The molecular structure of NGR1 was confirmed through the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Next, the potential targets of NGR1 were predicted using the PharmMapper database (https://www.lilab-ecust.cn/pharmmapper/), and all target names were verified with the UniProt database (https://www.uniprot.org/). In the GeneCards database (https://www.genecards.org/), the DisGeNET database (https://www.disgenet.org/search/), and the OMIM database (https://omim.org/), “tendinopathy” was used as a keyword to search, download, and integrate all related genes. The intersection between the NGR1 targets and tendinopathy-related genes was used to identify therapeutic targets for NGR1 using the search term “Venn”. A network diagram showing the relationship among NGR1, tendinopathy, and the therapeutic targets was constructed using Cytoscape 3.2.1 software. Then, the topological parameters, including centrality and node degree of the network, were calculated using the network analysis tool to evaluate the significance of the nodes. The protein–protein interaction (PPI) networks of the therapeutic targets were constructed using the STRING database (https://string-db.org/) with the following conditions: “minimum required interaction score = 0.7” and “hide disconnected nodes in the network.” The DAVID v6.8 database (https://david.ncifcrf.gov/) was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses.

Key targets were screened based on three parameters (betweenness, closeness, and degree) in the PPI network of therapeutic targets. Subsequently, molecular docking was performed between NGR1 and the top 5 potential target proteins using AutoDock Vina software. The 3D structures of the target proteins were downloaded from the PDB database (https://www.rcsb.org), and the. mol2 format files of NGR1 were obtained from the TCMSP database. Criteria for selecting target protein crystals included: (1) 3D protein structure determined by X-ray crystallography; (2) crystal resolution less than 2 Å; and (3) reliable genotype protein analysis. Water molecules and original ligands were removed from the target protein using PyMOL. Then, the target protein was imported into AutoDock Tools v1.5.6 for hydrogenation, charge calculation, and combination with non-polar hydrogens, and the result was stored in. pdbqt format. The size of the Grid Box was set to cover the binding area of the target protein and NGR1. Finally, AutoDock Vina was run using CMD command characters for molecular docking, and PyMOL was used to visualize the results.

8-week-old male Sprague–Dawley (SD) rats were purchased from Beijing Longan Laboratory Animal Breeding Center (Ethics Code. BJLongan-2025–0003, Jan. 16, 2025.). The housing conditions for rats are maintained in an SPF environment. The ambient temperature is maintained at 18°C–22°C, the relative humidity is maintained at 50%–60%, and the average lighting time is 10–14 h. Subsequently, the animals were randomly divided into 5 groups (n = 6, each), including Sham, collagenase-induced tendinopathy (CIT), CIT + NGR1 (1, 4, 8 μM) groups. The tendinopathy model was established following the procedure outlined in the pre-experiment (Supplementary Figure S1) and previous study (Liu et al., 2021). Briefly, following the successful administration of anesthesia, 50 μL type I collagenase (40 mg/mL) was administered via injection into the right Achilles tendon of the rat. The Sham group received equivalent saline injections instead of type I collagenase. After 1 weeks, all groups were treated differently. The rats were locally injected with a therapeutic drug or saline one time per week, depending on their group assignment (Sham: 50 μL saline; CIT: 50 μL saline; CIT + NGR1: 1, 4, 8 μM NGR1 dissolved in saline 50 μL). After 5 weeks, all animals were euthanized, and the right Achilles tendons were collected.

Rat tendon samples were fixed in 4% neutral-buffered paraformaldehyde (PFA; Solarbio, Beijing, China), embedded in paraffin, and sectioned continuously (5 μm thick). Hematoxylin and eosin (HE) as well as Masson staining were performed according to the protocols. Histological scoring followed the criteria developed by Stoll et al. (2011), where the intact group was assigned a score of 20 points. For immunohistochemical staining (IHC), the paraffin-embedded sections were incubated with 3% H₂O₂ for 15 min to inhibit endogenous peroxidase, followed by incubation with 10% goat serum for 1 h at 21 °C to block non-specific antigens. Next, the sections were incubated with specific primary antibodies against collagen I (Col1, ab270993; Abcam, CA, United States; 1:100), collagen III (Col3, ab7778; Abcam; 1:100), IL-6 (ab9324; Abcam; 1:100), matrix metalloproteinase 3 (MMP3, ab52915; Abcam; 1:100) overnight at 4 °C. Subsequently, the sections were incubated for 1 h at 21°C with horseradish peroxidase-conjugated secondary antibodies (PV-6001 and PV-6002, ZSGB-BIO). The integrated OD value of positive staining was evaluated using ImageJ software (National Institutes of Health, MD, United States).

