Skip to main content
NCBI home page
Springer logoLink to Springer
. 2025 Dec 11;41(1):161. doi: 10.1007/s10565-025-10117-4

Gentisic acid ameliorates lumbar disc herniation by regulating M1/M2 Polarization via the MAPK14/S100A9/Rac1/2 pathway

Shuoqi Li 1, Tiezhu Chen 2, Xiongjie Shen 3, Wanying Su 2, Xiaosheng Li 4,✉
PMCID: PMC12696013  PMID: 41372605

Abstract

Traditional Chinese medicine is gaining prominence in lumbar disc herniation (LDH) management, but the mechanisms of its active compounds and their molecular targets remain largely unclear. Herein, we aim to elucidate the therapeutic mechanism of Gentisic acid by investigating its role in regulating S100A9 in LDH. Clinical analysis reveals that S100A9 expression and inflammatory levels correlat positively with LDH severity. S100A9 is found to promote M1 macrophage polarization and impair dorsal root ganglion (DRG) neuronal activity. Mechanistically, Gentisic acid binds to MAPK14, downregulates S100A9 via MAPK14, and then suppresses M1 polarization, enhances neuronal autophagic flux, and improves neuronal viability through the S100A9/Rac1/2 pathway. In vivo experiments demonstrate that Gentisic acid ameliorates disc injury, improves neurological function, and alleviates pain in a rat LDH model, with efficacy comparable to celecoxib. These results suggest that Gentisic acid could alleviate LDH symptoms by modulating macrophage polarization and autophagy through the MAPK14/S100A9/Rac1/2 axis, offering a promising therapeutic strategy for LDH.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10565-025-10117-4.

Keywords: Lumbar disc herniation, Gentisic acid, S100A9, MAPK14, Rac1/2, M1/M2 polarization

Introduction

Lumbar disc herniation (LDH) is a common spinal neurologic disease (Yu et al. 2022). Its clinical manifestations mainly include persistent low back pain, radiating pain in the lower limbs, and dyskinesia, which seriously affects patients’ life quality (Zhang et al. 2023a). Current conservative and surgical treatments for LDH have some resistance and complications (Rogerson et al. 2019). Moreover, the molecular mechanisms of many LDH therapeutic drugs remain uncertain. Thus, there is an urgent need to explore more effective therapeutic agents and clarify their mechanisms.

Inflammatory response is a central pathological mechanism in the occurrence and development of LDH (Liu et al. 2016). An abnormal inflammatory environment disrupts the microecological balance of intervertebral disc tissues, promoting disc degeneration and nerve injury (Huang et al. 2018). As key participants in the inflammatory response, macrophages regulate inflammation through M1/M2 polarization (Zhou et al. 2019). M1-type macrophages primarily exert pro-inflammatory effects by releasing large amounts of pro-inflammatory cytokines (Wu et al. 2021). In contrast, M2-type macrophages possess anti-inflammatory and tissue repair functions (Li et al. 2025b). During LDH pathogenesis, an imbalance in M1/M2 macrophage polarization exacerbates inflammatory damage and promotes disease progression (Li et al. 2023b).

In inflammation modulation, mitogen-activated protein kinase 14 (MAPK14), S100 calcium-binding protein A9 (S100A9), and Ras-related C3 botulinum toxin substrate 1/2 (Rac1/2) are crucial regulators in cellular signaling cascades (Gu et al. 2025; Ye et al. 2025). As a core member of the MAPK signaling pathway, MAPK14 participates in regulating cell proliferation, differentiation, apoptosis, and inflammatory responses (Guo et al. 2025). S100A9 often forms a heterodimer called calprotectin (S100A8/A9) with S100A8, which exhibits enhanced biological functions (Colicchia et al. 2022). The S100A9 gene engages in inflammation and may serve as a key target for LDH treatment (Li et al. 2022). Rac1 and Rac2 are core members of the Rho GTPase family (Xiu et al. 2019). They can participate in regulating cell polarity and proliferation (Gao et al. 2021) and trigger the release of inflammatory mediators (Xu et al. 2023). These molecules may act synergistically to regulate the LDH inflammatory microenvironment and M1/M2 polarization, but a comprehensive understanding of the precise mechanisms remains unelucidated.

With growing evidence supporting its efficacy and safety, traditional Chinese medicine has emerged as an important complementary approach for LDH treatment (Qin et al. 2024). Yaobishu, a traditional Chinese medicine, is commonly used for LDH therapy (Li et al. 2024). Research shows that Gentisic acid is the active ingredient of Ligusticum chuanxiong Hort in Yaobishu (Li et al. 2022). Gentisic acid, a benzoic acid derivative present in various herbal ingredients, has diverse biological activities, including anti-inflammatory and antioxidant effects (Abedi et al. 2020). In diseases contexts, it also exerts hepatoprotective effects and reduces nephrotoxicity (Noei Razliqi et al. 2023; Saeedavi et al. 2023). However, the endogenous metabolism of Gentisic acid in LDH patients is incompletely clear.

In summary, this study aims to investigate the potential mechanism of Gentisic acid on LDH and its clinical application prospects. By analyzing LDH clinical samples and experimental models, we investigate the role of the Gentisic acid-MAPK14/S100A9 axis and Rac1/2 signaling in regulating M1/M2 immunity. This study may provide new strategies and insights for LDH treatment, ultimately leading to more effective clinical options for patients.

Materials and methods

Clinical samples

Clinical samples are obtained from Hunan Provincial People’s Hospital. According to Pfirrmann classification and Magnetic Resonance Imaging (MRI) T2-weighted images (T2WI) (Supplementary Figure S1), eight patients of grade Ⅲ and eight patients of grade Ⅳ are collected. The inclusion and exclusion criteria for all patients are as follows. Inclusion criteria: 1. Patients diagnosed with LDH undergoing lumbar discectomy or percutaneous endoscopic discectomy. 2. Available visual analog scale (VAS) scores (0–10) and Oswestry disability index (ODI) indices at predefined preoperative and postoperative intervals. 3. Access to nucleus pulposus (NP) paraffin sections and magnetic resonance imaging (MRI) scans. 4. Provision of informed consent for the use of tissue samples in clinical research. Exclusion criteria: 1. LDH secondary to tumors, infections, or other non-degenerative pathologies. 2. Comorbidities influencing macrophage infiltration or polarization. 3. History of steroid injections within 6 months preoperatively. 4. Revision surgery at the same disc level. 5. Age greater than 75 years. 6. Pfirrmann grade V degeneration at the involved segment. Additionally, according to the Pfirrmann classification, NP tissues from LDH patients are obtained as the LDH-Ⅲ and LDH-Ⅳ groups. The study is conducted following informed consent from all participants and with approval from the Ethics Committee of Hunan Provincial People's Hospital. Table 1 lists the basic clinical parameters of the participants.

Animals

Since Sprague–Dawley rats are more likely to exhibit nerve root compression, pain, and corresponding clinical symptoms when simulating LDH (Huang et al. 2018), we select SD rats for the animal experiments. Specific pathogen-free (SPF) Sprague–Dawley rats (6 weeks old, male, non-genetically modified) are ordered from Hunan SJA Laboratory Animal Co., Ltd. A total of 65 rats were used in this study. After the rats are acclimatized for one week, formal experimental manipulations begin to be scheduled. The rats are divided into 5 groups: Sham, LDH, Gentisic acid, EHT1864, and Celebrex groups, with 12 rats in each group. Additionally, 5 other rats are used for primary cell isolation.

The rat LDH models are prepared as follows: Following anesthesia with 2% isoflurane gas (792632, Sigma-Aldrich, Saint Louis, MO, USA), fine rongeurs are used to perform a right-sided L5 hemi-laminectomy in rats, which allows exposure of the L5-6 intervertebral discs. The L5-6 discs are then punctured to a 4 mm depth from their surface using sterile needles. In contrast, rats in the sham group undergo no additional procedures after the right-sided L5 hemi-laminectomy. After 5 weeks of LDH surgery, the rats are treated for 21 days as follows: they are gavaged with either Gentisic acid (100 mg/kg, twice daily) or Celebrex (16 mg/kg/day), or injected intraperitoneally with EHT1864 (35 mg/kg, once every two days). The Sham and LDH groups are gavaged with equal volumes of saline.

