CDDP was gifted by Tianjin Tasly Pharmaceutical Group Co. Ltd. (Tianjin, China). Doxium (calcium dobesilate, CaD) was purchased from Ebewe Pharma Ges. m. b. H. Nfg. KG (Ontara, Austria). Shandong Bausch Lomb Freida Pharmaceutical Co. Ltd provided the carbomer eye drops (Jinan, China). Santen Pharmaceutical (China) Co. Ltd. provided the tropicamide phenylephrine eye drops (Suzhou, China). Fluorescein sodium injection was provided by Alcon Laboratories (Texas, United States). Biosino Biotechnology and Science Inc provided the glycosylated hemoglobin (HbA1c) kits (Beijing, China). Mouse intercellular cell adhesion molecule-1 (ICAM-1), interleukin-1 beta (IL-1β), interleukin-6 (IL-6), and monocyte chemoattractant protein-1 (MCP-1) enzyme-linked immunosorbent assay kits were obtained from Beijing Soleibao Technology Co. Ltd. (Beijing, China). Terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) and immunohistochemistry kits were purchased from Beyotime Biotechnology Co. LTD (Shanghai, China). The Bax (ab32503) and Bcl-2 (ab182858) primary antibodies were obtained from Abcam (Cambridge, United Kingdom). The remainder of the reagents were bought from Sigma-Aldrich (St Louis, MO, United States).

CDDP was prepared by Tianjin Tasly Pharmaceutical Group (Supplementary Table S1). Briefly, Salvia miltiorrhiza Bunge and Panax notoginseng (Burkill) F.H.Chen were decocted with water and filtered, and the filtrate was concentrated. Add ethanol, leave it to stand and take the supernatant to concentrate. Take appropriate amount of polyethylene glycol and melt it with heat and add the concentrated paste obtained above and ground Cinnamomum camphora (L.) J.Presl, drop it into cooled liquid paraffin to get to the Dripping Pills. Separation was performed on Waters Acquity UPLC™ HSS T3 column (2.1 mm × 100 mm, 1.8 μm). The mobile phase was 80% acetonitrile solution containing 0.02% phosphoric acid (A) and 0.02% phosphoric acid (B). The elution gradient was 0–1.6 min, 9%–22% A, 91%–78% B; 1.6–1.8 min, 22%–26% A, 78%–74% B; 1.8–8.0 min, 26%–39% A, 74%–61% B; 8.0–8.4 min, 39%–9% A, 61%–91% B; 8.4–10.0 min, 9% A, 91% B; the flow rate was 0.4 mL/min. The CDDP standard used in this study complied with the Pharmacopoeia of the people’s Republic of China (2020 edition) and was analyzed for quality control of CDDP, and the relevant fingerprints are shown in Supplementary Figure S2.

GemPharmatech Co., Ltd provided 45 male db/db mice aged 20 weeks and 15 male db/m mice aged 20 weeks (Nanjing, China) and raised in SPF-class laboratory housing at the Institute of Medicinal Plant Development (IMPLAD), Chinese Academy of Medical Sciences. The mice lived in a temperature-controlled (temperature: 22°C ± 1°C) facility and could eat and drink freely. They acclimatized in this facility for 1 week before the study began. The 15 db/m mice were used as the controls (I), and the db/db mice were randomly allocated to three groups, namely, the model (II), CDDP (500 mg/kg, i.g.) (III), and positive control CaD (227.5 mg/kg, i.g.) (IV) groups. The body weights of the mice were measured every week. Groups III and IV received orally administered CDDP and CaD for 6 weeks (Figure 1A), whereas groups I and II received saline.

All animal experiments were approved by the Experimental Animal Ethics Committee of the Institute of Medicinal Plant Development, Peking Union Medical College, in accordance with the National Guidelines for the Care and Use of Animals (SLXD-20190711001).

All mice were fasted overnight after CDDP treatment, and HbA1c levels were determined using a detector (SANNUO, China) according to the manufacturer’s recommendations.

Following anesthesia with 1% pentobarbital sodium and 10% sumianxin injection, the mice received tropicamide phenylephrine eye drops to enlarge their pupils and carbomer eye drops to shield them. Fluorescein sodium (10%) was then injected intraperitoneally, and a digital fundus camera was used to identify FFA 2 minutes later (Retinal Imaging System, OptoProbe Research Ltd., Burnaby, Canada). Fluorescent images of retinas were collected and the vessel densities were calculated using AngioTool software (version 1.8.0) according to our latest study (Gao et al., 2020).

Mice were anesthetized with 1% pentobarbital sodium and 10% sumianxin injection. Tropicanamide phenylephrine eye drops were used to dilate the pupils and protected by carbomer eye drops. Mice were then placed on a bench, and their position was fixed according to the head of optic nerve. Retinal OCT was carried out utilizing a retinal imaging system (isOCT, 4D-ISOCT Optoprobe, Burnaby, Canada), and software (Version 2.0) from OptoProbe Research Ltd. was used to determine the thicknesses of the retinal layers.

