Female C57BL/6 mice (8–10 weeks old, 20–22 g weight) were supplied by the Animal Center of Anhui Medical University, China. The animals were acclimatized and housed under standard specific pathogen-free conditions (22 °C, 30–50% humidity, 12-h light/dark cycle; 4 mice/cage) with free access to food and water. After final assessments, all mice were humanely euthanized with carbon dioxide. Animal surgeries were performed under the same approved protocol as a prior study investigating a different therapeutic compound and mechanism.

Mice were randomly assigned to the sham operation group, vehicle group, or ATD treatment group. SCI was induced using a severe compression model. Briefly, after anesthesia induction with intraperitoneal 1% pentobarbital sodium (MERCK, Rahway, NJ, USA, CAS:57-33-0), mice underwent a T9–T10 laminectomy to expose the spinal cord. The exposed cord was then subjected to a 5-second compression using a calibrated Dumont #5 forceps (Fine Science Tools, Heidelberg, Germany, Cat1125220). The muscle and skin layers were then sutured sequentially. Postoperatively, all mice received antibiotic treatment to prevent infection and manual bladder expression twice daily until spontaneous urination was restored. Mice in the sham group underwent a laminectomy without spinal cord compression. The ATD treatment group received daily intraperitoneal injections of 20 or 40 mg/kg ATD (MCE, New Jersey, USA, HY-N0238) starting immediately after injury and continuing until day 7 post-injury. The vehicle group received an equivalent volume of the solvent to control for potential solvent-related effects. All surgical procedures were performed by a single, experienced, and independent surgeon to ensure consistency and minimize procedural variability.

Hindlimb motor function was assessed using the Basso Mouse Scale (BMS) score, the slant board test and footmark analysis test, as described previously. Mice were placed in an open field, and their hind limb movement was observed before operation and at 1, 3, 7, 14, 21, and 28 days post-injury (dpi). The footprint analysis test was conducted at 28 dpi to further evaluate the recovery of motor function. The front and back paws were soaked in green and red dyes, respectively, and the animals were placed on white paper to walk. All assessments were conducted in a blinded fashion by two independent, experienced investigators to ensure objectivity.

After the accomplishment of the necessary assessments, all mice were euthanized using carbon dioxide. The flow rate of CO₂ should be controlled at 50% of the volume of the euthanasia chamber per minute. The animals should be exposed to CO₂ for at least 5 minutes. Mice were transcardially perfused with 0.9% saline and then 4% paraformaldehyde (PFA) (Servicebio, Wuhan, Hubei, China, G1101). The spinal cord segment encompassing the injury epicenter was dissected and post-fixed overnight in 4% PFA at 4 °C. After sucrose gradient cryoprotection, the specimens were embedded in paraffin wax. Subsequently, longitudinal sections were cut at 5 µm for the following histological analyses. For hematoxylin and eosin (H&E) staining, sections first underwent hematoxylin staining for 5 minutes, were sequentially rinsed with double-distilled water, and were then subjected to eosin counterstaining for 1 minute. Digital images were acquired under a light microscope. Sections designated for Nissl staining were stained using a Nissl stain solution (Beyotime, Shanghai, China, C0117) at 37 °C for a duration of 10 minutes, briefly washed with double-distilled water, differentiated with 95% ethanol for 5 seconds, and then dehydrated and cleared as appropriate. Representative images were captured using a light microscope VS200 Panoramic Tissue Quantitative Analysis System (OLYMPUS, Tokyo, Japan, VS200). Positive Nissl staining was identified by the presence of distinct, dark purple granules (Nissl bodies) within the neuronal cytoplasm. Viable motor neurons with large, clear nuclei and rich Nissl substance were counted.

