Bleomycin sulfate was purchased from Selleck (Shanghai, China). Primary antibodies against GAPDH, fibronectin, collagen I, α-SMA, E-cadherin, N-cadherin, vimentin, Smad-2, Phospho-Smad2 (p-Smad2), Smad-3, and Phospho-Smad3 (p-Smad3) were sourced from ABclonal (Wuhan, China). The BCA protein assay kit was obtained from Beyotime Bio (Shanghai, China). High-Sig ECL western blotting substrate was purchased from Tanon (Shanghai, China). The MTT Cell Proliferation and Cytotoxicity Assay Kit was acquired from Solarbio (Beijing, China). ELISA kits for IL-6, IL-1β, and TNF-α were obtained from Boster (Wuhan, China). The Superoxide Dismutase (SOD) Activity Assay Kit, Malondialdehyde (MDA) Content Assay Kit, and Reduced Glutathione (GSH) Content Assay Kit were provided by Nanjing Jiancheng Bioengineering Institute (Nanjing, China). Ultra-micronized PEA was obtained from Epitech Group SpA (Saccolongo, Padua, Italy). Recombinant Human TGFβ1 protein was purchased from R&D Systems (Minnesota, United States). The TGF-β/Smad pathway activator, SRI-011381, was sourced from MedChemExpress (New Jersey, United States). All cell culture reagents were obtained from Gibco (Grand Island, NY, United States) and Solarbio (Beijing, China). Trizol reagent was sourced from Takara (Dalian, China). The PrimeScript RT Reagent Kit and the SYBR Premix Ex Taq Kit were purchased from Yeasen (Shanghai, China).

The raw sauerkraut juice was serially diluted in sterile physiological saline and plated onto de Man, Rogosa, and Sharpe (MRS) agar using the streak plate method. Following incubation until distinct white colonies appeared, individual colonies were randomly selected and subjected to two successive rounds of purification. The bacterial isolates were taxonomically classified through 16S ribosomal DNA (rDNA) sequencing with subsequent EzBioCloud database alignment (Yoon et al., 2017; Chalita et al., 2024), identifying three distinct strains: Lactiplantibacillus sp. LP03 (LP03), Levilactobacillus brevis LB06 (LB06), and Loigolactobacillus coryniformis LC02 (LC02). Following confirmation of LP03’s anti-fibrotic efficacy against BLM-induced pulmonary fibrosis in mice, the strain was deposited in the China Center for Type Culture Collection (CCTCC) with the official designation Lactiplantibacillus sp. ZYHL3 (CCTCC No. M 2025127).

To evaluate gastrointestinal survival capacity—a critical characteristic of probiotics—we assessed acid tolerance (pH 2.0–5.0 for 4 h, simulating typical gastric retention time) and bile salt resistance (0.03%–0.3% bile salt for 4 h, representing small intestinal transit duration). These experiments followed optimized protocols, incorporating modifications to pH ranges, bile salt concentrations, and incubation times to more closely simulate physiological conditions (Zhao et al., 2023, 2025; Xiong et al., 2024). All three Lactobacillus strains maintained > 50% viability under stringent conditions (pH 2.0 and 0.3% bile salt) in triplicate experiments (see Supplementary Figure 1).

All animal procedures were conducted in strict accordance with the ethical standards approved by the Animal Ethics Committee of the First Hospital of Jilin University (Approval No. JDYY20240346). Experimental protocols were authorized by the relevant licensing authority and complied fully with institutional regulations and the ARRIVE guidelines.

