Screening of Anti-Inflammatory Components of Qin Jin Hua Tan Tang by a Multivariate Statistical Analysis Approach for Spectrum-Effect Relationships

Qing Jin Hua Tan Tang (QJHTT) exerts therapeutic effects in patients with chronic obstructive pulmonary disease (COPD) by alleviating inflammation. However, the anti-inflammatory components of QJHTT have not yet been reported. Our study aimed to screen the active anti-inflammatory components of QJHTT using a multivariate statistical analysis approach for spectrum-effect relationships. Different polar fractions of QJHTT were prepared using ethanol, ethyl acetate, and n-butanol to analyze the phytochemical components. Phytochemical fingerprints were generated using ultrahigh-performance liquid chromatography. In total, 24 peaks were observed in ten batches of QJHTT extracts. The anti-inflammatory activity was evaluated using a xylene-induced ear-swelling mouse model. Additionally, the spectrum-effect relationship between the relative areas of the 24 peaks and pharmacological activity was investigated using multivariate statistical analysis. The potential anti-inflammatory ingredients obtained from the screening (multivariate statistical analysis) will be validated for their anti-inflammatory effects and mechanisms utilizing a lipopolysaccharide-induced macrophage inflammation model. QJHTT ethanol extract 1 exhibited good anti-inflammatory activity. Peaks 11, 12, 13, 14, and 16, which were closely correlated with anti-inflammatory activity, were identified as meranzin, baicalin, baicalein, chrysin-7-O-β-D-glucuronide, and wogonoside, respectively. The anti-inflammatory activities of meranzin, baicalin, baicalein, and wogonoside were verified in vitro. These four bioactive components significantly inhibited the secretion of inflammatory factors in the lipopolysaccharide-stimulated macrophage cell line. This research successfully screened the QJHTT anti-inflammatory active ingredient group. Meranzin, baicalin, baicalein, chrysin-7-O-β-D-glucuronide, and wogonoside were predicted to be the anti-inflammatory active ingredient groups of QJHTT.


Introduction
Qing Jin Hua Tan Tang (QJHTT), a traditional Chinese medicine (TCM) prescription (CMP), is widely used to treat phlegm heat cough syndrome, which is characterized by cough, pulmonary carbuncle, and pulmonary distension. In modern medicine, symptoms of phlegm heat cough syndrome correspond to those of chronic obstructive pulmonary disease (COPD) [1]. QJHTT is composed of the following 11 herbs: Scutellariae Radix, Gardeniae Fructus, induced inflammation in pulmonary epithelial cells (A549 cells) [4,5]. Isosteroid alkaloids isolated from Fritillariae Cirrhosae Bulbus exerted anti-inflammatory effects in RAW 264.7 cells by regulating MAPK phosphorylation [6]. In 2018, QJHTT was selected for the first batch of the "Classical Chinese Medicine Classical Directory," as it exhibited potent pharmacological activities against COPD. e anti-inflammatory activity of QJHTT has been previously reported. However, the components contributing to the anti-inflammatory activity of QJHTT have not been elucidated.
Activation of the inflammatory response can contribute to the progression of chronic diseases, such as COPD and metabolic diseases [7]. COPD is associated with local pulmonary inflammation and systemic inflammation, which are mainly localized in the peripheral airways and lung parenchyma [8]. Inflammation-induced fibrosis and peripheral airway collapse lead to airway obstruction, which contributes to the development of cough and asthma [9]. e major treatment for COPD is the administration of anti-inflammatory agents. NF-κB, which is a vital regulator of the inflammatory response, modulates the transcription of IL-6 and TNF-α [10]. QJHTT is a classical TCM for the treatment of COPD; however, the lack of clarity regarding the active ingredients significantly limits the clinical application of QJHTT.
e traditionally used dosage form is a decoction characterized by large doses of medicine and inconvenience in carrying. erefore, screening of the active group of QJHTT can overcome these disadvantages and be helpful for secondary development. In contrast to those in singleherbal medicines, bioactive components in multiherbal formulations are identified based on the correlation between different original single-herbal chemical fingerprints and efficacy indices. QJHTT comprises 11 traditional Chinese herbal medicines. Hence, the different origins of 11 herbal correlation analyses would be laborious work to identify the anti-inflammatory ingredients of QJHTT. Herein, we constructed fingerprint profiles of the different polar sites of QJHTT extracts, evaluated their anti-inflammatory activity, and screened their active anti-inflammatory ingredients using a spectrum-effect relationship analysis. To date, spectrum-effect relationship analysis has been successfully used to screen the active ingredients in single-herbal medicine and CMP [11]. e spectrum-effect relationship between chemical fingerprints of CMP and the bioactivity index was examined using multivariate statistical analysis [12,13]. Modernization of CMP analysis has been achieved through the application of multivariate statistical models, such as gray relational analysis (GRA) and partial least squares regression (PLSR), to analyze the spectrum-effect relationship [14]. In addition, redundancy analysis (RDA) was first applied to the spectrum-effect relationship, which maps the ingredients and bioactivity index onto the same two-dimensional ordination diagram, allowing for a visual representation of the correlation between ingredients and the bioactivity index. Systematic screening of QJHTT anti-inflammatory active ingredients was conducted using a combination of three statistical analysis methods.
In this study, the polar components of QJHTT were extracted using different solvents. e anti-inflammatory components of QJHTT could be rapidly and efficiently screened based on the correlation between the concentration of different polar components and their anti-inflammatory activities. Furthermore, the potential anti-inflammatory components were validated using an in vitro screening test.

