A549 pulmonary epithelial cells were purchased from the Bioresource Collection and Research Center (#60074, BCRC, Hsinchu, Taiwan). The complete medium for cells was Ham's F-12k medium (ACE Biolabs, Taoyuan, Taiwan), containing 7 mm glucose and 10% fetal bovine serum (FBS; Gibco, Waltham, Massachusetts, USA). Cells were maintained at 37 °C with 5% CO₂ and humidification.

To generate obese microenvironments in vitro, A549 cells were treated with palmitic acid (PA; Sigma, St. Louis, USA) in accordance with obesity-related conditions (18). PA was prepared in absolute ethanol and then conjugated with 1% bovine serum albumin (BSA; Sigma) in Ham's F-12k media in a 37 °C incubator for 2 h. After sterilization using a filter, the PA stock was stored at −20 °C. The working concentration of PA was 500 μm. Additionally, lipopolysaccharide (LPS; Sigma) was used as another model of obesity, as a higher level of endotoxin has been reported in obesity (20), and it generally stimulates acute inflammation (19). The working concentration of LPS was 10 μg/ml. Moreover, TGF-β (PeproTech, Thermo Fisher Scientific, Cranbury, New Jersey, USA) was used to induce lung fibrosis in vitro, with a working concentration of 10 μg/ml. LPS and TGF-β were prepared and stored at −20 °C for future use. The lunasin peptide was chemical synthesized with a purity more than 95% (Synpeptide Co., Ltd., Shanghai, China). The working concentrations of lunasin were 5 and 50 μm. Cells were cultured with obesity-related conditions, with or without lunasin for 24 h.

A549 cells were treated with serial doses of lunasin, PA, or LPS for 48 h. After removing the media, the cells were incubated in 0.5 mg/ml 3-(4, 5-Dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT, Sigma) in DMEM at 37 °C for 3 h. After aspirating, DMSO was added, and formazan crystals were solubilized on the shaker. The absorbance was determined at 540 nm using a spectrophotometer (Molecular Devices, CA, USA). Cell viability was calculated using the formula and is presented as percentages.

A549 cells were cultured in the presence or absence of 500 μm PA or 10 μg/ml LPS and treated with lunasin for 48 h. Then, supernatants were collected for cytokine analysis. IL-6, MCP-1, and TGF-β secretions were analyzed via ELISA, in accordance with the manufacturer's protocol (BioLegend, San Diego, CA, USA) and a previous study (19). Briefly, the plates were coated with capture antibodies overnight, then blocked and subsequently incubated with standards or samples. After washing, the plates were processed with detection antibodies, horseradish peroxidase-conjugated streptavidin, and the substrate. Absorbance was measured using a spectrophotometer (Molecular Devices, CA, USA), and the concentration was calculated based on the standard curve.

Cells were seeded at 1.1 × 10⁵ cells/well in 24-well plates overnight until cell confluence reached 80%. Cells were pre-treated with lunasin for 4 h, and cellular monolayers were scratched using a 1,000 μl pipette tip. Then, the cultured media were replaced with either 0.5% FBS/F-12 medium, 5 ng/ml TGF-β, or 100 ng/ml recombinant mouse leptin (R&D, Minneapolis, MN, USA), with or without lunasin, for 48 h. Imaging was performed 0, 24, and 48 h post-scratch using a microscope at 100 × magnification (Leica). Migrated areas were quantified using Image J, and the results are presented as the Migration index. Additionally, the healing area under the curve (AUC) was calculated according to biopharmaceutical references.