The rats are housed in cages measuring 30 cm × 30 cm × 15 cm, one cage per rat. Moreover, rats are housed in an environment of 22 °C, 45% humidity, and a 12-h light/dark cycle. Additionally, we use standardized experimental animal feed to raise the rats and monitor their health every day during the experimental period. To evaluate the biosafety after administration, we observe the rats' body weight, daily food/water intake, and activity levels every day throughout the experiment. Compared with the control group, no significant changes (e.g., weight loss, reduced activity, lethargy, or diarrhea) are observed in the drug-treated groups. At the end of the experiment, the rats are euthanized by cervical dislocation and blood, NP tissues, intervertebral disc tissues, and dorsal root ganglion (DRG) tissues are collected from the rats. The animal experiments are approved by the Ethics Committee of Hunan Provincial People's Hospital (2020KYLSD29H).

Detailed procedures for pathological staining of animal tissues are summarized in Supplementary Methods. The timeline of the in vivo experiment and the surgical images are shown in Supplementary Figures S2 and S3.

Isolation and characterization of rat peritoneal macrophages

Rat peritoneal macrophages are prepared via repeated intraperitoneal injection of culture medium, extraction, and differential wall affixation. After euthanasia, rats are placed in the supine position. All surgical instruments are subjected to strict aseptic treatment (autoclaving at 121 °C for 30 min), and the abdominal skin of rats is disinfected with iodophor. Then, 10 mL of DMEM serum-free medium (AW-M003, Abiowell, Changsha, China) is injected into the lower left corner of the abdominal cavity. During the lavage process, sterile phosphate-buffered saline (PBS) preheated to 37 °C is used to avoid tissue irritation and reduce microbial introduction. The abdomen is gently massaged for 3 min and left to stand for 5 min. Subsequently, the abdominal cavity is carefully opened with sterile scissors, and the abdominal fluid is extracted with a syringe, placed in a sterile centrifuge tube. The collected lavage fluid is first filtered through a 70 μm sterile cell strainer before being centrifuged to collect cellular precipitates. The cells are washed several times with DMEM serum-free medium, then resuspended and cultured in DMEM medium containing 15% fetal bovine serum (FBS, 10099141, Gibco, GrandIsland, NY, USA) and 1% penicillin/streptomycin (P/S, SV30010, Beyotime, Changsha, China). The 1% penicillin/streptomycin in the medium further eliminates potential contaminants. Additionally, the cells are identified by CD68 immunofluorescence, with purity exceeding 90%.

Isolation and characterization of rat DRG neurons

After sterilization, the ganglion and attached nerve fibers are removed from the medial side of the rat spinal canal in an ultra-clean table and placed in DMEM-F12 medium (AW-M006, Abiowell) under ice bath conditions. The attached nerve fibers and surrounding connective tissue membrane are carefully cut off with iris scissors and tweezers, then the ganglia are digested with trypsin for 20 min (shaken every 5 min). The digestion is terminated with 10% FBS. Next, the ganglia are pipetted to isolate individual nerve cells, which are finally resuspended and cultured in neuron-specific medium (iCell-0013, iCell, Shanghai, China) containing 50 × B27 (17504044, Gibco). Additionally, the cells are identified as positive via Neuron-specific enolase (NSE) immunofluorescence staining.

Cell treatment

After successful isolation and characterization of rat peritoneal macrophages and DRG neurons, they are subjected to the following treatments.

In Experiment 1, to investigate the effect of S100A9 on the polarization of rat peritoneal macrophages, we knock down and overexpress S100A9 in cells to verify its effect on macrophage polarization itself. We also divide rat peritoneal macrophages into Control, interferon-γ/lipopolysaccharide (IFN-γ/LPS), IFN-γ/LPS + S100A9, IL-4, and IL-4 + S100A9 groups. The Control group cells are incubated normally. IFN-γ (20 ng/mL 11276905001, Sigma-Aldrich)/LPS (100 ng/mL, L2630, Sigma-Aldrich) and IL-4 (20 ng/mL, PA5-79950, Thermo Fisher Scientific, Waltham, MA, USA) are utilized to induce M1/M2 macrophage differentiation, respectively (Atta et al. 2023; Mohamed Elashiry et al. 2021). During differentiation, S100A9 (1.5 µg/mL, PA5-82145, Thermo Fisher Scientific) is added to induce cells for 24 h (Neumann et al. 2025). Furthermore, macrophages from the IL-4 and IL-4 + S100A9 groups are co-cultured with DRG neurons for 48 h to expose S100A9’s effect on DRG neurons (Franz et al. 2022).

In Experiment 2, to verify the effect of Gentisic acid on MAPK14 protein, we categorize DRG neurons into oe-NC, oe-MAPK14, oe-NC + Gentisic acid, and oe-MAPK14 + Gentisic acid groups. The oe-NC and oe-MAPK14 group cells are transfected with oe-NC and oe-MAPK14, respectively. Additionally, the cytotoxicity and concentration of Gentisic acid (490–79-9, Selleckchem, Houston, TX, USA) are characterized and screened. On this basis, 1.6 μg/mL Gentisic acid is selected to treat the cells for 48 h.

In Experiment 3, to explore the effects of Gentisic acid and MAPK14/S100A9 on the viability of DRG neurons, we separate DRG neurons into Control, IFN-γ/LPS, oe-NC, Gentisic acid, Gentisic acid + oe-NC, Gentisic acid + oe-MAPK14, S100A9, S100A9 + Gentisic acid, S100A9 + CLI-095, S100A9 + EHT1864, si-NC, si-MAPK14, and si-MAPK14 + S100A9 groups. Among these, DRG neurons are transfected with oe-NC, oe-MAPK14, si-NC, or si-MAPK14. Cells are intervened by Gentisic acid (1.6 μg/mL), S100A9 (50 μM), CLI-095 (1 µM) (Piovan et al. 2022), or EHT1864 (10–6 mol/L) (Bruder-Nascimento et al. 2019) for 48 h.

In Experiment 4, to investigate the effect of Gentisic acid on S100A9-induced M1/M2 polarization, we divide rat peritoneal macrophages into IL-4, IL-4 + Gentisic acid, IL-4 + S100A9, and IL-4 + S100A9 + Gentisic acid groups. IL-4 (20 ng/mL) induces M1/M2 polarization for 24 h. During cell differentiation, S100A9 (1.5 µg/mL) and Gentisic acid (1.6 µg/mL) are added for 48 h of intervention. Additionally, macrophages from the four groups are co-cultured with DRG neurons for 48 h to explore Gentisic acid’s effect on DRG neurons.

Additional details of cell experimental reagents and operation precautions are provided in Supplementary Methods.

Western blot

Samples are initially treated with RIPA (R0010, Solarbio, Beijing, China) to extract total proteins. The proteins are then separated via sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a nitrocellulose membrane. The membrane is blocked with 5% skim milk for 1.5 h, then incubated with primary antibodies, followed by 1.5 h of incubation with secondary antibodies (details of primary and secondary antibodies are shown in Table 2). Lastly, the membrane is treated with ECL reagent (K-12049-D50, Advansta, Menlo Park, CA, USA) and subjected to analysis in an imaging system. β-actin is employed as the internal reference and probed on separate blots.

Bioinformatics analysis

The SMILES ID of Gentisic acid is obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/; the database ensures data accuracy through crowdsourced curation and cross-validation with scientific literature), and its drug targets are retrieved from SEA (http://sea.bkslab.org/), PharmMapper (http://www.lilab-ecust.cn/pharmmapper/), and Batman-TCM (http://bionet.ncpsb.org.cn/batman-tcm/) databases (quality control: the three databases use ligand-based virtual screening algorithms with validated prediction models, and only targets with prediction scores above the respective database thresholds are retained). After correction and deduplication via the Uniprot database (https://www.uniprot.org/; targets are standardized to UniProtKB accessions, and obsolete or non-human entries are excluded), a total of 274 targets are obtained. Using "lumbar disc herniation" as the keyword, disease-related targets are retrieved from CTD (https://ctdbase.org/), NCBI (https://www.ncbi.nlm.nih.gov/), and OMIM (https://www.disgenet.org/) (quality control: human gene search is conducted with the keyword "lumbar disc herniation") databases, and 669 relevant genes are identified after merging and deduplication. Subsequently, Venny 2.1 (https://bioinfogp.cnb.csic.es/tools/venny/) is used to construct a Venn diagram, screening 35 common targets between the drug and disease as predictive targets. These targets are uploaded to the String database (https://string-db.org/cgi/input.pl) to build a protein–protein interaction (PPI) network (organism: Homo sapiens; confidence score > 0.4) and visualized via Cytoscape software (https://cytoscape.org/). Furthermore, R 4.1.2 software (with packages such as clusterProfiler) is employed to perform Gene Ontology (GO) enrichment analysis (biological processes, molecular functions, cellular components) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis on the common targets. Results are filtered by adjusted p-value < 0.05 and visualized as bar charts and bubble plots. Finally, component-disease-pathway-target network files are imported into Cytoscape 3.8.0 to draw pathway network diagrams, revealing the multi-component and multi-target action characteristics.

