LC-MS/MS analysis and evaluation of the anti-inflammatory activity of components from BushenHuoxue decoction

Abstract Context: BushenHuoxue decoction (BSHXD) is a Chinese medicine prescription, which is composed of nine Chinese medical materials, used to treat osteoarthritis (OA). Objective: This study develops sensitive and convenient LC-MS/MS methods to analyze chemical components from BSHXD, and assess the anti-inflammatory activities thereof. Materials and methods: The chemical composition from BSHXD water extract was qualitative analyzed by high-performance liquid chromatography coupled with electrospray ionization quadrupole time-of-flight mass spectrometry (HPLC-ESI-Q-TOF-MS). Twelve reference compounds were analyzed by UPLC-ESI-MS/MS. Anti-inflammatory activities of target components were assessed by ELISA at 20 and 100 μg/mL. Results: It is the first time that 88 compounds were qualitatively identified from BSHXD, of which 12 with potential in treating OA according to the literature were quantified. Within BSHXD the contents of quercetin, isopsoralen, icarisideII, osthole, and isoimperatorin increased remarkably compared with those in single herb which make up BSHXD, the contents were 0.1999, 0.4634, 0.0928, 0.5364, and 0.1487 mg/g. ELISA data displayed that BSHXD and the five compounds mentioned inhibited the expressions of TNF-α, IL-6 and NO released from LPS-stimulated RAW264.7 cell, with maximum inhibition rates of 104.05% (osthole, 100 μg/mL), 100.03% (osthole, 100 μg/mL), and 93.46% (isopsoralen, 20 μg/mL), respectively. Discussion and conclusion: Content changes of 12 compounds in BSHXD and single herbs which comprise the prescription were measured and analyzed. Contents of five compounds increased may be explained by solubilization between drugs and chemical reaction. ELISA results reported that the increased contents of the five compounds could inhibit expression of the inflammatory factors.

OA which is also called hypertrophic arthritis or degenerative arthritis was frequently occurred and refractory. It is expressed by arthralgia, stiffness, and deformity (Zhang et al. 2007;Juhl et al. 2014). Depending on the basic theories of Chinese medicine, it is a type of bone obstruction disease. The principal pathological symptom of OA is lesion of cartilage tissue. Cytokine plays an important role in the pathogenesis of OA by promoting catabolism of cartilage matrix (Schable 2014;Yang et al. 2014;Cornejo et al. 2015).
High-performance liquid chromatography coupled with electrospray ionization quadrupole time-of-flight mass spectrometry (HPLC-ESI-Q-TOF-MS) was established for qualitative analysis on chemical compounds of BSHXD. Eighty-eight compounds were confirmed by comparing its retention time and MS spectrum with the corresponding reference compound. Through the review of literature, 12 active compounds with potential in treating OA and inhibiting chondrocyte apoptosis were found from these 88 compounds. UPLC-ESI-TQ-MS method was also established to simultaneously determine and compare the contents of these 12 compounds in BSHXD and in single herb decoction. The results showed that the contents of five compounds, which were quercetin, isopsoralen, icariside II, osthole, and isoimperatorin, increased remarkably in BSHXD. To evaluate the antiinflammatory activity and explore the therapeutic mechanism of BSHXD against OA, the effect of above five compounds on TNF-a, IL-6, and NO released by macrophage was assayed.
Methanol and acetonitrile were of HPLC grade and purchased from Hanbang Technology Co., Ltd. (Jiangsu, China). Ultra-pure water was obtained by the EPED super-purification system (Nanjing EPED Co., Ltd, Nanjing, China). All other chemicals and solvents used in this study were of analytical grade.  (Thermo Scientific SERIES II  WATER JACKET), and enzyme-labelling instrument (Thermo Scientific, Waltham, MA).
The quantitative analysis was performed on a Waters-Xevo TQ, tri-stage quadrupole mass spectrometer system (Waters, Milford, MA) equipped with an ESI source, and the ESI source was set in positive and negative mode. The scanning mode was established in MRM mode. The capillary voltage was 3 kV, ion source temperature was 150 C, the dry gas flow was 1000 L/h, dry heater was 550 C, the cone gas flow was 50 L/h, and the collision gas flow was 0.15 mL/min.