Cells were pre-treated with lunasin for 4 h and then cultured with 500 μm PA or 10 μg/ml LPS with or without lunasin for 24 h. Cells were then collected and extracted using radio immunoprecipitation assay (RIPA) buffer (Bioman, Taiwan) containing protease inhibitors (Targetmol, Boston, USA) and phosphatase inhibitors (Targetmol). Cell lysates were adjusted for protein loading according to a BCA protein assay (T-PRO, Taiwan) and denatured in a 100 °C water bath for 10 min. Then, samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) and electrotransferred to a PVDF membrane. Briefly, membranes were blocked in gelatin-NET buffer and then incubated with primary antibodies (Ab), containing anti-surfactant protein (SP)-D (abcam), anti-phospho-nuclear factor kappa B (NF-κB) p65 (phosphor Ser536; Santa Cruz, California, USA), anti-NF-κB p65 (Santa Cruz), and anti-GAPDH (abcam, Cambridge, UK), as well as secondary peroxidase AffiniPure goat anti-mouse or anti-rabbit antibody (Jackson, Pennsylvania, USA). After washing, targeted proteins were visualized using a LumiLong with chemiluminescence detection (Tprobio, Taiwan). The image was quantified using ImageJ (version 1.51j8; SciJaVa Common Library).

Cells were set at 2.5 × 10⁴ cells/ml in a μ-Slide eight Well (ibidi, Gräfelfing, Germany) and pre-treated with lunasin for 4 h. The cells were then cultured with 5 ng/ml TGF-β or 100 ng/ml leptin, with or without lunasin, for 48 h. After washing, cells were fixed in 4% paraformaldehyde (Sigma) for 10 min and permeabilized in 0.2% Triton-X 100 (Sigma) for 15 min. After washing, the slides were blocked with 5% normal serum (Jackson ImmunoResearch, West Grove, Pennsylvania, USA) in 0.3% Triton-X 100 and processed with E-cadherin (Cell signaling, Danvers, Massachusetts, USA) and vimentin (BioLegend) primary antibodies and CF™488A-conjugated and CF^TM 594-conjugated fluorescein secondary antibodies (Sigma). Nuclei were stained with the NucBlue™ fixed cell ready Probes™ reagent (Invitrogen™, California, USA). Images were taken at 200 × magnification using a fluorescence microscope (Leica, Wetzlar, Germany) and quantified using ImageJ (1.51j8; SciJAVA Common Library) by counting three random fields under a microscope.

Male C57BL/6JNarl mice were purchased from the National Laboratory Animal Center (Taipei, Taiwan). The experimental animal protocol was approved (#111-00118, Institutional Animal Care and Use Committee, National Taiwan University, Taipei, Taiwan), and animal care was conducted in accordance with animal welfare, ethical guidelines, and the 3R, as outlined in the National Research Council's Guide for the Care and Use of Laboratory Animals. Mice were housed in groups of three in polycarbonate cages measuring 30 × 20 × 15 cm under a 12-h light/dark cycle, with the light period occurring from 7:00 a.m. to 7:00 p.m. The temperature was maintained at 24 °C ± 2 °C and the humidity was kept at 60 ± 10%. After a one-week acclimatization period, 6-week-old mice were divided into two groups: a high-fat and high-fructose (HF) group and an HF diet was added the lunasin-enriched soy protein isolated (Archer Daniels Midland, Chicago, USA), at a lunasin concentration of 483 mg/kg in the diet (HFL; n = 6/group). The table of dietary composition is included in Supplementary material 1, and the actual lunasin content was specified in the footnote. The lunasin-enriched soy protein isolated was prepared and replaced the part of casein in the diet. A sample size of six mice per group was selected based on a similar high-fat diet study (21). Ethical constraints were also considered to meet scientific requirements, while following the 3Rs principles, and randomization to minimize variability and ensure reliable data (22). In 1999, the FDA approved a health claim stating that 25 g/day of soy protein may reduce the risk of coronary heart disease (23). The lunasin concentration in HFL was equivalent twice the FDA-recommended daily intake of soy protein. The experimental HF diet, in which 45% calories came from fat and 27% calories came from fructose, mimics diet-induced obesity. Mice were fed the experimental diets ad libitum from 6 weeks old until 22 weeks old. The animals were monitored daily for overall health, activity levels, and food intake. Body weight was measured weekly to assess their wellbeing and identify any potential distress. At the end of the experiment, mice were humanely euthanized in accordance with approved guidelines to minimize pain and distress. Euthanasia was performed by CO₂ inhalation at a flow rate of 30% of the chamber volume per minute, and CO₂ flow was maintained for at least 1 min after respiratory arrest. After exposure, mice that were dead or deeply anesthetized, followed by cervical dislocation to ensure irreversible death. Lung and spleen samples were collected and stored at −80°C for the future analysis.