A total of 8 weeks-old male C57BL/6J mice (20–22 g) were purchased from Jilin Qianhe Technology Industrial Co., Ltd. (Changchun, Jilin, China) and housed under specific pathogen-free conditions at the animal research facility of the First Hospital of Jilin University (Changchun, Jilin, China). The mice were acclimatized for 1 week before beginning the experiments and were given ad libitum access to food and water (Zhu et al., 2021a,b). Following acclimatization, mice were randomly allocated into five groups (n = 13 per group): (1) Control (Con) group: normal saline treatment; (2) Model group (BLM group): BLM-induced pulmonary fibrosis; (3) LP03 group: BLM + LP03; (4) LB06 group: BLM + LB06; and (5) LC02 group: BLM + LC02. To establish a pulmonary fibrosis model, mice were administered 5 mg/kg of BLM via endotracheal injection under anesthesia with tribromoethanol, following established protocols (Peng et al., 2020; Ding et al., 2022; Li G. et al., 2023; Zeng et al., 2024). Mice in the Con group received an intratracheal instillation of an equal volume of saline solution. The dosage and frequency of our Lactobacillus intervention were adapted from existing studies on probiotics in disease treatment and appropriately modified based on the specific context of our research (Duan et al., 2023; Shen et al., 2024; Vasconcelos et al., 2025). Seven days prior to BLM administration, 10⁹ CFU of LP03, LB06, or LC02, each suspended in 200 μL of saline, were administered by oral gavage once daily, 5 days a week, continuing until day 21 post-BLM administration. Mice in the Con and BLM groups were gavaged with an equal volume of saline at the same time. On day 21, the mice were sacrificed, and lung tissues, bronchoalveolar lavage fluid (BALF), blood, and colonic feces were collected for further analysis.

To further investigate the effects of the differential metabolite PEA on BLM-induced pulmonary fibrosis, mice were assigned to five experimental groups (n = 15 per group): (1) Con group; (2) BLM group; (3) PEA (3 mg/kg) group: BLM + PEA (3 mg/kg); (4) PEA (10 mg/kg) group: BLM + PEA (10 mg/kg); and (5) PEA (30 mg/kg) group: BLM + PEA (30 mg/kg). To enhance oral bioavailability, ultra-micronized PEA at various concentrations was dissolved in 1.5% carboxymethyl cellulose in accordance with previous studies (Annunziata et al., 2020; Lama et al., 2021). PEA was administered once daily via oral gavage from day 1 to day 20. On day 21, mice were humanely euthanized via intraperitoneal injection of sodium pentobarbital (50 mg/kg), and samples were subsequently collected for further analysis.

A549 cells (human type II alveolar epithelial cell line) and MRC5 cells (human embryonic lung fibroblast line) were obtained from the American Type Culture Collection (ATCC, Manassas, VA, United States). Both cell lines were cultured in DMEM medium containing 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin, and maintained at 37 °C in a 5% CO₂ atmosphere. Upon reaching approximately 80% confluency, cells were digested using 0.25% trypsin/EDTA and seeded into 6-well plates (Song et al., 2016, 2017). The following day, cells were serum-starved in DMEM containing 1% FBS for 12 h and then treated with different compounds (TGFβ1, PEA, SRI-011381).

The left lungs of mice were fixed in 4% paraformaldehyde for 24 h, embedded in paraffin, and sectioned into 4 μm thick slices. For H&E staining, sections were deparaffinized using xylene, rehydrated through graded ethanol solutions, and stained with hematoxylin to stain nuclei blue or purple, followed by eosin to stain cytoplasm, collagen, and other structures pink or red. After staining, the slides were dehydrated, cleared, and mounted. For Masson’s Trichrome staining, after deparaffinization and rehydration, the sections were stained with Weigert’s iron hematoxylin to stain nuclei black, followed by Biebrich scarlet acid fuchsin to stain cytoplasm and muscle fibers red. The sections were then differentiated with phosphomolybdic acid, followed by staining with aniline blue to visualize collagen fibers as blue or green. After final dehydration and clearing, the slides were mounted and photographed using light microscopy. All imaging was performed by an assessor blinded to the treatment conditions.

The enzymatic activities of superoxide dismutase (SOD) and the levels of malondialdehyde (MDA) and reduced glutathione (GSH) in lung tissue homogenates were measured according to the manufacturer’s protocols.

Cell viability was assessed using the MTT Cell Proliferation and Cytotoxicity Assay Kit. A549 and MRC5 cells were seeded into 96-well plates (5 × 10³ cells per well) and incubated for 12 h. Cells were treated with increasing concentrations of PEA (0, 0.5, 1, 2, 4, 8, 16, 32, and 64 μM) for 24 h. Following treatment, the medium was removed, and 90 μL of fresh medium was added, followed by 10 μL of MTT solution. After 4 h of incubation at 37 °C, the supernatant was removed, and 110 μL of Formazan solubilization solution was added to each well. The absorbance was measured at 490 nm using a microplate reader.