For immunofluorescence, mice were transcardially perfused with 4% paraformaldehyde, and the spinal cord tissues were harvested and processed for paraffin embedding. Prior to staining, the sections underwent deparaffinization in xylene and sequential rehydration through a graded alcohol series to aqueous conditions. Antigen retrieval was performed using citrate buffer (pH 6.0), followed by blocking with 5% bovine serum albumin (BSA) (Servicebio,Wuhan, Hubei, China, GC305010) in phosphate-buffered saline (PBS) for 1 h at room temperature to reduce nonspecific binding. To label astrocytes and neurons, sections were incubated overnight at 4 °C with rabbit anti-Glial Fibrillary Acidic Protein (GFAP) antibody (1:1000; Abcam, Cambridge Biomedical Campus, Cambridge, UK, ab68428) and rabbit anti-TUJ1 antibody (1:1000; Abcam, ab18207), respectively. Sections then received fluorescent secondary antibodies for detection: green-fluorescent Alexa Fluor 488 goat anti-rabbit Ig (Gat 1:50 from Elabscience, Wuhan, Hubei, China, E-AB-1055) and red-fluorescent Cy3 goat anti-mouse IgG (Gat 1:50 from Elabscience, E-AB-1011). Nuclei were counterstained with 4^′,6-diamidino-2-phenylindole (DAPI) (Beyotime, Shanghai, China, P0131) Fluorescent signals were visualized and images captured using a Leica DM-6B fluorescence microscope (Leica, Wetzlar, Hesse, Germany, DM-6B). For quantification of positive cells or stained areas, digital images were analyzed using ImageJ software (1.54r, NIH, Bethesda, MD, USA). Quantification data are presented as mean $\pm$ Standard Deviation (SD) (n = 6 mice per group). For each animal, three non-adjacent sections were analyzed, and from each section, three random fields within the lesion border zone were quantified and averaged.

Mice were transcardially perfused for tissue fixation, and paraffin-embedded tissue sections were subsequently prepared. Tissue sections were deparaffinized and rehydrated through a xylene-alcohol series, followed by antigen retrieval in citrate buffer. Subsequently, nonspecific binding was blocked by incubation with 5% BSA for 1 hour at room temperature. Apoptosis was then assessed by terminal deoxynucleotidyl transferase-mediated 2^′-Deoxyuridine-5^′-Triphosphate (dUTP) nick-end labeling (TUNEL) staining, which was carried out accordingly to the manufacturer’s instructions (Beyotime, C1089).

Primary fibroblasts derived from mouse spinal cord were acquired from Procell. Cells were expanded in the provided complete growth medium (Procell, Wuhan, Hubei, China, CM-M162) according to the supplier’s instructions. All cultures were kept in a humidified incubator at 37 °C with 5% CO₂. Primary fibroblasts were characterized by vimentin immunofluorescence staining, confirming a purity of $>$90%. The cells were confirmed to be free of contamination from human immunodeficiency virus type 1 (HIV-1), hepatitis B virus (HBV), hepatitis C virus (HCV), mycoplasma, as well as bacterial, fungal, and yeast pathogens. Cells were pretreated with ATD at a concentration of 100 µM for 24 h, with the compound dissolved in 0.1% Dimethyl Sulfoxide (DMSO) as vehicle control. Subsequently, fibroblasts were stimulated with TGF-$\beta$ (100 ng/mL, MedChemExpress, Monmouth Junction, NJ, USA, HY-P70648) for an additional 24 h to induce activation.

We seeded the fibroblasts into 96-well plates (5 $\times$ 10³ cells in 100 µL per well) and cultured them for 24 h at 37 °C in a humidified 5% CO₂ atmosphere. Following drug treatment, 10 µL of CCK-8 solution (Beyotime, Shanghai, China, C0037) was added to each well, and the cells were further incubated for 2 h. Following incubation, the optical density at 450 nm was recorded using a LUX multimodal microplate reader (Thermo Fisher Scientific, Waltham, MA, USA, VLBL00GD2). Cell viability was then expressed as a percentage relative to the control groups.

For Western blot analysis, proteins were extracted from cells or tissue samples using RIPA lysis buffer (Beyotime, P0013B). Following lysis, the protein concentration was quantified with a BCA assay kit (Beyotime, P0010) prior to further processing. After separation by SDS-PAGE (20 µg protein per lane) and transfer to PVDF membranes, the membranes were incubated in a blocking buffer (5% non-fat milk in Tris-Buffered Saline with Tween® 20 (TBST) (Servicebio, Wuhan, Hubei, China, G0001)) for 2 h at room temperature. This was followed by an overnight incubation at 4 °C with the respective primary antibodies. Membranes then received incubation with the appropriate HRP-conjugated secondary antibodies for 2 hours at room temperature. Detailed information for all antibodies involved in WB: FN (1:1000, Immunoway, SuZhou, JiangSu, China, YM8309), Collagen (1:1000, Immunoway, YT6135), $\alpha$-SMA (1:1000, Immunoway, YM8040), GAPDH (1:1000, Immunoway, YM8016), TGF-$\beta$ (1:1000, Immunoway, YM8257), P-SMAD2/3 (1:1000, Immunoway, YM8550), SMAD2/3 (1:1000, Immunoway, YT4332), SMAD4 (1:1000, Immunoway, YM8370), SMAD7 (1:1000, Immunoway, YN2330), Goat Anti Rabbit IgG(H+L) (HRP) (1:10,000, Immunoway, RS0002). Immunoreactive bands were detected using WesternBright ECL HRP substrate (Advansta, San Jose, CA, USA, K-12045-D20) at room temperature. To quantify protein expression, the optical density of each band was measured using ImageJ and normalized to its corresponding loading control.

