This study used female C57BL/6JNifdc mice (18–22 g, 6–8 weeks). The mice were purchased from Beijing Huafukang Biotechnology Co., Ltd. (Beijing, China, License No.: SCXK (Jing) 2024-0003) and were inhabited a temperature-controlled environment maintained between 22 and 24 °C and relative humidity levels spanning 60–80%, subjects benefited from unrestricted sustenance and hydration access amidst a 12-h photoperiod schedule. The animal study was approved by the Animal Welfare and Ethics Committee of Tianjin Medical University General Hospital (Approval No.: IRB2022-DWFL-201). All surgical procedures were performed in a sterile environment. The mouse T10 spinal cord contusion model was established using a modified Allen method adapted from rat spinal cord contusion models (Hua et al., 2024). Mice were subjected to isoflurane-induced anesthesia, succeeded by a longitudinal incision penetrating the skin and fascia, centered upon the T10 level. Muscular retraction then exposed the T9–T10 spinous processes, with the T10 lamina’s removal divulging the underlying spinal cord. The T9 and T10 spinous processes found stabilization via a spinous process fixator, setting the stage for spinal cord contusion induction utilizing the NYU Impactor-III. The Omav intervention arm received 100 μL (10 mg/kg) Omav intraperitoneally, whereas controls received a PBS volume equivalent. Omav administration commenced half an hour post-SCI establishment and extended daily for a seven-day regimen, with dosage calibration derived from preceding investigations (Jiang et al., 2022; Creelan et al., 2017; Liu et al., 2016).

Extraction of bone marrow from the femur and tibia of male C57BL/6 mice inaugurates the experimental protocol. Subsequent incubation of these cells in an induction medium containing DMEM/F-12, FBS, penicillin/streptomycin, and M-CSF facilitates their differentiation into M0 macrophages over 7 days, establishing a control group baseline. Further exposing M0 macrophages to LPS elicits a pro-inflammatory response, henceforth referred to as the LPS group. Conversely, an LPS and Omav co-treatment regimen for 24 h modulates macrophage polarization, instigating an anti-inflammatory cascade; these cells are designated as the SCI plus Omav group.

Utilizing BMS facilitates the evaluation of hind limb motor function restoration in murine subjects subsequent to SCI. Specifically, two trained experimenters conduct a 4-min evaluation of the mice in an empty room. Scores range from 0, indicating no motor function, to 9, representing normal function (Basso et al., 2006).

Eight weeks after SCI, electrophysiological activity of the hind limbs of mice was assessed using an electrophysiological device (Ran et al., 2023). Under the cloak of anesthesia, mice were prepped for electrophysiological assessment, commencing with strategic placement of two stimulating electrodes—one situated posteriorly on the cranium and the other aligned with the cervical region to document and juxtapose the peak amplitude and latency of MEPs across diverse murine cohorts.

Eight weeks after SCI surgery, the CatWalk XT system, comprising both hardware and software components, To objectively dissect the paw imprints, motor functionality, and interlimb synchrony of mice (Hua et al., 2024; Creelan et al., 2017), we harnessed the CatWalk gait analysis system. Bathed in the seclusion of a dark and quiet ambiance, mice embarked on a training regimen, traversing the illuminated glass corridor at least thrice. Once acclimatized, high-speed camera surveillance chronicled each unique paw impression as mice ambulated across the walkway. Data collection and analysis were performed using the CatWalk XT software, including calculation of the regularity index (RI) and average speed. The RI, defined as the ratio as a ratio of standard step sequence instances to aggregate paw placements, as gait analysis in SCI unveils, a direct correlation emerges between accurate gait sequences and regularity index escalation during movement. Additionally, stride length, a pivotal gait parameter, frequently succumbs to a marked decline post-SCI.