Proteins were extracted from tissue or cell lysates using RIPA buffer containing phosphatase and protease inhibitors. After centrifugation, protein concentrations were quantified using the BCA assay. Equal amounts of protein (25 μg) were separated by SDS-PAGE and transferred to PVDF membranes (Song et al., 2021; Luo et al., 2023). After blocking, the membranes were incubated overnight at 4 °C with primary antibodies, followed by incubation with HRP-conjugated secondary antibodies. Protein levels were visualized using enhanced chemiluminescence reagents, and densitometric analysis was performed using Image J software.

Bronchoalveolar lavage was performed with cold saline. The concentrations of IL-6, IL-1β, and TNF-α in the BALF were measured using ELISA kits according to the manufacturer’s protocols.

Lung tissues were initially fixed in 4% paraformaldehyde, followed by a series of dehydration steps prior to paraffin embedding. Sections (4 μm) were deparaffinized, rehydrated, and subjected to antigen retrieval using citrate buffer at 95 °C for 10 min. After antigen retrieval, the sections were incubated in endogenous peroxidase blocking solution at 37 °C for 30 min to inhibit non-specific binding. Subsequently, the sections were immunostained with primary antibodies against vimentin, incubating overnight at 4 °C. Following primary antibody incubation, the sections were treated with biotin-conjugated secondary antibodies at 37 °C for 30 min. Immunoreactivity was visualized using a DAB working solution, and the sections were counterstained with hematoxylin before mounting with a permanent mounting medium. Images of the stained sections were captured using light microscopy, with the imaging process conducted by an assessor blinded to the experimental groups.

For the transcriptional analysis of lung tissues or cultured cells, total RNA was extracted using TRIzol reagent, following the manufacturer’s protocol. The RNA was then reverse transcribed into complementary DNA (cDNA) using the PrimeScript RT Reagent Kit. Gene expression levels were quantified via qPCR using the SYBR Premix Ex Taq Kit, with cDNA as the template. GAPDH was employed as an internal control. Relative mRNA expression levels were calculated using the 2^ΔΔCt method. The primer sequences used in this study are provided in (see Supplementary Table 1).

Immediately after collection, all fecal samples were frozen at −20 °C and transported on ice to the laboratory. Bacterial DNA was extracted at Novogene Bioinformatics Technology Co., Ltd. using Tiangen kits, following the manufacturer’s protocol. To profile the mouse gut microbiome, the V4 region of the 16S rDNA was amplified and sequenced using universal primers targeting this region, which is conserved across most bacterial species (F: 5′-GTGCCAGCMGCCGCGGTAA-3′; R: 5′-GGACTACHVGGGTWTCTAAT-3′; a unique 6 bp barcode for each sample was incorporated in the reverse primer) (Sonnenburg et al., 2016). Single amplification reactions were performed in 25 μL volumes with 50 ng of template DNA. The PCR conditions were as follows: an initial denaturation at 94 °C for 4 min, followed by 30 cycles of 94 °C for 30 s, 54 °C for 30 s, and 72 °C for 30 s, with a final extension at 72 °C for 5 min. The PCR products were pooled in equimolar concentrations and sequenced on the Illumina HiSeq2500 platform. Alpha- and beta-diversities of the microbiota were analyzed using QIIME 1.9.1 software. The Shannon diversity index and the Simpson diversity index were used as measures of α-diversity, while β-diversity was assessed using weighted UniFrac distances and visualized through principal coordinate analysis (PCoA). Linear discriminant analysis (LDA) effect size (LEfSe) was performed to identify taxa with significant differential abundance among experimental groups. Briefly, a non-parametric Kruskal–Wallis test was used to detect significantly different features, followed by LDA to estimate the effect size of each feature. Only taxa with an LDA score (log10) > 2 and p < 0.05 were considered significant. All samples were paired-end sequenced on the Illumina HiSeq X Ten platform (insert size 350 bp, read length 150 bp) at Novogene Bioinformatics Technology Co., Ltd.