Molecular docking analysis

Molecular docking analysis of Gentisic acid with the MAPK14 protein is conducted using VINA 1.1.2 software, which utilized a semi-empirical free energy field to predict receptor-ligand binding energies. PyMOL 2.3 (Schrodinger, New York, NY, USA) is applied to visualize the ligands bound to the receptor-binding pocket and to forecast the receptor-ligand interactions and energies.

Drug affinity responsive target stability (DARTS)

First, MAPK14 protein (0.5 μg/mL) is incubated with Gentisic acid (1.6 μg/mL) at room temperature for 30 min. Next, diluted proteases (10 mg/mL) at different proportions (1:3200, 1:1600, 1:800, 1:400) are added and incubated for 10 min, then the reaction is terminated by adding a supersampling buffer. Finally, samples are obtained after adding 5 × sample buffer and heating at 95 °C for 10 min, and these samples are analyzed via western blot to detect the binding of MAPK14 and Gentisic acid.

Siegal neurological score

Siegal recommends a six-grade grading system for assessing the neurologic function. The grades are summarized in Table 1 below. Rat behavior is observed on days 0, 3, 7, 14, 21, and 28 after modeling. Neurological scores are used for statistical analyses. The points corresponding to the levels are also shown in Table 4.

Thermal withdrawal thresholds (TWL)

TWL is evaluated using the plantar tester thermal stimulation method: After a 15 min acclimatization period, radiant heat is set at 60 °C and induced by aiming the beam at the metatarsal surface of the hind paw through a glass plate. When rats feel pain and retract their paws, the beam is automatically turned off. The cutoff time is set at 40 s to prevent tissue injuries. Each paw of each rat is tested 5 times at 5 min intervals, and the test is performed under double-masked conditions.

Paw mechanical withdrawal thresholds (PWT)

First, the fibrillar pain meter is calibrated. A needle is pricked into the sole of the rat’s left hind paw. Positive responses include paw retraction, paw lifting, or hissing (distinguished via the rat’s voluntary movements). If a positive reaction is exhibited, a tinier needle is chosen. If no positive response is observed, a larger needle is used (maximum strength: 15 g). Each test is repeated 5 times at approximately 2-min intervals. Finally, if more than 3 positive reactions are recorded, the reaction with the lowest intensity is considered the rat’s PWT, and the test is performed under double-masked conditions.

Statistical analysis

Statistical analysis is performed with GraphPad Prism8 software (8.0.2.263, San Diego, CA, USA): in vitro experiments include 3 replications and in vivo experiments include 5 replications. Measurement data are expressed as mean ± standard deviation; for samples conforming to normal distribution, two-group comparisons are conducted via T-test and multiple-group comparisons via One-way ANOVA, followed by Tukey's post hoc tests. Variations are regarded as statistically significant when p < 0.05.

Ethics approval and consent to participate

The study involves human samples and is conducted only after two prerequisites are met: informed consent is obtained from all participants, and the study protocol is approved by the Ethics Committee of Hunan Provincial People's Hospital. Notably, informed consent is also secured from every individual participant included in the study. Additionally, the animal experiments are approved by the same ethics committee (2020KYLSD29H).

Results

S100A9 is upregulated in LDH clinical samples

Firstly, we collect NP tissues from LDH patients to explore the S100A9 protein expression and inflammation level. The clinical sample information is provided in Table 1. As is shown in Fig. 1A, The TNF-α and S100A9 levels in the LDH-Ⅳ group are higher than those in the LDH-Ⅲ group, while the IL-4 levels are lower than those in the LDH-Ⅲ group, suggesting that inflammation and S100A9 are elevated in LDH-Ⅳ patients. Moreover, we find that S100A9 is positively correlated with TNF-α and negatively correlated with IL-4 (Fig. 1B). Additionally, as an ion channel protein associated with neuropathic pain, TRPV1 (Oh et al. 2023) is expressed at a higher level in LDH-Ⅳ than in LDH-Ⅲ (Fig. 1C). Toll-like receptor 4 (TLR4) and p65 also are closely connected to inflammatory regulation (Jiang et al. 2023). Compared to the LDH-Ⅲ group, TLR4 and p65/p-p65 levels are upregulated in the LDH-Ⅳ group (Fig. 1D). These results demonstrat that more severe LDH in patients may be accompanied by higher S100A9 expression and inflammation.

Fig. 1.

Fig. 1

Open in a new tab

S100A9 is upregulated in LDH clinical samples. A TNF-α, IL-4, and S100A9 levels in NP tissues are measured by Western blot. B Spearman's coefficient analyzes the correlation of S100A9 with TNF-α and IL-4. C TRPV1 expression in NP tissue is detected by Western blot. D TLR4 and p65/p-p65 expression in NP tissues are tested by Western blot. *p < 0.05 vs. LDH-Ⅲ. n = 8

S100A9 induces polarization of primary rat peritoneal macrophages to affect the viability of DRG neurons

Next, we explore the effect of S100A9 on the polarization of primary rat peritoneal macrophages in vitro. Transfection efficiency is validated, with S100A9 expression significantly decreased in knockdown groups and increased in the oe-S100A9 group compared to their respective controls (Fig. 2A). Flow cytometry shows the proportion of M1 macrophages (CD68+CD86+) is lowest in the sh-S100A9 group and highest in the oe-S100A9 group, while M2 macrophages (CD68+CD206+) show the opposite trend. Moreover, the M1/M2 ratio is reduced by S100A9 knockdown and increased by overexpression (Fig. 2B). CD68 is highly expressed in primary rat peritoneal macrophages, confirming successful macrophage identification (Fig. 2C). Similarly, exogenous addition of S100A9 promotes IFN-γ/LPS-induced M1 differentiation and inhibits IL-4-induced M2 differentiation, and increases M1/M2 levels under these treatments (Fig. 2D). Since CD86 and CD206 are markers of M1 and M2 activation, respectively (Li et al. 2023a). We find that S100A9 promotes IFN-γ/LPS-induced CD86 expression and inhibits IL-4-induced CD206 expression, further validating its role in regulating macrophage polarization (Fig. 2E). These results indicate that S100A9 promotes M1 polarization and inhibits M2 polarization in macrophages. Subsequently, we verify the effect of S100A9-induced M1/M2 polarization on DRG neurons. As a molecular marker for neurons and neuroendocrine cells, NSE (Lauwers et al. 2022) is highly expressed in primary rat DRG neurons, suggesting successful isolation (Fig. 2F). Then, we co-cultured macrophages from the IL-4 and IL-4 + S100A9 groups with DRG neurons. CCK-8 results exhibit that S100A9 may inhibit the proliferation of DRG neurons via inducing macrophage polarization (Fig. 2G). Additionally, S100A9 increases the expression of TRPV1, TLR4, and p65/p- p65 in DRG neurons, suggesting it induces pro-inflammatory functions in these neurons (Fig. 2H-J). These results reveal that S100A9 could induce M1/M2 polarization to affect the viability of DRG neurons.

Fig. 2.

Fig. 2

Open in a new tab

S100A9 induces polarization of primary rat peritoneal macrophages to affect the viability of DRG neurons. A The transfection efficiency of S100A9 knockdown/overexpression is verified by RT-qPCR and Western blot. B The polarization levels of M1 and M2 (macrophages) are evaluated by flow cytometry. *p < 0.05 vs. sh-NC. #p < 0.05 vs. oe-NC. C CD68 expression in primary rat peritoneal macrophages is identified by immunofluorescence. D The polarization levels of M1 and M2 (macrophages) are verified by flow cytometry. E CD86 and CD206 levels in macrophages are measured by Western blot. *p < 0.05 vs. Control. #p < 0.05 vs. IFN-γ/LPS. &p < 0.05 vs. IL-4. F NSE expression in primary rat DRG neurons is examined by immunofluorescence. G The proliferation of DRG neurons is measured by CCK-8. H-I TRPV1 expression in DRG neurons is detected by immunofluorescence. J TLR4 and p65/p-p65 expression in DRG neurons are determined by Western blot. &p < 0.05 vs. IL-4. n = 3