Mice were maintained in darkness overnight. The following day, 1% pentobarbital sodium was administered intraperitoneally to mice to make them unconscious and 10% sumianxin in normal saline. Tropicanamide phenylephrine eye drops were used to dilate the pupils and protected by carbomer eye drops. A gold-wire electrode was positioned on the corneal surface of the right eye as a positive pole and another gold-wire electrode was positioned as a negative pole in the flew. The ground was provided via a needle electrode in the tail. The unstimulated left eye was covered with a dark patch. A visual electrophysiological device was used to measure the retinal ERG (OPTO-III, Optoprobe, Burnaby, Canada).

The eyeballs of the mice were harvested, and the retinas were separated under a stereomicroscope. The retinas were dehydrated and embedded in paraffin after being fixed with 4% paraformaldehyde. Paraffin sections (5 µm) were obtained using a Leica RM2235 tissue slicer (Leica, Germany), and were stained using hematoxylin and eosin (HE) as described in our previous study (Luo et al., 2015). A digital slide image scanning method was used to obtain retinal histopathology pictures (Aperio CS2, Leica, Germany).

The levels of serum sICAM-1, IL-6, and MCP-1 were determined by ELISA, in accordance with the provided protocols.

The extraction of total RNA, reverse transcription to cDNA, RT-PCR and data analysis were performed as described in our previous study (Luo et al., 2015). The primers were designed using premier primer Software 6.0 (Canadian Premier Life Insurance Company, Toronto, Canada) and the sequences are shown in Table 1.

The IHC assay was performed according to the DAB kit instructions. Briefly, 5-μm retinal sections were deparaffinized in xylene and ethanol at various concentrations (100%, 100%, 90%, 80%, and 70%). The sections were then applied for antigen restoration and incubated with primary antibodies at 4 °C for 12 h. Following washing, sections were incubated for 1 h at 37°C with secondary antibodies. DAB was used for visualization and the images were collected using the Leica’s Aperio CS2 digital slide image scanning device, and analyzed by Image Pro Plus (version 6.0; Media Cybernetics, Maryland, United States).

Following the manufacturer’s recommendations, the TUNEL assay was used to find retinal cell apoptosis. The TUNEL assay was performed according to our previous study (Gao et al., 2020). Briefly, retinal paraffin sections were deparaffinized, incubated with proteinase K for 30 min and with 3% H₂O₂ for 10 min. Sections were then incubated for 1 h at 37°C with a biotin-labeled TUNEL reaction mixture, followed by 30 min at 25°C with streptavidin-HRP working solution. Finally, sections were developed using DAB dye. The number of positive cells was determined after images were captured using a digital slide image scanning system (Aperio CS2, Leica, Germany).

Data are presented as mean ±standard error (SEM) using the GraphPad Prism software (version 6.0; San Diego, CA, United States). One-way analysis of variance (ANOVA) and Tukey’s post hoc test were used for comparison of multiple groups and p < 0.05 was considered statistically significant.

To determine whether CDDP influences DR, we first monitored the features of retinal morphological. As shown in Figures 1A, B, the retinas were edematous in the model group with disordered arrangement of the retinal cells, whereas these changes were significantly ameliorated after CDDP treatment. Body weight and blood glucose levels are two fundamental indices of diabetes. As shown in Figures 1C, D, CDDP treatment had no obvious influence on body weight and HbA1c levels, which indicated that the anti-DR effect of CDDP is independent of its hypoglycemic effect.

Retinal microangiopathy is an early feature of DR; therefore, we evaluated vascular permeability and neovascularization using FFA, and OPs waves by ERG. Figures 2A, D show that there were increased numbers of vessels mice in the DR model group compared to the control group. However, the vessel numbers were significantly reduced following CDDP treatment relative to the DR model group. Moreover, CDDP markedly increased the OPs wave amplitude but had no effect on the b/a ratio (Figures 2B, C, E). Taken together, these results show that CDDP and CaD both protect against vasculopathy in DR mice.

Retinal neuropathy is another early feature of DR; thus, we analyzed the retinal thickness of each retinal layer using OCT. As depicted in Figures 3A, B, the retinal GCL + IPL, IS/OS, and RPE layers in the model group were markedly reduced in comparison with the controls. However, following CDDP treatment, the thickness of the GCL+IPL layer was markedly increased, while no obvious effects were seen on the IS/OS and RPE layers. This suggests that CDDP protects against neuropathy in mice with early DR.

Inflammation is a key factor in DR pathogenesis (Forrester et al., 2020). Therefore, we measured the expression of inflammatory factors. CDDP markedly reduced the levels of IL-1β, IL-6, and TNF-α in retinal tissue (Figure 4A), accompanied by reduced expression of MCP-1, IL-6 and sICAM-1 in the sera (Figure 4B), suggesting that CDDP exerted its anti-DR effects by counteracting inflammation.

Excessive VEGF causes pathological vascular permeability and neovascularization, both of which are characteristic of DR (Campochiaro and Akhlaq, 2021). As indicated in Figures 5A, B, CDDP significantly inhibited VEGF expression in the retinas of DR mice.

Retinal cell apoptosis is an early feature of DR and was thus evaluated. Figures 6A, B indicate that relative to the controls, increased number of TUNEL-positive cells were observed in the model group, which was reduced following CDDP treatment. And as shown in Figures 6C, D, increased levels of Bcl-2 together with reduced Bax were further indications of the anti-apoptotic effects of CDDP.