Seven days after SCI, mice were deeply anesthetized with isoflurane, and the heart was perfused with phosphate-buffered saline (PBS). Once the liver appeared pale, perfusion was continued with 4% paraformaldehyde to fix the tissues. The spinal cord tissue was then dissected and fixed in 4% paraformaldehyde for 1 day. The tissue was subsequently dehydrated by sequential immersion in 10, 20, and 30% sucrose solutions, each for 1 day. Following desiccation, the spinal cord parenchyma was enshrouded in OCT compound (SAKURA, 4583) and sequestered in a frozen state at −20 °C. Employing a cryostat (CM30505; Leica), the tissue was sculpted into cryosections of 10 μm breadth. The slices, once stabilized with 4% paraformaldehyde for 30 min, were rinsed in PBS and thereafter subjected to a 1-h blockade and permeabilization using 5% fetal bovine serum (FBS) with 0.25% Triton X-100. Subsequently, the preparations were co-incubated with primary antibodies targeting CD68 (1:300, Abcam, ab53444), iNOS (1:200, Abcam, ab178945), and CD206 (1:200, Abcam, ab64693), Iba1 (1:200, ab283319) and afterward revealed with their cognate secondary antibodies. Chromatic delineation of nuclei was achieved with DAPI for 10 min, and imagery was acquired through a Leica fluorescence microscope (Leica TL LED, Germany). Quantitative scrutiny ensued via ImageJ software.

Immerse cells in a 4% PFA solution for fixation, succeeded by a blocking solution of 10% goat serum and 0.5% Triton X-100 for 1 h at ambient temperature. Subsequent to this, subject cells to a prolonged 4 °C incubation with primary antibodies CD68, iNOS, and CD206. Following this incubation period, introduce corresponding secondary antibodies for an additional hour at room temperature. Finally, stain nuclei with DAPI for 10 min, capture images using a Leica fluorescence microscope (Leica TL LED, Germany), and perform quantitative analysis using ImageJ software.

After preparing the paraffin sections, incubate the sections at 60 °C within the unwavering climes of an oven for a four-hour interval precedes a successive immersion in xylene I and xylene II, each enduring for a 10-min span each. Next, immerse the sections in ethanol solutions at different concentrations (100, 95, 90, 80, and 70%) for 5 min each. Rinse the sections with distilled water for 1 min, stain with hematoxylin (Solarbio, #G1120, Beijing, China) for 2 min, and differentiate using 1% hydrochloric acid solution. Immerse sections in distilled water for a 2-min rinse, then apply a 0.5% eosin solution for staining. Execute a sequential dehydration protocol, transitioning sections through 70, 80, 90%, and ultimately 100% ethanol solutions. Subsequently, subject slides to 5-min immersions in xylene I and xylene II, culminating in neutral resin encapsulation. Conclude the procedure by capturing high-resolution images using an automated tissue in situ multi-label quantitative analyzer (Vectra Polaris, PerkinElmer).

Isolated BMDM cells underwent 70 μm cell filtration, yielding a single-cell suspension primed for incubation with FITC anti-rat CD11b and PE anti-rat CD86 antibodies over 30 min. Succeeding this, a 15-min exposure to cell-fast™ Fix/Perm Buffer Set preceded an additional 30-min incubation with CD206 (mannose receptor) antibody in dimmed, ambient conditions. Thereafter, cells encountered goat anti-mouse IgG H&L for another round of incubation. Concluding this sequence, samples were harvested via BD LSRFortessa flow cytometer and their data underwent FlowJo-mediated analysis.

Protein quantification transpired via a BCA assay (Thermo Scientific), establishing the foundation for subsequent 12.5% SDS-PAGE-mediated target protein segregation and PVDF membrane transference. Upon PVDF membrane blocking with 5% non-fat milk in TBST, an overnight 4 °C incubation with GAPDH, iNOS, and CD206 antibodies ensued, followed by secondary antibody interaction under optimized parameters. Protein band visualization emerged through ECL reagents, producing images documented with the ChemiDoc XRS system. ImageJ software-directed quantitative analyses provided a comprehensive evaluation of protein expression patterns.

On the 3rd day post-SCI, the concentrations of inflammatory cytokines TNF-α, IL-1β, and IL-6 within spinal cord tissue homogenates were ascertained utilizing respective IL-6, IL-1β, and TNF-α ELISA kits. The spinal cord tissue homogenate supernatant was collected and processed to extract various cytokines. Following the kit instructions, the samples were diluted and centrifuged, then biotinylated antibodies from the kit were added to a 96-well plate and incubated at 37 °C for 1 h, followed by washing the plate three times. Then, enzyme conjugate working solution was added, and the plate was incubated at 37 °C for 30 min. A quintet of washes heralded substrate incorporation, succeeded by a 15-min incubation epoch at 37 °C. Culminating in stop solution admixture, the optical density values of individual wells were instantaneously gauged via a microplate reader at a 450 nm wavelength. Each group contained three samples, and each sample was tested three times to ensure the reliability of the experimental results.