Preparation of sample solutions for LC-MS/MS analysis
Angelicae Pubescentis Radix 400 g, Taxilli Herba 400 g, Achyranthis Bidentatae Radix 600 g, Epimedii Folium 600 g, Angelicae Sinensis Radix 600 g, Chuanxiong Rhizoma 600 g, Paeoniae Radix Alba 600 g, Polygoni Cuspidati Rhizoma Et Radix 600 g, and Arisaematis Rhizoma Preparatum 600 g were mixed together. The mixture was decocted with 30 L water three times (1, 1, and 0.5 h). The filtrates from each decoction were consolidated and concentrated to 1000 mL, and added 1100 mL 95% ethanol, suspension was centrifuged at 5000 rpm for 10 min after 48 h standing. The supernatant was dried on a water bath, and dissolved 0.0001 g dried supernatant in 50% methanol to 100 mL, then filtered through 0.22 lm membrane filter to produce stock solutions. The reference compounds were accurately weighed and dissolved in 50% methanol to produce stock reference solutions. The above stock solutions was stored at 4 C and brought to room temperature before use.

Cell culture
The murine macrophages RAW264.7 were cultured in 1640 culture medium with 10% FBS in an incubator containing 5% CO 2 at 37 C. The cells were digested with trypsin when they grew to an appropriate amount, and placed into the wells of a 48-well plate at 200 lL/well (1 Â 10 5 cells/well) and cultured for 24 h in the sterilized incubator.

Grouping and administration
LPS was diluted with 1640 culture medium to 500 ng/mL, BSHXD and compound samples were dissolved in DMSO at 1 Â 10 5 lg/mL and diluted to 100 and 20 lg/mL with the culture medium. Above samples were added to each of control group, LPS group, and sample groups (LPS þ sample 20 lg/mL, LPS þ sample 100 lg/mL) in an amount of 300 lL. The 48-well plate was incubated another 24 h in the sterilized incubator.

Sample detection
The supernatant was collected after centrifuging at 10,000 rpm for 2 min at 4 C. The supernatant was treated and the absorbance was measured at 450 nm (TNF-a, IL-6) and 540 nm (NO) using an enzyme-labelling instrument according to the instruction of the manufacturer, respectively. Cytokine concentrations were calculated by a calibration curve prepared by standard concentrations as X-axis, and OD values as Y-axis.

Data analysis
, C is the cytokine concentration. All samples were assayed in triplicate. The results were presented as means ± standard deviations (SD). Statistical analyses were performed using a one-way analysis of variance ANOVA test (SPSS v.16.0, SPSS Inc., Chicago, IL), followed by Student's two-tailed unpaired t-test. p < 0.05 was considered as the statistically significant.

Optimization of ion condition
In order to determine appropriate ion condition for Q-TOF and TQ mass spectrometer, all the analytes were detected under affuse mode, and the fragment ions were automatically collected under MRM scanning mode, with optimal cone voltage and collision energies.

Selection of mobile phases
A series of experiments were carried out with different mobile phases, for example methanol/water, methanol/acetonitrile-water, methanol/acetonitrile-0.1% formic acid-water, methanol/ acetonitrile-0.05% aqueous formic acid. It turned out that the best chromatographic peak was obtained when using methanol-0.1% aqueous formic acid solution and 0.1% aqueous formic acid solution-acetonitrile as the mobile phases for HPLC-ESI-Q-TOF-MS and UPLC-ESI-TQ-MS, respectively.

Qualitative analysis
Qualitative analysis on compounds in BSHXD was achieved by HPLC-ESI-Q-TOF-MS, the total ion flow chart of positive and negative ion modes is shown in Figure 1. These peaks showed different molecular ion the MS 2 spectra, which exhibited a fragmentation pathway. According to measured molecular weight, theoretical molecular weight, fragment ion, elemental analysis, and compared with relevant literature data, 88 compounds from BSHXD were detected and confirmed. The mass error for molecular ions was in ±5 ppm. The detailed spectral data are presented in Table 1. The results provided some evidence of the material basis for the BSHXD.