Gentisic acid binds to MAPK14 and downregulates S100A9 levels

Research reveals that Gentisic acid (the active ingredient in Yaobishu) may exert therapeutic effects on LDH rats (Li et al. 2022), but its precise mechanism remains unclear. To elucidate its targets, Venn diagram analysis identifies 35 shared targets between Gentisic acid and LDH (Fig. 3A), among which MAPK14 is found to interact with S100A9 (Fig. 3B). Protein–protein interaction (PPI) network analysis (Fig. 3C) and functional enrichment (GO/KEGG, Figs. 3D-3E) demonstrated that these targets are primarily enriched in inflammatory responses and MAP signaling pathways. Network pharmacology further constructs a "Gentisic acid-LDH-pathway-target" regulatory network (Fig. 3F), suggesting that Gentisic acid might modulate S100A9 expression via MAPK14. Molecular docking results indicate high binding affinity between Gentisic acid and MAPK14 protein (binding energy < −5 kcal/mol), with the docking model showing stable binding to MAPK14’s active pocket (Fig. 3G-H). DARTS assay combined with WB validation confirms direct binding of Gentisic acid to MAPK14 (Fig. 3I). Cellular experiments demonstrate no significant cytotoxicity of Gentisic acid (0.08–8 μg/mL) on DRG neurons (Fig. 3J), prompting the selection of 1.6 μg/mL for subsequent experiments. Overexpression of MAPK14 upregulates mRNA and protein expression of MAPK14 and S100A9, while Gentisic acid reverses the effect of oe-MAPK14 and decreases protein expression of MAPK14 and S100A9 (Fig. 3K-L). These results prove that Gentisic acid likely exerts its anti-inflammatory effects by directly binding to and inhibiting MAPK14 activity, thereby downregulating S100A9 expression.

Fig. 3.

Fig. 3

Fig. 3

Open in a new tab

Gentisic acid binds to MAPK14 and downregulates S100A9 levels. A A Venn diagram (via Venny 2.1, link: https://bioinfogp.cnb.csic.es/tools/venny/; quality control: preprocessed by deduplication/Uniprot standardization) is constructed to display the 35 interacting targets between Gentisic acid (274 targets from SEA/PharmMapper/Batman-TCM) and LDH (669 targets from CTD/NCBI Gene/OMIM, restricted to direct disease associations). B Another Venn diagram is created to depict the interaction between the 35 identified targets and S100A9, which successfully pinpoints MAPK14 as the crucial target. C Protein–protein interaction (PPI) network of the 35 targets is built by String database (link: https://string-db.org/; quality control: Homo sapiens, confidence score > 0.4) and visualized by Cytoscape 3.8.0 (link: https://cytoscape.org/), containing 35 nodes and 262 edges. D The common targets of the drug and disease are subjected to Gene Ontology (GO) enrichment analysis, including biological processes (BP), molecular functions (MF), and cellular components (CC). Items with an adjusted p-value < 0.05 are filtered. A total of 1520 biological processes, 44 molecular functions, and 35 cellular components are enriched for the intersection targets. E The common targets of the drug and disease are subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Items with an adjusted p-value < 0.05 are filtered, and a total of 156 signaling pathways are enriched. F Component-disease-pathway-target regulatory network intuitively demonstrates the multi-component and multi-target action characteristics of active ingredients in traditional Chinese medicine during disease treatment. In the diagram, blue represents compounds, pink represents targets through which compounds act on diseases, green represents the top 20 most significant pathways, and yellow represents diseases. G Molecular docking analysis about Gentisic acid and MAPK14 protein of human origin. H Molecular docking analysis about Gentisic acid and MAPK14 protein of rat origin. I The binding of Gentisic acid and S100A9 protein is verified by DARTS assay. &p < 0.05 vs. Contorl. J CCK-8 detection of Gentisic acid cytotoxicity. K The relative expression of MAPK14 and S100A9 is measured by RT-qPCR. L The protein expression of MAPK14 and S100A9 is determined by Western blot. *p < 0.05 vs. oe-NC. #p < 0.05 vs. oe-NC + Gentisic acid. n = 3

Gentisic acid improves DRG neuronal cell viability by targeting MAPK14/S100A9 to modulate THP-1 polarization

The mechanism of Gentisic acid on DRG neuronal cell viability is further validated. Initial qRT-PCR and western blot analyses confirm successful MAPK14 overexpression (Fig. 4A). Notably, Gentisic acid downregulates endogenous MAPK14 and S100A9 expression in macrophages, whereas this effect is significantly attenuated in MAPK14-overexpressing cells (Fig. 4B). In THP-1 cells, LPS/IFN-γ stimulation induces M1 polarization, evidenced by increased CD68+CD86+ cell populations and elevated CD86 protein levels. Gentisic acid effectively attenuates these changes, but oe-MAPK14 partially counteracts Gentisic acid’s inhibitory effects (Figs. 4C-4D). When DRG neurons are cultured with conditioned media from these macrophage groups, Gentisic acid-treated media enhances neuronal viability, reduces TRPV1 expression, and decreases TLR4 and p-p65 levels. Importantly, the neuroprotective effects are partially reversed when neurons are exposed to media from MAPK14-overexpressing macrophages, indicating the action of Gentisic acid may depend on the MAPK14/S100A9 axis (Figs. 4E-G). These findings collectively demonstrate that Gentisic acid improves DRG neuronal viability by suppressing M1 macrophage polarization through the MAPK14/S100A9 signaling pathway.

Fig. 4.

Fig. 4

Open in a new tab

Gentisic acid improves DRG neuronal cell viability by targeting MAPK14/S100A9 to modulate THP-1 polarization. A The transfection efficiency of oe-MAPK14 is verified by RT-qPCR and Western blot. B The levels of MAPK14 and S100A9 are detected by Western blot. C The proportion of CD68+CD86+ cells is detected by flow cytometry. D The expression of CD86 in macrophages is measured by Western blot. E The proliferation of DRG neurons is evaluated by CCK-8. F The expression of TRPV1 in DRG neurons is detected by immunofluorescence. G The protein levels of TLR4, p65, and p-p65 in DRG neurons are detected by Western blot. *p < 0.05 vs. Control. #p < 0.05 vs. NC. &p < 0.05 vs. Gentisic acid + oe-NC. n = 3

Gentisic acid ameliorates autophagy in DRG neurons through the MAPK14/S100A9 pathway

Next, we explore the molecular mechanisms by which gentisic acid regulates autophagy in DRG neurons. We also apply the TLR4 inhibitor CLI-095 and the Rac1/2 inhibitor EHT1864 to explore whether gentisic acid exhibits therapeutic equivalence to these pharmacological agents. CCK-8 assays demonstrate that Gentisic acid, CLI-095, and EHT1864 all reverse the reduced cell proliferation induced by S100A9 (Fig. 5A). In contrast to the Control group, S100A9 increases TRPV1, GTP-Rac1 + Rac2/total-Rac1 + Rac2, and GTP-RhoA/total RhoA expression in DRG neurons. However, compared to the S100A9 group, Gentisic acid, CLI-095, and EHT1864 all reduce TRPV1 expression as well as attenuate Rac1 and Rac2 activation (Fig. 5B, C). Electron microscopic observations show that the intervention of Gentisic acid, CLI-095, and EHT1864 all restore S100A9-induced autophagic flow block in DRG neurons (Fig. 5D). Moreover, Gentisic acid, CLI-095, and EHT1864 also enhance LC3II/I and Beclin-1 expression and reduce p62 levels (Fig. 4E). Notably, MAPK14 expression is inhibited only by Gentisic acid (Fig. 4F). Co-immunoprecipitation confirms the direct interaction between MAPK14 and S100A9 (Fig. 5G). Subsequently, we successfully knock down MAPK14 and select MAPK14#1 with a higher knockdown efficiency for subsequent experiments (Fig. 5H). Knocking down MAPK14 partially reverses the effects of LPS/IFN-γ, and reduces the proportion of CD68+CD86+ macrophages, as well as the expression levels of CD86 and S100A9 (Fig. 5I-J). When DRG neurons are treated with the conditioned medium from these macrophages, the medium from the MAPK14 knockdown group promotes neuronal proliferation (Fig. 5K), decreases TRPV1 expression (Fig. 5L-5M), inhibits the activation of the TLR4/p53 pathway (Fig. 5N), and the activation of Rac1/2/RhoA (Fig. 5O). Exogenous addition of S100A9 partially reverses the protective effects conferred by the aforementioned MAPK14 knockdown (Fig. 5K-O). These results indicat that Gentisic acid regulates the expression of S100A9 through MAPK14, inhibits the M1 polarization of macrophages, and thereby improves the autophagic state and viability of DRG neurons via the TLR4/p53 and Rac1/RhoA signaling pathways.

Fig. 5.