Untargeted metabolite analysis in serum samples was performed using liquid chromatography-mass spectrometry (LC-MS). Chromatographic separation was conducted on a Thermo Scientific UltiMate 3000 HPLC system equipped with an ACQUITY UPLC BEH C18 column (100 mm × 2.1 mm, 1.8 μm, Waters, United Kingdom) for reversed-phase separation. The column was maintained at 35 °C, and the mobile phase was composed of solvent A (water with 0.1% formic acid) and solvent B (acetonitrile with 0.1% formic acid) at a flow rate of 0.4 mL/min. The gradient elution program was as follows: 0–0.5 min, 5% B; 0.5–7 min, 5% to 100% B; 7–8 min, 100% B; 8–8.1 min, 100% to 5% B; 8.1–10 min, 5% B. A 4 μL injection volume was used for each sample. Metabolites were detected using a high-resolution tandem mass spectrometer (Q-Exactive, Thermo Scientific) operating in both positive and negative ion modes. Precursor ion spectra (m/z 70–1,050) were acquired at a resolution of 70,000, with a target automatic gain control (AGC) of 3e6 and a maximum injection time of 100 ms. Data acquisition was conducted in top 3 mode in data-dependent acquisition (DDA), and fragment ion spectra were collected at a resolution of 17,500, with a target AGC of 1e5 and a maximum injection time of 80 ms.

The wound healing assay was performed as described before (Song et al., 2022). Briefly, A549 and MRC5 cells were seeded in 6-well culture plates and allowed to grow to confluence. Once confluent, a wound was created by scratching the center of the well using a 200 μL pipette tip. The cells were then washed three times with PBS to remove detached cells and debris. The cells were subsequently cultured in DMEM containing 1% FBS, with or without the addition of 10 μM PEA or 10 ng/mL SRI-011381, for an additional 24 h. Microscopic imaging was performed to capture the wound area before and after treatment. The wound healing rate was quantified using ImageJ software (National Institutes of Health, Bethesda, MD, United States) by calculating the difference in scratch area at 0 and 24 h. Wound healing rate (%) = 100 × (scratch area at 0−scratch area at 24 h)/scratch area at 0 h. This experiment was performed in triplicate.

Data were analyzed using GraphPad Prism 8.0 and expressed as mean ± standard deviation (SD). One-way ANOVA followed by Tukey’s multiple comparison test was used for statistical comparisons. Kaplan-Meier survival analysis was employed to assess survival data. A p-value of < 0.05 was considered statistically significant.

Initially, we explored the effects of these three Lactobacillus strains, representing different genera, on BLM-induced pulmonary fibrosis in mice. Figure 1A illustrates the schematic diagram of the in vivo study design, investigating the effects of various Lactobacillus strains on the BLM-induced pulmonary fibrosis mouse model. Figure 1B shows that LP03 significantly reduced the mortality rate in BLM-treated mice, whereas LB06 and LC02 had negligible effects on survival, highlighting LP03’s distinctive and potential role in improving the prognosis of pulmonary fibrosis. The mice’s body weight was recorded on the first day of the experiment and on the day of euthanasia, showing that all three genera alleviated BLM-induced weight loss, with the LP03 group demonstrating the most significant improvement (Figure 1C). The lung coefficient, defined as the ratio of lung weight to body weight, is a key parameter in animal studies for assessing the extent of pulmonary edema or fibrosis (Zuo et al., 2022; Pan et al., 2024). Compared to the control group, the lung coefficient in the BLM group increased significantly, but subsequently decreased markedly following LP03 administration. In contrast, treatment with LB06 and LC02 had no effect on the lung coefficient (Figure 1D). Importantly, H&E and Masson’s staining of lung tissue revealed that LP03 administration, but not LB06 or LC02, alleviated inflammatory cell infiltration, reduced alveolar wall thickness, and decreased collagen deposition in the lungs of BLM-treated mice (Figures 1E, F). The Szapiel score (Figure 1G) and Ashcroft score (Figure 1H) were used to quantitatively assess the degree of alveolitis and fibrosis, respectively, across the different groups. These findings suggest that LP03 intervention can mitigate the severity of BLM-induced pulmonary fibrosis.

