2.1 Materials and chemicals

Liquiritigenin apioside and licorice-saponin G2 were isolated in our laboratory, and their structures were unambiguously identified by nuclear magnetic resonance (NMR) and MS methods. The other reference standards were purchased from National Institutes for Food and Drug Control., Chengdu Puruifa Medical Technological Co., Ltd (Chengdu, China), Lemeitian Co. Ltd (Chengdu, China), Shanghai yuanye Bio-Technology Co., Ltd (Shanghai, China) and Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). Detailed information regarding the reference standards used in this study is presented in Table S1.

JZOL was provided by Jiangsu Kanion Pharmaceutical Co. Ltd. (No. 181026, Lianyungang, China). Scutellariae Radix, Fritillariae Ussuriensis Bulbus, Rhei Radix et Rhizoma, Bovis Calculus Artifactus, and Glycyrrhizae Radix et Rhizomawere were purchased from the Ji’an Medical Material Market (Jiangxi, China). All herbal medicines were identified by Prof. GuangXiong Zhou of Jinan University, and voucher specimens were deposited at Institute of Traditional Chinese Medicine & Natural Products, college of pharmacy, Jinan, University, Guangzhou, China.

LC-MS grade acetonitrile: Fisher Scientific, LC-MS-grade formic acid: Sigma-Aldrich, Water: Watsons. Methanol, and ethanol used for sample extraction were of analytical grade. Details of other materials are recorded in the corresponding methods.

2.2 Sample preparation

JZOL was directly evaporated with a rotary evaporator and then diluted to a concentration of 10 mg/mL. Subsequently, 2 mL of these solutions were transferred into a separate clean tube, dried under a nitrogen gas stream at room temperature, and then suspended in 2 mL of water. For the Solid-phase extraction (SPE, Vac 3 cc, 200 mg Phenomenex strata C18-E cartridges from Torrance, CA) process: the cartridge were pre-conditioned with 3 mL of methanol, followed by 3 mL of water before application. The centrifuged supernatants were loaded onto the SPE cartridges and washed with 2 mL of water. Afterward, an elution step was performed using 4 mL of methanol, and the eluted solution was collected. Supernatant aliquots of 2 µL centrifuged supernatant were then injected into the LC-MS system.

2.3 Animals and drug administration

Healthy male Sprague Dawley rats, weighing 200 ± 20 g, were purchased from the experimental animal center of Guangdong province (Guangzhou, China). All animals were randomly assigned to groups and maintained under standard condition (23 ± 2 °C and 55–60 % relative humidity). The rats were given free access to water and standard chow for a week before the experiment. Next, the drug-treated group (n = 7) was administered JZOL at the dose of 3.31 g/kg/day (equal clinical dose). Blood samples (0.3 mL) were collected from the jugular vein and placed in heparinized tubes at 0.25, 0.5, 1.0, 2.0, 4.0, 6.0, 8.0, 12.0, 24.0, 36.0, 48.0 and 72.0 h after the drug administration. After centrifuging (14,000 rpm, 4 °C) the blood sample for 10 min, the obtained plasma was stored at −80 °C.

2.4 Pretreatment of biological samples

A volum of 100 µL plasma sample was spiked with 50 µL of IS and 400 µL of 0.5 % formic acid in acetonitrile for protein precipitation. Following a 2-minute vortexing and subsequent centrifugation (14,000 rpm, 20 min), the supernatants were dried to completion using nitroge. The residue was reconstituted in 100 µL methanol and vortexed for 2 min. After centrifuged (14,000 rpm, 20 min), an aliquot of 2 µL was injected into the LC-MS system for analysis.

2.5 UPLC-Q/TOF-MS analysis

UPLC analyses: ACQUITY^TM UPLC I-Class system, binary solvent system, and an automatic sample manager. Q/TOF-MS analyses: Waters SYNAPT™ G2 mass spectrometer (Waters, Manchester, UK), and an ESI interface. The detail information of optimal conditions was presented in supplementary material.

2.6 UPLC-QqQ-MS analysis

Components were detected under a Xevo TQ-XS mass spectrometer (Waters Corp., Milford, MA, USA) equipped with electrospray ionization (ESI) interface. Detailed information regarding the optimal conditions can be found in the supplementary materials. The most appropriate precursor ion, daughter ion, cone voltage, and collision energy were optimized and displayed in Table S2.

