QFLT decoction consisted of CaSO₄·2H₂O (Gypsum fibrosum) (Shigao, SG) 30 g, Nervilia fordii (Hance) Schltr. (Orchidaceae, Nervilia plicatae herba) (Qingtiankui, QTK) 10 g, Trichosanthes kirilowii Maxim. (Cucurbitaceae, Trichosanthis pericarpium) (Gualoupi, GLP) 15 g, Scutellaria baicalensis Georgi (Lamiaceae, Scutellariae radix) (Huangqin, HQ) 15 g, Fritillaria thunbergii Miq. (Liliaceae, Fritillariae thunbergii bulbus) (Zhebeimu, ZBM) 15 g, Houttuynia cordata Thunb. (Saururaceae, Houttuyniae herba) (Yuxingcao, YXC) 20 g, Prunus armeniaca L. (Rosaceae, Armeniacae semen amarum) (Kuxingren, KXR) 10 g, Platycodon grandiflorus (Jacq.) A. DC. (Campanulaceae, Platycodonis radix) (Jiegeng, JG) 15 g, Phragmites australis (Cav.) Trin. ex [Poaceae, Phragmitis rhizoma] (Lugen, LG) 15 g and Glycyrrhiza uralensis Fisch. ex DC. (Fabaceae, Glycyrrhizae radix et rhizoma) (Gancao, GC) 10 g. These components were purchased from the First Affiliated Hospital of Guangzhou Medical University. Before the experiment, the Chinese medicine extract was made by Wuhan Kangle Pharmaceutical Co., Ltd. A total of 155 g (raw materials) QFLT decoction was soaked in 1,500 ml distilled water for 1 h, then decocted for 2 h (CaSO₄·2H₂O first decocted for 30 min) and filtered. The filter residues were soaked in 750 ml of distilled water, boiled for 2 h, and filtered again (Liu et al., 2014). Then, we obtained about 200 ml of filtrates each time. And the two filtrates were combined to obtain 400 ml of QFLT solution. The preparation method was consistent with the process used for patients in medicine/hospital. The combined filtrates were concentrated to 100 ml, then freeze-dried to obtain the extract (extraction rate 26%), and stored at −20°C. The dried powder was dissolved in water to prepare oral QFLT for mice with three concentrations of 0.4, 0.8, and 1.6 g/ml, and stored at −4°C.

For analysis, the chemical composition of QFLT was determined using the Thermo Dionex Ultimate 3000 UPLC system (Thermo Fisher, Waltham, MA, United States). 1 g QFLT freeze-dried powder was added to 10 ml chromatographic methanol, and ultrasonically extracted for 1 h. The extract was centrifuged at 12,000 rpm for 15 min and then passed through a 0.22 μm microporous membrane. The QFLT sample was separated on a Waters Acquity UPLC BDS C18 column (2.4 × 150 mm, 2.74 μm). The sample injection volume was 5 μL. In the mobile phase, the mixture of acetonitrile (A) and 0.1% formic acid (B) as solvent performed to elute at a flow rate of 0.30 ml/min. Column temperature was set at 45°C. The gradient elution was 0–3 min, 95% B; 3–45 min, 95%–25% B; 45.1–50 min, 95% B. Mass conditions were set as follows: spray voltage (+), 3.5 kV; spray voltage (−), 3.0 kV; capillary temperature, 320°C; sheath gas flow, 30 Arb; aux gas flow, 10 Arb; aux gas heater temperature, 400°C; and S-Lens RF Level, 55%. Finally, the total ion chromatogram was obtained under the positive and negative ion mode. Subsequently, according to published indicators of quality, chromatographic behavior and fragment ion mass, the chemical compounds in QFLT were identified. The data were processed using Xcalibur software version 2.7.

We collected the putative targets of QFLT using SwissTargetPrediction (http://www.swisstargetprediction.ch/), BATMAN-TCM (http://bionet.ncpsb.org.cn/batman-tcm/), STITCH v5.0 195 (http://stitch.embl.de/) and Drugbank v5.1.7 196 (https://go.drugbank.com/). Gene names corresponding to protein names were identified by Uniprot (https://www.uniprot.org/).