At 14 days post-injury, mice were perfused, and a 5-mm segment of spinal cord tissue centered precisely on the injury epicenter was dissected under a surgical microscope. This segment, encompassing the primary lesion and peri-lesion area, was immediately snap-frozen for subsequent total RNA extraction. RNA was extracted from tissues, cells, and exosomes with TRIzol reagent (Gibco, Grand Island, NY, USA, 15596026CN). For cDNA synthesis, the Superscript III RT Reaction Mix (Invitrogen, Carlsbad, CA, USA, 18080085) was used. Quantitative PCR was performed with SYBR Green PCR Master Mix on a 7900 Fast Real-Time PCR System (7900, Applied Biosystems, Foster City, CA, USA). Relative gene expression was determined by normalizing to GAPDH and applying the 2^{-Δ⁢Δ⁢CT} calculation. The primer sequences were listed below:

Fn: 5^′–CCATTCCACCTTACAACAC–3^′, 5^′–CAAGCCAGACACAACAAT–3^′;

Col1$\alpha$1: 5^′–GTGAGACAGGCGAACAAG–3^′, 5^′–CCAGGAGAACCAGGAGAA–3^′;

GAPDH: 5^′–TGTCTCCTGCGACTTCAACA–3^′, 5^′–GGTGGTCCAGGGTTTCTTACT–3^′.

All statistical analyses were performed with SPSS (version 16.0; IBM, Chicago, IL, USA) and GraphPad Prism (version 8.0, GraphPad Software, Boston, MA, USA). Quantitative data are expressed as means $\pm$ SD. Differences were considered statistically significant based on a two-tailed Student’s t-test (for two-group comparisons) or one-way ANOVA with Tukey’s post hoc test (for multi-group comparisons), as appropriate. For non-parametric data, the Mann–Whitney U test (two-group comparison) or Kruskal-Wallis test with Dunn’s post hoc test (multiple comparisons) was used. A p-value 0.05 was defined as the threshold for statistical significance.

To evaluate the therapeutic efficacy of ATD after SCI, mice underwent surgical induction of SCI followed by intraperitoneal administration of ATD post-operation (Fig. 1A). A series of behavioral assessments were subsequently performed to evaluate functional recovery, with the experimental timeline summarized in Fig. 1B. Recovery of motor function represents a critical endpoint for assessing therapeutic outcomes. Therefore, locomotor performance was systematically evaluated using the Basso Mouse Scale (BMS) at multiple time points post-injury. Mice in the ATD-treated groups exhibited significantly improved motor function starting on day 14 after SCI, with the most pronounced recovery observed in the high-dose group (Fig. 1C,D). Consistent with these findings, swimming test results demonstrated enhanced hindlimb motor coordination and propulsion in the ATD treatment groups compared to the vehicle group, particularly in the high-dose cohort (Fig. 1E,F). Footprint analysis confirmed that ATD treatment led to improved gait symmetry and stride patterns, indicating enhanced functional recovery of both hindlimbs (Fig. 1G–I). Collectively, these data demonstrate that ATD promotes robust recovery of motor function following SCI in mice.

Fig. 1.

Atractylodin significantly improve the motor function of mice after SCI. (A) Chemical structure of ATD. (B) Schematic representation of the timeline for behavioral experiments conducted at various time points following SCI. (C) Results from the BMS test showing the recovery of hindlimb motor function after SCI (n = 12 animals per group). (D) Results from the Santing board test showing recovery of hindlimb motor function after SCI (n = 6 in each group). (E,F) Results from the swimming test showing recovery of hindlimb motor function after SCI (n = 6 in each group). (G–I) Results from footprint analysis showing the recovery of hindlimb motor function after SCI (n = 6 in each group). Data shown are the mean $\pm$ SEM, *p 0.05, **p 0.01, ***p 0.001. SCI, spinal cord injury; ATD, atractylodin; BMS, Basso Mouse Scale; SEM, Standard Error of the Mean.