Animals.
Male Kunming mice (4-6 weeks with an average body weight of 25-30 g) were obtained from the Center of Laboratory Animal Science of Guangdong (quality exequatur number: SCXK 20180002). e animals, which were housed under standard temperature, humidity, and light conditions, had free access to water. e mice were allowed to fast for 12 h prior to the experiments. All procedures were approved by the Institutional Animal Use Committee (Ethics Committee of ZSSOM on Laboratory Animal Care, No. 2020-0112).
Mass spectrometry was performed in both positive and negative ion modes with the following parameters: gas temperature, 300°C; drying gas flow rate, 10 L/min; sheath gas temperature, 365°C; sheath gas flow rate, 12 L/min; nebulizer gas pressure, 45 psi. Potential active compounds were identified using a Q-TOF analyzer. e data were collected and processed using Mass Hunter software (version B.07.00).

Methodological
Validation. EAWE-1 was repetitively injected with 20 µL five times to confirm the precision of the instruments. e RSD values of the major chromatograph peak area were all less than 2.0%, and the RSD values of the relative retention times of the major chromatograph peaks were all below 2.0%, suggesting that the instrument had good precision. EAWE-1 was analyzed at 0, 2, 4, 8, 12, and 24 h to evaluate the stability of the sample at room temperature. e RSD values of the major chromatograph peak area at different times were all less than 5.0%, suggesting that the sample was stable within 24 h. Five batches of EAWE-1 were prepared following the method described in Section 2.3. e RSD values of the major chromatograph peak area of the five batches of EAWE-1 were all below 5%, suggesting that the method had good repeatability.

Anti-Acute Inflammatory Effect of Different Polar QJHTT
Extracts. Mice were randomly divided into the following groups: model, positive control, and extract-treated groups. e mice were administered the target drugs by gavage once a day for seven consecutive days. e treatment regimens for different groups (7 mice/group) were as follows: model group, administered with 0.5% sodium carboxymethylcellulose (CMC-Na) (0.2 mL/25 g bodyweight/day); positive control group, administered with dexamethasone acetate (0.54 mg/kg bodyweight/day); low-dose extract-treated group, administered with extracts at a dose of 200 mg/kg bodyweight/day; high-dose extract-treated group, administered with extracts at a dose of 400 mg/kg bodyweight/day. At 1 h after drug administration, 20 μL xylene was applied to the right ear (but not the left ear) of the mice in all treatment groups for 30 min. e mice were sacrificed by cervical dislocation under ether anesthesia, and two ear punches were collected. e difference in the weight of the left and right ear punches was calculated using the weight difference (WD). e swell inhibition rate (SIR) was calculated using the following formula [15]: SIR % � (1 − WD of the extract-treated group/WD of the model group) × 100.

