Gastrodin (Cat no. B21243) was purchased from Shanghai Yuanye Bio-Technology Co., Ltd (Shanghai, China). Human TGF-β1 (Cat no. AF-100-21C) was obtained from PEPROTECH (Rocky Hill, NJ, United States). Minimum Essential Medium (MEM) Alpha Medium was purchased from Corning (Canton, NY, United States). Fetal bovine serum (FBS), trypsin-EDTA, and bicinchoninic acid (BCA) protein assay reagent kit were purchased from Thermo Fisher Scientific (Waltham, MA, United States). Antigen repair solution (Cat no. MVS-0066) and UltraSensitive™ SP (mouse/rabbit) immunohistochemistry (IHC) (Cat no. KIT-9720) and DAB (Cat no. DAB-0031) kits were purchased from Maixin Biotechnology (Fuzhou, Fujian, China). Masson (Cat no. G1340) and Sirius red (Cat no. PH1099) staining kits were purchased from Solarbio Science and Technology (Beijing, China). Polyvinylidene fluoride (PVDF) membranes were obtained from Millipore (Billerica, MA, United States). Antibodies against α-SMA (Cat no. 19245S) and p-Smad2 (p-Smad2; Cat no.18338S) were purchased from Cell Signaling Technology (Danvers, MA, United States). Antibodies against α-SMA (Cat no. 40482), p-Smad2 (Cat no. 13429), p-Smad3 (Cat no. 12838), Smad2(Cat no. 41442), Smad3(Cat no. 41445) and anti-mouse (Cat no. L3032) and anti-rabbit (Cat no. L3012) secondary antibodies were purchased from Signalway Antibody (College Park, MD, United States). Antibodies against collagen I (Cat no. 14695-1-AP), collagen III (Cat no. 22734-1-AP), and fibronectin (Cat no. 15613-1-AP) were purchased from Proteintech (Rosemont, IL, United States). Antibodies against TGF-β1 (Cat no. ABP52598) and cell counting kit-8 (Cat no. ktc011001) were obtained from Abbkine Scientific (Wuhan, Hubei, China). Antibodies against p-Smad3 (p-Smad3; Cat no. ab52903) and goat polyclonal secondary antibody to rabbit IgG-H&L (Alexa Fluor^® 647, Cat no. ab150079) were purchased from Abcam (Cambridge Science Park, Cambridge, UK). Primary and secondary antibody dilution buffers were supplied by Beyotime Biotechnology (Shanghai, China).

For animal studies, gastrodin was dissolved in double-distilled water (ddH₂O) to obtain a dose of 3.5 mg/kg/day based on the average body weight of rats. The dose were selected based on unpublished preliminary study in mice (5 mg/kg/day) adjusting for body surface area and calculating by the following formula: The dose for rat = the dose for mouse (5 mg/kg/day)*body weight of mouse (0.02 kg)* conversion factor (7)/body weight of rat (0.2 kg) = 3.5 mg/kg/day. For cell experiments, the stock solution of gastrodin (1 M) was prepared with dd H₂O with completely medium to indicate concentrations before use.

Ten 4-week-old female spontaneously hypertensive rats (SHRs) and five female Wistar Kyoto (WKY) rats were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd (Beijing, China). After 2–3 days of adaptive feeding before the experiment, SHRs were randomly categorized into the model (SHR) and treatment (SHR + Gastrodin) groups, with five rats in each group. The five WKY rats served as the control group (WKY). The SHR + Gastrodin group was intragastrically administered with gastrodin daily (3.5 mg/kg/day; 1 ml for each rat), whereas both the WKY and SHR groups were intragastrically administered with equal volumes of dd H₂O for 10 weeks. All animals were maintained under specific pathogen-free conditions at a constant temperature of 24°C ± 2°C and relative humidity of 50%–60% with a 12-h light–dark cycle. Water and food were provided and were freely accessible to the animals throughout the experiment. The animal experimental protocols were approved by the Animal Care and Use Committee of Fujian University of Traditional Chinese Medicine. The procedures were performed in strict accordance with local and international codes of ethics.

