Chemical Identification and Antioxidant Screening of Bufei Yishen Formula using an Offline DPPH Ultrahigh-Performance Liquid Chromatography Q-Extractive Orbitrap MS/MS

Chronic obstructive pulmonary disease (COPD) has high morbidity and mortality and presents a threat to human health worldwide. Numerous clinical trials have confirmed that Bufei Yishen formula (BYF), an herbal medicine, can alleviate the symptoms of COPD by reducing oxidative stress-mediated inflammation. However, the active components of BYF remain unclear. We developed an efficient ultrahigh-performance liquid chromatography Q-Extractive Orbitrap mass spectrometry method to identify the composition of BYF and determine its antioxidant profile through an offline screening strategy based on 1,1-diphenyl-2-trinitrophenylhydrazine (DPPH)-liquid chromatography-mass spectrometry. In total, 189 compounds were identified in BYF extract, including 83 flavonoids, 24 lignans, 20 alkaloids, 15 saponins, 11 terpenoid, 10 saccharides, eight lipids, seven organic acids, two coumarins, two amino acids, and seven other compounds. Among them, 79 compounds were found to have a potential antioxidant activity. In vitro validation indicated that the free radical scavenging activities of rosmarinic acid and calycosin were similar to that of the positive control (DPPH IC50 = 25.72 ± 1.02 and 147.23 ± 25.12 μg/mL, respectively). Furthermore, calycosin had a high content in serum after the oral administration of BYF, indicating that calycosin might be the major antioxidant compound in BYF.


Introduction
Chronic obstructive pulmonary disease (COPD) is a major chronic disease with high morbidity and mortality, especially among the elderly and smokers [1]. According to the World Health Authority and Global Initiative for Chronic Obstructive Lung Disease, bronchodilators and corticosteroids along with nonpharmacologic therapies such as pulmonary rehabilitation are frequently used to treat COPD. Although these therapies can reduce exacerbations and alleviate symptoms, there is little evidence to suggest that they can suppress the progression of COPD. Various recent clinical trials have suggested that herbal medicines have the potential to improve symptoms, reduce the frequency of acute exacerbation, and improve the quality of life for COPD patients [1]. Bufei Yishen formula (BYF) is an oral prescription for COPD that has proven clinically effective for COPD control [2]. e use of BYF and BYF combined with other therapies (e.g., acupoint sticking, electroacupuncture, and Tongsai granules) have shown beneficial effects in terms of lung function, clinical symptoms, quality of life, and acute exacerbation frequency in patients with stable COPD [3][4][5].
In a COPD rat model exposed to cigarette smoke and bacteria, BYF also ameliorated airway inflammation and remodeling [6][7][8][9][10][11]. We previously demonstrated that the mechanism of BYF against COPD might involve reducing inflammatory cytokines and oxidative stress, regulating immune response and lipid metabolism [12][13][14][15], restoring the 17/Treg balance by activating adenosine 2a receptor [16], modulating the activities of STAT3 and STAT5 in COPD rats [17], and suppressing interleukin expression and/or secretion [18]. However, the effective substances of BYF are not clear at present, which forms a bottleneck problem in the further development of the preparation. It is well known that the ingredients of traditional Chinese medicine are complex, the unclear of effective substances make it difficult to select the biomarkers for quality control.
Oxidative stress triggering sustained inflammatory response is a major contributing factor in COPD [19,20]. A system analysis integrating transcriptomics, proteomics, and metabolomics showed that the target proteins of BYF against COPD are glutamate-cysteine ligase, glutathione reductase, G6PD, glutathione S-transferase P, glutathione S-transferase A1/2, GSTM1/2, and SOD1, which are predominantly enriched in oxidative stress-related pathways [15]. We speculated that the antioxidant profiling of BYF might help uncover the effective substances of BYF.
At present, "separation-activity verification" is the main strategy for screening antioxidant substances in herbal medicines. In this approach, as many natural products as possible are isolated from the herbal medicine, and their activities are evaluated through antioxidant assays. However, the separation step in this method is time-consuming. Ultrahigh-performance liquid chromatography (UHPLC) coupled with high-resolution mass spectrometry (HRMS) has shown great potential for the rapid identification of antioxidants in natural products [21][22][23]. It is based on the hypothesis that the reaction of antioxidants with 1,1diphenyl-2-trinitrophenylhydrazine (DPPH) will significantly reduce the concentrations of compounds with potential antioxidant activity. Due to the accurate mass measurement provided by UHPLC-HRMS, the antioxidants in herbal medicines can be easily screened and identified. In this work, an efficient UHPLC Q-Extractive Orbitrap MS/ MS method was developed to elucidate the chemical composition of BYF. e antioxidants were identified by offline DPPH-UHPLC Q-Extractive Orbitrap MS/MS and their concentrations were detected in rat serum after the oral administration of BYF. Figure 1 shows a schematic diagram of the experimental design. is study combines rapid antioxidant screening based on the chromatography-activity relationship with an evaluation of drug absorption to indicate the antioxidant substances contained in BYF. is allows to quickly identify the antioxidants that really effective in vivo, and the proposed strategy also provides reference for the screening of antioxidants of other traditional Chinese medicines. Paeoniae radix Rubra (PRR). Extraction of AR, LF, EF, PRR, and P were conducted by 12 L of water per gram of crude materials by two times. After filtering, the filtrate was concentrated. Reflux extraction of GRR, FTB, CRP, AJH, SCF, PF, and CF were conducted by 10 L of 70% ethanol per gram of crude materials by 2 times. After filtering, the ethanol was recovered from the filtrate and combined with the abovementioned concentrate. e combined concentrate was further concentrated into a thick paste with a relative density of 1.18-1.22 (60°C). Finally, the per-gram dry extract obtained was equivalent to 3.81 g of raw medicinal herb.