Fig. 5

Fig. 5

Open in a new tab

Gentisic acid ameliorates autophagy in DRG neurons through the MAPK14/S100A9 pathway. A The proliferation of DRG neurons is examined by CCK-8. B TRPV1 expression in DRG neurons is tested by immunofluorescence. C The expression of GTP-Rac1 + Rac2/total-Rac1 + Rac2 and GTP-RhoA/total RhoA in DRG neurons is examined by Western blot. D The morphology of autophagic vesicles in DRG neurons is visualized by electron microscopy. Red arrows represent autophagosomes, yellow arrows represent damaged mitochondria, and green arrows represent normal mitochondria. Scale bar = 1 μm. E The levels of LC3II/I, Beclin-1, and p62 in DRG neurons are assessed by Western blot. F The expression of MAPK14 protein is quantified by Western blot. *p < 0.05 vs. Control. #p < 0.05 vs. S100A9. ns represents no significant difference. G The interaction between MAPK14 and S100A9 proteins is validated through Co-immunoprecipitation. H The transfection efficiency of si-MAPK14 is evaluated by RT-qPCR and Western blot. @p < 0.05 vs. si-NC. I The proportion of CD68+CD86+ cells is examined by flow cytometry. J The levels of CD86 and S100A9 in macrophages are measured by Western blot. &p < 0.05 vs. Control. @p < 0.05 vs. si-NC. K The proliferation of DRG neurons is evaluated by CCK-8. L-M The expression of TRPV1 in DRG neurons is visualized by immunofluorescence. N The expression of TLR4, p65, and p-p65 in DRG neurons are analyzed by Western blot. O The levels of GTP-Rac1 + Rac2/total-Rac1 + Rac2 and GTP-RhoA/total RhoA in DRG neurons are determined by Western blot. &p < 0.05 vs. Control. @p < 0.05 vs. si-NC. $p < 0.05 vs. si-MAPK14 + S100A9. n = 3

Gentisic acid inhibits S100A9-induced M1/M2 polarization to improve the viability of DRG neurons

To investigate the specific effect of Gentisic acid on S100A9, we apply IL-4 to induce macrophage differentiation and intervened by adding S100A9 and Gentisic acid during the differentiation process. Flow cytometry reveals that Gentisic acid reverses the effect of S100A9 on M1 and M2 macrophage differentiation (Fig. 6A). Moreover, the intervention of Gentisic acid reduces the M1/M2 ratio when compared to the IL-4 + S100A9 group (Fig. 6B). Gentisic acid also reverses the elevated CD86 levels and decreased CD206 levels in macrophages caused by S100A9 (Fig. 6C). The above findings suggest that Gentisic acid could regulate S100A9-induced M1/M2 polarization. Then, we co-culture the macrophages treated as above with DRG neurons to assess their effects on neuronal function. Figure 6D shows that Gentisic acid could enhance the proliferation of DRG neurons even in the presence of S100A9 interference. Furthermore, in comparison to the IL-4 + S100A9 group, Gentisic acid decreases the expression of TRPV1, TLR4, and p-p65/p65 in DRG neurons, suggesting that Gentisic acid alleviates the inflammation of DRG neurons (Fig. 6E-F). These results reveal that Gentisic acid could regulate S100A9-induced M1/M2 polarization to improve the viability of DRG neurons.

Fig. 6.

Fig. 6

Open in a new tab

Gentisic acid inhibits S100A9-induced M1/M2 polarization to improve the viability of DRG neurons. A The polarization levels of M1 and M2 macrophages are determined by flow cytometry. B Changes the M1/M2 ratio. C The levels of CD86 and CD206 in macrophages are measured by the protein assay kit. D The proliferation of DRG neurons is tested by CCK-8. E TRPV1 expression in DRG neurons is detected by immunofluorescence. (F) TLR4, p65, and p-p65 expression in DRG neurons are examined by Western blot. *p < 0.05 vs. IL-4. #p < 0.05 vs. IL-4 + S100A9. n = 3

Gentisic acid ameliorates intervertebral disc tissue injuries in LDH rats

To further investigate the therapeutic effects of Gentisic acid on LDH, we construct rat LDH models and perform interventions with Gentisic acid, EHT1864, and Celebrex (a non-steroidal anti-inflammatory drug for pain relief and inflammation reduction) (Kim et al. 2022). Behavioral studies show that the LDH models increase the Siegal neurological score of rats. After 5 weeks postoperatively, Gentisic acid, EHT1864, and Celebrex all reduce this score, indicating they could alleviate the neurological injuries of the LDH rats (Fig. 7A). For pain perception (assessed via TWL and PWT), LDH group rats have lower TWL and PWT than the Sham group. After 5 weeks postoperatively, the three interventions all alleviate these abnormalities (Fig. 7B). Histologically, the model group exhibits characteristic intervertebral disc degeneration. Hematoxylin–eosin (H&E) staining reveals a disrupted NP structure, evident by proteoglycan loss, abnormal cell morphology, and inflammatory infiltration. This is further confirmed by Safranin O-fast green staining, which shows severe proteoglycan depletion and cellular injury in the NP, such as shrunken morphology and a reduced number of cells. In contrast, all three interventions significantly alleviate histopathological injuries in both the nucleus pulposus and the annulus fibrosus and exhibit a protective effect on disc cells (Fig. 7C, D). Additionally, the three interventions reverse LDH-induced elevated TRPV1 expression in rat dorsal root ganglia (Fig. 7E). These results demonstrate that Gentisic acid could ameliorate intervertebral disc tissue injuries in LDH rats. Its therapeutic efficacy is comparable to the clinical drug Celebrex, while the Rac1 inhibitor EHT1864 also shows a similar protective effect, suggesting a potential role for Rac1 signaling in the process ameliorated by Gentisic acid.

Fig. 7.

Fig. 7

Open in a new tab

Gentisic acid ameliorates intervertebral disc tissue injuries in LDH rats. A Siegal neurological score in rats. B TWL and PWT in rats. C H&E staining observation of intervertebral disc pathology in rats. D Safranin O-solid green staining observation of intervertebral disc cell injuries in rats. E Western blot detection of TRPV1 expression in rat dorsal root ganglia. *p < 0.05 vs. Sham. #p < 0.05 vs. LDH. n = 12 rats/group

Gentisic acid improves intervertebral disc tissue lesions in LDH rats through the MAPK14/S100A9/Rac1/2 signaling pathway

Next, we further explore the molecular mechanisms by which Gentisic acid ameliorates LDH in rats. Western blot exhibits that Gentisic acid, EHT1864, and Celebrex (the three interventions) decrease S100A9 and MAPK14 expression while suppressing Rac1/2/RhoA in LDH rats (Fig. 8A-B). Electron microscopy observation shows that the three interventions ameliorate autophagic flow blockage in NP tissues (Fig. 8C). Moreover, they reverse LDH-induced effects, upregulating LC3II/I and Beclin-1 expression while reducing p62 levels (Fig. 8D). ELISA results show that the LDH group has increased S100A9 and pro-inflammatory factor TNF-α levels, with decreased anti-inflammatory factor IL-4. After intervention with Gentisic acid, EHT1864, and Celebrex, TNF-α and S100A9 levels decrease, while IL-4 levels recover (Fig. 8E). For M1/M2 immunity in LDH rats, the three interventions reverse LDH-induced elevated CD86 levels and reduced CD206 levels, suggesting their involvement in M1/M2 immune regulation (Fig. 8F). Additionally, they reverse LDH-induced elevated TLR4 and p65/p-p65 expression (Fig. 8G). Consistent with the aforementioned results, the therapeutic effect of Gentisic acid is comparable to that of EHT1864 and Celebrex. These results reveal that Gentisic acid may ameliorate M1/M2 polarization and mitochondrial function in intervertebral disc tissues via the MAPK14/S100A9/Rac1/2 signaling pathway to alleviate symptoms.

Fig. 8.

Fig. 8

Fig. 8

Open in a new tab

Gentisic acid improves intervertebral disc tissue lesions in LDH rats through the MAPK14/S100A9/Rac1/2 signaling pathway. A Western blot detection of S100A9 expression in intervertebral disc tissues. B Western blot detection of GTP-Rac1 + Rac2/total-Rac1 + Rac2 and GTP-RhoA/total RhoA expression in intervertebral disc tissues. C Electron microscopic observation of the number of mitochondrial injuries in intervertebral disc tissues. Yellow arrows represent normal mitochondria. Red arrows represent injured mitochondria. Green arrows represent autophagosomes. Scale bar = 1 μm. D Western blot detection of LC3II/I, Beclin-1, and p62 levels in intervertebral disc tissues. E The kit detection of TNF-α, IL-4, and S100A9 levels in intervertebral disc tissues. F Western blot detection of CD86 and CD206 levels in intervertebral disc tissues. G Western blot detection of TLR4, p65, and p-p65 expression in intervertebral disc tissues. *p < 0.05 vs. Sham. #p < 0.05 vs. LDH. n = 12 rats/group

Discussion

LDH is a common chronic lower back disease, and its treatment remains challenging [30]. To improve LDH treatment, we demonstrate the potential mechanism of Gentisic acid in LDH and confirm its effectiveness in alleviating LDH by regulating the MAPK14/S100A9/Rac1/2 signaling pathway and M1/M2 immune activation. These findings may provide useful clues for exploring new LDH treatment avenues.