All experimental data were collected and processed by Waters Masslynx™ software 4.1 and the Quanlynx program (Waters, Milford, MA, USA).

2.7 Method validation of pharmacokinetic research

The method validation was carried out according to the 2018 Food and Drug Administration (FDA) guidelines on bioanalytical method validation in terms of selectivity, linearity, accuracy, matrix effect, precision, and stability. The detail information of method validation was presented in supplementary material.

2.8 The candidate Q-markers selection

Twenty-two components, whose structures are all well-defined by the reference standards, covering the characteristics of multiple structural types (flavonoids, triterpenoid saponins, alkaloids, bile acids, anthraquinones and others), higher content (11.56 ∼ 867.40 μg/mL), and bioavailable constituents in vivo (absorbed in blood) were selected as candidate Q-markers of JZOL based on the researches of qualitative and quantitative chemical analysis, as well as PK study. In detail, the selection of the 22 components was a comprehensive compilation, consisting of 13 components that could be measured for in vitro content, and 21 prototypes (Table S3, Fig. S1) (Table S3, Fig. S1) that were accurately identified with reference standard for in vivo verification.

2.9 Network pharmacology studies

The structures of selected candidate Q-markers in JZOL were prepared in SMILES format, and imported into the SwissTargetPrediction (https://www.swisstargetprediction.ch/) for the prediction of targets in “homo sapiens” species (Gfeller et al., 2014). The target set was created to explore potential interactions between protein targets and the compounds in JZOL. As the Drugbank database (https://www.drugbank.ca/) provides comprehensive profiles of protein targets associated with clinical symptoms (Wishart et al., 2018), the protein targets involved in the symptom of “inflammation” were systematically searched and derived as a target set of inflammation. The target sets involved in the regulation of Gypsum Fibrosum and Chloriti Lapis was respectively derived from the STITCH server (https://stitch.embl.de/) (Kuhn et al., 2014) based on the main ionic components including Ca²⁺, SO4^2-, Fe²⁺, Mg²⁺, K⁺, Na⁺ (Ikarashi et al., 2012; Chen et al., 2022). These target sets were compared and overlapped to get the common targets for further analysis (Amala et al., 2023).

The key targets involved in the interactions of 22 compounds and the symptom of “inflammation” were collected for bioinformatics analysis. The Search Tool for the Retrieval of Interacting Genes (STRING) database (https://string-db.org/) is used to input all essential targets for protein–protein interaction (PPI) prediction (Szklarczyk et al., 2021). The detail information was supported in supporting information. The regulation effect of 22 compounds of JZOL as well as the main ionic components of Gypsum Fibrosum and Chloriti Lapis were characterized by their involved key targets, while the corresponding signaling pathways were further analyzed by gene set enrichment analysis (GSEA) (version 3.0, https://software.broadinstitute.org/gsea/). The top 20 items with the most significant were derived for further analysis.

The importance of targets that involved in the interactions of 22 compounds and the symptom of “inflammation” were further evaluated by molecular docking using Glide 6.6 of Schrödinger Software Suite (Schrödinger, LLC, New York, NY, 2015). The 3D crystal structures for 46 unique targets were retrieved from the Protein Data Bank database (https://www.rcsb.org/) or the AlphaFold protein structure database (https://alphafold.ebi.ac.uk/). The target structures were initially readied by introducing hydrogen atoms, rectifying absent side chains, and conducting minimization using the Protein Preparation Wizard module. Thereafter, the binding pocket for each target was defined according to the location of the co-crystalized ligand or the best site predicted by the fpocket software.

The 3D structures of 22 compounds were generated and optimized using the LigPrep module, targeting a pH range of 7.0 ± 2.0. The standard precision (SP) protocol was utilized to conduct docking simulations involving 46 targets and 22 compounds. The best docking score for each target-compound couple in the docking runs was selected for analysis, while the docking scores for those with no binding results were set as 0 (Table S4). All other parameters were set as default values.