Raw base call (BCL) files generated on the Illumina NovaSeq 6000 platform were converted to demultiplexed FASTQ files according to sample-specific barcodes. Adapter removal and quality filtering were performed using fastp (v0.23.4), which trimmed residual adapter sequences and discarded reads containing low-quality bases or excessive ambiguous nucleotides (N), yielding high-quality clean reads. These clean reads were subsequently aligned to the mouse reference genome (Mus musculus, GRCm39) using HISAT2 (v2.2.1). Gene models were defined according to the Ensembl Mus musculus gene annotation (Ensembl release 104, GTF format). For each sample, gene- and transcript-level expression was quantified with StringTie (v2.2.3) and reported as fragments per kilobase of transcript per million mapped reads (FPKM), and a corresponding gene-level raw count matrix was generated for downstream statistical analysis in R software.

The raw count matrix was imported into DESeq2 (v1.40) for normalization and differential expression analysis. To reduce noise from extremely lowly expressed genes, we first removed genes with fewer than 10 raw counts in at least two samples, resulting in 15,680 genes retained for downstream analyses. Library size normalization was performed using the median-of-ratios method implemented in DESeq2, and gene-wise dispersion parameters were estimated under the negative binomial framework. Variance-stabilizing transformation (VST) was then applied to the normalized counts to obtain an expression matrix with approximately homoscedastic variance across the dynamic range.

Principal component analysis (PCA) was conducted on the centered VST matrix using the base R function prcomp. The first two principal components were visualized with ggplot2 (v3.4) as a scatterplot, and 95% confidence ellipses were overlaid to illustrate within-group variability. For differential expression testing, a simplified design formula ~ condition was used in DESeq2 to fit negative binomial generalized linear models. Wald statistics were calculated for group contrasts, and log₂ fold changes (log₂FC) were shrunk using the apeglm method to improve effect-size estimation, particularly for lowly expressed genes. Genes with an adjusted p value (Benjamini–Hochberg–corrected, p_adj) < 0.05 and an absolute log₂FC > 2 were considered differentially expressed (DEGs). Volcano plots were generated with ggplot2, and clustered heatmaps were produced using pheatmap (v1.0.12) based on the top 140 DEGs, with rows centered to highlight relative expression changes; red and blue side bars indicated upregulated and downregulated genes, respectively.

Enrichment of significant DEGs was performed using clusterProfiler v4.6. GO-BP: enrichGO (q < 0.05, BH correction); redundancy was removed using simplify. Subsequently, a network was constructed using GOSemSim v2.23, igraph v1.5, and ggraph v2.1, based on Wang semantic similarity >0.3, with node size ∝ − log₁₀(p). KEGG: enrichKEGG (p < 0.05); displayed as a bubble chart, where bubble size represents the number of genes and color indicates −log₁₀(p_adj). GSEA: all genes were ranked by log₂FC; fgsea (multilevel mode) was used to run MSigDB C2 (mouse homolog). Pathways with |NES| > 1.5 and p_adj < 0.05 were considered significant and presented as a ridgeline-dot combination plot (ridgeline shows gene distribution, dot-plot encodes NES size/direction).

We calculated the Z-scores of 11 classic pathways, including MAPK, PI3K, and NF-κB, using the PROGENy model implemented in decoupleR v2.0. The analysis was performed with the top 1,000 genes, specifying the organism as mouse, and Z-scores were computed. The resulting data were visualized as a heatmap using ComplexHeatmap; hierarchical clustering was conducted using complete linkage criteria based on Euclidean distance.