Reagents and
BYF extract (0.2 mg) was extracted by ultrasonication with 20 mL of 70% methanol for 30 min. e extract was filtered, and the filtrate was centrifuged at 13000 rpm for 10 min at 4°C. e supernatant was stored at −20°C before analysis. All reference standards and internal standards (IS) were dissolved in methanol at concentrations of 10 μg/mL.
For antioxidant quantification, stock solutions of the standard references and IS were prepared in methanol at concentrations ranging from 0.515 to 1.23 mg/mL. A series of working solutions of mixed reference standards were obtained by further dilution with methanol. Calibration standards were prepared by spiking 10 μL of the standard solutions into 90 μL of blank biological samples to obtain final concentrations of 0.0124-1080 ng/mL. All working solutions were stored at −80°C. Calibration curves were acquired by plotting the peak area ratio (y) of each compound to the IS against the corresponding concentration of each compound (x). e acceptance criterion of a calibration curve was a correlation coefficient (r) of 0.99 or better along with relative errors for each point within ±15%.

Preparation of Serum Samples.
e rats were randomly divided into the normal group and BYF group, with six rats in each group. e rats in the BYF group were gavaged twice per day with 18.28 g/kg/d BYF for 7 d. e normal group was given the same amount of normal saline by gavage for 7 d. e rats were fasted for 1 d before blood collection. Blood was collected from the orbital vein at 10 min, 30 min, 1 h, 2 h, and 4 h after the last gavage, and centrifuged at 3000 rpm for 10 min at 4°C. Finally, the serum was separated and stored at −80°C for later use.
To qualitative analyze the BYF components in rat serum after oral administration, we randomly mixed five serum samples from each group collected at different times after administration (20 μL for each sample, 100 μL in total). For the quantitative analysis of BYF components in rat serum, 100 μL of the serum samples collected at different times after BYF administration was taken and analyzed.
A protein precipitation procedure was used to extract BYF components. An 100-μL aliquot of serum was spiked with 300 μL of acetonitrile (containing 200 ng/mL IS and 6 ng/mL ascorbic acid). e mixture was vortex-mixed for 3 min, allowed to stand at 4°C for 5 min, and centrifuged at 4°C and 18534 g for 10 min. Next, 300 μL of the supernatant was transferred into a new tube and evaporated to dryness using an Integrate SpeedVac System ( ermoFisher Scientific Corporation, USA). e residue was redissolved in 50 μL of 50% acetonitrile, and a 35-μL aliquot of the supernatant was collected for analysis.
For the quantitative analysis of antioxidants in rat serum, the samples were loaded onto an Agilent ZORBAX-Extend-C18 LC column (4.6 × 50 mm, 1.8 μm) at 30°C. Mobile phase A was composed of water and 0.1% formic acid. Mobile phase B was composed of acetonitrile and 0.1% formic acid. e flow rate was 0.5 mL/min. e gradient elution condition were as follows: 5% B (0-2 min); linear gradient from       . e injection volume was 5 μL. e mass spectrometer was equipped with a heated electrospray ionization probe. e spray voltage was set at 3500 V for positive ion mode and 2800 V for negative ion mode. e flow rates of the sheath gas and aux gas were 40 and 10 Arb, respectively. e capillary temperature was 325°C, and the aux gas heater temperature was 300°C. Full scans from m/z 100 to 1500 were performed in the Orbitrap at a resolution of 70 K for quantification. e AGC target value was 3 × 10 6 , and the maximum injection time was 200 ms. Parallel reaction monitoring (PRM) mode was used for fragmentation identification and quantification of BYF metabolites. e target MS2 scan in PRM mode was conducted at a resolution of 17.5 K with an isolation width of e reduction in the peak area compared with the control group indicated the DPPH radical scavenging activity of the compounds in BYF.

Determination of Antioxidant Activities.
In order to determine the antioxidant activity of potential antioxidants, DPPH radical scavenging assay, ABTS radical scavenging activity, and ferric-reducing antioxidant power (FRAP) assay were conducted.