Lung tissues were separated for two experiments. Firstly, lung tissue was cut into 0.1 mg pieces and added to RIPA buffer at a concentration of 0.1 mg/ml to facilitate homogenization. The homogenate was then centrifuged at 13,000 × g at 4 °C for 20 min, and the supernatant was collected for cytokine analysis. In another experiment, lung tissue was cut into 0.05 mg and cultured in 1 ml of Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco, Waltham, Massachusetts, USA) with 1% fetal bovine serum (Gibco) in 24-well plates in the presence or absence of 1 μg/ml LPS (Sigma) for pro-inflammatory stimulation. After 24 h of culturing, the supernatants were harvested for cytokine analysis. The method was based on, and modified according to a previous study (24). Cytokine levels were adjusted based on the protein concentration in individual lung tissues. The cytokines IL-6, TNF-α, and TGF-β were assayed according to the manufacturer's protocols (BioLegend).

The spleen was collected, ground up, and processed with erythrocyte lysis and ammonia chloride-tris buffer to release signal cells. After washing with Hank's balanced saline solution (HBSS; Sigma), splenocytes were placed at 5 × 10⁶ cells/ml RPMI-1640 medium (Gibco) with 10% mouse serum replacement (Protide Pharmaceuticals, Illinois, USA) in 48-well plates and activated using 10 μg/ml LPS and concanavalin A (ConA, Sigma) for 48 h. Then, supernatants were collected for cytokine analysis. The cytokines under study were analyzed according to the manufacturer's protocols and included IL-2, IL-4, IL-6, IL-17A, TNF-α, and interferon (IFN)-γ (BioLegend).

All data used for this study was available in Supplementary material 3. In the cellular study, data were obtained from at least three independent experiments. In the animal study, data from five to six mice were presented as the mean ± standard error of the mean (SEM). Statistical analysis was performed via Student's t-test using Statistical Product and Service Solutions (SPSS, version 25; Armonk, NY, USA). A statistically significant difference was considered for values of P < 0.05.

A549 cells were treated with serial doses of lunasin, PA, and LPS for 24 h, and cell viability was detected via MTT (Figures 1A–C). Cultured supernatants were analyzed for the pro-inflammatory cytokine IL-6 (Figures 1D–F) to confirm the toxicity and stimulatory effects of the reagents. Treatments with lunasin at 1 to 100 μm and LPS at 0.1 to 20 μg/ml did not affect A549 cell viability, suggesting that these treatments did not exhibit toxicity; however, PA at concentrations greater than 200 μm significantly reduced cell viability. Regarding the pro-inflammatory effect, A549 cells treated with PA at concentrations higher than 500 μm and LPS at concentrations higher than 1 μg/ml showed a significant increase in IL-6 production, but 50 μm lunasin reduced IL-6 secretion (P < 0.05). Therefore, in future experiments, treatment doses were selected at 5 and 50 μm of lunasin, corresponding to low and high doses, and are also similar to previous studies on its anti-inflammatory properties (17, 19, 25). We used 500 μm of PA and 10 μg/ml of LPS to mimic obese conditions, referring to these concentrations, which represent an inflammatory response characterized by an increase in IL-6. Additionally, the selected concentrations of PA and LPS were similar to those in other related studies (26, 27).