Spectrum-Effect Relationship Analysis
2.7.1. GRA. GRA was performed using the Data Processing System (DPS V15.10) to examine the correlation between the compounds and pharmacodynamics. e relative areas of the 24 peaks (P1-P24) were standardized before performing the GRA. e standardized data were selected as the subsequences, and the SIR levels were defined as the parent subsequence [16]. e resolution coefficient (ξ) was 0.5, after calculating the gray relational coefficients and grades for 24 peak areas.

PLSR Analysis. PLSR analysis was performed using
Simca-P13.0 (Umetrics, Sweden) to establish the statistical correlation between multiple dependent and independent variables. In this study, 24 chromatographic peaks were used as independent variables, while SIR was used as a dependent variable to generate a regression model. e correlation between the peak areas and SIR was analyzed using the PLSR model [17].

RDA for Screening Anti-Inflammatory Bioactive
Components. RDA was performed to further demonstrate the direct correlation between the component peak area and bioactivity. is is a direct gradient analysis method that enables a statistical evaluation of the correlation between one or a set of variables and another set of multivariate data [18,19]. In this study, RDA was performed using CANOCO 4.5, to analyze the correlation between the 24 peak areas and efficacy indices (anti-inflammatory activity).

UHPLC Analysis and Peak Identification.
e chromatograms of 10 samples (WE-1, WE-2, EAWE-1, EAWE-2, NBWE-1, NBWE-2, AE-1, AE-1, EAAE-1, and EAAE-2) were obtained using UHPLC. e mass spectrum peaks in both positive and negative ion modes were identified based on the accurate mass and precursor ions along with the corresponding secondary fragment ion. e bioactive phytochemical components of the 10 samples with potent anti-inflammatory activity were identified. Standard reference controls for the identified components were used for further analysis. e cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin solution in a humidified chamber at 5% CO 2 and 37°C.
2.9.3. ELISA Test. RAW264.7 cells (5 × 10 4 cells/well) were cultured in 96-well plates for 12 h and treated with meranzin, baicalin, baicalein, or wogonoside (20 µM) for 2 h. Cells in the model group were treated with an equal volume of basal medium. Next, the cells were stimulated with 1 µg/mL LPS for 24 h. e LPS-induced production of TNF-α and IL-6 in the culture supernatant was analyzed using an ELISA kit according to the manufacturer's instructions [20].

Western
Blotting. RAW264.7 cells (1 × 10 6 cells/well) were cultured in six-well plates for 12 h and treated with meranzin, baicalin, baicalein, or wogonoside (20 µM) for 2 h. Cells in the model group were treated with an equal volume of basal medium. Next, the cells were stimulated with 1 µg/mL LPS for 24 h. e cells were washed twice with ice-cold phosphate-buffered saline (pH 7.5) and subjected to nuclear and cytoplasmic protein extraction using a nuclear extraction kit (Beyotime, China). Protein concentration was determined using a bicinchoninic acid protein assay kit (Beyotime, China). Equal amounts (10 µg) of protein were subjected to sodium dodecyl sulfatepolyacrylamide gel electrophoresis on a 10% gel. e resolved proteins were transferred onto polyvinylidene difluoride membranes.

Statistical Analysis.
All data are expressed as the mean ± standard deviation. Statistical significance was analyzed by one-way ANOVA and a least significant difference (LSD) test for multiple comparisons. All statistical analyses were performed using SPSS (version 12.0; IBM, USA).