Following SCI, astrocytes and fibroblasts collaborate to form a dense glial-fibrotic scar surrounding the lesion core, which acts as a major barrier to axonal regeneration. This study further investigated the impact of ATD on glial-fibrotic scar formation after SCI (Fig. 2A). Quantitative real-time PCR (qPCR) analysis revealed that ATD treatment significantly downregulated the mRNA expression levels of key fibrosis-related markers, Fn and Col1$\alpha$1, with a clear dose-dependent effect (Fig. 2B,C). Immunofluorescence staining demonstrated that at 14 dpi, ATD administration markedly reduced the protein expression of fibrotic components including FN, Collagen, and $\alpha$-smooth muscle actin ($\alpha$-SMA), along with a significant decrease in the fibroblast-specific marker PDGFR$\beta$ (Fig. 2D–K). Furthermore, fluorescence-based analysis at 28 days post-treatment showed substantially lower fibronectin expression and a reduced area of fibrotic scarring in the ATD-treated groups (Fig. 2L,M). Compared to the vehicle control group, ATD treatment significantly diminished the GFAP-positive glial scar region at 28 dpi. Notably, the structural integrity of the glial-fibrous scar interface was disrupted, with loss of tight apposition between glial and fibrotic compartments, indicating impaired scar continuity (Fig. 2L,N). Collectively, these findings demonstrate that ATD effectively narrows the spatial extent of the glial-fibrotic scar boundary and reduces the overall scar volume after SCI, thereby creating a more permissive microenvironment for axonal regrowth. Moreover, histopathological evaluation by H&E staining revealed no apparent abnormalities in vital organs, including the heart, liver, spleen, lung, and kidney, after 28 days of ATD treatment (Fig. 3A–E), suggesting a favorable safety profile under the experimental conditions used. To comprehensively evaluate the in vivo biosafety of ATD, we performed serum biochemical analysis on blood samples collected at 7 dpi. Key markers of hepatic and renal function—alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (CRE)—were quantified across all experimental groups (Fig. 3F–I). Compared to the Sham group, the SCI+Vehicle group showed no significant increase in any of these markers, indicating that the injury itself did not induce overt hepatic or renal dysfunction at this time point. Crucially, mice treated with either dose of ATD (20 or 40 mg/kg) exhibited serum levels of ALT, AST, CRE, and BUN that were statistically indistinguishable from both the Sham and SCI+Vehicle controls (Fig. 3F–I). All measured values remained within established normal physiological ranges. These data demonstrate that ATD administration at the therapeutic doses used in this study did not elicit detectable hepatotoxicity or nephrotoxicity.

Fig. 2.

Atractylodin inhibits fibrous scar formation following SCI. (A) Schematic illustration of fibrous glial scar formation in the injured spinal cord of mice following SCI modeling. (B) Quantitative gene expression analysis of Fn at the lesion center in the Vehicle, ATD20, and ATD40 groups at 14 dpi (n = 6 animals per group). (C) Quantitative gene expression analysis of Col1$\alpha$1 at the lesion center in the Vehicle, ATD20, and ATD40 groups 14 days after SCI (n = 6 animals per group). (D,E) Immunofluorescence staining and quantitative assessment of FN distribution around the injury site in the Vehicle, ATD20, and ATD40 groups at 14 dpi (n = 6 animals per group; scale bar = 100 µm). (F,G) Immunofluorescence staining and quantitative analysis of collagen deposition around the injury site in the three experimental groups at 14 dpi (G) (n = 6 animals per group; scale bar = 100 µm). (H,I) Immunofluorescence staining and quantification of $\alpha$-SMA expression around the injury region in the Vehicle, ATD20, and ATD40 groups at 14 dpi (n = 6 animals per group; scale bar = 500 µm, enlarged view scale bar = 50 µm). (J,K) Immunofluorescence staining and quantitative evaluation of PDGFR$\beta$ in peri-lesion areas of all groups at 14 dpi (n = 6 animals per group; scale bar = 500 µm, enlarged view scale bar = 50 µm). (L–N) Immunofluorescence staining and quantitative analysis of LAMININ and GFAP-positive areas around the injury site in the Vehicle, ATD20, and ATD40 groups at 28 dpi (n = 6 animals per group; scale bar = 500 µm, enlarged view scale bar = 100 µm). Data shown are the mean $\pm$ SEM, *p 0.05, **p 0.01, ***p 0.001.