(1) DPPH radical scavenging assay. e DPPH radical scavenging assay was performed on a spectrophotometer microplate reader from ermoFisher Scientific (Vantaa Finland) using multiwell plates as a previously published method described [30]. e DPPH solution was diluted by methanol to 0.1 mM as a working solution. e reaction was initiated by mixing 50 μL of test solution with 150 μL of DPPH working solution and incubated in dark at room temperature for 30 min. Monitoring of the absorbance at 517 nm was carried out after the reaction was completed. e scavenging capacity of samples were calculated by experimental scavenging capacity (ESC) using equation (1) as follows: where Abs sample is the absorbance value of the sample (DPPH solution plus antioxidant) at each time interval and Abs blank is the absorbance value of the blank (methanol plus antioxidant(s)). Abs control is the absorbance value of control (methanol plus DPPH solution). e value of 50% inhibition (IC 50 ) was calculated by the graph plotting sample concentration and inhibition percentage.
(2) ABTS radical scavenging activity. e ABTS radical scavenging activity of the crude extracts was determined using the method described by Zhou et al. [30] with minor modifications. Aqueous ABTS (7 mM) was mixed with 2.45 mM aqueous potassium persulfate (1 : 1, v/v), and the solution was left to react for 16 h at room temperature in the dark.
e ABTS•+ solution was diluted with absolute ethanol to an absorbance at 734 nm of 0.70 ± 0.02 to obtain an ABTS•+ radical working solution. en, 160 μL of the ABTS•+ radical working solution was mixed with 40 μL of test solutions, and the mixture was incubated for 6 min. e absorbance of the mixture was measured at 734 nm. e ABTS radical scavenging assay was performed on a spectrophotometer microplate reader. e ABTS radical scavenging activity was calculated according to the following equation: ABTS radical scavenging activity % where A sample � the absorbance at 734 nm with sample and A blank � the absorbance at 734 nm without sample. e IC 50 value was calculated and represents the concentration necessary to reduce the maximum response of the ABTS by half.
(3) FRAP assay. e FRAP assays were performed by a total antioxidant capacity assay kit with the PRAP method according to manufacturer's instruction (Beyotime Biotech Inc, Shanghai, China). Briefly, 180 μL FRAP working solution was mixed with 5 μL extract of BYF, or 5 μL distilled water as blank control, or 5 μL 0.15-1.5 mM FeSO 4 standard solution (dissolved in distilled water) as standard curve. e absorbance of the mixture was measured at 593 nm after incubation at 37°C for 3-5 minutes. e total antioxidant capacity of the sample was calculated according to the standard curve. For FRAP method, the total antioxidant capacity of the extract is expressed by the concentration of FeSO 4 standard solution with equivalent antioxidant capacity.
As shown in Table 1, 189 chemical constituents were identified in BYF based on the library; their MS/MS spectra were matched with online databases and/or published references. Structurally, the main components of BYF were flavonoids (83 compounds), lignans (24 compounds), and alkaloids (20 compounds). Other identified components included 15 saponins, 11 terpenoids, 10 saccharides, eight lipids, seven organic acids, two coumarins, two amino acids, and seven other compounds. Among the identified compounds, 37 were identified by comparison with the retention times and MS spectra of standards; the MS/MS spectra of these compounds are shown in Supplementary Materials Figures S1-S37.  and their glycosides including icariin, epimedin A,  and compounds 17, 54, 55, 60, 61, 63, 68, 70, 76, 77, 81, 83,  87, 93, 98, 99, 106, 110, 111, 113, 114, 120, 122, 127, 132, 133, and 166. ese flavonoids have the characteristic isobutenyl neutral loss of 56 Da as a diagnostic ion (Table 1). For flavonoid glycosides, the common neutral losses of 162 and 146 Da were due to the presence of glucosyl and rhamnosyl groups. For example, in Figure 3 [24] was used for the structural identification of flavonoid aglycones in this work, especially those whose structures were not completely determined. e other flavonoids were mainly from FTB, GRR, LF, PF, and AR.