To explore whether lunasin influences inflammatory mediators under obese conditions, cells were cultured in high-glucose, PA or LPS conditions to mimic obese micro-environments and treated with TGF-β to induce fibrosis. Because long-term obesity is often accompanied by insulin resistant and hyperglycemia (28). Therefore, a high-glucose condition was also included in this experimental design. The cultured supernatant was analyzed for the production of the cytokines IL-6, MCP-1, and TGF-β via ELISA. Firstly, to confirm whether glucose levels influenced cellular inflammation, cells were cultured in normal-glucose (7 mm) or high-glucose (35 mm) media and challenged with PA, LPS, or TGF-β at the same time. Pro-inflammatory IL-6 and MCP-1 cytokines were significantly increased under obesity-related challenge (Figures 2A, B). In high-glucose media, IL-6 and MCP-1 levels were increased through TGF-β and LPS stimulation, respectively. While TGF-β secretion was induced by LPS stimulation, it was not affected by glucose levels (Figure 2C). Therefore, the inflammatory cytokines IL-6 and MCP-1 were affected by high glucose, but this was not observed for the fibrotic mediator TGF-β.

Lunasin treatment reduced IL-6 and MCP-1 production induced by PA and suppressed MCP-1 secretion triggered by LPS in A549 cells cultured in normal-glucose media (Figures 2D, E). At a low dose of 5 μm, lunasin also decreased TGF-β production in response to LPS stimulation and slightly inhibited PA-induced TGF-β compared with the control group in cells cultured in high-glucose media (P = 0.073; Figure 2F). Collectively, lunasin significantly reduced the production of the pro-inflammatory mediators IL-6 and MCP-1, as well as the profibrotic mediator TGF-β, in obese-mimicking microenvironments, independent of glucose levels. Therefore, subsequent experiments were conducted using normal glucose combined with PA or LPS. Regarding the validity of the lunasin dosage, a low concentration of 5 μm lunasin markedly reduced IL-6, MCP-1, and TGF-β production, indicating that this level was optimal for achieving anti-inflammatory activity in A549 cells. However, a higher concentration of 50 μm slightly reduced cell viability, although the effect was not significant.

To investigate whether lunasin modulates SPs and provides protective effects, A549 cells were treated with lunasin under spontaneous, PA, or LPS conditions. The expression of SP-A, SP-D, and their possible signaling molecule NF-κB was detected (Figure 3). The SP-D level in A549 cells was inhibited by LPS stimulation, whereas it was significantly restored under lunasin treatment compared with LPS only, as determined via ELISA (P < 0.05; Figure 3A); however, this effect was not observed for SP-A (Figure 3B). According to Western blot analysis, 50 μm lunasin consistently increased SP-D production under spontaneous conditions. Additionally, LPS reduced the level of SP-D protein, while 5 μm lunasin tended to increase its expression under LPS stimulation, although without a significant difference (Figures 3C, E). The discrepancy regarding effective doses between ELISA and Western blot analysis likely reflects differences in the detection of secreted vs. intracellular proteins and the duration of treatment.

In an advanced investigation, the molecular signaling of SP-D following lunasin treatment, including NF-κB and its phosphorylation level, was analyzed and quantified (Figures 3D, F). LPS challenge significantly enhanced the phosphorylation of NF-κB, and treatment with 5 μm lunasin reduced this phosphorylation (P = 0.054), but no such observation was made for the PA model. Based on this evidence, lunasin significantly enhanced SP-D protein expression by suppressing NF-κB phosphorylation, which subsequently inhibited the production of inflammatory cytokines in A549 cells under obesity-related conditions.

To assess whether lunasin interferes with fibrosis in A549 cells, wound-healing assays were performed under TGF-β and leptin stimulation to mimic an obese microenvironment. PA treatment did not influence cell migration; thus, it was excluded as a fibrosis model, and the data are not shown. The wound healing assay was analyzed to assess cell migration and also as an indicator of cell fibrosis. Confluent cell monolayers were scraped, and the area of wound closure was evaluated at 0, 24, and 48 h (Figure 4A). The cell layer of the control group showed a significant distance in the scraping. Compared with controls, TGF-β and leptin significantly promoted cell migration, as reflected in the migration index and AUC (Figures 4B, C). However, lunasin treatment did not alter wound healing under either TGF-β or leptin stimulation after scraping compared with the control group.