UHPLC Chromatogram of Different Polar Fractions of QJHTT Extracts.
e UHPLC-UV chromatogram of NBWE-1 is shown in Figure 1. Analysis of the UHPLC-UV chromatograms (360 nm) of the ten QJHTT extracts (Figure 2) revealed variations in the phytochemical components. In total, 24 peaks were identified in all 10 extracts. e relative retention times and areas of the 24 peaks are shown in Table S1.

Effect of QJHTT Extracts on Xylene-Induced Ear Swelling in Mice.
e weight of the ear punch (with a diameter of 6 mm) was measured after the administration of QJHTT or dexamethasone acetate following xylene treatment. As shown in Table 2  e correlation coefficient is classified in three grades (>0.8 means strong correlation, 0.7-0.8 means moderate correlation, and <0.7 means low correlation) [21]. As shown in Table 3, the coefficient of correlation of peaks 13, 16, 14, 11, and 12 with the antiinflammatory activity of QJHTT was higher than 0.7, which indicated a moderate correlation. e coefficient of correlation of peaks 9,6,21,7,24,19,4,17,5,1,20,8,2,18,10,22, and 3 with the anti-inflammatory activity of QJHTT was between 0.6 and 0.7, which indicated a low correlation. ese findings indicate that QJHTT contains multiple components that synergistically exert beneficial effects. Peaks 11,12,13,14, and 16 were the top four ranked peaks correlated with anti-inflammatory activity and hence may be the major bioactive components of QJHTT.

PLSR Analysis.
In the PLSR model analyzing 24 chromatographic peak areas and SIR, the first four principal components explained the variances of the x and y variables, which were 96.3% and 97.2%, respectively, and the variance of Q2 was 0.823.

Four Potential Anti-Inflammatory Components Inhibit the Nuclear Translocation of NF-κB.
NF-κB mediates the inflammatory process in COPD by enhancing the transcription of proinflammatory cytokine-encoding genes [24]. In this study, the effects of meranzin, baicalin, baicalein, and wogonoside on the nuclear translocation of NF-κB were examined. As shown in Figures 10(a) and 10(b), treatment with meranzin, baicalin, baicalein, and wogonoside alleviated the LPS-induced downregulation of cytoplasmic levels of NF-κB. All components, except wogonoside, increased the nuclear levels of NF-κB (Figures 10(a) and 10(d)). Furthermore, the level of IκB-α in the meranzin-treated, baicalin-treated, baicalein-treated, and wogonoside-treated cells was markedly higher than that in the LPS-treated group. is indicated that the degradation of IκB-α was inhibited by meranzin, baicalein, and wogonoside (Figures 10(a) and 10(c)). ese findings  suggest that meranzin, baicalin, and baicalein may inhibit the nuclear translocation of NF-κB by inhibiting the degradation of IκB-α. Baicalin, baicalein, and wogonoside inhibition of NF-κB activation inhibits LPS-induced inflammatory factor production, which is consistent with results reported in previous studies [25][26][27].