Alkaloids.
A total of 20 alkaloids were identified in BYF. Among them, 18 alkaloids were from only FTB. FTB mainly contains steroidal alkaloids such as peimisine and peimine [41].
ere are few characteristic fragments of steroidal alkaloids, in which only neutral loss of H 2 O can be found. erefore, these structures are confirmed by comparing the retention time with the standard. FTB also contains some alkaloid glycosides such as sibelicin glycoside and yibeinoside A. e common neutral loss of 162 Da was attributed to the presence of a glucosyl group (Table 1).

Screening of Antioxidant Components
Using the Offline DPPH-UHPLC Q-Extractive Orbitrap MS/MS. As mentioned above, BYF can significantly alleviate the symptoms of COPD in clinical practice. We have also done some research on the mechanism of BYF in treating COPD, the most important is BYF treatment could effectively inhibit the inflammatory response of the lungs [12,13]. In COPD rats, BYF significantly inhibited the expression of IL-1β, IL-6, TNF-α, and sTNFR2 induced by cigarette smoke and bacterial infection exposures.
e inhibition of BYF on inflammatory response in rat COPD model may be through restoring the 17/Treg balance by activating adenosine 2a receptor [16] and modulating the activities of STAT3 and STAT5 [17]. 17/Treg imbalance is considered to be important of COPD development. In COPD patients, the 17/ Treg cell balance shifts toward 17 cells, which triggers inflammatory responses in the airways and lungs and exacerbates alveolar destruction by producing interleukin-17 [17]. On the one hand, regulating oxidative stress is also an important mechanism of BYF regulating inflammation. We used transcriptomics and proteomics finding that the target proteins of BYF against COPD are enriched in oxidative stress-related pathways [15], ese will further inhibit the inflammatory response related to oxidative stress. erefore, we studied the antioxidant activity of BYF and its antioxidants in order to reveal its effective substances.

Total Antioxidant Capacity of BYF.
We evaluated the total antioxidant capacity of BYF by DPPH, ABTS, and FRAP assays. Table 2 shows that the IC 50 values of BYF in the DPPH and ABTS assays were 1136.36 ± 148.03 and 602.35 ± 81.26 μg/ mL, respectively. In addition, BYF showed high total antioxidant capacity of FRAP (0.51 ± 0.04 mM). us, it is necessary to further screen the active components of BYF.