In String v11.0, we searched these putative targets using the multiple proteins option. We set the organism to be homo sapiens and selected a high confidence level (0.700). We performed PPI network visualization and further analysis using Cytoscape v3.7.2. In addition, the cytohubba app of Cytoscape was utilized to analyze the PPI network.

We collected disease gene targets for ALI/ARDS from two databases, GeneCards (https://www.genecards.org/) and CTD (https://ctdbase.org/). The search terms “acute lung injury” or “acute respiratory distress syndrome” were employed to acquire ALI/ARDS-related targets. After deletion of redundant items, the Venn diagram showing interaction between QFLT and ALI/ARDS was created.

GO/KEGG Pathways enrichment analysis relied on the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) database by DAVID v6.8, in which p < 0.05 was considered statistically significant.

Seven-week-old adult male mice C57BL/6J of average weight 18–22 g were purchased from Weitong Lihua Laboratory Animal Technology Co., Ltd. [Animal License: SCXK (Jing) 2016-0006] in Beijing, China. They were raised in an environment matching SPF-condition for several days. The number of animals in each group was estimated based on previous studies of LPS-challenged ALI animal models (Fu et al., 2019; Amatullah et al., 2021). The methods of animal experiments followed the ethical codes of Beijing University of Chinese Medicine. The experiments have been approved by the Animal Care and Use Committee of Beijing University of Chinese Medicine (BUCM-4-2019102105-4119).

Mice were randomly divided into six groups with eight mice in each group, including the control, LPS, low-dose, medium-dose, high-dose, and control-high-dose groups. After a 12-h fasting, mice were treated with LPS (5 mg/kg) (Sigma-Aldrich, St. Louis, Missouri) in the LPS group and 0.9% NaCl in the control group (Zhou et al., 2017; Zhong et al., 2019; Huang et al., 2020) One week before intratracheal administration, mice were treated with low-dose QFLT (4 g/kg), medium-dose QFLT (8 g/kg), high-dose QFLT (16 g/kg), control-high-dose QFLT (16 g/kg) once a day as a protective treatment. The control and LPS groups were orally administered equal volumes of 0.9% NaCl.

Dosage of QFLT decoction was determined according to conversions from clinical daily adult dosages. The daily dosage of QFLT for mice is 8 g/kg (the extract) for mice, equivalently, the daily dose of QFLT for adult is 40.3 g (the extract), which is calculated based on formula converting dosage of human into the mouse according to the respective body surface areas of the Chinese Medicine Pharmacology Research Technology. In order to explore the dose-effect relationship, doubled up as the high-dose group (16 g/kg), and doubled down as the low-dose group (4 g/kg). Therefore, this study set the dose of QFLT (4, 8, and 16 g/kg) to treat ALI mice.

Twenty-four hours after LPS administration, mice were euthanized using sodium pentobarbital injection. Bronchoalveolar lavage fluid (BALF) and lung tissues were collected for further studies. BALF was stained using Wright-Giemsa solution (Solarbio, Beijing, China) to sort and count the cells. The upper left lung was excised and fixed with 4% neutral formaldehyde for conventional hematoxylin and eosin (H&E) staining to reflect the pathological morphological changes in the lung tissue of LPS-challenged mice. ALI H&E sections of 4–5 μm in thickness were examined and assessed by professionals under light microscopy. Lesion severity was scored as: 0 (normal), 0.5 (mild), 1 (mild), 2 (moderate), 3 (severe), based on indicators including alveolar wall congestion, emphysema, bronchial epithelial cell degeneration, inflammatory cell infiltration and lymphocyte proliferation. The right lower lung’s wet/dry (W/D) weight ratio was calculated to evaluate lung edema.