At the end of experiment, tissues were collected after 10 weeks of gastrodin treatment. Then, renal tissue samples were fixed with 4% paraformaldehyde solution for 48 h, dehydrated, and embedded in paraffin. Subsequently, the tissues were cut into 4-µm sections and stained with hematoxylin for 50 s and eosin for 2 s (H&E staining) to visualize the pathological changes in the renal tissues. The slides were observed under a light microscope (Leica, Wetzlar, Germany) at ×400 magnification. The semi-quantitative percentage of the damaged area was scored as follows: 0, none; 0.5, <10%; 1, 10%–25%; 2, 25%–50%; 3, 50%–70%; and 4, >75%.

Renal tissues were harvested, fixed with 4% paraformaldehyde for 48 h, and embedded in paraffin. The tissues were then sliced into sections of 4-μm thickness for Masson’s trichrome staining. Renal tissue sections were successively immersed in Masson complex for 5 min, acidified with ethanol differentiation solution for 1 s, Masson blue solution for 1 min, Ponceau S red magenta solution for 5 min, distilled water for 1 min, phosphomolybdic acid solution for 1 s, and aniline blue solution for 50 s. The blue pixel content of the images was photographed to quantify renal fibrosis. Five different views were selected under a light microscope in each slide to estimate the values of the integral optical density and total area using the ImageJ software (open source Java image processing program available at https://imagej.nih.gov/ij/). Later, the collagen content was calculated as the percentage of positive area relative to the total area.

The renal tissues were sectioned at 6 μm and conventionally dewaxed in water and then were stained with Sirius red staining drops for 1 h, rinsed with running water, then stained with Mayer hematoxylin for 8 min, and rinsed with water for 10 min. The slides were viewed under a Leica DM2700P polarizing microscope (Leica, Wetzlar, Germany) at ×400 magnification. For each rat, the area with collagen fibers was assessed under five microscope fields using Image-Pro-Plus software (Media Cybernetics, Rockville, MD, United States).

The renal tissues were dewaxed and then subjected to heat-induced antigen retrieval in citrate buffer (pH 6.0). The tissues were incubated with the following primary antibodies: α-SMA (1:1,000), collagen I (1:600), collagen III (1:600), TGF-β1 (1:200), p-Smad2 (Cat no. 13429; 1:100), p-Smad3 (Cat no. 12838; 1:75), Smad2(Cat no. 41442; 1:100) or Smad3(Cat no. 41445; 1:100). After overnight incubation at 4°C, the slides were washed with phosphate-buffered saline (PBS) and later incubated with biotinylated secondary sheep anti-mouse/rabbit IgG followed by conjugated horseradish peroxidase-labeled streptavidin. Subsequently, the slides were incubated with diaminobenzidine as the chromogen and counterstained with diluted Harris hematoxylin. The slides were visualized under a Leica DM4000B intelligent automated optical microscope (Leica) and a light microscope at ×400 magnification. Five fields of view were randomly selected for each slide, and the average percentage of positively stained cells in each field was counted using ImageJ software.

The GeneCards was used to retrieve the targets using the term, hypertensive renal injury. Further, DisGeNET was employed for the extraction of targets via “renal fibrosis (RF), kidney/renal injury (KI) terms”. Gastrodin targets were collected using pharmamapper and Swiss databases.

A search tool for the retrieval of interacting genes (STRING) (https://cn.string-db.org/, Version 11.5) was used to obtain the Protein-protein interaction (PPI) network data of gastrodin and renal fibrosis-related common targets. The interaction network of “gastrodin-renal fibrosis-target” was constructed by using the software of Cytoscape (Version 3.8.2).

Database for annotation, visualization and integration discovery (DAVID) database was employed for the enrichment of gastrodin and renal fibrosis-associated common targets into multiple functional processes and KEGG pathways.