Antioxidant Screening of BYF.
DPPH is a stable free radical with an odd electron. DPPH is commonly used to assess the radical scavenging activity of antioxidants; it is capable of accepting one or more hydrogen atoms from an antioxidant, resulting in an unconjugated structure with reduced MS response, which can be detected by HRMS [42,43]. Moreover, the use of DPPH saves time and labor compared to other free radicals such as ABTS [44].
is antioxidant screening strategy based on the change in MS signal can be divided into online and offline modes. Online screening requires two HPLC pumps, one for chromatographic separation and the other for delivering DPPH solution.
e chromatographic fraction and DPPH react online in the pipeline. is method is rapid but has relatively poor stability [44]. erefore, the more stable and sensitive offline mode was used in this work. In offline mode, the herbal medicine extract was fully reacted with DPPH, and the reaction solution was injected into the mass spectrometer for antioxidant detection.
In an offline experiment, the concentration ratio of DPPH in the extract will significantly affect the efficiency of antioxidant screening [45]. A relative excess of DPPH will not affect the free radical scavenging ability of the active components. However, when the DPPH concentration is insufficient, the free radical scavenging ability cannot be detected [46]. We optimized the DPPH concentration ( Figure 4) and found that 10 mM DPPH was most suitable to screen the free antioxidant components. e components with peak intensity decreased more than 20% were considered as potential antioxidants, which are summarized in Table 3.

Antioxidant Activities of the Potential Antioxidants.
To verify the antioxidant activities of the potential antioxidants determined above, we measured the free radical scavenging ability of 13 potential antioxidants with available reference standards by DPPH and ABTS assay. As shown in Table 2, 4 compounds showed high free radical scavenging ability for DPPH and or ABTS. Among them, rosmarinic acid had a strongest scavenging activity in DPPH assay (IC 50 � 25.72 ± 1.02 μg/mL), and rosmarinic acid and calycosin both showed strong scavenging activity in ABTS assay (IC 50   International Journal of Analytical Chemistry phenolic acids and flavonoids in BYF play a major role in the antioxidative activity. Rosmarinic acid has been reported to alleviate oxidative lung damage and airway inflammation based on its strong antioxidant activity [47][48][49]. Rosmarinic acid can also decrease the population of inflammatory cells; reduce the levels of proinflammatory cytokines such as IL-4, IL-5, and IL-13; upregulate IFN-c secretion; upregulate the activities of SOD, GPx, and CAT; increase Cu/Zn SOD; and significantly downregulate ROS production and the expressions of NOX-2 and NOX-4 in lung tissues [48]. Calycosin has also shown good antioxidant activity [50] and can ameliorate various lung injuries including sepsis, cecal ligation, and puncture by regulating oxidative stress-mediated inflammation in vivo and augmenting superoxide dismutase and glutathione [51]. Since oxidative stress and inflammation are the main pathogeneses of COPD, we suspect that these components are important antioxidants in BYF for the treatment of COPD.

Analysis of Antioxidants in Rat Serum.
e antioxidants in BYF may not show the expected antioxidant activity in vivo because of their poor absorption after oral administration. To investigate whether the potential antioxidants might be present in vivo, rats were orally administered with a high dosage of BYF extract. For consistency with the efficacy experiment, the serum was collected at 10 min, 30 min, 1 h, 2 h, and 4 h after the last BYF administration (after 7 d of continuous oral administration of BYF extract). Based on the retention time and HRMS spectra, we identified 79 compounds in rat serum after oral BYF administration: 34 flavonoids, 14 lignans, 7 alkaloids, one saponin, three organic acids, four saccharides, three lipids, four terpenoids, two amino acids, one coumarin, and six other compounds. Among them, 26 were identified by comparison with the standard materials (Table S2). e total ion chromatograms of rat serum at 1 h after the administration of BYF extract are shown in Supplementary Materials Figure S38. e 13 main components of BYF in rat serum were quantified; endogenous compounds and compounds with insufficient contents were not measured. e quantitative results are shown in Figure 2(b) and Figure S39, and the standard curves and linear ranges are shown in Table S3. Among the BYF components detected in serum, schisandrin B, schisantherin A, and schisantherin B had the highest contents. Among the validated potential antioxidants (Table 2), hesperidin, and naringenin were detected in rat serum (Table 3 and  Table S2); however, they are in trace amounts and the concentrations were not obtained. Rosmarinic acid was not found in serum after oral administrated of BYF, us, although rosmarinic acid showed the best antioxidant activity, it may not be the major active component in BYF due to its poor absorption or low content. Calycosin most likely to be responsible for the antioxidant effect of BYF in vivo, because it showed a high content in serum. e serum concentration-time curve of calycosin is shown in Figure 2( Figure 4: MS Intensity reduction for compounds in BYF after reaction with DPPH at different concentrations (500 μM, 1 mM, 2 mM, 5 mM, and 10 mM). e compound numbers marked on the X-axis refer to Table 1.