Primary rat tenocytes were isolated from tendon fragments dissected from the Achilles’s tendon of 6-week-old Sprague–Dawley (SD) rat. The rat tendon fragments were mechanically sliced into 2–5 mm³ pieces and enzymatically digested with 1% type I collagenase (C0130; Sigma-Aldrich, MO, United States) at 37°C on a shaking incubator for 1 h. After digestion, the cells were resuspended in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) with 20% (v/v) fetal bovine serum (FBS; HyClone, Logan, UT, United States) containing 1 g/L penicillin–streptomycin (Invitrogen, CA, United States) at 37°C in a humid environment with 5% CO₂. Starting with the initial medium exchange, the concentration of FBS was reduced to 10%, with subsequent changes of the medium occurring bi-daily. Upon achieving a confluence of 85%–90%, the cells should be passaged. Tenocytes at passage three (P3) are designated for use in various experimental assays.

Cell proliferation and viability were assessed using the Cell Counting Kit-8 (CCK-8). Tenocytes were initially seeded in 96-well plates at a density of 1 × 10³/well. Subsequently, various concentrations of NGR1 (Purity: HPLC ≥98%; CAS No.: 80,418–24–2; Sichuan Weikeqi Biotechnology Co., Ltd., China.) were introduced after tenocyte adhesion and incubated for 24 h. Afterward, CCK-8 reagent (10 μL/well) was added and allowed to incubate for 2 h. Use a microplate reader to measure the absorbance at 450 nm wavelength to assess cell viability.

Cell viability is calculated by optical density (OD) using the formula: O D ¯ experimental − O D ¯ blank O D ¯ control − O D ¯ blank × 100 %

Tenocytes were starved for 6 h without FBS and subsequently treated with LPS (250 ng/mL; L2880-10 MG; Sigma-Aldrich, MO, United States) and NGR1 (1, 5, 10 μg/mL) for 24 h. Total RNA was extracted from primary cultured tenocytes using the RNeasy Plus Mini Kit (Cat. No. 74136, QIAGEN). Purified RNA (2 μg) was reverse-transcribed using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Boston, MA, United States). The real-time qPCR was performed on the Applied Biosystems StepOnePlus Real-Time PCR System (Foster City, CA, United States) using SYBR Green PCR Master Mix (Toyobo, Japan). The expression levels of target mRNA were normalized to those of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) RNA. Relative gene expression was calculated using the 2^−ΔΔCT method and expressed as fold change. Each real-time qPCR assay was performed in triplicate, with at least three different biological replicates. Primer sequences are listed in Supplementary Table S1.

The tenocytes were first rinsed with PBS, then fixed with 10% neutral-buffered formalin at 21°C for 30 min. The cells were treated with Triton X-100 (Beyotime Biotechnology, Beijing) for 10 min to penetrate the cell membrane, followed by blocking with goat serum (Beyotime Biotechnology) for 1 h to prevent nonspecific binding. The cultured cells were incubated overnight at 4°C with primary antibodies against Col1 (ab270993; Abcam; 1:100). Subsequently, the cells were washed three times with PBS and incubated at 21°C for 1 h with Alexa Fluor 488-conjugated goat anti-rabbit secondary antibodies (A-11008, Thermo Fisher Scientific; 1:200). Nuclei were stained with DAPI (Beyotime Biotechnology, Jiangsu) for 10 min. Finally, the samples were rinsed with PBS and observed under a confocal microscope (Olympus Life Science, Tokyo, Japan).

IL-6 and TNF-α production in the supernatants was quantified by ELISA kits according to the manufacturer’s protocol. The IL-6 and TNF-α ELISA kits (PI328 and PT516) were from Beyotime Biotechnology (Jiangsu, China).

Tenocytes were starved for 6 h without FBS and subsequently treated with LPS (250 ng/mL) and NGR1 (1, 5, 10 μg/mL) for 4 days. Then, tenocytes were lysed in radioimmunoprecipitation assay (RIPA) lysis buffer containing protease inhibitors and/or a phosphatase inhibitor cocktail. The protein concentration was determined using the Bradford Protein Assay Kit (Beyotime, China). The total cell lysates were prepared in lysis buffer (150 mM NaCl, 1% Nonidet P-40, 50 mM Tris, 5 mM NaF), separated by SDS polyacrylamide gel electrophoresis (PAGE), and transferred to a polyvinylidene fluoride (PVDF) membrane. After blocking with 5% BSA in 0.1% Tween 20 TBS (TBST), the membranes were incubated with the corresponding primary antibodies overnight at 4 °C. After washing three times with TBST, the membranes were incubated with secondary antibodies at 21°C for 1 h and visualized using the BIORAD ChemiDoc XRS + system.