The most enriched fibrosis-associated targets (TGF-β1 and SMAD2) were selected for molecular docking with gastrodin. The 3D protein structures of TGF-β1 (PDB: 5VQP) and SMAD2 (PDB:1KHX) were retrieved from the protein databank (PDB). Moreover, pocket sites of these proteins were identified by implementing Prankweb online server. Then, the energy of protein structures was minimized by using the dock prep program. Gastrodin and protein structures were optimized and converted into PDBQT format for docking purposes. Furthermore, PyRx software was used for docking by choosing proposed active or pocket sites of the proteins, and proteins-gastrodin docking interactions were visualized by discovery studio.

The rat kidney fibroblast cell lines (NRK-49F) were obtained from the BeNa (Beijing, China). Cells were cultured in MEM supplemented with 10% FBS and 1% penicillin/streptomycin (100 IU/mL and 100 μg/mL) at 37°C under a 5% CO₂ atmosphere. The NRK-49F cells were incubated in a serum-free MEM for 4 h and then categorized into the control, TGF-β1, and TGF-β1 + Gastrodin (25, 50, or 100 μM) groups. The cells in the TGF-β1 and TGF-β1+ Gastrodin groups were stimulated with TGF-β1 (5 ng/mL), and those in the TGF-β1 + Gastrodin groups were treated with gastrodin (25, 50, or 100 μM). The cells in the control and TGF-β1 groups were treated with an equal volume of dd H₂O for a total of 24 h.

Cell viability was assessed using the CCK-8 assay according to the manufacturer’s instructions. Briefly, 3 × 10³ cells were seeded into each well of 96-well plates for 24 h. Later, the cells were stimulated with TGF-β1 and treated with or without gastrodin (25, 50, or 100 μM) treatment for 24 h. At the end of the experiment, 10 µL of CCK-8 solution was added to each well, and the absorbance was measured at 450 nm using a microplate reader (Multiskan FC, Thermo Fisher Scientific) after 2 h of incubation in the dark. The viability of the untreated cells was taken as 100%.

The NRK-49F cells were treated with TGF-β1 (5 ng/mL) alone or TGF-β1 in combination with gastrodin (25, 50, or 100 μM) for 24 h. The cells were fixed with 4% paraformaldehyde for 30 min, followed by permeabilization with 0.25% Triton X-100 in PBS for 5 min at room temperature. After blocking with 10% goat serum for 60 min, the slides were immunostained with anti-fibronectin (1:200), anti-α-SMA (1:200), anti-p-Smad2 (Cat no.18338S; 1:600), or anti-p-Smad3 (Cat no. ab52903; 1:200). The cells were then washed thrice with PBS and incubated with goat polyclonal secondary antibody to rabbit IgG-H&L (diluted 1:200) for 1 h. After washing with PBS, the cells were stained with hoechst for 10 min and then washed twice with PBS. Finally, the stained cells were visualized under a confocal microscope (UltraVIEW ®Vox, PerkinElmer, Santa Clara, CA, United States ). The images were acquired at ×600 magnification. The fluorescence intensity of the stained cells was calculated using Volocity software (PerkinElmer).

The NRK-49F cells were cultured in 6-well plates to a density of 0.8 × 10⁵ cells per well. After being cultured for 24 h, the NRK-49F cells were incubated in a blank medium for approximately 6 h. Then, the cells were stimulated with TGF-β1 and treated with or without gastrodin (25, 50, or 100 μM) for 48 h. Cell lysis buffer containing protease inhibitor and PMSF was added to the renal tissues, which were then ground using a low-temperature grinder, kept on ice for 30 min, and centrifuged at 4°C and 14,000 g for 20 min. The protein concentration was estimated using the BCA assay. A total of 50 μg of protein from each sample was separated on 10% SDS-PAGE gel, transferred to PVDF membranes, and blocked with 0.5% bovine serum albumin for 2 h. This step was followed by incubation with the primary antibodies (overnight at 4°C), including rabbit anti-collagen I (1:1,000), mouse anti-α-SMA (1:10,000), rabbit anti-fibronectin (1:1,000), and rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:5,000). Subsequently, the membrane was washed with TBST and then incubated with anti-mouse (1:10,000) and antirabbit (1:5,000) secondary antibodies at room temperature for 1 h. At the end of incubation, the membranes were washed with TBST and detected using an electrochemiluminescence kit. The expression of GAPDH protein was used as an internal control. The protein levels were analyzed with ImageJ software. The gray value of each protein band was normalized to that of the corresponding GAPDH band. The expressions of these proteins were normalized to those of the control group, which was set as 1.

