JGF composition was manufactured by the Branch of Linsen Chinese and Kunming, Taipei City Hospital (Taipei, Taiwan). JGF consists of five medicinal herbs (Scheme 1): 10 g of Forsythia suspensa (Thunb.) Vahl (Monarch; 君; cat no.: 057A-0), 8 g of Scutellaria baicalensis Georgi (Minister; 臣; cat no.: 051-1), 6 g of Bupleurum Chinese DC. (Assistant; 佐; cat no.: 009-1), 6 g of Magnolia officinalis var. biloba Rehder & E.H. Wilson (Assistant; 佐; cat no.: 085K-1) and 3 g of Agastache rugose (Fisch. & C.A. Mey.) Kuntze (Guide; 使; cat no.: 089-1). All medicinal herbs were purchased from the GMP-certified pharmaceutical company (Sun Ten Pharmaceutical Co., Ltd. Taiwan). The preparation and quality control of JGF followed those in a previous study (Ping et al., 2022). Briefly, after decoction, the water-extract was filtered through filter paper to remove insoluble residues, then pre-frozen at −80°C overnight, followed by vacuum freeze-drying. The HPLC fingerprint profile of JGF from Sun Ten Pharmaceutical Co., Ltd. was established using reverse phase-HPLC with a diode-array detector (Supplementary Figure S1).

Murine macrophage-like RAW264.7 and alveolar macrophage MH-S cells were purchased from the Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan). RAW264.7 and MH-S cells were respectively cultured in Dulbecco’s modified Eagle’s medium (DMEM, GIBCO-Life Technologies) and Roswell Park Memorial Institute (RPMI) 1640 medium (GIBCO-Life Technologies) at 37°C under a mixture of 95% air and 5% CO₂. The medium was supplemented with 5% heat-inactivated fetal bovine serum (FBS, HyClone, Marlborough, MA), 3.7 g/L NaHCO₃, 100 units/mL penicillin and 100 units/mL streptomycin (Biological Industries, Cromwell, CT). Attached RAW264.7 and MH-S cells were detached using Trypsin-EDTA (Invitrogen, Co., Carlsbad, CA).

The viability of RAW264.7 and MH-S cells was measured using the crystal violet assay as in previous studies (Ping et al., 2022; Lin et al., 2023). Initially, cells were seeded in 12-well plates at a density of 5 × 10⁴ cells per well and incubated overnight. Then, cells were randomly inoculated with or without LPS (1 μg/mL). Later, various concentrations of JGF (0–600 μg/mL) were added to the cell cultures for 24 h. After treatment, suspended cells were washed with phosphate-buffered saline (PBS). Then, the attached (live) cells were stained with 1% crystal violet (dissolved in 30% ethanol) for 1 h and removed. Next, cells were washed with deionized water, dried and dissolved with 30% acetic acid to release the crystal violet for detection. The ELISA reader was used to measrure the O.D (optical density) at the absorbance wavelength of 570 nm.

RAW264.7 and MH-S cells were cultured in 96-well plates at a density of 1 × 10⁵ cells/well and stimulated with LPS (100 ng/mL; E. coli O55:B5; Sigma Chemical L 2630, St. Louis, MO, USA) in the presence or absence of JGF at different concentrations (0–600 μg/mL) for 24 h. NO production was determined using the Griess assay as described previously (Lin et al., 2023). Griess reagent was added to the culture medium, followed by a 10-min incubation. A standard curve was generated using NaNO₂. Finally, absorbance was measured at 540 nm using a microplate reader. In each experiment, the amount of NO produced by LPS stimulation was defined as 100%.

RAW264.7 and MH-S cells were seeded in 96-well plates at a density of 2 × 10⁴ cells/well for 24 h. The cells were treated with various concentrations of JGF (0–600 μg/mL) and/or LPS (100 ng/mL) for another 24 h. The supernatants were collected for measuring IL-6 and TNF-α concentrations using an enzyme-linked immunosorbent assay (ELISA) kit (BioLegend, San Diego, CA, USA) as described previously (Lin et al., 2023). The optical densities, indicative of IL-6 and TNF-α concentrations, were measured by TECAN Sunrise TMELISA Reader (Tecan Group Ltd., Männedorf, Switzerland) at wavelengths of 450 nm and 550 nm (reference absorbance). The levels of IL-6 and TNF-α after treatment with LPS alone were designated to be 100%.

MH-S cells were seeded in 6-well plates at a density of 5 × 10⁵ cells/well for 24 h, and then incubated in the fresh medium containing only JGF (600 μg/mL) or LPS (100 ng/mL), or JGF + LPS for another 24 h. After stimulation, the supernatants were collected and centrifuged at 1,000 × g for 5 min at 4°C, followed by cytokine array (C3 and C4) analysis (RayBiotech Life, Inc. GA 30092, USA).

