Metformin hydrochloride (D9351) was purchased from Solarbio Company; glucose (50–99-7) was obtained from China National Pharmaceutical Group Chemical Reagent Co., Ltd.; the reverse transcription cDNA kit (A5001) was sourced from Promega, United States; the Superoxide Dismutase (CuZn-SOD/Mn-SOD) assay kit (hydroxylamine method) (E-BC-K022-M), Catalase (CAT) colorimetric assay kit (E-BC-K031-M), Malondialdehyde (MDA) colorimetric assay kit (E-BC-K025-M), and ATP content chemiluminescent assay kit (E-BC-F002) were purchased from Wuhan Elabscience Biotechnology Co., Ltd.; the Zebrafish Fructosamine (FRA) ELISA kit (KT5402-A) was obtained from Jiangsu KET Biological Technology Co., Ltd.; the Fast antibody dye-based quantitative PCR premix (MQ10401S) was purchased from Mona Biotechnology Co., Ltd.; primers were synthesized by Shanghai Shenggong Biotechnology Co., Ltd. Cleaved-caspase 3 antibody (GB11532), Alexa Fluor 488 labeled goat anti-rabbit IgG (GB25303); One-step TUNEL cell apoptosis detection kit (TMR red fluorescence) (G1502-50T) were purchased from Wuhan Servicebio Biotechnology Co., Ltd.

We obtained the following: The Roche Blood Glucose Meter Excellence Precision Model (Roche Diagnostics Products Shanghai Co., Ltd.); ZebraTower Automated Zebrafish Analysis System Adult Observation Tower (ViewPoint, France); Fishbio-d5 Zebrafish Double-Row Five-Layer Circulating Rearing Rack (Shanghai Fexi Biotechnology Co., Ltd.); Medifuge™ Small Desktop Centrifuge (Thermo Fisher Scientific, United States); Cytation3 Multifunctional Microplate Reader (BioTek, United States); T100™ Real-Time Fluorescent Quantitative PCR (Real-time qPCR) System (Bio-Rad, United States); HM355S Paraffin Microtome (Thermo Fisher Scientific, United States); Microscopic Imaging System (Motic, China).

Male AB strain zebrafish, aged 6–7 months, were purchased from the National Zebrafish Resource Center and maintained in a zebrafish breeding system. Rearing conditions included pH (6.8–8.5), temperature (24–28°C), alkalinity (50–100 mg/L), salinity (0.5–2.0 g/L), dissolved oxygen (>4 mg/L), and a light/dark cycle of 14/10 h. The zebrafish were fed twice daily. All procedures were approved by the Ethics Committee of Hunan University of Chinese Medicine (Approval No. LLBH-202311100013).

Seven to eight month old AB strain zebrafish were randomly selected and bred in a breeding tank, fed with Artemia. Based on the results of acute toxicity testing, the concentrations were set as low, medium, and high doses at 50, 100, and 200 μg/mL, respectively. Additionally, positive control drug metformin and model and control groups were included. The modeling lasted for 14 days. A total of 240 adult zebrafish aged 6–7 months were randomly assigned to the control group, model group, and three PSP dose groups (50, 100, and 200 μg/mL), as well as the metformin group (30 μg/mL), with 40 fish in each group. The control group was maintained in a standard aquatic environment, while the model group, PSP groups, and metformin group were exposed to a 2% glucose solution (10 fish/1500 mL). They were fed with three times the standard amount of Artemia daily, continuously exposed for 28 days to induce a hyperglycemic model. Starting from day 14, the low, medium, and high doses of PSP and metformin groups received continuous treatment for 14 days.

During the acute toxicity assay, larvae were maintained in 6-well plates at a density of 10 embryos per well (3 mL solution/well). Water quality (pH, dissolved oxygen, temperature) was monitored daily, and 50% of the solution was replaced every 24 h to ensure optimal conditions. The groups were set as control, 50 μg/mL PSP, 100 μg/mL PSP, 200 μg/mL PSP, and 400 μg/mL PSP. The experiment lasted for 96 h, with observations and recordings of zebrafish hatching and developmental conditions taken every 8 h. The PSP concentrations (50–400 μg/mL) were selected using a tiered approach: lower doses (50–200 μg/mL) aligned with preliminary tests and environmental exposure data, while the highest dose (400 μg/mL) served as a conservative safety margin for extreme toxicity assessment (16–18).

Prior to body weight and blood glucose measurements, zebrafish were fasted for 12 h. The zebrafish to be tested were collected, anesthetized with an ice bath, surface water was removed, and they were placed on a tray. Body weight was recorded and measured both before modeling and prior to blood collection. The tail was amputated for blood collection, and blood glucose levels were quickly measured using a Roche blood glucose meter. Zebrafish tissue was collected and homogenized in an appropriate volume of physiological saline under ice bath conditions. The mixture was centrifuged at 3000 rpm for 10 min, and the supernatant was collected. Fructosamine was measured using undiluted supernatant per ELISA kit protocol.