This study also validates the endogenous metabolic mechanism of Gentisic acid in LDH. Salicylic acid exerts its effects dependent on the MAPK cascade signaling pathway (Li et al. 2025a). As an endogenous metabolite of salicylic acid capable of binding to MAPK14, Gentisic acid’s metabolic homeostasis is likely regulated by the MAPK14/S100A9 axis. Under LDH conditions, abnormal activation of the MAPK14/S100A9 axis may disrupt the metabolic balance of Gentisic acid and exacerbate inflammatory injury. In this study, Gentisic acid could modulate this axis, suggesting that Gentisic acid may also exert therapeutic effects by restoring endogenous metabolic homeostasis, rather than solely relying on the pharmacological effects of exogenous drugs.

S100A9 is a calcium-binding protein of the S100 family (Chen et al. 2023b) that associates with inflammation, immune response, cell proliferation, cell migration, and tissue repair (Xia et al. 2024). Research demonstrates that S100A9 exacerbates nerve injury by inducing neuroinflammation and activating inflammatory vesicles (Wang et al. 2024). In this study, we find S100A9 levels are positively correlated with disease severity and inflammation in LDH patients, consistent with previous studies showing S100A9 exerts pro-inflammatory effects in inflammatory diseases.

Furthermore, research indicates that S100A9 binds to TLR4 and activates NF-κB to diminish M2 macrophage development while inducing pro-inflammatory functions in these cells (Franz et al. 2022). We confirm S100A9 induces M1/M2 macrophage polarization. After co-culturing macrophages and DRG neurons, S100A9 exerts pro-inflammatory effects on DRG neurons, further verifying the critical link between S100A9 and inflammation.

Several drugs can exert their effects through S100A9. For instance, A S100A9 inhibitor alleviates Sepsis-induced liver injuries (Zhang et al. 2023b), and melatonin ameliorates atherosclerosis by inhibiting S100A9-mediated vascular inflammation (Chen et al. 2023a). Given this, we investigate the molecular network connecting Gentisic acid and S100A9. Venn diagram shows the common target of Gentisic acid and S100A9 is MAPK14. As a key member of the MAPK signaling pathway, MAPK14 participates in inflammation regulation in many diseases (Madkour et al. 2021). In acute respiratory distress syndrome, MAPK14 activates the S100A8/S100A9 pathway to improve resident alveolar macrophage self-renewal and reduce inflammation (Ye et al. 2025). Both S100A9 inhibitors and melatonin act directly on S100A9 to inhibit its expression. In comparison, our study demonstrates that Gentisic acid binds to MAPK14 and may inhibit S100A9 expression by blocking MAPK14-mediated activation of S100A9.

From the perspective of inflammation regulation, the Rac1/2 pathway influences the macrophage polarization toward M1 or M2 phenotypes by regulating cytoskeletal rearrangement and reactive oxygen species production (Fu et al. 2023; Ngo et al. 2021). In this study, the Rac1/2 pathway is shown to interact with the MAPK14/S100A9 axis. Binding of gentisic acid to MAPK14 suppresses Rac1/2 overactivation, resulting in reduced M1 macrophage polarization. On the other hand, autophagy plays an essential role in the maintenance of neuronal homeostasis and stress response (Cheng et al. 2025). The Rac1/2 pathway participates in regulating autophagy (Ding et al. 2024), influencing the expression of autophagy-related proteins (LC3II/I, Beclin-1, and p62) and autophagosome formation (Lu et al. 2020; Xu et al. 2021). Under LDH pathological conditions, aberrant Rac1/2 activation may disrupt autophagic flux in DRG neurons. However, Gentisic acid modulates Rac1/2 activity through the MAPK14/S100A9 axis, which improves autophagy, promotes the clearance of damaged organelles and misfolded proteins, and enhances neuronal survival. Therefore, Gentisic acid exerts its therapeutic effects by regulating macrophage polarization and neuronal autophagic function, potentially offering an entirely new mechanism of action and innovative approaches for LDH treatment.

Research shows that reducing inflammation-related signals effectively alleviates pain in LDH rat models (Wu et al. 2024). For instance, desmethoxycurcumin mitigates inflammatory responses in LDH via MAPK and NF-κB pathways (Lu et al. 2022), while TAK-242 alleviates pain behaviors in LDH rats by blocking TLR4/NLRP3 inflammasome activation (Hu et al. 2024). Our research also confirmed that Gentisic acid repairs intervertebral disc tissue damage in LDH rat models through inflammatory regulation. It is worth noting that our current research results mainly focus on the regulatory role of Gentisic acid in the MAPK14/S100A9/Rac1/2 axis (a pathway related to both inflammation and neuronal function). MAPK14 exhibits molecular crosstalk with NF-κB or NLRP3 (Akaras et al. 2025). Thus, Gentisic acid may also synergistically inhibit LDH-related inflammation by interfering with the phosphorylation of NF-κB p65 or the assembly of the NLRP3 inflammasome. This would help further expand and clarify the molecular mechanism network of Gentisic acid in LDH. Future research will further explore these potential cross-pathway effects, providing more comprehensive evidence for the therapeutic application of Gentisic acid in LDH.

Moreover, Gentisic acid exhibits therapeutic effects on LDH rats that are comparable to those of Celebrex. Celebrex, a first-line medication for LDH, is known to alleviate pain through mechanisms potentially involving gut microbiota-mediated activation of the Wnt/β-catenin pathway (Li et al. 2024). While the mechanism of Gentisic acid may differ, its efficacy is supported by its documented biological activities, including the modulation of autophagy (Hosseinzadeh et al. 2024) and angiogenesis (Angulo et al. 2015), underscoring its potential as a multi-faceted therapeutic agent for LDH.

However, this study has some limitations. The in vitro system employed in this study cannot fully replicate the complex in vivo microenvironment of the intervertebral disc. The exclusion of female rats also limits the generalizability of our findings to both sexes. Moreover, the lack of an annulus-puncture only control group and/or an NP to root control group restricts the ability to fully disentangle the specific contributions of mechanical damage (from annulus puncture) and inflammatory responses (from nucleus pulposus exposure) to LDH pathology, and future studies are proposed to incorporate these groups for more nuanced mechanistic insights. Furthermore, additional in vivo experiments still need to be conducted in the future to further verify the regulatory role of the MAPK14/S100A9/Rac1/2 pathway in macrophage polarization. This is crucial to advance its clinical application and provide more effective treatment options for clinical therapy.

In summary, we conclude that Gentisic acid could ameliorate LDH by regulating the MAPK14/S100A9/Rac1/2 signaling pathway and the M1/M2 immune activation. Gentisic acid may have the potential to become a promising drug for treating LDH. This study may provide new ideas for LDH treatment and could offer important clues for understanding its pathogenesis.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 593 KB) (593.3KB, docx)
Supplementary file2 (DOCX 365 KB) (365.3KB, docx)

Acknowledgements

Thank the doctors and nurses in the department for their help in the collection of clinical samples.

Authors' contributions

Shuoqi Li: Data curation, Formal Analysis, Investigation, Methodology, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing. Tiezhu Chen and Xiongjie Shen: Data curation, Formal Analysis, Investigation, Validation, Writing – review & editing. These two authors (Tiezhu Chen and Xiongjie Shen) contributed equally to this study. Wanying Su: Data curation, Writing – review & editing. Xiaosheng Li: Conceptualization, Supervision, Writing – review & editing.

Funding

The work was supported by [Research and Development Program of Hunan Province of China] (Grant number [2022SK2024]).

Data availability

The datasets generated during and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Declarations

Ethical approval

The study involved human samples was approved by the Ethics Committee of Hunan Provincial People's Hospital. Informed consent was obtained from all individual participants included in the study. The study was approved by the Ethics Committee of Hunan Provincial People's Hospital (2020KYLSD29H).

Consent to participate

The study involved human samples was conducted after informed consent was obtained from all participants.

Consent to publish

Not applicable.

Clinical trial number

Not applicable.