Cells were cultured in 6-cm dishes (Corning, USA) at a density of 5 × 10⁵ cells/well and divided into groups of cells treated with JGF (200 μg/mL) alone, and LPS (100 ng/mL) plus JGF (200 μg/mL). After treatment, cells were rinsed with cold PBS and lysed in a specific lysis buffer (Yeh et al., 2022). Then, whole cell lysates were centrifuged (13,000 × g, 10 min, 4°C) and the supernatants were collected for Western blot analysis. The supernatants were quantified using Bradford assay (Bio-Rad, Hercules, CA). The cell extracts (21 μg) were then separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF (0.22 μm) membranes. The PVDF membranes were blocked with 10% bovine serum albumin (60 min, room temperature). After blocking, samples were incubated at 4°C overnight with the primary antibodies mentioned below. The indicated molecules were then detected using secondary antibodies. Antibodies against iNOS, p-JAK2 (Tyr1007/1008), p-STAT3 (Tyr705), and β-actin antibodies were purchased from GeneTex, Inc. (Hsinchu, Taiwan). Anti-p-ERK1/2, p-P38 and p-JNK1/2 (Thr183/Tyr185) antibodies were purchased from Cell Signaling Technology, Inc. (MA, USA). Anti-COX-2 antibody was purchased from Santa Cruz Biotechnology, Inc. (TX, USA). In this experiment, blots were visualized using GE Amersham Imager 600. β-actin and α-tubulin expressions act as internal controls.

Male C57BL/6 mice (8–10 weeks old) were purchased from the National Laboratory Animal Center (Taipei, Taiwan) and nurtured at the animal facility of the National Yang Ming Chiao Tung University. The LPS-induced inflammatory animal model mouse model was modified as described previously (Ou et al., 2023; Nguyen et al., 2021). As recommended by TM experts, the dose of JGF for humans was 30 mg/kg orally, and the calculated oral dose for mice was approximately 370 mg/kg. JGF (400 mg/kg) was administered orally (p.o.) 1 h before as well as 1 and 6 h after intraperitoneal injection (i.p.) of LPS (20 mg/kg). JGF was prepared in deionized water (ddH₂O), and LPS was prepared in normal saline and filtered through a 0.22 μm membrane. Mice were allocated (n = 5 per group) to groups of CTL (ddH₂O + saline), LPS (ddH₂O+ LPS 20 mg/kg), JGF (JGF 400 mg/kg + saline), and JGF + LPS (JGF 400 mg/kg + LPS 20 mg/kg). Blood collection was at 1 h post-LPS injection for IL-6 and TNF-α measurement. At 12 h post-LPS injection, the mice were weighed and then sacrificed using carbon dioxide. Serum, bronchoalveolar lavage fluid (BALF) and abdominal fluid were gathered for the IL-6 TNF-α measurement.

The statistical differences between the mean values of each group were analyzed by one-way ANOVA with post-hoc Tukey tests and t-test using GraphPad Prism 8.4. P values less than 0.05 were regarded as statistically significant. Each experiment was repeated three times or as indicated. All data are expressed as mean ± standard deviation (SD).

To examine the immunomodulatory effect, RAW264.7 and MH-S were treated with various concentrations of JGF (0–600 μg/mL) for 24 h. As shown in Figures 1A,B, JGF showed insignificant cytotoxicity to either cell line except at a high concentration (600 μg/mL), at which JGF eradicated 35% and 20% of RAW264.7 and MH-S cells, respectively. The half-maximal inhibitory concentration (IC₅₀) of JGF in those cells exceeded 600 μg/mL. Subsequently, the anti-inflammatory activity of JGF on LPS-stimulated macrophages was evaluated. These cell cultures were co-treated with various concentrations of JGF (0–600 μg/mL) and LPS (100 ng/mL) for 24 h. JGF could not sustain the survival rate under LPS stimulation, while low concentrations of JGF (50 and 100 μg/mL) rescued slightly MH-S cell viability (Figures 1C,D).

To evaluate the immune response of macrophages to JGF, we performed an array to determine the cytokine profiles in macrophages with or without LPS stimulation. The signal intensities were extracted and normalized (Supplementary Tables S1, S2). The indicated cytokine expressions varied among groups. More than 50 molecules exhibited altered expression patterns in JGF-treated MH-S macrophages (Figures 2A,C). Notably, in the JGF (600 μg/mL) treatment group, cytokines, including MMP-3, VEGF, IL-1β, and IL-2, were visibly elevated compared with the control (CTL) group (Figures 2A,B). However, macrophage-derived chemokine (MDC), macrophage inflammatory protein-3 Alpha (MIP-3α), and monocyte chemotactic protein-5 (MCP-5), which amplify the host response to inflammation (Katou et al., 2001; Dieu-Nosjean et al., 2000; Sarafi et al., 1997), were dramatically reduced following JGF treatment (Figure 2C). Additionally, pro-inflammatory factors such as TNF-α, IL-6, IL-12, MIP-1α were downregulated in MH-S cells after JGF treatment, suggesting that JGF may exhibit immunomodulatory properties in resting macrophages.