Following the method described by Yu et al. (19), behavioral testing was conducted on zebrafish at the end of the 28-day intervention period. Zebrafish were observed using the ZebraTower Adult Observation Tower to record their locomotor behavior. The total swimming distance over a 60-s period and the swimming speed (high speed >5 cm/s, moderate speed 2–5 cm/s, low speed <2 cm/s) were recorded and statistically analyzed using the zebrafish behavioral analysis system. Adult zebrafish subjected to behavioral analysis were subsequently used for blood glucose measurement to ensure cohort consistency.

After behavioral testing, zebrafish were anesthetized with an ice bath. The pancreas was dissected according to the zebrafish anatomical diagram and homogenized in an appropriate volume of PBS under ice bath conditions. Tissues were homogenized in PBS (1:9w/v) and centrifuged (3,000 rpm, 10 min). Supernatants were diluted 1:10 for SOD/CAT and 1:5 for MDA/ATP. The activities of SOD, CAT, MDA, and ATP content were measured following the instructions provided with the biochemical assay kits.

Zebrafish pancreatic and intestinal tissues were fixed in 4% paraformaldehyde, embedded in paraffin, sectioned, and baked. The sections were deparaffinized, stained with hematoxylin for 10 min, differentiated with hydrochloric alcohol, blued, stained with eosin for 2 min, and subjected to gradient dehydration. The sections were then cleared with xylene and mounted. Morphological and structural changes in the pancreatic tissue were observed under an optical microscope. Image analysis was performed using ImageProPlus 6.0 software.

Zebrafish pancreatic tissue was collected, and the tissue from every five zebrafish was placed in a 1.5 mL EP tube for total RNA extraction using the Trizol method. After measuring the RNA concentration using a multifunctional microplate reader, cDNA was synthesized using a reverse transcription kit. The reaction program was as follows: 95°C for 5 min; 95°C for 30 s; 58°C for 30 s, for 40 cycles. GAPDH was used as the reference gene, and data analysis was performed using the 2^-△△Ct method. The primers used were designed and synthesized by Sangon Biological Engineering (Shanghai) Co., Ltd. (Table 1).

The active ingredients of P. sibiricum were retrieved from the TCMSP database¹, with screening criteria set at oral bioavailability (OB) ≥ 30% and drug-likeness (DL) ≥ 0.18 to ensure pharmacological relevance. Potential targets of the compounds were predicted using Swiss Target Prediction², with a probability threshold of >0.5 to enhance prediction reliability.

Diabetes-related targets were obtained from the GeneCards database³ by searching for “diabetes,” and only targets with a relevance score ≥ median were retained to prioritize high-confidence associations. Common targets between P. sibiricum and diabetes were identified through Venn analysis⁴.

The protein–protein interaction (PPI) network was constructed using STRING⁵ with a confidence score cutoff ≥ 0.4 to ensure robust interactions. Functional enrichment analysis, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, was performed using DAVID⁶, with a significance threshold of p < 0.05. Networks were visualized using Cytoscape 3.10.2.

The structure was obtained by searching “Polygonatum sibiricum polysaccharides” in SciFinder⁷ and downloading the corresponding SDF files. The PDB files for “AMPK” and “SIRT1” were downloaded from the PDB database⁸. The PDB files of the relevant proteins and the SDF files of the small molecules were input into the CB-Dock2 website⁹ for molecular docking, and the Vina scores and docking diagrams were generated. The resulting Vina scores were uploaded to the Microbioinformatics platform¹⁰ for further analysis.

Paraffin sections were dewaxed with xylene, and treated with DNase- and proteinase K-free solution (20 mg/L) for 30 min at room temperature. After washing with PBS three times, TUNEL detection solution was added, followed by incubation at room temperature in the dark for 60 min. The sections were washed with PBS three times, and mounted with an anti-fade reagent containing DAPI. Imaging was performed using a panoramic scanner, and three fields of view were selected from each sample to calculate the TUNEL-positive cell rate.

Paraffin sections were baked, permeabilized, and subjected to antigen retrieval. After blocking, primary antibody incubation was performed overnight at 4°C. The following day, the sections were incubated with fluorescent secondary antibody for 1 h in the dark. After staining, the slides were mounted with an anti-fluorescence quenching agent containing DAPI. Imaging was performed using a whole-slide scanner, and the average fluorescence intensity of each group was analyzed using Image J.