Statistical analysis was performed using the SPSS statistical program (SPSS/PC+, version 22.0, Chicago, IL, United States). The results were presented as the mean value ±standard deviation (SD). One-way analysis of variance was used to compare the differences among three or more groups when the data were normally distributed. Then the LSD was used when the variance was chi-square, and the Games Howell was used when the variance was not chi-square. The non-parametric Kruskal–Wallis test was used to compare the differences among three or more groups when the data were not normally distributed, selected Pairwise Comparisons in the view option. Differences associated with p < 0.05 were considered statistically significant.

H&E staining revealed histopathological changes in the renal cortex after gastrodin treatment (Figures 1A, B). Compared with the WKY kidneys, the SHR kidneys exhibited obvious contraction and thickening of the glomerular atrophy, glomerular basement membrane (red arrow), and dilated tubules (black arrow). Gastrodin significantly alleviated hypertension-associated glomerular contraction and basement membrane thickening in the lesion area of SHR kidneys. The degree of collagen fiber formation in the renal tissues of SHRs was measured using Masson’s trichrome staining and Sirius red staining. Masson’s staining revealed that collagen fiber deposition in the interstitial space of the renal tubule was more extensive in the SHR group than that in the WKY group (Figures 1C, D; *p < 0.05 vs the WKY group), which was attenuated after gastrodin treatment (Figures 1C, D; #p < 0.05, vs the SHR group). The Sirius red staining images showed that the collagen fibers in the renal tissues were significantly higher in the SHR group than those in the WKY group (Figures 1E, F; *p < 0.05 vs the WKY group), although this was reduced after gastrodin treatment (Figures 1E, F; #p < 0.05, vs the SHR group).

We explored the potential effects of gastrodin on renal injury by analyzing the expression of α-SMA using IHC staining of the renal tissue. The results revealed that the renal tissue of SHRs showed significant upregulation of α-SMA (Figures 2A, B; *p < 0.05 vs the WKY group), which was reversed after gastrodin treatment (Figures 2A, B; #p < 0.05 vs the SHR group). To further examine the type of collagen fiber formation, we used IHC staining to analyze the expressions of collagen I and collagen III in the renal tubule interstitium. SHRs demonstrated significant upregulation of collagen I and collagen III (Figures 2C–F; *p < 0.05 vs the WKY group), which was reversed after gastrodin treatment (Figures 2C–F; #p < 0.05 vs the SHR group). Thus, gastrodin reduced collagen deposition in the renal tissues of SHRs.

The intersection of targets from different databases is represented into Circos diagram. Gene searches were performed using these terms, renal fibrosis (RF), renal/renal injury (KI), hypertensive renal injury and gastrodin using pharmamapper and stitch databases, as well as the Swiss database (Figure 3A). The overlapping gene analysis was performed and 40 common genes were found (Figure 3B). We entered these 40 common targets in the STRING database and obtained the network relationship data of target interaction under the threshold of a minimum required interaction score >0.4. Each edge represents the interaction between protein and protein (Figure 3C). The list of common 40 targets is given in the Supplementary Table S1.