Compounds selection for quantitative analysis
As shown in the literature, catechin could reduce inflammation and slow cartilage breakdown (Adcocks et al. 2002), paeoniflorin could down regulate the levels of TNF-a and myeloperoxidase, and reduce the production of IL-6 in LPS-simulated mouse macrophage RAW264.7 cells ). Other research indicated that paeoniflorin inhibited intercellular adhesion molecule-1 expression in LPS-treated U937 cells and TNF-a-stimulated human umbilical vein endothelial cells by suppressing the activation of the NF-jB pathway (Jin et al. 2011). Hyperoside could significantly decrease the mRNA expression and production of IL-1b, IL-6 in stimulated HMC-1 cells (Han et al. 2014). Ferulic acid may offer beneficial effects against osteoarthritis (Li et al. 2011), and ferulic acid has  chondroprotective effects on hydrogen peroxide-stimulated chondrocytes through depressing hydrogen peroxide-induced pro-inflammatory cytokines and metalloproteinase gene expression at the mRNA level (Chen et al. 2010). Polydatin has efficacious anti-inflammatory activity by attenuating the phosphorylation of ERK1/2, JNK, and p38 (Lou et al. 2015).
Quercetin could ameliorate all markers of inflammation (Gardi et al. 2015). Research suggested that MAPK signaling factors were involved in inflammation, quercetin inhibited the MAPK signal factors in macrophages, and quercetin also inhibited the secretion of the inflammatory cytokines IL-1b, IL-6, and stimulated the anti-inflammatory cytokine IL-10 (Seo et al. 2015). Based on the research, proinflammatory cytokines in the cartilage and synovium will stimulate their own production and induce chondrocytes to produce some abnormal biomechanical forces, such as proteases, chemokines, and nitric oxide, which will result in an imbalance between the chondrocyte anabolic and catabolic pathways, and ultimately leads to progressive joint destruction. Resveratrol keeps chondrocyte from apoptosis and reverses the catabolic state of chondrocytes in OA pathway (Dave et al. 2008). Due to its antiapoptotic, anti-inflammatory, and antioxidant properties, resveratrol have anti-osteoarthritic effects (Shen et al. 2012). Inflammatory cytokine IL-1b is one of the key inflammatory factors in intervertebral disc degeneration, psoralen could remit the degeneration of intervertebral disc chondrocyte induced by IL-1b (Yang et al. 2015), it also significantly suppressed T helper 2 cytokines of IL-4, IL-5, and IL-13 by ConA-stimulated D10 cells without inhibitory effect on cell viability (Jin et al. 2014). Isopsoralen was being used for its central inhibitory activities and inhibitory role in cell proliferation and antimicrobial (Liu et al. 2013). Icariside II exhibits anti-inflammatory activity, but its molecular pathways in human cells are poorly understood . Osthole can prevent isoprenalin-induced myocardial fibrosis in mice, and the mechanisms perhaps related to the reduction of TGF-b1 expression (Chen et al. 2011), osthole could also enhance osteoclasts apoptotic and inhibit the bone resorption through RANK þ RANKL/TRAF6/Mkk/JNK signal pathway (Ming et al. 2012). TNF-a is a major inflammatory cytokine that mediates immune responses and systemic inflammation. Isoimperatorin inhibits TNF-a-induced expression of VCAM-1, therefore, isoimperatorin can be used for the treatment of pathologic inflammatory disorders (Moon et al. 2011). Through the review of the literature, we selected above described 12 active compounds with potential in treating OA and inhibiting chondrocyte apoptosis from these 88 compounds above, which were identified by HPLC-ESI-Q-TOF-MS for qualitative analysis. Chemical structures of these 12 compounds are shown in Figure 2.