2.10 Development of multi-factor analysis to balance and screen Q-markers with the multiple characteristics in JZOL

A multi-factor analysis mode was developed to balance the Q-markers related multiple characteristics and to screen the Q-markers in JZOL. The candidate Q-markers with clear herbal origin were defined as king, minister, assistant, and guide based on the chemical profile characterization and TCM compatibility theory. In addition, the network pharmacology was applied to provide a bridge to link the chemical constituents in TCM prescriptions with the corresponding targets to gain comprehensive insight into the therapeutic mechanism of multi-components. Consequently, a multi-factor analysis mode was developed by SPSSPRO (https://www.spsspro.com/mydata/index) platform, and the properties of compatibility contribution (the candidate Q-markers from King, Minister, Assistant, or Guide herb are defined as 4, 3, 2, and 1, respectively), content, pharmacokinetics (C_max and AUC_0-∞), and network pharmacology (degree value) were converted into five dimensions. However, there may be missing values among multiple dimensions. For example, although some components are highly exposed in TCM prescriptions, they may not be absorbed in vivo. Hence, the supplement of the missing value follows these principles: if the candidate Q-marker cannot be absorbed in vivo, its value is filled with 0; if the candidate Q-marker is difficult to quantify accurately, its value is filled with the corresponding average; the missing value of content is filled with the minimum value of content. All the data are then standardized as 0–1 points from minimum to maximum according to Eq. (1). The maximum variance method is used to carry out factor rotation. The KMO value (>0.5) and Bartlett's test (p ≤ 0.05) are applied to judge whether factor analysis is suitable. The number of extracted factors is determined as the main common factor through the gravel map. Based on the determined common factors at all levels, the ingredients' use obtains an eigenvalue matrix, and by calculating the weight of different common factors, the composite score (Eq.2–3) is ultimately calculated for the sample and ranked.

V is the variable of corresponding characteristic data of candidate JZOL Q-marker; D is the standardized value of variables; F₁-F_P is the score of each common factor, which is used to calculate the overall score of total factors (F); $f$ means factor coefficient, and z is observed variable, which is calculated by a linear combination of several independent factors and a unique variable (Schreiber et al., 2021). X is the weight of different common factors.

2.11 Confirmation of Q-markers based on anti-inflammatory assays

In a humidified environment at 37 °C with 5 % CO2, RAW 264.7 cells were cultured and introduced into DMEM supplemented with 10 % fetal bovine serum (FBS, Gibco, New York, USA). Cell viability was assessed using the [3-(4,5-dimethyl-thiazol-2-yle) 2,5-diphenyltetrazolium bromide] (MTT, Aladdin, Shanghai, China) colorimetric assay. RAW 264.7 cells were incubated for 24 h in 96-well plates at a concentration of 2 × 105 cells per well. Then the components were dissolved in the medium (containing 1 µg/ml LPS) and given to the test wells. The negative control was treated with complete medium (containing 0.1 % DMSO) and the positive control was treated with 1 µg/ml LPS. After 24 h treatment under normal conditions, the supernatant was collected and analyzed by a commercial NO test kit (Nanjing Jiancheng Biological Engineering Research Institute, China). The IL-6, IL-1β, and PGE₂ analysis was measured by ELISA kit (IL-6: Multisciences Biotechnology, Zhejiang, China; IL-1β: R&D system, Minneapolis, MN; PGE₂: Enzo Life Sciences, Farmingdale, NY) according to the manufacturer's instructions.

Data: Mean ± SD from three independent experiments. Statistical analysis: GraphPad PRISM v7.04, ANOVA followed by Dunnett's post hoc test, P < 0.05 considered significant.

3.1 Chemical profile of JZOL by UPLC-Q/TOF-MS analysis

The chemical constituents in JZOL were systematically investigated by searching online databases or Internet search engines (Chemical Abstracts Service (CAS) database, Massbank, Web of Science, and ChemSpider). As a result, a total of 92 compounds, including 37 flavonoids (23 flavones, 4 chalcones, 4 flavanones, 4 isoflavones, and 2 isoflavanones), 14 triterpenoid saponins, 12 alkaloids, 11 bile acids, seven diterpenoids, six anthraquinones, and five other types of compounds were identified or tentatively characterized in JZOL (Table 1; Fig. 2). Out of these, 28 peaks were further confirmed by comparison with reference standards (Fig. S2). The base peak intensity (BPI) profiles of JZOL in both negative and positive ion modes are shown in Fig. 3. The origins of the identified compounds were determined by comparing the base peak chromatograms of JZOL with those of individual single herbs (Fig. S3 and S4). Fig. 4 illustrates the detailed fragmentation and proposed fragment pathways for representative compounds. Fig. 4 displays the intricate fragmentation patterns and proposed fragment pathways of representative compounds, while additional in-depth information can be found in the Supporting Information.