Proteins were analyzed using antibodies against Col 1 (ab270993; Abcam, CA, United States; 1:1,000), Col 3 (ab7778; Abcam; 1:1,000), tenomodulin (Tnmd, ab203676; Abcam; 1:1,000), P65 (ab32536; Abcam; 1:1,000), IκBα (ab32518; Abcam; 1:1,000), GAPDH (4,970, Cell Signaling Technology). Anti-mouse (ZB-2305, HRP-conjugated) and antirabbit (ZB-2301, HRP-conjugated) secondary antibodies were purchased from ZSGB-BIO (Beijing, China; 1:1,000).

We performed RNAseq analysis on rat tenocytes using the NovelBrain Cloud Analysis Platform. In the experimental setup, tenocytes in the control group remained untreated. The model group received treatment with LPS, while the NGR1 group was treated with a combination of LPS and NGR1 (10 μg/mL). After a 2-day treatment, total RNA was extracted from the rat tenocytes using TRIzol reagent. cDNA libraries were constructed for each pooled RNA sample (per rat) using the VAHTS™ Total RNA-seq (H/M/R) Kit. Differential gene and transcript expression analyses were performed using TopHat and Cufflinks, while HTseq was used to count the gene and lncRNA counts. The FPKM method was employed to determine gene expression levels. We applied the DESeq algorithm to identify differentially expressed genes (DEGs). Significant analysis was performed using P values and false discovery rate (FDR) analysis. DEGs were considered significant if they had an log ⁡ 2 FoldChange > 1 and an FDR <0.05. GO analysis was performed to elucidate the biological implications of the differentially expressed genes, encompassing biological processes (BP), cellular components (CC), and molecular functions (MF). GO annotations were downloaded from NCBI (http://www.ncbi.nlm.nih.gov/), UniProt (http://www.uniprot.org/), and Gene Ontology (http://www.geneontology.org/). Pathway analysis was conducted to identify significantly influenced pathways in which the DEGs were involved, according to the KEGG database. Fisher’s exact test was used to identify significantly influenced GO categories and pathways. The threshold of significance was defined by the p value.

The pain response in experimental animals was evaluated using a hot plate test. Briefly, the animals were placed on a hot plate analgesia meter (Ugo Basile, Italy) maintained at 55°C. The latency from initial hind paw contact to the onset of nociceptive behaviors (e.g., paw shaking, jumping, or licking) was recorded. Each animal underwent three trials at 15-min intervals, and the mean response latency was calculated as the final pain threshold. All behavioral assessments were performed by observers blinded to experimental group assignments.

Data were analyzed using GraphPad Prism 8.0.1 (244) software (San Diego, United States). Differences between multiple groups were assessed using ANOVA with Tukey’s post hoc test, and differences between two groups were assessed using an unpaired t-test. All data are presented as means ± SD, and statistical significance was established at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***). All the statistical analysis was independently conducted by two authors (J. W. and Y. L.) whom were blinded with the experimental protocol under the consideration of data reliability.

In total, 52 therapeutic targets were obtained through the intersections between NGR1 targets (n = 307) and tendinopathy targets (n = 377) (Figure 2A). Based on these data, the NGR1−tendinopathy− target network was constructed and visualized (Figure 2B). We constructed a PPI network (Figure 2C) to investigate the interactions among the targets, which were determined based on the degree of connectivity of each target in the network. These targets exhibit a high degree of interaction within the PPI network, including tumor necrosis factor (TNF, degree = 43), interleukin-6 (IL-6, degree = 42), serine/threonine kinase 1 (ATK1, degree = 40), endothelial growth factor receptor (EGFR, degree = 36), interleukin-1B (IL-1B, degree = 34), and Matrix metalloproteinase 9 (MMP9, degree = 33) (Supplementary Table S2). GO enrichment analysis was conducted to evaluate the common biological functions of the 52 targets (Figure 2D). KEGG enrichment analysis showed 147 related pathways with a significance threshold of p < 0.05. Among them, the PI3K/AKT signaling pathway and proteoglycans in cancer showed the highest correlation (Figure 2E). Concurrently, we constructed a core target-pathway relationship network to elucidate gene functions across distinct signaling pathways (Figure 2F).