Quantitative analysis
UPLC-ESI-TQ-MS technology was used to determine the contents of catechin, paeoniflorin, hyperoside, ferulic acid, polydatin, quercetin, resveratrol, psoralen, isopsoralen, icariside II, osthole, and isoimperatorin. LC-MS/MS chromatogram of the 12 compounds in BSHXD is given in Figure 3, and their contents in the single herbs and decoction were summarized in Table 2. The standard curves and linear ranges of these 12 compounds are shown in Table 3. The precision and the accuracy were validated by the determination of the peak areas of compounds of interest during the preparation procedure. Relative standard deviation (RSD) of each compound was lower than 3%, the results showed that the method displays good precision and accuracy for each compound. The stabilities of these 12 compounds were tested at 0, 2, 4, 6, 12, and 24 h, RSD values of the peak areas were all no more than 3%, that revealed that the compounds were stable within 24 h. BSHXD sample was made into six solutions, inject 2 lL each time, and MS chromatography was used to determine the contents of each compound. The average content of catechin was 0.0411 mg/g, paeoniflorin was 4.2455 mg/g, hyperoside was 0.1199 mg/g, ferulic acid was 0.0576 mg/g, polydatin was 0.7589 mg/g, quercetin was 0.2001 mg/g, resveratrol was 0.1567 mg/g, psoralen was 0.8863 mg/g, isopsoralen was 0.4598 mg/g, icariside II was 0.0919 mg/g, osthole was 0.5274 mg/g, and isoimperatorin was 0.1490 mg/g, and RSD values were all less than 3%, indicating that the method was stable.

ELISA results
TNF-a, IL-6, and NO play an important role in cells, tissues, and organs, respectively. Within cartilage, pro-inflammatory cytokines such as TNF-a auto-catalytically stimulate its production and induce chondrocytes to produce additional catabolic mediators that abnormal biomechanical forces will lead to progressive joint destruction (Abramson & Yazici 2006). IL-6-signal transducer may conduce to the posttraumatic development of osteoarthritis (Liu et al. 2015), and IL-6 can be considered as a marker of nerve injury and proinflammatory cytokines which produced by joint tissue (Malek et al. 2015). Imbalance of catabolic and anabolica factors including cytokines and NO could result in OA (Chevalier et al. 2013). In that case, TNF-a, IL-6, and NO were utilized to explore the mechanism of BSHXD in treating OA. Changes of released inflammatory mediators' concentration were observed by ELISA. It is shown that quercetin, isopsoralen, icariside II, osthole, isoimperatorin, and BSHXD have different effects in inhibiting the release of TNF-a, IL-6, and NO. Quercetin (100 lg/mL), isopsoralen (100 lg/mL), icariside II (100 lg/mL), osthole (20 and 100 lg/mL), isoimperatorin (100 lg/mL), and BSHXD (100 lg/mL) had a significant inhibition effect on the release of TNF-a, p < 0.01; quercetin (20 and 100 lg/mL), isopsoralen (100 lg/mL), icariside II (100 lg/mL), osthole (20 and 100 lg/mL), and isoimperatorin (20 and 100 lg/mL) had a significant inhibition effect on the release of IL-6, p < 0.01; five compounds (20 and 100 lg/mL) and BSHXD (100 lg/mL) had a remarkable inhibition effect on the release of NO, p < 0.01. The results showed that the monomers hold generally stronger inhibition effect than BSHXD. Tables 4-6 describe the detailed inhibition results of these components of decoction on TNF-a, IL-6, and NO released by RAW264.7 cell after induction of LPS.

Conclusion
According to LC-MS/MS analysis, 88 compounds from BSHXD were confirmed. Twelve compounds which have a potential role in treating OA were selected and quantified. By comparing the contents of 12 compounds in BSHXD and single herbs, we found that five of them increased significantly. Therefore, the antiinflammatory activity in vitro was tested. ELISA was used to detect the effect of quercetin, isopsoralen, icariside II, osthole, isoimperatorin, and BSHXD on TNF-a, IL-6, and NO released by macrophage, we found that the compounds had some or remarkably inhibitory effect on the former cytokines, which may demonstrate the possible reason and mechanism of BSHXD in treating OA. In traditional Chinese medicine theory, the fact that different herbs used in combination can enhance the therapeutic efficacy compared with those when they were used separately is called 'Xiang Xu'. In BSHXD, nine herbs which have different therapeutic effects were decocted together, and the results of UPLC-ESI-TQ-MS showed the contents of compounds of interest increased or decreased, which may be due to certain chemical reactions occurred among the chemical constituents in the herbs. That means that the contents of some compounds which have potential therapeutic effects on OA were higher in decoction than in single herb that may give rise to reinforcement of therapeutic effects on OA.