Conflicts of interest

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Tiezhu Chen and Xiongjie Shen these two authors (Tiezhu Chen and Xiongjie Shen) contributed equally to this study.

References

  1. Abedi F, Razavi BM, Hosseinzadeh H. A review on gentisic acid as a plant derived phenolic acid and metabolite of aspirin: comprehensive pharmacology, toxicology, and some pharmaceutical aspects. Phytother Res. 2020;34:729–41. 10.1002/ptr.6573. [DOI] [PubMed] [Google Scholar]
  2. Akaras N, Şimşek H, İleritürk M, Küçükler S, Gür C, Kandemir FM. Carvacrol mitigates mercury chloride induced neurotoxicity by regulation of NRF-2/HO-1/NF-κB expression. J Trace Elem Med Biol. 2025;91:127715. 10.1016/j.jtemb.2025.127715. [DOI] [PubMed] [Google Scholar]
  3. Angulo J, Cuevas P, Cuevas B, El Youssef M, Fernández A, Martínez-Salamanca E, et al. Diacetyloxyl derivatization of the fibroblast growth factor inhibitor dobesilate enhances its anti-inflammatory, anti-angiogenic and anti-tumoral activities. J Transl Med. 2015;13:48. 10.1186/s12967-015-0413-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Atta AA, Ibrahim WW, Mohamed AF, Abdelkader NF. Microglia polarization in nociplastic pain: mechanisms and perspectives. Inflammopharmacology. 2023;31(3):1053–67. 10.1007/s10787-023-01216-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bruder-Nascimento T, Callera GE, Montezano AC, Belin de Chantemele EJ, Tostes RC, Touyz RM. Atorvastatin inhibits pro-inflammatory actions of aldosterone in vascular smooth muscle cells by reducing oxidative stress. Life Sci. 2019;221:29–34. 10.1016/j.lfs.2019.01.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chen L, Wang X, Liu C, Deng P, Pan L, Yang L, et al. Melatonin ameliorates atherosclerosis by suppressing S100a9-mediated vascular inflammation. Eur J Pharmacol. 2023a;957:175965. 10.1016/j.ejphar.2023.175965. [DOI] [PubMed] [Google Scholar]
  7. Chen Y, Ouyang Y, Li Z, Wang X, Ma J. S100A8 and S100A9 in cancer. Biochimica et Biophysica Acta (BBA). 2023;1878:188891. 10.1016/j.bbcan.2023.188891. [DOI] [PubMed] [Google Scholar]
  8. Cheng Z, Gan W, Xiang Q, Zhao K, Gao H, Chen Y, et al. Impaired degradation of PLCG1 by chaperone-mediated autophagy promotes cellular senescence and intervertebral disc degeneration. Autophagy. 2025;21:352–73. 10.1080/15548627.2024.2395797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Colicchia M, Schrottmaier WC, Perrella G, Reyat JS, Begum J, Slater A, et al. S100A8/A9 drives the formation of procoagulant platelets through GPIbα. Blood. 2022;140:2626–43. 10.1182/blood.2021014966. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ding X, Gao X, Ren A, Xu J, Jiang X, Liang X, et al. Sevoflurane enhances autophagy via Rac1 to attenuate lung ischaemia-reperfusion injury. Chem Biol Interact. 2024;397:111078. 10.1016/j.cbi.2024.111078. [DOI] [PubMed] [Google Scholar]
  11. Franz S, Ertel A, Engel KM, Simon JC, Saalbach A. Overexpression of S100A9 in obesity impairs macrophage differentiation via TLR4-NFkB-signaling worsening inflammation and wound healing. Theranostics. 2022;12:1659–82. 10.7150/thno.67174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fu H, Zhang P, Zhao XD, Zhong XY. Interfering with Rac1-activation during neonatal monocyte-macrophage differentiation influences the inflammatory responses of M1 macrophages. Cell Death Dis. 2023;14:619. 10.1038/s41419-023-06150-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gao L, Kong L, Zhao Y. The regulatory role of Rho GTPases and their substrates in osteoclastogenesis. Curr Drug Targets. 2021;22:1064–70. 10.2174/1389450121666200925150446. [DOI] [PubMed] [Google Scholar]
  14. Gu C, Liu Y, Lv J, Zhang C, Huang Z, Jiang Q, et al. Kurarinone regulates Th17/Treg balance and ameliorates autoimmune uveitis via Rac1 inhibition. J Adv Res. 2025;69:381–98. 10.1016/j.jare.2024.03.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Guo F, Song F, Chen Z, Niu N, Sun L, Yan M, et al. Mesenchymal stem cell-derived exosomal CBLB ameliorates infantile pneumonia progression probably by ubiquitinating MAPK14. J Inflamm. 2025;22:23. 10.1186/s12950-025-00450-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hosseinzadeh Y, Ghasemzadeh Rahbardar M, Mehri S, Razavi BM, Hosseinzadeh H. Protective effect of aspirin and gentisic acid, a plant-derived phenolic acid, on acrylamide-induced neurotoxicity by inhibiting apoptosis and autophagy. Naunyn Schmiedebergs Arch Pharmacol. 2024. 10.1007/s00210-024-03498-6. [DOI] [PubMed] [Google Scholar]
  17. Hu Y, Zhao X, Chen M, Zhou F, Zhang X, Chen C, et al. Massage ameliorates lumbar disc herniation-related radicular pain in rats by suppressing TLR4/NLRP3 inflammasome signaling transduction. J Orthop Surg (Hong Kong). 2024;32:10225536241238638. 10.1177/10225536241238638. [DOI] [PubMed] [Google Scholar]
  18. Huang SJ, Yan JQ, Luo H, Zhou LY, Luo JG. IL-33/ST2 signaling contributes to radicular pain by modulating MAPK and NF-κB activation and inflammatory mediator expression in the spinal cord in rat models of noncompressive lumber disk herniation. J Neuroinflammation. 2018;15:12. 10.1186/s12974-017-1021-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Jiang M, Wang K, Huang Y, Zhang X, Yang T, Zhan K, et al. Quercetin alleviates lipopolysaccharide-induced cell oxidative stress and inflammatory responses via regulation of the TLR4-NF-κB signaling pathway in bovine rumen epithelial cells. Toxins (Basel). 2023. 10.3390/toxins15080512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kim J, Seo BB, Hong KH, Kim SE, Kim YM, Song SC. Long-term anti-inflammatory effects of injectable celecoxib nanoparticle hydrogels for Achilles tendon regeneration. Acta Biomater. 2022;144:183–94. 10.1016/j.actbio.2022.03.033. [DOI] [PubMed] [Google Scholar]
  21. Lauwers M, Frikha A, Fage D, Duterme C, De Vos N, Cotton F, et al. Development and validation of a simple correction method for the measurement of neuron-specific enolase in hemolyzed serum samples. Scand J Clin Lab Invest. 2022;82:12–7. 10.1080/00365513.2021.2009021. [DOI] [PubMed] [Google Scholar]
  22. Li X, Li S, Zang Z, He Y. Yaobishu regulates inflammatory, metabolic, autophagic, and apoptosis pathways to attenuate lumbar disc herniation. Oxid Med Cell Longev. 2022;2022:3861380. 10.1155/2022/3861380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li X, Shi H, Zhang D, Jing B, Chen Z, Zheng Y, et al. Paeonol alleviates neuropathic pain by modulating microglial M1 and M2 polarization via the RhoA/p38MAPK signaling pathway. CNS Neurosci Ther. 2023;29:2666–79. 10.1111/cns.14211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li XC, Luo SJ, Wu F, Mu QC, Yang JH, Jiang C, et al. Investigation of macrophage polarization in herniated nucleus pulposus of patients with lumbar intervertebral disc herniation. J Orthop Res. 2023;41:1335–47. 10.1002/jor.25480. [DOI] [PubMed] [Google Scholar]
  25. Li S, Chen T, Zhou Y, Li X. Palmitic acid and trans-4-hydroxy-3-methoxycinnamate, the active ingredients of Yaobishu formula, reduce inflammation and pain by regulating gut microbiota and metabolic changes after lumbar disc herniation to activate autophagy and the Wnt/β-catenin pathway. Biochimica et Biophysica Acta (BBA). 2024;1870:166972. 10.1016/j.bbadis.2023.166972. [DOI] [PubMed] [Google Scholar]
  26. Li C, Wang K, Lei C, Zou Y, Yang S, Xiang F, et al. Β-aminobutyric acid-induced resistance in postharvest peach fruit involves interaction between the MAPK cascade and SNARE13 protein in the salicylic acid-dependent pathway. J Exp Bot. 2025a;76:1202–29. 10.1093/jxb/erae448. [DOI] [PubMed] [Google Scholar]