The relative mRNA expression levels of fibronectin, collagen I, and α-SMA—key biomarkers of fibrosis—were assessed by RT-qPCR. The results showed that BLM-induced toxicity significantly upregulated the transcription of all three genes, an effect that was notably reversed by LP03 administration, but not by LB06 or LC02 treatments (Figures 1I–K). Furthermore, western blot analysis revealed that only LP03 treatment significantly attenuated the BLM-induced upregulation of fibronectin, collagen I, and α-SMA at the protein level among the groups (Figure 1L–O). These findings were consistent with the mRNA expression results, further corroborating LP03’s notable anti-fibrotic effect in mice. Collectively, these results suggest that administration of LP03 can mitigate BLM-induced pulmonary fibrosis in vivo.

Accumulating evidence has established EMT as a pivotal mechanism driving pulmonary fibrosis pathogenesis (Kim et al., 2006; Ding et al., 2023; Jin et al., 2024; Ni et al., 2024; Zhou S. et al., 2024). In the present study, we demonstrated that BLM significantly upregulated the protein expression of mesenchymal markers, such as vimentin and N-cadherin, in lung tissue, an effect that was notably reversed by LP03 treatment (Figures 2A, B, D). Conversely, LP03 treatment reversed the BLM-induced reduction in the epithelial marker E-cadherin (Figures 2A, C). A similar trend was observed at the mRNA level, further supporting the Western blot findings (Figures 2E–G). Additionally, the expression of vimentin was further assessed by IHC. As shown in Figure 2H, vimentin levels were elevated in the BLM group, a phenomenon that was reversed following LP03 treatment. Collectively, these results confirm that LP03 effectively mitigates EMT induced by BLM in vivo.

Furthermore, we assessed oxidative stress levels by measuring the activity of SOD and the concentrations of MDA and GSH in lung tissue, given the pivotal role of oxidative stress in the pathogenesis of pulmonary fibrosis (Assayag et al., 2021; Liu et al., 2023). SOD is an essential antioxidant enzyme that catalyzes the conversion of superoxide radicals (O2) into hydrogen peroxide (H2O2), thereby protecting cells from oxidative damage. MDA is a product of lipid peroxidation and its levels correlate with cellular oxidative damage. GSH is a tripeptide that acts as a critical intracellular antioxidant, maintaining cellular redox homeostasis. In the present study, the BLM challenge resulted in a significant increase in MDA concentration, whereas LP03 treatment partially restored MDA levels. Conversely, the BLM challenge induced a reduction in pulmonary GSH levels and SOD activity, both of which were notably elevated following LP03 treatment (Figures 2I–K). These findings provide compelling evidence that LP03 exerts a protective effect against oxidative stress in the context of pulmonary fibrosis.

To assess the inflammatory response in BLM-induced pulmonary fibrosis, the expression of inflammatory cytokines in the BALF was measured (Figures 2L–N). Following BLM administration, levels of the inflammatory cytokines TNF-α, IL-6, and IL-1β were significantly elevated. However, after treatment with LP03, the concentrations of TNF-α, IL-6, and IL-1β were notably reduced, highlighting the protective effect of LP03 against the inflammatory response in BLM-induced pulmonary fibrosis.

The TGF-β/Smad signaling pathway plays a critical role in promoting EMT and is a central pro-fibrotic axis responsible for the aberrant expression and deposition of extracellular matrix (ECM) components in pulmonary fibrosis (Ding et al., 2023; Li G. et al., 2023; Du et al., 2024). To further elucidate the mechanisms underlying the anti-fibrotic effects of LP03, we examined its impact on the TGF-β1/Smad2/3 signaling pathway in a BLM-induced pulmonary fibrosis mouse model. As anticipated, western blot analysis showed significantly elevated levels of TGF-β1 and increased phosphorylation of Smad2 and Smad3 in the BLM group compared to the control group. Notably, treatment with LP03 markedly suppressed the expression of TGF-β1 and the phosphorylation of Smad2/3, indicating effective inhibition of the TGF-β1/Smad signaling cascade (Figure 3).