First, we used biomaRt (Ensembl 104) to convert mouse gene symbols to human HGNC uppercase symbols. Next, we identified the intersection with the LM22 gene signature matrix, which contains at least 510 genes. We then called CIBERSORT.R with the LM22 gene signature matrix, performing 1,000 permutations and disabling quantile normalization, to estimate the relative proportions of immune cell types. We retained 10 cell types with abundance greater than 0.5% and plotted a heatmap normalized by rows, along with a stacked bar chart. For inter-group comparisons, we used a two-tailed unpaired t-test with Benjamini–Hochberg multiple testing correction.

Statistical scrutiny transpired via GraphPad Prism 9.2.0 software (GraphPad software, San Diego, CA, USA). Student’s t-test in concert with one-way analysis of variance (ANOVA), interlaced with Tukey’s post hoc assay, constituted the substratum for between-group evaluations, with the demarcations of significance ordained as follows: p < 0.05, *p < 0.01, **p < 0.001, ***p < 0.0001. Each experimental endeavor underwent at least three replications to ensure robustness. Subsequently, results manifested as the standard error of the mean (SEM).

Employing BMS scores as a metric of motor function recuperation post-SCI in mice, we discovered that Omav intervention corresponded with significantly enhanced BMS scores from 21 days post-injury (dpi) onwards, compared to untreated SCI controls (Figure 1B). Cat Walk gait analysis further elucidated superior hind limb coordination among SCI mice receiving Omav therapy (Figures 1C,E–J). To appraise nerve conduction enhancement, electrophysiological assessments were conducted on the subjects. MEP data unveiled shortened latency and amplified amplitude in SCI mice administered Omav (Figures 1D,K–L), suggesting improved neural signal transmission. Histological examinations through H&E staining illuminated reduced lesion dimensions in Omav-treated SCI mice (Figure 1M). Collectively, these findings corroborate Omav’s potential in fostering functional restoration following SCI. To investigate the long-term effects of Omav on functional recovery after SCI, we established three experimental groups: sham, SCI, SCI+Omav. The recovery of mice was monitored over an 8-week period post-injury (Figure 1A).

First, principal component analysis (PCA) was performed on the sequencing results of six bulk RNA-seq SCI samples (Figure 2A), revealing that the first principal component (PC1) explained about 84% of the variation in expression.

Further DEGs analysis (Figures 2B,C) identified a total of 422 significant DEGs with a screening threshold of |log₂FC| > 2.0 and FDR < 0.05, including 195 upregulated genes and 287 downregulated genes. Functional analysis revealed that downregulated genes were concentrated in immune-related transcription regulation (e.g., Bach2os/Bach2) and inflammation-induced apoptosis (e.g., Gpr17); while upregulated genes were mainly involved in matrix-related proteins, driving angiogenesis, fibroblast adhesion, wound healing (e.g., Ccn1/Cyr61), and tissue repair-related functional modules (e.g., Csrnp1). Heatmap clustering analysis (Figure 2C) showed that upregulated and downregulated genes formed two distinct modules, which could completely differentiate SCI and SCI + Omav group samples under unsupervised conditions.

GO biological process (GO-BP) enrichment network analysis (Figure 2D) and gene set enrichment analysis (GSEA) (Figure 2E) results indicated that Omav significantly inhibited immune response activation, such as myeloid leukocyte activation and macrophage migration, among other inflammation-related pathways. Meanwhile, neurotransmitter transport/secretion and cell adhesion-related pathways showed mild activation. Omav simultaneously suppressed immune-inflammatory pathways and cell cycle pathways, showing that its anti-inflammatory and anti-proliferative effects support synaptic repair. This suggests that Omav treatment inhibits immune-inflammatory pathways and cell cycle pathways, indicating that it’s anti-inflammatory and anti-proliferative effects complement synaptic repair. It also reflects that the spinal cord tissue after Omav treatment may be undergoing a functional state transition from a high metabolic, oxidative stress-active inflammatory state to a tissue repair-oriented state.

KEGG signaling pathway enrichment analysis (Figure 2F) further validated the above results, finding that Omav treatment significantly downregulated antigen processing and presentation functions and chemokine signaling pathways; while significantly upregulating cell adhesion molecules (CAMs), IL-17 signaling, and phenylalanine metabolism signaling pathways. These changes suggest that Omav may exert a dual regulatory role in both anti-inflammation and promoting antioxidant effects.