Contrary to the CTL group, pro-inflammatory cytokines such as IL-6, GM-CSF, G-CSF, Pro-MMP-9, IL-12, VEGFR3, MIP-2, MDC, RANTES, KC, CRG-2, and IL-1β were significantly expressed after LPS (100 ng/mL) treatment (Figure 2A), confirming LPS as a potent inflammatory stimulus. Interestingly, VEGF levels were further elevated in LPS-stimulated macrophages upon JGF treatment (Figure 2D), suggesting that JGF may influence LPS-induced macrophage polarization (Lai et al., 2019; Wheeler et al., 2018). We further the effects of JGF on LPS-induced pro-inflammatory molecules. As shown in Figure 2E, JGF countered the LPS detrimental effect by decreasing GM-CSF, TNF-α, MDC, and Pro-MMP-9 (Katou et al., 2001; Parajuli et al., 2012; Wang et al., 2009). In short, these findings demonstrated that JGF did modulate the expression of cytokines in macrophages and may function as a potent immunomodulatory agent.

We further used KEGG (Kyoto Encyclopedia of Genes and Genomes) and GO (Gene Ontology) platforms to investigate the immunoregulatory profiles of JGF under basal and inflammatory conditions. As shown in Figure 3A, the enrichment network and dot plot showed that JGF activated a wide range of immune and inflammatory signaling pathways. Notably, the cytokine–cytokine receptor interaction, IL-17 signaling, and TNF signaling pathways were significantly enriched. Additional intracellular pathways such as the MAPK and JAK-STAT signaling pathways were upregulated, suggesting that JGF may broadly enhance innate immune signaling. In contrast, the key pro-inflammatory pathways, including cytokine–cytokine receptor interaction, TNF, IL-17, JAK-STAT, and PI3K-AKT pathways, were downregulated in the immunosuppressive role of JGF under LPS-induced inflammatory stress (Figure 3B). Interestingly, viral protein interaction with cytokine and cytokine receptor was shown in JGF alleviated LPS-mediated signaling, which may be correlated to JGF as an agent for COVID-19 (Ping et al., 2022). Together, JGF may function as a bidirectional immune modulator.

JGF exhibited different macrophagic responses. Induction of NO, TNF-α, and IL-6 by LPS served as markers of inflammatory activation in these cells (Guha and Mackman, 2001; Greenhill et al., 2011). To assess whether JGF itself triggers an inflammatory phenotype, macrophages were treated with JGF alone. As shown in Figure 4A, both Griess and ELISA assays revealed that JGF upregulated NO levels in RAW264.7 and MH-S cells concentration-dependently. However, even at the highest concentration tested (600 μg/mL), JGF-induced NO levels remained approximately ten-fold lower than those elicited by LPS (100 ng/mL). IL-6 and TNF-α expressions in RAW264.7 were slightly elevated by JGF administration, whereas MH-S cells showed trivial change (Figures 4B,C). To investigate the potential involvement of MAPK signaling pathways in JGF and LPS-induced NO production, three specific MAPK inhibitors were used: PD98059 (ERK inhibitor), SP600125 (JNK inhibitor), and SB203580 (p38 inhibitor). As shown in Supplementary Figure S2, although JGF-induced NO production was relatively low (approximately 10% of LPS-induced levels), individual MAPK inhibitors showed slight suppressive effects. Notably, the combination of all three inhibitors resulted in a more pronounced inhibition. These results suggested that JGF may provoke innate immune adaptation (Figure 3A).

In addition, JGF had a broad inhibitory effect on LPS-induced inflammatory factors (Figure 2). Therefore, we deciphered whether JGF could reverse NO, TNF-α, and IL-6 changes in macrophages pre-treated with LPS. As shown in Figure 4D, JGF significantly decreased NO emission from LPS-stimulated macrophages in a concentration-dependent manner. The half-maximal effective concentration (EC₅₀) of JGF against LPS-induced NO production on RAW264.7 and MH-S cells were 87.07 and 71.84 μg/mL, respectively. JGF also inhibited TNF-α secretion on LPS-stimulated MH-S cells but scarcely impacted RAW264.7 (Figure 4E). The EC₅₀ to TNF-α on MH-S cells was 488.71 μg/mL. Correspondingly, JGF reversed the effect of LPS on the IL-6 level (Figure 4F), in which the EC₅₀ for RAW264.7 and MH-S cells were 167.01 and 175.82 μg/mL, respectively. Thus, despite slightly activating inflammatory markers when solely administered, JGF could adequately inhibit the acute LPS-mediated reactions in macrophages.