Discussion
A single TCM contains hundreds of phytochemical components, while CMP is composed of two or more kinds of TCM. us, identification of the active components in these complex formulations is a major challenge. Recently, the   spectrum-effect relationship has been used to identify the active components of TCM formulations. In this study, the anti-inflammatory components of QJHTT were screened using GRA, PLSR, and RDA. e findings of this study revealed that multiple components exhibited anti-inflammatory activities. Moreover, the different active components of CMP exhibited variable pharmacological activities. Peaks 11,12,13,14, and 16 were strongly correlated with antiinflammatory activity.
is result was validated using all three spectrum-effect analysis methods. Comparing the results of GRA and PLSR of our research data with the published reference, as shown in Table S4, the consistency of the GRA and PLSR results in this study is better than in other similar published articles. In this study, both the GRA and PLSR results indicated that peaks 11, 12, 13, 14, and 16 were highly correlated with anti-inflammatory activity. However, in four of the other five published spectrum-effect relationship analysis articles, the GRA and PLSR results were not consistent.
erefore, multivariate statistical analysis is a reliable method for screening the active components of QJHTT. Peaks 11,12,13,14, and 16 are potential anti-inflammatory components.
TCM formulations are natural combinations of many components. Screening of the active components in the TCM formulation can aid in identifying a new combination of drugs to improve the therapeutic efficacy of the formulation. Spectrum-effect relationship analysis can aid in screening a combination of drugs. Combination drugs have recently been proposed for the secondary development of TCM [28]. Based on the results of this study, we would further investigate the synergistic anti-inflammatory effects of the compounds meranzin, baicalin, baicalein, chrysin-7-O-beta-D-glucuronide, and wogonoside. Drug combinations can achieve better efficacy and fewer side effects due to the synergistic effect compared to single components [29]. Considering that combination drugs are composed of two or more drugs, the screening of their appropriate drug combinations is a great challenge [30]. is study provides a foundation for drug combinations of screened active ingredients. In contrast to the traditional method of spectrumeffect analysis, where different original herbs are collected for spectrum-effect analysis, the screened potential active ingredients are less accurate because of their high similarity and minor differences in composition between samples. In this study, extraction with different polar solvents yielded most of the components in the original herbs. is made the content of the components vary significantly between groups, which is more conducive to screening potential active ingredients with higher accuracy.
QJHTT is a classic CMP used to treat pulmonary inflammation and lung fever-induced cough. However, the active components of QJHTT have not been identified, which has significantly limited its clinical application. In this study, the anti-inflammatory bioactive components of QJHTT were identified using a multivariate statistical analysis. Among the five potential components with antiinflammatory activity, meranzin exhibited anti-inflammatory activity in a xylene-induced ear-swelling mouse model [31]. Baicalin and baicalein have been reported to exhibit antioxidant and anti-inflammatory activities in diabetes, cardiovascular disease, and inflammatory bowel diseases by attenuating the activity of NF-κB and inhibiting the secretion of several inflammatory cytokines [32]. Meranzin is a component of Citri Grandis Exocarpium, while baicalin, baicalein, and chrysin-7-O-β-D-glucuronide are components of Scutellariae Radix. Additionally, wogonoside is a component of Scutellariae Radix. Fritillariae unbrgii Bulbus, Scutellariae Radix, and Citri Grandis Exocarpium are used to treat lung fever-induced cough (induced by the inflammation of the lungs and bronchial tubes). e antiinflammatory components predicted using spectrum-effect relationship analysis were experimentally validated in vitro. Meranzin, baicalin, baicalein, and wogonoside mitigated LPS-induced inflammation and enhanced the secretion of inflammatory factors, such as TNF-α, IL-4, and IL-6 in macrophages. Meranzin is the first reported inhibitor of LPS-induced inflammation in macrophages. Nuclear translocation of NF-κB (p65) from the cytoplasm promotes the transcription and secretion of inflammatory factors by binding to the promoter region of their genes. e degradation of IκB-α, an inhibitor of NF-κB, activates the two subunits of NF-κB, which translocate to the nucleus and initiate the expression of inflammatory factors [33]. ese results indicated that meranzin, baicalein, and wogonoside inhibited the nuclear translocation of NF-κB by inhibiting the degradation of IκB-α. Although baicalin inhibited the nuclear translocation of NF-κB, it did not markedly promote the degradation of IκB-α. ese results indicate the reliability of the potential active components obtained from the screening.

Conclusions
In this study, we successfully screened the active anti-inflammatory components of QJHTT using a multivariate statistical analysis method. e method involved analyzing the spectrum-effect relationship of different polar extracts of QJHTT using three correlation models. e active components (meranzin, baicalin, baicalein, chrysin-7-O-beta-Dglucuronide, and wogonoside) mediating the anti-inflammatory effects of QJHTT were identified. Meranzin, baicalin, baicalein, and wogonoside inhibited the nuclear translocation of NF-κB and consequently inhibited the activation of inflammatory pathways. Further studies are needed to explore the differences in the efficacy of anti-inflammatory components of CMP.