, the edema of cells in each layer was significantly reduced, the arrangement tended to be regular. In addition, the fluorescence staining showed in DM group, the intensities of marker proteins were all the strongest and the positive value was high, while they decreased and became moderate in the SAL group, which is different from the lowest fluorescence in the CON group. In network mechanisms, pharmacology network analysis acquired 6 genes by venny intersecting, and the results of PPI analysis showed that there was a close relationship among 6 genes, and the NLRP3 gene had the highest comprehensive score. The verification of molecular docking showed that CASP3, other NFE2L2, NFKB1, NLRP3, PARK2 and SIRT1 could combine with salidroside, and qPCR verified the change of NLRP3, NFEZL2 and NFKB1. Our results reported that SAL inhibited effectively diabetic retinal RGC pyroptosis, which is associated with NLRP3, NFEZL2 and NFKB1 and multiple pathways, indicating that SAL could be considered as a potential drug to treat and protect DR, and underlying network pharmacological mechanism is involving in anti-inflammatory signal pathway.

HE and Nissl staining

In this study, HE and Nissl staining showed that the retinal layers of the CON group were the most complete and neatly arranged, and ganglion cells arranged in a single layer, while the retinal cells of the diabetic group were disordered, the ganglion cell layer and the inner and outer nuclear layers had obvious vacuolar degeneration. Comparatively the addition of SAL reversed these changes. These suggested that SAL could effectively inhibit retinal thinning, reduce cell damage and neuron loss to improve DR damage. It has been reported that (Ji et al. [8]) the retinal tissue of normal control rats did not have pathological changes, while in the diabetic model rats, it showed the significant retinal changes. One study showed that [30] morphology of the retinal cell did not obviously change after 4 w in DM group, nevertheless, retinal thickness was significantly thinner and RGC numbers were significantly reduced at 4 w. What’s more, it also presented retinal cell of INL and ONL arrangement disorder. Moreover, some studies have found that SAL had an inhibitory effect on nerve damage caused by many diseases [7] and SAL inhibited the glutamate-induced apoptosis of rat hippocampal neurons [32]. Here, we confirmed the effect of SAL in the improvement of morphological character in retina of DM,, which is useful to the usage of SAL in clinic.

Immunofluorescence analysis

In our study, the fluorescence of DM group for observed genes was the strongest, but the addition of SAL lowered fluorescence intensities. Yin et al. [37] observed that NLRP3 and CAP1 were localized in the RGC layer and INL by using immunohistochemistry. At the same time, they also found that the expressions of CASP1, NLRP3, and their downstream mature molecules IL-18 and IL-1B were increased in the retina of DM rats. Moreover, it has been reported that salidroside inhibited NLRP3-dependent pyroptosis in different disease [1, 2, 33], while there was not a research related SAL to ameliorate DR by inhibiting pyroptosis in previous study. Our findings provided new evidences to understand the effect of SAL in anti-pyroptosis of RGC after DM.

Intersection genes and construction of PPI network

In this study, six intersecting genes were obtained by venny, and the results of PPI analysis showed that there was a close relationship among the six genes, and the NLRP3 gene had the highest comprehensive score. Previously, it has been known that the sirtuin (SIRT) family were involved in the development of various diseases such as neurodegeneration, cardiovascular pathologies, metabolic disorders, and cancer. SIRT1, 3, 5, and 6 were key enzymes in DR since they modulated glucose metabolism, insulin sensitivity, and inflammation [29]. Comparatively, in the present study, we showed that [34] NFE2L2 was an important component of the intracellular antioxidant machinery, in which, NFE2L2 could be considered as a target for treatment of diabetic complications. NF-KB is a nuclear transcription factor that can regulate the expression of various genes in inflammatory response, immune response, cell proliferation and apoptosis. The continuous activation of NF-KB increases the release of inflammatory factors in the inflammatory response, and regulates cell proliferation and apoptosis [22]. Numerous studies have demonstrated that the NLRP3 inflammasome plays an important role in the pathogenesis of various diseases [6], NLRP3 inflammasome retina caused by early hyperglycemia, and affecting the structure and function of the blood-retinal barrier [24]. Our results suggested that NLRP3 inflammasome and related proteins have been involved in the process of SAL administration after DR.

GO and KEGG analysis

In this study, we found that the first 9 pathways of BP were enriched. There were two MF that is DNA-binding transcription factor binding, protein domain specific binding. Whereas, KEGG signaling pathways are: Lipid and atherosclerosis, Alcoholic liver disease, Parkinson disease. Multiple genes network analysis to uncover the mechanism of drug action is a hotspot based on bioinformatics and network pharmacology [31, 40]. Recent research showed that [20] SAL flattened high glucose-induced injury in retinal pigment epithelium cells by activating PI3K/AKT and AMPK signaling pathways. What’s more, SAL could suppress the P2X7/NF-KB/NLRP3-mediated pyroptosis [2]. In addition, SAL could decrease the expression of TLR4, NF-κB, P-NF-κB, NLRP3, ASC, cleaved Caspase-1, cleaved GSDMD, IL-1β, and IL-18 by inhibiting TLR4/NF-κB/NLRP3/Caspase-1 signaling pathway [1]. Together, all results showed that SAL may regulate DR to inhibit pyroptosis via various signaling pathways combined with our findings.

Molecular docking and qRT-PCR validation

In our study, the verification of molecular docking showed that except CASP3, other NFE2L2, NFKB1, NLRP3, PARK2 and SIRT1 could combine with SAL, and qRT-PCR confimed the change of mRNA levels for NLRP3, NFE2L2, PARK2 and NFKB1.These results suggested that NFE2L2, NFKB1, NLRP3, PARK2 and SIRT1 might be proposed as new therapeutics to treat DR. Mortuza et al. showed that the expression of SIRT1 was decreased in human retinal endothelial cells with high glucose concentration [18]. Similarly, in retinal endothelial cells, hyperglycemia determines SIRT1 down regulation followed by a decrease of mitochondrial antioxidant enzymes levels through pathways controlled by p300 and Fork head box protein O1 [42]. These evidences pointed that SIRT1 was increased in DM groups. Comparatively, NFE2L2 had pivotal roles in many signaling pathways that were altered in the retina in diabetes, and were implicated in the development of DR [12]. Luo et al. [16] found that NFE2L2 were decreased in blood samples of DR patients and high glucose-treated human RPE and ARPE-19 cells. Whereas, the overexpression of NFE2L2 promoted proliferation and suppressed apoptosis and inflammation. This was exactly the opposite of the results of this experiment, which may be caused by different model condition, needing to be verified in later experiment.

Although, previous NF-KB studies have found that NLRP3 inflammasome was involved in the formation of pathological retinal neovascularization by establishing advanced DR animal models, it is lack evidence to show the relation between SAL and NLRP3 [3, 24].

Our results are the first time to show the effect of SAL in DR, which is associated with several molecular network that has been reported previously [9, 11, 28].

Conclusion

The main novel findings are that salidroside can significantly improve the morphological retinopathy in diabetic rats, especially for RGC, in which, the underlying mechanism is related to the regulation of (NLRP3, GSDMD, Caspase-1, IL-1β, IL-18, Of them, NLRP3, NFE2L2 and NFKB1 could be considered as the direct target of SA, so as to provide the protection for RGC in our experimental condition.