The Enzyme-linked immunosorbent assay kit (ELISA) theory utilized specific antibody-antigen interactions to detect antigen. We detected the secretion of inflammatory cytokines TNF-α, IL-6 and IL-1β in BALF using an ELISA kit (Proteintech, Wuhan, China). In addition, we collected the macrophages cell-free supernatants for the detection of TNF-α, IL-6, and IL-1β secretion. The optical density was measured at 450 nm.

To quantify cytokines RNA expression in mice lungs stimulated with LPS, the levels of TNF-α, IL-1β, IL-6 mRNA were examined using qPCR. We used the NCBI database for primers design and purchased primers from Sangon Biotech (Shanghai, China). The primers were presented as follows. Data were normalized by the expression of GAPDH mRNA in each sample. TNF-α (Forward primer: GAT​C-GGT​CCC​CAA​AGG​GAT​G, Reverse primer: CCA​CTT​GGT​GGT​TTG​TGA​GTG), IL-1β (Forward primer: AGA​GTC​CCC​AAC​TCA​TCT​CCT, Reverse primer: AAG​TCC​CTA​GGT​TGG​GCT​TG), IL-6 (Forward primer: GAC​AAA​GCC​AGA​GTC​CTT​CAG​A, Reverse primer: TGT​GAC​TCC​AGC​TTA​TCT​CTT​GG), GAPDH (Forward primer: CTC​TGG​TGG​CTA​GCT​CAG​AAA, Reverse primer: CCC​TGT​TGC​TGT​AGC​CGT​AT).

The multiple oxidation products in mice lungs and A549 cells were measured using the MDA assay kit (TBA method), SOD assay kit (hydroxylamine method), GSH-Px assay kit (colorimetric method) and GSH kit, respectively (Nanjing, China). The ROS was evaluated using a Fluorometric Intracellular ROS kit (Beyotime, China).

Macrophages (MH-S) (Manassas, VA, United States) were incubated in DMEM containing 10% FBS fetal bovine serum, 100 units/mL Penicillin G and 100 μg/ml Streptomycin Sulfate solution at 37°C in a 5% CO₂ atmosphere. The cells were then cultured in a serum-free medium for 3 h and stimulated with LPS (5 μg/ml) in the presence or absence of QFLT. After 24-h incubation, the cultured medium was used for cytokine detection with the ELISA kit.

The human lung type II epithelial cell line A549 (Stem Cell Bank, Chinese Academy of Sciences, Shanghai, China) was incubated in a DMEM medium containing 10% FBS and 1%PS. Cells were grown in a 96-well plate to 70%–80% confluency overnight and then in a serum-free medium for 3 h. After QFLT (0.1, 1, 10 μg/ml) pretreatment for 24 h, cells were incubated with LPS (800 μg/ml) for 24 h to detect cell viability by CCK-8.

All data referenced above experiments were calculated as the mean ± SD. Each assay was performed for at least three independent and repeatable experiments. The statistical analysis and graphics were performed using GraphPad Prism 8.0. One-way ANOVA and Student’s two-tailed t-test were carried out. p < 0.05 was considered to indicate a statistically significant difference between groups.

Aqueous extract of QFLT was analyzed by UHPLC-MS to identify the chemical compounds (Figure 2). Based on the retention time and molecular ion peaks, compared with published data from TCMSP database (http://tcmspw.com/tcmsp.php), accurate quality, chromatographic behavior, and fragment ion mass, 47 chemical components were identified, including 35 flavonoids, four alkaloids, four organic acids, three cyanogenic glycosides, and one coumarin, (Table 1).

Compound targets collection relied on SwissTargetPrediction, BATMAN-TCM, STITCH and Drugbank database. After excluding duplicate values, 627 potentially relevant targets of QFLT were obtained. The detailed information about the interaction between potential targets and the 47 identified compounds was showed in Supplementary Data Sheet S1. Disease targets (Supplementary Data Sheet S2) collection identified 28,080 ALI-related targets and 17,773 ARDS-related targets in the CTD database and 6783 ALI-related targets and 3381 ARDS-related targets in the GeneCards database. Of these, 2547 overlapping targets may be considered important ALI/ARDS-related targets (Figure 3A). Figure 3B showed that 2547 targets for disease and 627 targets for QFLT had 328 overlaps (Supplementary Data Sheet S3), which may be the key targets for QFLT in treating ALI/ARDS.