Fig. 3.

Atractylodin exhibits favorable biocompatibility in vivo. (A–E) Histopathological analysis (H&E staining) of major organs was performed on day 28 after SCI. No significant pathological alterations were observed in the heart (A), lungs (B), liver (C), kidneys (D), or spleen (E) across the sham, Vehicle, ATD20, and ATD40 groups. All tissue sections displayed normal architecture and cellular morphology, with no signs of inflammation, necrosis, or tissue damage (n = 6 animals per group; scale bar = 100 µm). (F–I) Evaluation of systemic biosafety via serum biochemistry. At 7 days post-injury, key markers for hepatic and renal function were measured. Serum levels of (F) alanine aminotransferase (ALT), (G) aspartate aminotransferase (AST), (H) blood urea nitrogen (BUN), and (I) creatinine (CRE) showed no significant differences among the sham, Vehicle, ATD20, and ATD40 groups (n = 6 animals per group). ns p $>$ 0.05.

We investigated the effects of ATD on the dynamic regulation of neurons and axons following SCI. H&E staining and Nissl staining were performed on spinal cord tissue sections to assess structural integrity and neuronal morphological changes after injury. The results revealed extensive neuron loss and cavity formation at the lesion epicenter in SCI mice. The Vehicle group displayed characteristic features of severe neurodegeneration, such as neuronal pyknosis, Nissl body disintegration, and nuclear dissolution. These pathological features were substantially alleviated by ATD treatment (Fig. 4A–D). Next, TUNEL staining combined with NeuN immunofluorescence was conducted to evaluate neuronal apoptosis and survival. In the sham group, neurons were regularly aligned and showed minimal apoptotic activity. In stark contrast, the SCI group suffered substantial neuronal loss, disorganized cytoarchitecture, and extensive apoptosis. However, ATD treatment markedly increased the number of surviving NeuN-positive neurons in the peri-lesion areas (Z1–Z3) compared to the Vehicle group (Fig. 4E–H). Collectively, these results indicate that ATD potently inhibits neuronal apoptosis and promotes neuronal survival following SCI. To investigate whether ATD further contributes to axonal regeneration by attenuating fibrotic scarring, we performed immunofluorescence staining for TUJ1 and NF at 28 dpi. Quantitative analysis revealed that four weeks of ATD treatment significantly expanded the area of TUJ1-positive labeling relative to the Vehicle group (Fig. 4I,J), implying the promotion of axonal sprouting. Furthermore, the ATD group exhibited a marked increase in the number of NF-positive axons penetrating the lesion core (Fig. 4K,L), indicating successful axonal regeneration across the scar boundary.

Fig. 4.

Atractylodin inhibits neuronal apoptosis and promotes axonal regeneration following SCI. (A) Representative H&E-stained images of the lesion area at 28 days post-SCI in Sham, Vehicle, ATD20, and ATD40 groups (scale bar = 500 µm). (B) Corresponding quantitative analysis of lesion size from panel (A). (n = 6 animals per group). (C) Nissl staining was used to evaluate the morphological integrity and distribution of Nissl-positive neurons in the injured spinal cord tissues across all experimental groups at 28 dpi (scale bar = 100 µm, enlarged view scale bar = 20 µm). (D) Quantitative analysis of the number of Nissl-positive neurons from panel (C) (n = 6 animals per group). (E) Representative images of TUNEL assay detecting apoptotic cells in the lesion area (scale bar = 500 µm, enlarged view scale bar = 100 µm). (F) Quantification of TUNEL-positive cells from panel (E) (n = 6 animals per group). (G) Immunofluorescence labeling of NeuN⁺ neurons was performed to assess neuronal survival (scale bar = 500 µm). (H) Quantification of NeuN⁺ neuron counts from panel (G) (n = 6 animals per group). (I) Representative double immunofluorescence images of TUJ1+ axons (green) and GFAP+ astrocytes (red) (scale bar = 500 µm, enlarged view scale bar = 20 µm). (J) Quantification of the relative positive areas of TUJ1 and GFAP from panel (I). (K) Immunofluorescence images showing NF+ neurofilaments (regenerating axons) in the injured spinal cord (scale bar = 500 µm). (L) Quantified fluorescence intensity of NF+ signals from panel (K). (n = 6 animals per group). Data shown are the mean $\pm$ SEM, *p 0.05, **p 0.01, ***p 0.001, ns p $>$ 0.05.