Tissues (100 mg) were homogenized in RIPA lysis buffer and incubated at 4°C, followed by sonication and centrifugation at 12,000 rpm for 15 min (centrifuge radius: 5 cm). The supernatant was collected, and protein concentration was measured using a BCA protein assay kit. A loading buffer with a protein concentration of 2 g/L was prepared and heated at 100°C for 20 min. A total of 20 μg of protein was loaded onto a 10% SDS-PAGE gel, with electrophoresis run at 80 V for 30 min, followed by 120 V for 60 min. Proteins were transferred to a 0.45 μm PVDF membrane by wet transfer for 120 min. The membrane was blocked with 5% skim milk at room temperature for 1 h. The primary antibody against cleaved-caspase3 (1:8000) was incubated overnight at 4°C, followed by incubation with the corresponding secondary antibody (1:10,000) for 1 h at room temperature. Protein bands were detected using ECL chemiluminescent substrate and analyzed with Image Lab software for densitometric analysis.

Experimental data are expressed as mean ± standard deviation, and statistical graphs were generated using Prism GraphPad 8.0. Data analysis was performed using SPSS 25.0. For multiple group comparisons, one-way ANOVA was used. If variances were homogeneous, the LSD method was applied; if variances were heterogeneous, the Tamhane’s T2 method was used. A p-value of less than 0.05 was considered statistically significant.

Zebrafish embryos were exposed to PSP from fertilization until hatching was completed. The results showed that exposure to 50, 100, 200, and 400 μg/mL PSP did not affect the normal hatching process of the embryos. The morphology of the zebrafish larvae across all treatment groups was consistent with that of the control group, and no abnormal behaviors were observed (Figure 1A). The hatching rate was 100% in all groups, with slight differences in the timing of hatching observed, though all embryos successfully hatched within 96 h post-fertilization (Figures 1B,C).

The total swimming distance of zebrafish in the model group was significantly increased compared to the control group (p < 0.01), while the time spent swimming at moderate speed was notably reduced (p < 0.01), and the time spent swimming at high speed was significantly increased (p < 0.01). In contrast, the PSP-treated groups (all doses) and the metformin group showed a reduction in total swimming distance compared to the model group, though these differences were not statistically significant (p > 0.05). Additionally, the time spent swimming at moderate speed was significantly increased (p < 0.05, p < 0.01), while the time spent swimming at high speed was significantly decreased (p < 0.01) in the PSP-treated groups and metformin group (Figure 2).

Body weight measurements taken before and after treatment revealed a significant increase in the model group compared to baseline (p < 0.01), while no significant changes in body weight were observed in the other groups (p > 0.05). The body mass index (BMI) was significantly higher in the model group compared to the control group (p < 0.01). In contrast, the BMI of zebrafish in the PSP-treated groups (50, 100, and 200 μg/mL) and the metformin group showed significant reductions compared to the model group (p < 0.01) (Figures 3A,B).

Blood glucose levels and fructosamine content were significantly elevated in the model group compared to the control group (p < 0.01). Treatment with PSP at 50 μg/mL, 100 μg/mL, 200 μg/mL and metformin significantly reduced both blood glucose levels and fructosamine content (p < 0.01), suggesting that PSP exerts a clear hypoglycemic effect (Figures 3C,D).

Biochemical analysis showed that the ATP content, SOD and CAT activities in the model group were significantly decreased (p < 0.01), whereas MDA activity was significantly increased (p < 0.01). In the 50 μg/mL PSP-treated group, ATP content and CAT activity remained unchanged (p > 0.05), but SOD activity was significantly increased (p < 0.01), and MDA activity was significantly decreased (p < 0.05). In the 100 μg/mL, 200 μg/mL PSP-treated groups and metformin group, ATP content, SOD and CAT activities were significantly increased (p < 0.05, p < 0.01), while MDA activity was significantly decreased (p < 0.01), indicating that PSP demonstrates strong antioxidant properties (Figures 3E–H).

The HE staining showed distinct morphological differences between the groups. In the control group, pancreatic tissue exhibited normal architecture with intact acinar cells containing abundant zymogen granules, and well-defined pancreatic islets with clear boundaries and no signs of inflammation. In the model group, severe pancreatic damage was observed, including loss of zymogen granules in acinar cells, decreased islet volume, blurred boundaries, and a marked reduction in the number of islet cells. Treatment with PSP at 50, 100, and 200 μg/mL, as well as metformin, significantly alleviated these pathological changes. PSP treatment partially restored pancreatic architecture, including partial restoration of acinar cell granules, moderately increased islet volume, and mild islet cell proliferation. Although some acinar cells still exhibited granule loss and blurred boundaries, the overall tissue structure was notably improved (Figure 4A).