The GO analysis was performed using DAVID database to clarify the mechanism of gastrodin’s behavior on hypertensive renal injury. After entering these 40 common targets were significantly enriched into several biological processes, cellular processes, and molecular functions using (FDR cutoff 0.05) and were categorized separately. The top 40 processes of each GO term were represented in figures and details were given in Supplementary Table S2. In hypertension pathogenesis, these targets were involved in vasculature development, pos. reg. of cell population proliferation and wound healing, reg. of epithelial cell proliferation, epithelial cell proliferation, and blood vessel development. TGF-β1 is involved in the formation of these biological processes (Figure 4A). The main cellular component also consisted of secretory granules, vesicle lumen, and secretory vesicle (Figure 4B). In addition, the category of molecular functions included signaling receptor binding, receptor-ligand activity, signaling receptor activator activity, and growth factor activity (Figure 4C). KEGG pathway enrichment analysis by the DAVID database using FDR cutoff 0.05 and identified multiple signaling pathways MAPK signaling pathway and FoxO signaling pathway, notably multiple inflammation or oxidative stress associated pathways, including TNF signaling pathway, Th17 cell differentiation, Toll like receptor, NF-kappa B signaling pathways, and so on (Figure 4D; Supplementary Table S2). Furthermore, quantitative comparison of all 40 proteins in GO terms and pathways is shown (Figure 4E). TGF-β1 and Smad2 proteins were abundantly enriched into 126 and 65 biological processes out of 371 processes respectively. In pathways analysis, TGF-β1 and Smad2 were quantitatively enriched into 64 and 49 pathways out of 135 pathways respectively.

The TGF-β1 and SAMD2 proteins were found abundantly enriched into most GO categories and pathways. Therefore, these targets were considered renal fibrosis potential targets and docked with gastrodin in order to provide a deeper insight into the binding interactions of the inhibitor (gastrodin) at the active sites. Docking energies and related docking residue sites are shown in Table 1. It is generally believed that the lower the docking energy between ligands and proteins, the stronger their bond. The results analysis showed that TGF-β1-gastrodin had a high binding affinity −5.1 kcal/mol (Figure 5A). Furthermore, best docking binding energy −6.6 kcal/mol was determined between SMAD2 and gastrodin (Figure 5B).Thus, this molecular docking results suggest that gastrodin might be one of the potential inhibitor molecule for renal fibrosis.

Owing to the essential role of the TGF-β/Smad pathway in the development of hypertensive renal fibrosis, we explored the underlying mechanism of the renal protective effect of gastrodin using IHC. The results of IHC staining revealed that the renal tissue of SHRs exhibited a significant increase in TGF-β1, and ratio of p-Smad2/Smad2, p-Smad3/Smad3, which were attenuated after gastrodin treatment (Figures 6A–F; *p < 0.05 vs the WKY group; #p < 0.05 vs the SHR group). These results suggested that gastrodin treatment attenuates the activation of the TGF-β1/Smad2/3 signaling pathway in renal tissues of SHRs.

We examined the effects of the combination of gastrodin and TGF-β1 on cell viability and noted that 100 μM gastrodin and 5 ng/mL TGF-β1 did not affect the cell viability (Supplementary Figures S1A, B; Figure 7A; *p < 0.05 vs the control group). Therefore, 100 μM gastrodin and 5 ng/mL TGF-β1 were used for the subsequent in vitro experiments. Immunofluorescence analysis indicated that gastrodin treatment partially reversed the TGF-β1-induced increase in the expressions of α-SMA and fibronectin (Figures 7B, C; *p < 0.05 vs the control group; #p < 0.05 vs the TGF-β1 group) in the NRK-49F cells. Moreover, determination of the expression of the fibrosis-related proteins using western blot analysis revealed that TGF-β1 infusion significantly upregulated the expressions of α-SMA, fibronectin, and collagen I, which were all reversed after gastrodin treatment (Figures 7D, E; *p < 0.05 vs the control group; #p < 0.05 vs the TGF-β1 group).

Consistently, TGF-β1 significantly upregulated the p-Smad2 and p-Smad3 (both mainly located in the nucleus) in the NRK-49F cells, which were reduced after gastrodin treatment (Figures 8A, B; *p < 0.05 vs the control group; #p < 0.05 vs the TGF-β1 group).