Following the analysis of PPI data, six key targets (IL-6, IL-1β, MMP3, MMP13, MMP9, and TNF-α) were selected for further investigation. Subsequent molecular docking with the ligand NGR1 revealed significant interactions, as detailed in Figure 2G, which illustrates the three-dimensional structures of the proteins and the specific interactions at their binding sites. The binding energies recorded for all interactions were below −5 kcal/mol, suggesting robust and stable docking configurations (Figure 2H).

The healing process of tendinopathy was evaluated by morphological analysis including macro view, HE and Masson staining, and IHC. Macroscopically, it is evident that the CIT group is characterized by a substantial amounts of inflammatory hyperplasia (pale and deep yellow tissues) covering the tendon structures. Meanwhile, HE and Masson staining of tendons revealed that the collagen fibers in the healthy Achilles tendon tissue of the sham group were orderly arranged. In contrast, the control group’s Achilles tendon tissue exhibited typical tendinopathy pathological features, such as disorganized arrangement of collagen bundles, fragmentation of collagen fibers, fat infiltration and the ingrowth of neovessels, which was observed a gradation of therapeutic reversal in the low, medium, and high dosage groups of NGR1. IHC staining revealed that NGR1 exhibited a dose-dependent therapeutic effect on tendinopathy, including reducing the expression of the inflammatory cytokine IL-6, downregulating Col3 and MMP3 expression, and upregulating Col1 expression at 5 weeks (Figure 3). The high-dose (8 μM) NGR1 group exhibited the best therapeutic effects. This indicates that NGR1 plays a therapeutic role in the early inflammatory stage of tendinopathy with 5 weeks. In addition, the hot plate test showed that the pain response of rat in the CIT group was more sensitive than other group after treatment (Figure 4C).

Histological staining results showed that local administration of NGR1 has no negative impact on healthy tendon tissue (Supplementary Figure S2A). Additionally, evaluation of heart, liver and kidney samples showed no significant differences compared with the control group, confirming the absence of systemic side effects (Supplementary Figure S2B–D). Consequently, NGR1 emerges as a safe therapeutic agent capable of effectively alleviating tendinopathy progression in rat via local administration.

The tenocytes were first identified. Immunofluorescence results showed high expression of Col 1 in the cultured cells, and light microscopy revealed the characteristic spindle-shaped morphology, consistent with tenocyte features (Figure 4A). CCK-8 test results established a safe concentration range for NGR1 (0–10 μg/mL), considered the therapeutic dose of the in vitro study (Figure 4B). According to the real-time qPCR results, NGR1 enhanced the expression of Col 1a1, scleraxis (Scx), and Tnmd (Figure 4D), findings that are corroborated by Western blot analysis (Figure 4C). NGR1 effectively suppressed the expression of pro-inflammatory cytokines IL-1β, IL-6, and TNF-α, while also downregulating MMP3, 9, and 13 (Figure 4E). Furthermore, NGR1 reduced the production of IL-6, and TNF-α (Figure 4F) in the supernatants.

RNA sequencing analysis revealed that the Pearson correlation coefficients for samples within each group exceed 0.98, demonstrating a high degree of correlation among them. The analysis also identified 11,567 common targets between the NGR1 and Model group, among which Col 1a1, Col 3a1, and IL-6 were the core genes. (Figure 5A). Differential expression analysis revealed that, in comparison to the control group, the model group exhibits upregulation in 1,128 genes and downregulation in 915 genes. When comparing the NGR1 group to the model group, there are increases in expression levels of 28 genes and decreases in 192 genes (Figures 5B,C). Cluster analysis indicated that expression levels of MMP2, 3, 9, 10, 11, 12, and 13 were elevate in the model group compared to the control group. However, these levels decreased after treatment with NGR1, suggesting that NGR1 may effectively manage tendinopathy by modulating extracellular matrix metabolism (Figures 5D,E). Based on the GO analysis results, the most significant term in the BP category is immune system process; in the CC category, extracellular region shows the highest significance, followed closely by “extracellular space”; in the MF category, signaling receptor activator activity exhibits the highest significance level, followed by molecular transducer activity (Figure 5F). The KEGG analysis results suggested that the signaling pathways potentially involved in the treatment of tendinopathy with NGR1 include the TNF signaling pathway, NF-κB signaling pathway, and PI3K-Akt signaling pathway, among others (Figure 5G).