Intestinal tissue analysis revealed substantial differences between groups. In the control group, intestinal tissue showed a uniform distribution of villi, smooth boundaries, and no signs of inflammation. The villi were tall and slender, with intact columnar epithelial cells and goblet cells, and the lamina propria showed no immune cell infiltration. In the model group, the intestinal tissue exhibited severe damage, including villus loss, bifurcation, ruptured villus tips, and significant injury to the columnar epithelial cells. Extensive neutrophil and mononuclear cell infiltration indicated a marked inflammatory response. Treatment with PSP at 50, 100, and 200 μg/mL, as well as metformin, significantly improved the intestinal tissue morphology. PSP-treated groups exhibited fewer bifurcations and ruptures of the villi, clearer villus boundaries, and a marked reduction or absence of inflammatory cell infiltration, suggesting that PSP helps mitigate hyperglycemia-induced intestinal damage and inflammation (Figure 4B).

Through network pharmacology analysis, we identified a total of 442 targets associated with P. sibiricum and 16,075 targets related to diabetes, with 389 common targets (Figure 5A and Supplementary Tables S1–S3). These common targets underwent protein–protein interaction (PPI) analysis (Figure 5B and Supplementary Table S4), which revealed key biological processes including “AMPK signaling pathway,” “insulin resistance,” and “insulin signaling pathway” (Figure 5C and Supplementary Table S5). Key signaling pathways identified in the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis included the “inflammatory response,” “negative regulation of apoptotic process,” “insulin-like growth factor receptor signaling pathway,” “insulin receptor signaling pathway,” and “cellular response to insulin stimulus” (Figure 5D and Supplementary Table S6).

Molecular docking simulations were conducted to explore potential interactions between PSP components and key targets involved in glucose metabolism, namely AMPK and SIRT1. Eight PSP-related structures were identified, including Polysucrose, 6-O-Tosyl-β-cyclodextrin, Polyglucose, Poly-d-galacturonic acid, Poly(mannuronic acid), Polymaltose, Galacturonic acid, and D-Mannuronic acid (Supplementary Figure S1). Docking simulations revealed that all eight PSP components exhibited strong binding affinities with AMPK and SIRT1, suggesting that they may play key roles in PSP’s hypoglycemic effects. Among these, 6-O-Tosyl-β-cyclodextrin, Polysucrose, Polymaltose, and Galacturonic acid displayed particularly strong binding interactions, suggesting their primary involvement in PSP’s biological activity (Figures 6A,B).

In comparison with the control group, the model group showed significantly upregulated mRNA expression of inflammatory markers (IL-1β, IL-6, TNF-α) and the mitochondrial damage gene Cycs (p < 0.01), and a significant downregulation of the glucose metabolism gene Pck2 (p < 0.01). Treatment with PSP at doses of 50, 100, and 200 μg/mL, as well as metformin, significantly reduced the mRNA expression of inflammatory factors and mitochondrial damage markers (p < 0.01) and significantly upregulated Pck2 expression (p < 0.01) (Figure 7A).

Compared to the control group, the model group exhibited significantly downregulated mRNA expression levels of AMPK, SIRT1, PCG-1α, Nrf1, and Nrf2 (p < 0.01). In the 50 μg/mL PSP-treated group, the mRNA expression levels of AMPK, Nrf1, and Nrf2 remained unchanged (p > 0.05), while those of SIRT1 and PCG-1α were significantly upregulated (p < 0.05, p < 0.01). In the 100 μg/mL, 200 μg/mL PSP-treated groups and the metformin group, the expression of AMPK, SIRT1, PCG-1α, Nrf1, and Nrf2 were significantly increased (p < 0.05, p < 0.01) (Figure 7B).

Caspase 3 mRNA expression was significantly elevated in the model group compared to the control group (p < 0.01), indicating extensive β-cell apoptosis. PSP treatment at low, medium, and high doses (50 μg/mL, 100 μg/mL, 200 μg/mL) dose-dependently reduced Caspase 3 expression by 36, 52, and 74%, respectively (p < 0.01). The high-dose PSP group showed similar efficacy to the metformin-positive control, which reduced Caspase 3 expression by 68% (p < 0.01) (Figure 8A). Furthermore, mRNA levels of Bcl-2 were significantly decreased, while BAX expression was significantly increased in the model group. PSP treatment mitigated these changes (p < 0.01).

Western blot analysis confirmed that cleaved Caspase-3 protein levels were significantly higher in the model group compared to the control, but PSP and metformin treatments effectively reduced these levels (p < 0.01) (Figure 8B). Fluorescence intensity analysis showed a marked reduction in cleaved Caspase-3 fluorescence in the PSP-treated groups compared to the model group (p < 0.01) (Figure 8C). TUNEL staining further confirmed that β-cell apoptosis was significantly reduced in the PSP-treated groups (p < 0.01) (Figure 8D).