Box plot analysis of key genes (Top 20) (Figure 2G) further clarified the changes in key genes after Omav treatment: the expression levels of Gpr17, Plau, and Thap2 significantly decreased, suggesting a reduction in expression after drug treatment, which may alleviate inflammation or glial responses. In contrast, Hif3a, Egr1, and Ccn1 significantly increased, indicating that their expression levels rose after drug treatment, closely related to hypoxic responses and injury repair, potentially promoting the repair process.

To gain a deeper understanding of the specific signaling pathways affected by Omav after spinal cord injury, we used the PROGENy tool to assess pathway activity (Figure 3A). The results showed that Omav treatment significantly reduced the activity of MAPK, PI3K, and NF-κB pathways in spinal cord tissue after SCI; while p53 and hypoxia pathways were mildly upregulated.

Subsequently, we analyzed immune cell infiltration using the CIBERSORT algorithm (Figures 3B,C), finding that the proportion of M1 macrophages in SCI spinal cord tissue significantly decreased after Omav treatment, whilst the contingent of M2 macrophages exhibited a marked augmentation; additionally, the proportion of regulatory T cells (Treg) also showed a slight increasing trend. This indicates that the changes in these pathways after Omav treatment are related to macrophage polarization.

Box plot analysis of immune cells (Figures 3D,E) further clarified the changes in immune cell distribution ratios after Omav treatment: the proportion of activated dendritic cells (DC.Actived) and M1 macrophages significantly decreased (p < 0.001), suggesting that inflammation in SCI was suppressed after Omav treatment; while M2 macrophages and regulatory T cells (Treg) significantly increased, indicating that Omav treatment can promote the local immune microenvironment to shift toward inflammation suppression and tissue repair. In addition, we tested serum levels of IL-1β, IL-6, and TNF-α using ELISA kits. It revealed that after SCI, the levels of pro-inflammatory cytokines increased, which were decreased after Omav treatment (Figures 3F–H). These trends align with the classical characteristics of macrophages transitioning from pro-inflammatory (M1) to tissue repair (M2). This underscores the necessity to explore the research value of reducing inflammatory response and regulating macrophage polarization after SCI with Omav treatment.

To explore the mechanism by which Omav repairs spinal cord injury by regulating macrophage polarization, we performed immunofluorescence staining on spinal cord tissue at 7 dpi. We stained for macrophage marker CD68 and M1 phenotype marker iNOS (Figure 4A). Our findings unveiled a substantial decline in CD68 and iNOS positive cell counts (Figures 4C,D). Concomitantly, we harnessed CD68 and CD206 immunolabeling to illuminate macrophage/populations and M2 phenotypes (Figure 4B). Subsequent analyses unveiled diminished CD68+ cell presence in SCI mice receiving Omav (Figure 4E), while CD206+ cell proportion experienced a notable upswing (Figure 4F). In corroboration, western blot assays echoed this mechanistic shift (Figure 4G). Findings highlighted a significant elevation in CD206 levels among the SCI animals treated with Omav cohort compared to controls (Figure 4H), alongside a marked reduction in iNOS expression levels (Figure 4I).

Subsequently, we stained for microglia marker Iba1 and M1 phenotype marker iNOS (Figure 5A). The results showed a significant reduction of Iba1 and iNOS positive cells (Figures 5C,D). On the other hand, we stained for microglia marker Iba1 and M2 phenotype marker CD206 (Figure 5B). The results demonstrated a significant decrease in Iba1+ cells in the SCI + Omav group (Figure 4E), while the proportion of CD206+ cells significantly increased (Figure 5F).

Collectively, these results evince Omav’s capacity to modulate macrophage/microglia polarization in favor of the neuroprotective M2 phenotype in vivo.

Primary BMDMs were extracted from adult male C57 mice (Figure 6A). An LPS-provoked BMDM in vitro inflammation paradigm was instituted to further substantiate the modulatory influence of Omav upon M1/M2 phenotypic disposition. Flow cytometric interrogation was employed to discern the M1 phenotype (CD86+) (Figure 6B), and the findings disclosed that Omav administration attenuated the M1 polarization of LPS-stimulated BMDMs (Figure 6C). Simultaneously, flow cytometry was used to detect the M2 phenotype (CD206) (Figure 6D), and the results indicated that Omav treatment promoted M2 polarization (Figure 6E).