As mentioned, JGF inhibited NO production in LPS-stimulated RAW264.7 and MH-S cells. The biological effects of JGF on pro-inflammatory mediators were examined using Western blot assay. iNOS and COX-2 are two critical molecules involved in the pathophysiology of inflammation (Surh et al., 2001). As shown in Figure 4G, JGF increased slightly the quantity of iNOS and COX-2 compared with that of LPS. Specifically, JGF-induced COX-2 expression more significantly in RAW264.7 cells than in MH-S cells, which may indicate why JGF induced less inflammatory responses in MH-S cells as shown in Figure 4H. On top of that, JGF repressed the intense expression of iNOS and COX-2 in LPS-stimulated RAW264.7 and MH-S cells (Figure 4I). Together, besides mildly enhancing immune ability, JGF significantly reduced LPS-driven overt inflammation.

The JAK/STAT pathway plays a significant role in cellular response to LPS toxification (Lin et al., 2022). To explore the potential intracellular effects of JGF on RAW264.7 and MH-S cells, both in the presence and absence of LPS stimulation, the activation status of JAK/STAT signaling was examined. As illustrated in Figure 5A, without LPS stimulation, JGF upregulated the JAK2 phosphorylation (Figure 5C) but has no significant effect on the activation of STAT3 (Figure 5D). Interestingly, in LPS-stimulated cells, co-treatment with JGF resulted in a marked reduction in STAT3 phosphorylation. The MAPK pathway, comprising JNK1/2, ERK1/2, and p38 cascades, is known to play a critical role in LPS-induced inflammatory responses (Kyriakis and Avruch, 2012). We first examined the phosphorylation of JNK1/2. As shown in Figures 5B,D, JGF induced phosphorylation of JNK1/2—particularly JNK1—in the absence of LPS. However, in the presence of LPS, JGF significantly suppressed JNK1 activation. In contrast, JGF did not affect JNK2 phosphorylation in MH-S cells. We further showed that inhibition of JNK effectively inhibited the phosphorylation of STAT3 induced by LPS stimulation in macrophages (Supplementary Figure S3). In addition, while JGF enhanced ERK1/2 and p38 phosphorylation under basal conditions, it failed to suppress their LPS-induced activation (Figures 5E,F). These findings suggest that JGF may activate macrophages under non-inflammatory conditions, yet mitigate excessive inflammatory responses triggered by LPS.

LPS-induced endotoxemia is a well-accepted in vivo model for conducting systemic inflammatory response syndrome (SIRS), leading to numerous lethal consequences in animals and increases in pro-inflammatory cytokines (Seemann et al., 2017). Accordingly, we established an LPS-challenged mouse model to assess the anti-inflammatory potential of JGF (Figure 6A). JGF treatment markedly mitigated body weight loss in LPS-exposed mice, with a relative improvement of approximately 3% (Figure 6B). In general, IL-6 and TNF-α are crucial mediators for regulating local and systemic acute inflammation (Nandi et al., 2010). To evaluate the effect of JGF on these cytokines, serum levels of IL-6 and TNF-α were initially measured. As shown in Figure 6C, LPS administration significantly increased circulating levels of both cytokines, whereas JGF treatment led to reductions of 41% and 35% in IL-6 and TNF-α, respectively. Notably, JGF alone did not elicit significant cytokine induction in serum after 1 hour of administration (Figure 6C). Interestingly, even after three administrations of JGF, serum IL-6 levels in mice remained largely unchanged, while TNF-α levels showed a slight increase (Figure 6D). Nevertheless, the overall expression of both cytokines remained low. Given that acute inflammation can progress to multi-organ dysfunction, particularly affecting the lungs and peritoneal cavity, IL-6 and TNF-α levels in bronchoalveolar and abdominal fluid samples were further measured (Chen et al., 2018). We found that JGF alone treatment could induce IL-6 levels but not TNF-α in both compartments (Figures 6E,F), which was consistent with our in vitro findings. Intriguingly, in mice challenged with LPS, subsequent treatment with JGF resulted in substantial decreases in IL-6 and TNF-α levels in lung and peritoneal fluids at 12 h, with reductions of 68% and 54% for IL-6, and 48% and 59% for TNF-α, respectively (Figures 6E,F). Together, these results indicated that JGF functions as an immunomodulatory agent predominantly by modulating macrophage activity in mice.