Supplementary Materials
e supporting data offered in supplementary materials are as follows: Table S1. Parallel reaction monitoring transitions for metabolites of Bufei Yishen formula included in the assay. Table S2. Chemical composition information of Bufei Yishen Formula Based on UHPLC-Q-Extractive Orbitrap MS. Table S3. Calibration curves, Linear ranges and LLOQs of the Bufei Yishen Formula compounds in serum. Figure  S1. e MS/MS spectra of the reference standard of quinic acid. Figure S2. e MS/MS spectra of the reference standard of Pyroglutamic acid. Figure S3. e MS/MS spectra of the reference standard of oxypaeoniflorin. Figure S4. e MS/ MS spectra of the reference standard of loganin. Figure S5. e MS/MS spectra of the reference standard of rhoifolin. Figure S6. e MS/MS spectra of the reference standard of Hesperidin. Figure S7. e MS/MS spectra of the reference standard of rosmarinic acid. Figure S8. e MS/MS spectra of the reference standard of diosmin. Figure S9. e MS/MS spectra of the reference standard of Peimisine. Figure S10.  Figure S11. e MS/MS spectra of the reference standard of Peiminine B. Figure S12. e MS/MS spectra of the reference standard of Epimedin A (Hexandraside F). Figure  S13.
e MS/MS spectra of the reference standard of Calycosin. Figure S14. e MS/MS spectra of the reference standard of Epimedin B. Figure S15. e MS/MS spectra of the reference standard of Epimedin C (Baohuside VI). Figure S16. e MS/MS spectra of the reference standard of Icariin. Figure S17. e MS/MS spectra of the reference standard of Ginsenoside Re. Figure S18.
e MS/MS spectra of the reference standard of Ginsenoside Rb1. Figure S19. e MS/MS spectra of the reference standard of Perillaldehyde. Figure S20. e MS/MS spectra of the reference standard of astragaloside Iv. Figure S21. e MS/ MS spectra of the reference standard of Naringenin. Figure S22. e MS/MS spectra of the reference standard of Apigenin. Figure S23. e MS/MS spectra of the reference standard of Formononetin. Figure S24. e MS/MS spectra of the reference standard of Ginsenoside Rg2. Figure S25. e MS/MS spectra of the reference standard of isosinensetin. Figure S26.
e MS/MS spectra of the reference standard of Methyl eugenol. Figure S27. e MS/MS spectra of the reference standard of D-ribophytosphingosine. Figure S28. e MS/MS spectra of the reference standard of schisandrol A. Figure S29. e MS/ MS spectra of the reference standard of Nobiletin. Figure  S30.
e MS/MS spectra of the reference standard of tangeretin. Figure S31. e MS/MS spectra of the reference standard of corosolic acid. Figure S32. e MS/MS spectra of the reference standard of Schisantherin B. Figure S33. e MS/MS spectra of the reference standard of Schisantherin A. Figure S34. e MS/MS spectra of the reference standard of Schisandrin A. Figure  S35.
e MS/MS spectra of the reference standard of Schisandrin B. Figure S36. e MS/MS spectra of the reference standard of Arbutin. Figure S37. e MS/MS spectra of the reference standard of Schisandrin C. Figure  S38. A total ion chromatogram of the mix rat serum after administrated of BYF extract at 1 h. Figure S39. e content of Bufei Yishen Formula components in the rat serum. (Supplementary Materials)