Close

Fig. 2

Chemical structures of components identified or characterized in JZOL (red color represented the confirmed compounds by comparison with reference standards). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Export to PPT

Close

Fig. 3

The basic peak ions chromatography of JZOL detected by UPLC-Q/TOF-MS (red color represented the confirmed compounds by comparison with reference standards). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Export to PPT

Close

Fig. 4

The detailed fragmentation and proposed fragment pathways of compounds (A) 16-liquiritin; (B) 30-baicalin; (C) 67-glycyrrhizic acid; (D) 23-peimisine; (E) 78- cholic acid; (F) 68-rhein.

Export to PPT

3.2 Method validation of pharmacokinetic research

Based on the analysis of exogenous substances by JZOL in rats, 21 accurately identified prototype components were detected in rat plasma (Zhang et al., 2022, Table S3), 11 of which showed high exposure and covered different structural types of the compounds. These compounds included baicalin (flavonoids), wogonoside (flavonoids), wogonin (flavonoids) and oyoxylin A-7-O-β-d-gluA (flavonoids) from Scutellariae Radix, liquiritin (flavonoids), liquiritigenin (flavonoids), isoliquiritigenin (flavonoids), isoliquiritin (flavonoids), glycyrrhizic acid (triterpenoid saponins) from Glycyrrhizae Radix et Rhizoma, rhein from Rhei Radix et Rhizoma, and peimisine (alkaloids) from Fritillariae Ussuriensis Bulbus.

A set of method validation tests of 11 analytes (Fig. S5) were conducted in the matrix, including specificity and selectivity, standard curve and linear range, precision and accuracy, extraction recovery and matrix effect, and stability. Specificity testing demonstrated no or negligible chromatographic interference to the analytes (Fig. S6). All analytes exhibited good linearity (r ≥ 0.99) within the test ranges (Table S5), and both the accuracy (RE: 3.33 %-8.68 %) and precision (RSD: −2.32 %-2.16 %) of the LLOQ met the requirements of the guidelines (Table S6), indicating that the established method was sufficiently sensitive for quantifying the 11 analytes in rat plasma. As shown in Table S7, the values of intra- and inter-day precision at three concentrations (LQC, MQC, HQC) were less than 15 %. And the accuracy (RE%) of 11 analytes ranged between −10.16 % to 14.05 % for intra-day and −7.51 %–10.01 % for inter-day, demonstrating satisfactory precision and accuracy for pharmacokinetic studies. The recoveries of the analytes varied from 84.79 % to 107.94 % (glycyrrhizic acid: 10.79 %) and the recovery of IS was 100.74 %, illustrating consistent recovery and precision (Table S8). The matrix effects of the analytes ranged from 85.46 % to 111.97 % with RSDs within 8.94 %, indicating no remarkable matrix effect in rat plasma (Table S8). The carry-over effect was negligible in this method, as insignificant peaks were observed in each channel of the blank plasma sample after ULOQ (Fig. S7). The RSD values of stability under different conditions (free-thaw cycles, room temperature for 8 h, −80 °C for 2 weeks, MS Auto–sampler for 24 h) were less than 12.54 % (Table S9), indicating that the analytes remained stable under analytical conditions. The summarized results prove the validity of the developed UPLC-MS/MS method for the pharmacokinetic analysis of 11 ingredients in rat plasma.

3.3 Pharmacokinetic studies

The validated UPLC-MS/MS method for analyzing 11 components in rat plasma was applied for the pharmacokinetic study following JZOL oral administration. The mean plasma concentration–time profiles of the 11 detected prototypes are shown in Fig. 5. Statistical analysis of the pharmacokinetic parameters was performed with WinNonlin 6.3 software, and the corresponding estimated pharmacokinetic parameters (t_1/2, T_max, C_max, AUC_0-∞, AUC_0-t, MRT_0-∞, MRT_0-t) are given in Table 2.

Close

Fig. 5

Comparison of serum concentration–time curves of 11 analytes in JZOL (n = 7).