To elucidate the underlying mechanisms by which LP03 mitigates BLM-induced pulmonary fibrosis, we investigated the gut microbiota composition across three experimental groups through 16S rDNA sequencing. Microbial α-diversity, evaluated using the Shannon and Simpson indices to assess community richness and evenness, showed no statistically significant differences across study groups (p > 0.05). Nevertheless, LP03 administration was found to partially restore the gut microbiota diversity to levels comparable to the Con group in BLM-treated mice (Figures 4A, B). Additionally, β-diversity was assessed using Principal Coordinates Analysis (PCoA) based on weighted UniFrac distances, which revealed a distinct separation between the Con and BLM groups. The microbiota of the LP03 group, however, exhibited an intermediate profile, indicating that LP03 mitigates the dysbiosis induced by BLM (Figure 4C). To explore specific alterations in microbial community structure among the groups, we analyzed and presented the relative abundance of the top ten phyla (Figure 4D). The analysis revealed that BLM treatment caused an imbalance in the phylum-level composition of the gut microbiota, most notably a decrease in the proportion of Firmicutes and an increase in Proteobacteria. However, LP03 treatment largely restored the structural composition of the gut microbiota. At the genus level, a heatmap depicting the relative abundance of the top thirty genera across the groups was generated (Figure 4E). This analysis revealed that BLM treatment reduced the abundance of beneficial genera such as Ligilactobacillus and Akkermansia, while increasing the abundance of harmful genera like Listeria and Acinetobacter. However, LP03 treatment effectively reversed these trends. Furthermore, LEfSe analysis identified distinct gut microbiota species with significantly increased abundance in each group, as shown in Figure 4F.

To investigate the mechanism by which LP03 alleviates pulmonary fibrosis in mice, we employed LC-MS to analyze the serum of the mice. Partial Least Squares Discriminant Analysis (PLS-DA) revealed significant differences in metabolite levels among the three experimental groups, incorporating data from both positive and negative ion modes (Figures 5A–C). To further elucidate the effects of LP03 on the modulation of serum metabolites in BLM-treated mice, we performed comparative analyses between the BLM and Con groups, as well as between the LP03 and BLM groups. The comparative analyses were based on thresholds for fold change (FC) > 1.5 and p-values < 0.05. As depicted in the volcano plot, 187 metabolites were found to be upregulated and 93 metabolites downregulated in the BLM group relative to the Con group (Figure 5D). Moreover, in the LP03 group compared to the BLM group, 100 metabolites were increased, while 170 metabolites were decreased (Figure 5E).

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis demonstrated that BLM administration induced significant alterations in metabolic pathways, particularly disrupting amino acid and lipid metabolism (Figure 5F). Additionally, LP03 treatment mainly modified lipid metabolism, with marked alterations in glycerophospholipid and fatty acid metabolic pathways (Figure 5G). We hypothesized that the therapeutic effects of LP03 in BLM-induced pulmonary fibrosis are mediated through serum metabolite alterations. To identify metabolites with potential antifibrotic activity, we performed hierarchical clustering analysis of the top 20 upregulated metabolites in LP03-treated mice compared to BLM controls (Figure 5H). Among these candidates, palmitoylethanolamide (PEA) emerged as a promising mediator, as this lipid metabolite has been previously established to exert antifibrotic effects in pulmonary, retinal, and hepatic fibrosis (Di Paola et al., 2016; Ohara et al., 2018; Ye et al., 2020). These results suggest that LP03 modulates host lipid metabolism and elevates systemic levels of PEA in the BLM-induced pulmonary fibrosis mouse model. Based on these findings, we propose that PEA mediates, at least in part, the attenuation of BLM-induced pulmonary fibrosis. Subsequent investigations into LP03’s antifibrotic mechanisms have consequently focused on PEA’s therapeutic role.