To investigate the anti-fibrotic effects of ATD on spinal cord fibroblasts, we first confirmed its biocompatibility. The CCK-8 assay revealed no significant cytotoxicity at a concentration of 100 µM (Fig. 5A). We then modeled fibrotic activation using TGF-$\beta$ stimulation. While TGF-$\beta$ significantly upregulated the mRNA expression of the key fibrosis-related genes Fn and Col1$\alpha$1, ATD treatment effectively suppressed this induction (Fig. 5B,C). At the protein level, Western blot analysis confirmed that TGF-$\beta$ markedly increased the expression of FN, Col1$\alpha$1, and $\alpha$-SMA, all of which were suppressed by ATD in a concentration-dependent manner (Fig. 5D–G). For all original Western blot figures of Fig. 5D, see Supplementary Material. Consistent with these findings, immunofluorescence staining demonstrated that TGF-$\beta$ induced robust fibroblast activation, characterized by increased $\alpha$-SMA expression and cytoskeletal reorganization, which are hallmark features of myofibroblast differentiation. This pro-fibrotic response was substantially attenuated by ATD treatment, as evidenced by reduced $\alpha$-SMA signal intensity (Fig. 5H,I). Taken together, these findings demonstrate that ATD effectively inhibits TGF-$\beta$-induced fibrotic activation of spinal cord fibroblasts.

Fig. 5.

ATD inhibits TGF-$\beta$-induced fibrotic activation in spinal cord fibroblasts. (A) Cytotoxicity of various concentrations of ATD was assessed using the CCK-8 assay (n = 3). (B,C) Quantitative real-time PCR (qPCR) analysis revealed that ATD treatment significantly suppressed TGF-$\beta$-induced mRNA expression of fibronectin (Fn) (B) and collagen type I alpha 1 (Col1$\alpha$1) (C) in mouse spinal cord fibroblasts (n = 3). (D–G) Western blot analysis showed reduced protein levels of FN (E), collagen (F), and $\alpha$-SMA (G) in ATD-treated cells compared to Vehicle controls, with quantification shown in panels E, F, and G respectively (n = 3). (H,I) Immunofluorescence staining confirmed downregulation of $\alpha$-SMA expression in ATD-treated fibroblasts, with representative images shown in (H) and quantitative analysis in (I) (n = 3; scale bar = 50 µm). Data shown are the mean $\pm$ SEM, *p 0.05, **p 0.01, ***p 0.001, ns p $>$ 0.05.

Following SCI, TGF-$\beta$ promotes the formation of fibrosis through activation of both canonical SMAD-dependent and non-canonical signaling pathways. In this study, TGF-$\beta$ stimulation was found to significantly increase the phosphorylation of SMAD2/3 in fibroblasts. ATD treatment markedly suppressed TGF-$\beta$-induced SMAD2/3 phosphorylation in a concentration-dependent manner (Fig. 6A–C). Within the TGF-$\beta$/SMAD signaling cascade, SMAD4 acts as a central mediator and SMAD7 functions as an inhibitory regulator of SMAD2/3 activity. Therefore, we examined the expression levels of SMAD4 and SMAD7 following ATD treatment. ATD significantly reduced SMAD4 protein expression, while having no significant effect on SMAD7 protein (Fig. 6D–G). For all original Western blot figures of Fig. 6A,D,F, see Supplementary Material. These results indicate that ATD attenuates the fibrotic activation of fibroblasts primarily by suppressing the canonical TGF-$\beta$/SMAD pathway.

Fig. 6.

ATD inhibits the TGF-$\beta$/SMAD signaling pathway in fibroblasts. (A) Representative western blot images of protein expression in TGF-$\beta$-induced mouse spinal cord fibroblasts. (B) Quantitative analysis of TGF-$\beta$ expression levels. (C) Quantitative analysis of phosphorylated SMAD2/3 (P-SMAD2/3) and total SMAD2/3 levels. Data are from three independent experiments (n = 3). (D,E) The expression of SMAD4 was similarly evaluated by Western blot and quantified (n = 3). (F,G) SMAD7 protein expression was also examined by Western blot and subjected to quantitative analysis (n = 3). Data shown are the mean $\pm$ SEM, ns p $>$ 0.05, **p 0.01, ***p 0.001.