To further evaluate the role of lunasin in fibrosis, EMT markers were examined in A549 cells. Vimentin (red), a mesenchymal marker, is highly expressed during EMT, whereas E-cadherin (green), an epithelium marker, is abundant in normal epithelial cells, as shown via immunofluorescence (Figures 5A, B). The quantification of fluorescence intensity confirmed that TGF-β and leptin stimulation induced fibrosis, with significantly increased vimentin expression and decreased E-cadherin expression (Figures 5C, D), validating the models. Lunasin treatments did not alter EMT markers expression or cell migration (Figures 4, 5), suggesting it didn't directly interfere fibrosis in A549 cell. However, lunasin reduced TGF-β production (Figure 2), speculating a potential effect at the early stage of fibrosis in A549 cells. Additional analyses such as histological assessments, Masson's trichrome staining, or α-SMA and collagen I staining will be needed in future studies substantiate this conclusion.

To evaluate whether the anti-inflammatory effects of lunasin observed in vitro can be reproduced in vivo, C57BL/6 mice were fed with an HF diet supplemented with a lunasin-enriched soy protein isolate. Lung and spleen tissues were collected to assess local and systemic immune responses. Lung tissues were used to assess the inflammation in the targeted area, while spleens showed the immune responses around the whole body (Figure 6A). At the end of the study, the lung weight (Figure 6B) and body weight of mice (Supplementary material 2) did not differ among groups, indicating that lunasin does not affect lung mass or weight gain in mice. However, lunasin ameliorated adipocyte size and hepatic steatosis, as indicated by H&E staining and quantitative analysis in mice fed a high-fat diet (Supplementary material 2), suggesting that lunasin modified obesity-related pathological changes. Cytokine IL-6, TNF-α, and TGF-β levels in lung homogenates from mice were analyzed via ELISA. Mice fed with the HF diet with lunasin supplementation showed significantly lower TNF-α and TGF-β levels (P < 0.05) and mildly influenced IL-6 (P = 0.089) compared with the HF group (Figures 6C–E). Consistently, in ex vivo lung culture supernatants from lung tissue, lunasin supplementation significantly reduced LPS-induced TNF-α production (P < 0.05) and slightly reduced TGF-β levels (P = 0.073; Figures 6G, H) in the HFL group compared with the HF group, while IL-6 was unchanged (Figure 6F). The level of MCP-1 in lung tissue was not affected by lunasin treatment (data not shown). Together, these results demonstrate that lunasin supplementation attenuates pulmonary inflammation by suppressing the pro-inflammatory cytokine TNF-α and the profibrotic mediator TGF-β in mice fed high fat and high fructose diet.

The immunomodulatory effects of lunasin were assessed by measuring the secretion of cytokines IL-6, TNF-α, IL-17A, IL-2, IFN-γ, and IL-4 in the supernatants of cultured splenocytes, i.e., the mitogens LPS-stimulated B lymphocytes and ConA-stimulated T lymphocytes. Both mitogens markedly increased cytokine secretion (Figure 7). The splenic weights of the mice were not different between the two groups. However, splenocyte numbers in the HFL group were lower than those in the HF group (P < 0.05; Figure 7A). In ConA-stimulated splenocytes in the HFL group, the production of IL-17A, IFN- γ, and IL-4 was decreased (P < 0.05), but the IL-2 level was slightly elevated compared to that in the HF group (P = 0.055; Figures 7B–G). In contrast, mice in the HFL group showed significantly higher IFN-γ secretion in LPS-stimulated splenocytes, while this level was lower in ConA-stimulated splenocytes. Interestingly, the HFL group exhibited increased levels of T helper (Th)1 cytokine IL-2 and reduced levels of Th2 cytokine IL-4. In order to determine the balance of Th1/Th2 cells, the ratio of IL-2/IL-4 and IFN-γ/IL-4 was calculated to confirm that lunasin supplementation significantly enhanced Th1 polarization (Figures 7H, I). In summary, lunasin regulated systemic immunity by promoting Th1 responses and suppressing pro-inflammatory IL-17A level in the spleens of mice fed with an HF diet, suggesting to ameliorate obesity-related chronic inflammation and immune decline.