Export to PPT

According to the pharmacokinetic profiles of the 11 ingredients from JZOL, the t_1/2 value was in the range of 1.39–50.93 h. Liquiritin, wogonoside, oroxylin A 7-O-β-d-glucuronide and baicalin were rapidly absorbed, with high C_max values (138.02 ± 40.22, 343.88 ± 223.12, 98.97 ± 63.50 and 1134.15 ± 550.36 ng/mL) and high AUC_0-∞ values (1086.90 ± 344.28, 3685.52 ± 869.65, 2164.23 ± 1398.77 and 9650.72 ± 5013.90 ng*h*mL⁻¹), indicating high exposure and peak concentrations in blood. The C_max values of wogonin, isoliquiritigenin, liquiritigenin, and isoliquiritigenin ranged from 1.67 to 8.73 ng/mL. Among them, isoliquiritigenin and liquiritigenin exhibited two peaks in the pharmacokinetic profile.

Rhein derived from Rhei Radix et Rhizoma demonstrated rapid but poor absorption from the gastrointestinal tract (T_max = 0.42 ± 0.13 h, t_1/2 = 9.86 ± 2.73 h, C_max = 18.98 ± 7.19 ng/mL, AUC_0-∞=89.85 ± 46.70 ng*h*mL⁻¹, MRT_0-∞=11.70 ± 4.10 h). In this research, the peak plasma concentration (C_max) of glycyrrhizic acid was 528.59 ± 109.52 ng/mL, and its AUC_0-t reached 2201.08 ± 1040.42. Peimisine had a longer peak time (T_max = 5.37 ± 2.92 h), a lower peak value (C_max = 1.59 ± 0.46 ng/mL), and a relatively longer in vivo residence time (MRT_0-∞=53.67 ± 42.07 h).

3.4 Network pharmacology

By integrating 13 candidate components selected by chemical profile with 21 components selected by metabolite profile and PK profile, 22 candidate Q-markers were selected. As shown in Fig. 6A, an overlap of 46 targets was simultaneously involved in the interactions with 22 Q-marker candidates and the regulation of inflammation, suggesting that these targets formed a network target group (Wang et al., 2022) for the anti-inflammatory effect of JZOL. To reveal the relationship among these targets, the protein–protein interaction (PPI) network was constructed as shown in Fig. 6B and sequent subnetwork extraction showed that 12 targets in Mode 1 (BCL2L1, MAPK8, PPARG, MAPK14, HSP90AA1, IKBKB, NOS2, PTGS2, NFKB1, CASP1, GAPDH, TLR9) have had the most significant interaction association and were likely to be the core of the network target group. The further pharmacological network analysis (Fig. 6C) also suggests that these targets from a broad interaction with all 22 compounds, with targets like PTGS2 and CASP1 potentially playing a more important role by interacting with more compounds than other targets. The further signaling pathway analysis also suggested that “signaling by interleukins” and “cytokine signaling in immune system” were the top two significant pathways (Fig. 6D) and were mainly associated with most of the targets in Mode 1 (Fig. S7). As for the regulatory effect of Gypsum Fibrosum and Chloriti Lapi, it’s indicated that nine targets (CXCR4, HPSE, LTB4R, PLA2G4A, PRKCA, PRKCQ, PRKCG, PTAFR, TRPV1) among the network target group (Fig. 6A) were involved in the interaction with the ionic composition in Gypsum Fibrosum and Chloriti Lapis. All nine targets, except HPSE, were recognized for Gypsum Fibrosum and Chloriti Lapis, probably due to the common ionic composition of Ca²⁺. These targets mainly played an auxiliary role by affecting pathways such as “signaling by GPCR” (Fig. S8), which is consistent with the Chinese medicine theory that Gypsum Fibrosum and Chloriti Lapis mainly assist other herbs in exerting their function.

Close

Fig. 6

The potential regulation effect of 22 Q-marker candidates as well as Gypsum Fibrosum and Chloriti Lapis in JZOL via the network pharmacology analysis. (A) The Venn diagram for the discovery of 46 key targets from different target sets, namely targets interacting with compounds, targets related to inflammation, targets regulated by G. Fibrosum and targets regulated by C. Lapis. (B) The main target interaction mode recognized by the protein–protein interaction analysis. (C) The pharmacological network involved in the 22 Q-marker candidates in JZOL and ions in G. Fibrosum and C. Lapis. (D) The top 20 significant signaling pathways by GSEA analysis.

Export to PPT

The binding affinities between 46 targets in the network target group and 22 Q-marker candidates of JZOL were evaluated further through molecular docking, as shown in Fig. 7A. It can be seen that all compounds, except Glycyrrhizic acid and Licoricesaponin G₂, formed a direct interaction with the network target group. These two compounds might exert indirect regulation with the targets related to Gypsum Fibrosum and Chloriti Lapis, as shown in Fig. 6C. Based on the cutoff of binding affinities, the Q-marker candidates showed different levels of importance in targeting the network target group, as indicated by the statistics of target numbers (Fig. 7B).