Based on the 328 overlapping targets, we constructed the compound-target network (Figure 4). The network consisted of 373 nodes and 1718 edges (Supplementary Data Sheet S4). The average number of neighbors, representing the average connectivity among network nodes, was 9.212, and network centralization and heterogeneity were 0.178 and 1.633, respectively, indicating the polypharmacology characteristic of TCM. These results indicated that nine herbs in QFLT acted synergistically to treat ALI/ARDS in a “multi-compound, multi-target” mode. Acacetin (degree 73, Neighborhood Connectivity 13), Baicalein (degree 71, Neighborhood Connectivity 12), Luteolin (degree 61, Neighborhood Connectivity 14), Liquiritigenin (degree 75, Neighborhood Connectivity 10), Wogonin (degree 71, Neighborhood Connectivity 13), and Isorhamnetin (degree 68, Neighborhood Connectivity 14) may be the important anti-inflammatory and anti-oxidative compounds of QFLT against ALI/ARDS. These indicated that QFLT’s anti-inflammatory and anti-oxidative pharmacological effects has correlation with regulation of disease targets based on multiple anti-inflammatory and anti-oxidative components.

To further explore the function of target protein interaction, the PPI network obtained from the STRING database was analyzed by cytohubba app of cytoscape. As shown in Figure 5, we identified a number of immune and inflammatory response (cytokine and pattern recognition receptors) and oxidative stress (oxidoreductases such as CYP450 and ALOX) related targets in the PPI network through gene annotation. The results suggested the potential pharmacological effects of QFLT in anti-infection, anti-inflammation, anti-oxidant, alleviation of cytokine storm and organism impairment.

To further explore the anti-inflammatory and anti-oxidative functions of QFLT in ALI/ARDS, we performed a potential GO and KEGG pathways enrichment analysis on the critical targets via DAVID. Figure 6A showed results of the top 15 GO analyses (adjusted p < 0.01), suggesting that many biological processes may be involved in ALI/ARDS treatment, including cytokine-mediated signaling pathway, positive regulation of phosphatidylinositol 3-kinase signaling, and positive regulation of protein kinase B signaling.

For concise presentation, the top 15 of KEGG pathways (adjusted p < 0.01) were presented on bubble charts after removing human disease pathways. As shown in Figure 6B, these pathways may be the crucial pathways for ALI/ARDS treatment. The important targets of immune and inflammatory response such as TNF, RELA, NFKB1, IKBKB, and TLR4 were enriched in the TNF signaling pathway and Toll-like receptor signaling pathway. The activation of TNF signaling pathway plays an integral role in the production of proinflammatory cytokines such as IL-1β, TNF-α, and IL-6 (Walczak, 2011). TLR4 is a susceptibility gene for ALI, and TLR4-TRIF-TRAF6 signaling was a critical pathway of ALI. The production of oxidants can trigger lung injury and stimulate cytokine production by lung macrophages via TLR4-TRIF, suggesting that oxidative stress and innate immunity playing critical roles in ALI (Imai et al., 2008). Therefore, we will further focus on TNF signaling pathway, Toll-like receptor signaling pathway and verify the downstream biomarkers (TNF-α, IL-1β, IL-6) in vivo and in vitro experiments.

The lung W/D ratio was significantly increased in the LPS group compared to the control group (p < 0.001). However, QFLT reversed this trend dose-dependently compared with the LPS group (p < 0.05). As shown in Figure 7B, QFLT decreased the inflammatory score of ALI lung tissue compared with the LPS group. For the H&E staining, as show in Figures 7A–C, the alveolar structure of the LPS group was severely damaged, alveolar cavity hemorrhage accompanied by inflammatory cell infiltration, pulmonary interstitial edema, and the alveolar septum was significantly thicker than the control group. QFLT significantly alleviated the pulmonary morphological damage challenged by LPS.