In an effort to further corroborate Omav’s modulatory influence on M1/M2 phenotypes, we resorted to immunofluorescence staining of macrophage identifier CD68 and M1 marker iNOS (Figure 7A). Consequential analyses unveiled a conspicuous decrease in relative average iNOS intensity within the LPS + Omav group (Figure 7C). Subsequently, we stained for macrophage marker CD68 and M2 phenotype marker CD206 (Figure 7B). The results showed that in the LPS + Omav group, the relative average intensity of CD206 significantly increased (Figure 7D). Additionally, we validated this mechanism through western blot (Figure 7E), a comprehensive analysis revealed Omav’s efficacy in suppressing M1 polarization within LPS-stimulated BMDMs, as evinced by a marked decline in iNOS expression levels (Figure 7F). Concurrently, a discernible elevation in CD206 levels served as an indicator of augmented M2 polarization (Figure 7G), substantiating Omav’s propensity to foster an anti-inflammatory macrophage phenotype.

To further study the effects of Omav on macrophages in the inflammatory response process after SCI, we conducted Omav drug intervention in the LPS-induced BMDMs in vitro inflammation model and analyzed RNA-seq data results. Principal component analysis results indicated clear separation between the LPS group and the LPS + Omav group in gene expression, consistent with the sequencing results of tissue samples. This emphasizes the significant impact of Omav treatment on macrophage inflammatory responses (Figure 8A).

Differential gene volcano plots and heatmaps clearly showed that under LPS-induced inflammatory conditions, several pro-inflammatory and immune activation genes were significantly downregulated after drug treatment (e.g., CCR5, C1q family), while anti-inflammatory or antioxidant protective genes (e.g., Mt1, Gclc, Gclm) were upregulated, indicating that Omav can effectively reduce inflammation caused by M1 macrophages and promote the tissue protection or repair of M2 macrophages. These gene changes reflect the effective regulation of the drug on immune responses and inflammatory processes (Figures 8B,C). GO-BP and GSEA analysis results indicated that the changes in gene expression of macrophages after drug (Omav) treatment were mainly concentrated in immune regulation, interferon signaling pathways, and cytoplasmic translation, suggesting that the regulation of these pathways or biological processes may be one of the important mechanisms by which the drug exerts anti-inflammatory and tissue repair effects (Figures 8D,E). KEGG pathway enrichment analysis results confirmed that Omav treatment significantly downregulated immune-inflammatory-related pathways (e.g., chemokine and cytokine pathways) while upregulating antioxidant and cell protection-related pathways (e.g., glutathione metabolism, ferroptosis pathways), demonstrating that Omav effectively improves the LPS-induced inflammatory state by regulating immune inflammatory responses and antioxidant protective pathways (Figure 8F).

By integrating RNA-seq data at the tissue level and BMDMs level (Figures 8G,H), we ultimately identified eight key target genes (Figure 8I). Among them, four genes were commonly upregulated in both datasets (Ceacam10, Gssos2, Gm14257, Ctsd4), while the other four were commonly downregulated (Ccr5, Ccl2, Rgs18, Dhrs3). The upregulated Ctsd4/CTSD can activate the lysosomal pathway of macrophages/microglia, promoting the clearance of myelin and necrotic cells, thereby reducing secondary injury; while the regulation of genes such as Ccr5 and Ccl2 is closely related to apoptosis, inflammation, and the inhibition of glial scars.

Subsequently, we performed quantitative analysis of CTSD, CCR5, and CCL2 using qRT-PCR. Compared to the SCI group, Omav treatment significantly upregulated the mRNA expression level of the lysosomal enzyme cathepsin D (CTSD) (Figure 7J). Correspondingly, the mRNA expression levels of CCL5 receptor 5 (Ccr5) and C-C chemokine ligand 2 (CCL2) were significantly decreased after Omav treatment compared to the SCI group (Figures 8K,L). The above cross-validation results indicate that Omav treatment significantly weakened the inflammatory response of macrophages, highlighting the central role of inflammation regulation in the pharmacological mechanism of Omav.