Close

Fig. 7

The affinity evaluation between 22 Q-marker candidates and 46 key targets by molecular docking. (A) The affinity (ΔG) distribution for 22 Q-marker candidates when interacting 46 key targets were evaluated by boxplot. (B) The importance of candidates analyzed by the number of interacting targets with different affinity cutoffs (0, −5, −7, −9 kcal/mol).

Export to PPT

3.5 Candidate Q-markers prediction based on multi-factor analysis

A table filled with five factors corresponding to 22 selected components was input to the SPSSPRO website for multi-factor analysis. Firstly, through pre-computation and applicability test, the statistic Kaiser-Meyer-Olkin (KMO) values were 0.645, which were greater than 0.6, and the p values of Bartlett's Test were equal to 0.000, reaching a very significant level. This indicates that the 22 candidate Q-markers covering five factors were suitable for factor analysis (Table S10, S11). Based on the result of the explanation rate of cumulative variance after rotation, the common factors were determined as three, which explained the information of 97.1 % of the original data (X1 = 56.11 %, X2 = 20.92 %, and X3 = 20.11 %). The three common factors were mainly attributed to C_max (PK result), compatibility, and network pharmacology, indicating their importance in distinguishing candidate Q-markers. According to Eq (2)–(3), the F value for each candidate Q-marker was then calculated between −0.70 to 2.87, and component 30 (baicalin, F = 2.87), component 39 (oroxylin A 7-O-β-d-glucuronide, F = 0.44), and component 38 (chrysin-7-O-β-d-glucuronide, F = 0.41) were the top three ingredients (Table 3). It is worth noting that the F score for baicalin is almost five times that of the second chemical, preliminarily proving the importance of baicalin in JZOL, as stated in the 2020 edition of the Pharmacopoeia of the People's Republic of China.

Based on the results of the multi-factor analysis, seven components (aloeemodin-8-O-β-d-glucopyranoside, baicalin, chrysin-7-O-β-d-glucuronide, oroxylin A 7-O-β-d-glucuronide, wogonoside, chrysophanol-8-O-β-d-glucopyranoside, skullcapflavone II) were selected as candidate Q-markers of JZOL with the F > 0.

3.6 Activity evaluation

The anti-inflammatory activities of the above seven candidate Q-markers were further verified through anti-inflammatory assays, with two chemicals (16 and 79) with F＜0 selected as negative controls. Firstly, all chemicals were screened at respective maximum non-toxic concentrations. The concentrations of compounds 13, 16, 30, 38, 39, 45, 46, 74 and 79 were 50, 100, 50, 3.13, 50, 12.5, 3.13, 12.5, and 3.13 µg/ml, respectively. As shown in Fig. 8, baicalin (30), oroxylin A 7-O-β-d-glucuronide (39) and wogonoside (46) showed significant anti-inflammatory activities in inhibiting the production of inflammatory factors (NO, IL-6, IL-1β, and PGE₂) in LPS-induced RAW264.7 cells. Compounds 13 and 38 could effectively reduce the release of NO and PGE₂; compound 45 has a significant effect on reducing the release of IL-1β and PGE₂ at a lower concentration range, and compound 74 showed substantial reduction in the release of IL-6 and NO. Although compounds 16 and 79 showed no obvious therapeutic effect on the constructed cell model and evaluation indicators in this experiment, it further demonstrated that the rationality of the strategy used in this study in balancing the various properties of the Q-markers simultaneously and in screening out suitable anti-inflammatory Q-markers from hundreds and thousands of ingredients in TCMps.

Close

Fig. 8

The anti-inflammatory activity evaluation of compounds 13, 16, 30, 38, 39, 45, 46, 74 and 79 with the LPS- induced RAW264.7 cell model (‾x ± s, n = 3). (A) statistical analysis of inhibition on NO production, (B) statistical analysis of inhibition on PGE₂ production, (C) tatistical analysis of inhibition on IL-1β production, (D) statistical analysis of inhibition on IL-6 production. Cells were exposed to three different administration concentrations (low, medium, high) of compounds and LPS. ^###p < 0.001 vs. control group and ^***p < 0.001, ^**p < 0.01, *p < 0.05 vs. LPS-treated group.

Export to PPT