Total cells, neutrophils and macrophages contained in BALF from mice exposed to LPS were counted and stained with Wright-Giemsa. As shown in Figures 8A–C, compared with the control group, the numbers of total cells, neutrophils and macrophages were significantly increased in the LPS group (p < 0.01), and QFLT pretreatment significantly decreased the number of inflammatory cells compared with LPS group (p < 0.05).

To determine whether QFLT regulated inflammatory and immune response in LPS-challenged mice, we collected BALF to examine proinflammatory cytokines in the alveolar microenvironment. As illustrated in Figures 8D–F, compared with the control group, secretions of TNF-α, IL-6, and IL-1β were elevated significantly in LPS group (p < 0.01), which was significantly inhibited by QFLT treatment (p < 0.05). In Figures 8G–I, levels of TNF-α, IL-6, and IL-1β mRNA significantly increased in LPS-challenged mice lung tissue (p < 0.01), and QFLT treatment significantly suppressed the mRNA level of these genes (p < 0.05).

Macrophages, as the primary contributors, taking essential responsibilities to potentially pathological inflammatory processes, can produce rapidly large amounts of inflammatory cytokines in response to danger signals (Hamidzadeh et al., 2017). To confirm the anti-inflammatory pharmacological effects of QFLT in LPS- activated macrophages, macrophages were incubated with or without LPS (5 μg/ml) in the absence or presence of different concentrations of QFLT (0, 0.1, 1, 10 μg/ml) for 24 h. The levels of TNF-α, IL-6, and IL-1β were detected by ELISA (Figures 9A–C). The results showed that compared with the LPS group, the levels of TNF-α, IL-6, and IL-1β were also significantly reduced by QFLT treatment in a dose-dependent manner (p < 0.05). In Figures 9D–F, we also further explored whether QFLT treatment was time dependent in changes of proinflammatory cytokines levels. The levels of TNF-α, IL-6, and IL-1β were measured in LPS-activated macrophages after 12-h and 24-h of QFLT treatment, respectively. The results showed that QFLT can decrease the levels of TNF-α, IL-6 and IL-1β in LPS-activated macrophages in a time-dependent manner (p < 0.05).

To further investigate the potential role of QFLT in ALI/ARDS as an antioxidant, we further examined multiple oxidative stress-related biomarkers. MDA can reflect the severity of oxidative stress damage, while SOD and GSH-Px can reduce tissue and cell damage caused by oxidative stress. Compared with the control group, the LPS group of mice showed increased MDA level (p < 0.001) but decreased SOD (p < 0.01) and GSH-Px level (p < 0.01). The results demonstrated that QFLT treatment significantly attenuated MDA generation and SOD and GSH-Px depletion (p < 0.05) (Figures 10A–C).

Oxidant production within lung can lead to extensive destruction of alveolar epithelial cells in ALI/ARDS (Ward, 2010). Therefore, we sought to determine the effects of QFLT treatment on LPS- activated alveolar epithelial A549 cells. To determine the proper LPS concentration on the oxidation model of alveolar epithelial A549 cells, the cells were incubated with different concentrations of LPS (0–1,000 μg/ml) for 24 h, then the cell viability was assessed by CCK-8 assay. Accordingly, A549 cells were in the absence or presence of different concentrations with QFLT (0, 0.1, 1, 10 μg/ml) pretreatment for 24 h and then incubated with/without LPS (800 μg/ml) for 24 h. The results suggested that QFLT pretreatment enhanced LPS-activated A549 cell viability in a concentration-dependent manner (p < 0.05). Reduced glutathione (GSH), as an enzyme-catalyzed antioxidant, plays an important role in alleviating oxidative tissue injury. Compared with the LPS group, QFLT treatment increased GSH content (p < 0.05). ROS generation was assessed in LPS-activated A549 cells indicating that QFLT pretreatment dramatically reduced ROS generation (Figures 10D–G).