Next, we investigated whether PEA mitigates BLM-induced pulmonary fibrosis. Figure 6A is a schematic diagram of the experimental design. We found that oral administration of PEA effectively improved the survival rate of mice in the pulmonary fibrosis model (Figure 6B), mitigated weight loss induced by BLM (Figure 6C), and reduced the lung coefficient (Figure 6D), exhibiting a clear dose-dependent effect. Moreover, histopathological analysis of lung tissue, using H&E and Masson’s trichrome staining, confirmed that PEA markedly alleviated pulmonary inflammatory cell infiltration and collagen fiber deposition in the mice (Figures 6E–H). Additionally, western blot analysis revealed that PEA significantly reduced the expression levels of fibronectin, collagen I, and α-SMA in a dose-dependent manner in the pulmonary fibrosis mouse model (Figures 6I–L). To further investigate whether PEA alleviates pulmonary fibrosis in mice by modulating the TGF-β1/Smad2/3 pathway and EMT, we performed western blot analysis to assess the expression levels of EMT-related markers, TGF-β1, and the phosphorylation status of Smad2/3. The results revealed that PEA significantly reduced the ratios of p-Smad2/Smad2 and p-Smad3/Smad3, as well as the expression level of TGF-β1, in a dose-dependent manner (Figures 6M–P). These findings indicate that PEA effectively suppresses the TGF-β1/Smad2/3 signaling pathway activated in BLM-induced pulmonary fibrosis. Furthermore, PEA significantly reversed the elevated expression levels of vimentin and N-cadherin, and the decreased expression of E-cadherin observed in the BLM group, suggesting that PEA efficiently inhibits BLM-induced EMT in vivo (Figures 6M, Q–S).

Subsequently, we investigated whether PEA attenuates fibrosis-associated gene expression and EMT in an in vitro model of pulmonary fibrosis, utilizing the human bronchial epithelial cell line A549 and the human lung fibroblast cell line MRC5. To assess the in vitro toxicity of PEA, A549 and MRC5 cell lines were exposed to a range of PEA concentrations. As shown in Figures 7A, B, PEA demonstrated no significant cytotoxicity to the A549 cell line at concentrations up to 32 μM, and no significant cytotoxicity to the MRC5 cell line at concentrations up to 16 μM. Based on the concentrations of TGF-β1 used in cell models of pulmonary fibrosis and PEA in in vitro experiments as reported in the literature, we selected a concentration of 5 ng/mL for TGF-β1 to stimulate the cells and 10 μM for PEA in the subsequent experiments (Ohara et al., 2018; Wang et al., 2020; Schiano Moriello et al., 2022; Niu et al., 2023; Ma et al., 2024). We then investigated the effect of PEA on the expression of fibrosis-related genes in the A549 and MRC5 cell lines under TGF-β1 stimulation. Our results demonstrated that PEA significantly reduced the expression levels of fibronectin, collagen I, and α-SMA in A549 (Figures 7C–F) and MRC5 (Figures 7C, G–I) cell lines, consistent with previous in vivo findings, thereby further confirming the potential antifibrotic effects of PEA.

To investigate whether PEA modulates the expression of fibrosis-related genes by inhibiting TGF-β1/Smad2/3 signaling-mediated EMT, we supplemented the TGF-β/Smad signaling pathway agonist SRI-011381 in TGF-β1- and PEA-treated cell lines. Western blot analysis demonstrated that PEA attenuated the phosphorylation of Smad2 and Smad3 induced by TGF-β1 in both A549 and MRC5 cell lines, while SRI-011381 reversed this effect of PEA (Figures 8A–E). Furthermore, in the A549 cell line, PEA reduced the expression of N-cadherin induced by TGF-β1 and increased the expression of E-cadherin, which was suppressed by TGF-β1. Notably, SRI-011381 reversed the effect of PEA on the expression of these EMT-related markers (Figures 8A, F, G). Additionally, PEA significantly reduced the mobility of TGF-β1-stimulated A549 and MRC5 cells, as shown by the wound healing assay, which was reversed by SRI-011381 (Figures 8H–K). These findings support the conclusion that PEA inhibits the acquisition of a mesenchymal phenotype in TGF-β1-stimulated alveolar epithelial cells and pulmonary fibroblasts by suppressing Smad2/3 phosphorylation. In summary, PEA inhibits TGF-β1-induced phosphorylation of Smad2/3, thereby suppressing EMT and the expression of fibrosis-related genes in vitro.