  27. Li G, Wan Y, Jiao A, Jiang K, Cui G, Tang J, et al. Breaking boundaries: chronic diseases and the frontiers of immune microenvironments. Med Res. 2025b;1:62–102. 10.1002/mdr2.70007. [Google Scholar]
  28. Liu ZH, Miao GS, Wang JN, Yang CX, Fu ZJ, Sun T. Resolvin D1 inhibits mechanical hypersensitivity in sciatica by modulating the expression of nuclear factor-κB, phospho-extracellular signal-regulated kinase, and pro- and antiinflammatory cytokines in the spinal cord and dorsal root ganglion. Anesthesiology. 2016;124:934–44. 10.1097/aln.0000000000001010. [DOI] [PubMed] [Google Scholar]
  29. Lu W, Lin J, Zheng D, Hong C, Ke L, Wu X, et al. Overexpression of microrna-133a inhibits apoptosis and autophagy in a cell model of Parkinson’s disease by downregulating Ras-related C3 botulinum toxin substrate 1 (RAC1). Med Sci Monit. 2020;26:e922032. 10.12659/msm.922032. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  30. Lu B, Chen X, Chen H, Li Q, Li H, Xu Y, et al. Demethoxycurcumin mitigates inflammatory responses in lumbar disc herniation via MAPK and NF-κB pathways in vivo and in vitro. Int Immunopharmacol. 2022;108:108914. 10.1016/j.intimp.2022.108914. [DOI] [PubMed] [Google Scholar]
  31. Madkour MM, Anbar HS, El-Gamal MI. Current status and future prospects of p38α/MAPK14 kinase and its inhibitors. Eur J Med Chem. 2021;213:113216. 10.1016/j.ejmech.2021.113216. [DOI] [PubMed] [Google Scholar]
  32. Mohamed Elashiry M, Tian F, Elashiry M, Zeitoun R, Elsayed R, Andrews ML, et al. Enterococcus faecalis shifts macrophage polarization toward M1-like phenotype with an altered cytokine profile. J Oral Microbiol. 2021;13:1868152. 10.1080/20002297.2020.1868152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Neumann A, Efimova I, Shteinberg I, Tron R, Gaskin A, Erez N, et al. S100A8/S100A9 impairs energy expenditure and whole body metabolism. Am J Physiol Endocrinol Metab. 2025;329:E341–54. 10.1152/ajpendo.00076.2025. [DOI] [PubMed] [Google Scholar]
  34. Ngo ATP, Parra-Izquierdo I, Aslan JE, McCarty OJT. Rho GTPase regulation of reactive oxygen species generation and signalling in platelet function and disease. Small GTPases. 2021;12:440–57. 10.1080/21541248.2021.1878001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Noei Razliqi R, Harooni E, Neisi N, Jafar Sameri M, Ahangarpour A. Hepatoprotective effects of gentisic acid through its anti-oxidant properties in nicotinamide-streptozotocin induced diabetic mice. Iran J Basic Med Sci. 2023;26:1409–15. 10.22038/ijbms.2023.70659.15359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Oh SJ, Lim JY, Son MK, Ahn JH, Song KH, Lee HJ, et al. TRPV1 inhibition overcomes cisplatin resistance by blocking autophagy-mediated hyperactivation of EGFR signaling pathway. Nat Commun. 2023;14:2691. 10.1038/s41467-023-38318-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Piovan A, Filippini R, Argentini C, Moro S, Giusti P, Zusso M. The effect of C-phycocyanin on microglia activation is mediated by Toll-like Receptor 4. Int J Mol Sci. 2022. 10.3390/ijms23031440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Qin X, Sun K, Xu W, Gao J, Jiang H, Chen W, et al. An evidence-based guideline on treating lumbar disc herniation with traditional Chinese medicine. J Evid Based Med. 2024;17:187–206. 10.1111/jebm.12598. [DOI] [PubMed] [Google Scholar]
  39. Rogerson A, Aidlen J, Jenis LG. Persistent radiculopathy after surgical treatment for lumbar disc herniation: causes and treatment options. Int Orthop. 2019;43:969–73. 10.1007/s00264-018-4246-7. [DOI] [PubMed] [Google Scholar]
  40. Saeedavi M, Goudarzi M, Fatemi I, Basir Z, Noori SMA, Mehrzadi S. Gentisic acid mitigates gentamicin-induced nephrotoxicity in rats. Tissue Cell. 2023;84:102191. 10.1016/j.tice.2023.102191. [DOI] [PubMed] [Google Scholar]
  41. Wang G, Huang K, Tian Q, Guo Y, Liu C, Li Z, et al. S100A9 aggravates early brain injury after subarachnoid hemorrhage via inducing neuroinflammation and inflammasome activation. iScience. 2024;27:109165. 10.1016/j.isci.2024.109165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wu H, Zheng J, Xu S, Fang Y, Wu Y, Zeng J, et al. Mer regulates microglial/macrophage M1/M2 polarization and alleviates neuroinflammation following traumatic brain injury. J Neuroinflammation. 2021;18:2. 10.1186/s12974-020-02041-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wu S, Chen Z, Wu Y, Shi Q, Yang E, Zhang B, et al. ADSC-exos enhance functional recovery after spinal cord injury by inhibiting ferroptosis and promoting the survival and function of endothelial cells through the NRF2/SLC7A11/GPX4 pathway. Biomed Pharmacother. 2024;172:116225. 10.1016/j.biopha.2024.116225. [DOI] [PubMed] [Google Scholar]
  44. Xia P, Ji X, Yan L, Lian S, Chen Z, Luo Y. Roles of S100A8, S100A9 and S100A12 in infection, inflammation and immunity. Immunology. 2024;171:365–76. 10.1111/imm.13722. [DOI] [PubMed] [Google Scholar]
  45. Xiu Y, Zhang H, Wang S, Gan T, Wei M, Zhou S, et al. cDNA cloning, characterization, and expression analysis of the Rac1 and Rac2 genes from Cynoglossus semilaevis. Fish Shellfish Immunol. 2019;84:998–1006. 10.1016/j.fsi.2018.11.006. [DOI] [PubMed] [Google Scholar]
  46. Xu Z, Li B, Du Y, Zhang R, Tong X, Li J, et al. Helicobacter pylori regulates ILK to influence autophagy through Rac1 and RhoA signaling pathways in gastric epithelial cells. Microb Pathog. 2021;158:105054. 10.1016/j.micpath.2021.105054. [DOI] [PubMed] [Google Scholar]
  47. Xu L, Ren H, Xie D, Zhang F, Hu X, Fang S, et al. Rac2 mediate foam cell formation and associated immune responses in THP-1 to promote the process of atherosclerotic plaques. Mol Immunol. 2023;163:196–206. 10.1016/j.molimm.2023.10.004. [DOI] [PubMed] [Google Scholar]
  48. Ye Y, Yang Q, Wei J, Shen C, Wang H, Zhuang R, et al. Rvd1 improves resident alveolar macrophage self-renewal via the ALX/MAPK14/S100A8/A9 pathway in acute respiratory distress syndrome. J Adv Res. 2025;67:289–99. 10.1016/j.jare.2024.01.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yu P, Mao F, Chen J, Ma X, Dai Y, Liu G, et al. Characteristics and mechanisms of resorption in lumbar disc herniation. Arthritis Res Ther. 2022;24:205. 10.1186/s13075-022-02894-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zhang AS, Xu A, Ansari K, Hardacker K, Anderson G, Alsoof D, et al. Lumbar disc herniation: diagnosis and management. Am J Med. 2023a;136:645–51. 10.1016/j.amjmed.2023.03.024. [DOI] [PubMed] [Google Scholar]
  51. Zhang Y, Wu F, Teng F, Guo S, Li H. Deficiency of S100A9 alleviates sepsis-induced acute liver injury through regulating AKT-AMPK-dependent mitochondrial energy metabolism. Int J Mol Sci. 2023b. 10.3390/ijms24032112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhou F, Mei J, Han X, Li H, Yang S, Wang M, et al. Kinsenoside attenuates osteoarthritis by repolarizing macrophages through inactivating NF-κB/MAPK signaling and protecting chondrocytes. Acta Pharm Sin B. 2019;9:973–85. 10.1016/j.apsb.2019.01.015. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary file1 (DOCX 593 KB) (593.3KB, docx)
Supplementary file2 (DOCX 365 KB) (365.3KB, docx)

Data Availability Statement

The datasets generated during and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Articles from Cell Biology and Toxicology are provided here courtesy of Springer

RESOURCES