Quantitative Identification of Antioxidant Basis for Dendrobium Nobile Flower by High Performance Liquid Chromatography-Tandem Mass Spectrometry

Dendrobium nobile is a beautiful orchid and a widely used medicinal plant. In vitro antioxidant assays suggested that D. noblie flower extracts showed significantly higher 2, 2′-azinobis-3-ethylbenzthiazoline-6-sulfonate (ABTS) and 1, 1-diphenyl-2-picrylhydrazyl (DPPH) scavenging rates and much more ferric-reducing power than those of root, stem, leaf and fruit. To better understand the antioxidant basis of D. nobile flower, high-performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS) was used for metabolic identification and quantification. Finally, there were 72 metabolites among the total of 712 identified components showed significant association (coefficient >0.8, p < 0.05) with ABTS scavenging rates, DPPH scavenging rates, and ferric-reducing power. The three enriched classes of flower metabolites, including amino acids and their derivatives, organic acids and their derivatives, and flavonoids, formed the main antioxidant basis. The significantly accumulated rutin, astragalin, isomucronulatol-7-O-glucoside, quercetin 4′-O-glucoside, methylquercetin O-hexoside, caffeic acid, caffeic acid O-glucoside, and p-coumaric acid (Log2(fold change) >2, p < 0.01, distribution in flower >0.1%) made a key contribution to the higher antioxidant activities in flower. The relative quantification results of HPLC-MS/MS were verified by the common quantification methods. The antioxidant basis revealed of D. nobile flower will be helpful in the production of healthy or beauty products.

is is helpful for further quantitative identification of some novel antioxidants in D. nobile. e in-vitro antioxidant activities of the extracts from different tissues of D. nobile will be firstly evaluated in this paper. HPLC-MS/MS is then employed for metabolic analysis.
e final co-analysis will indicate the main chemical basis for the respective antioxidant activities.

Plant Materials.
Fresh roots, stems, leaves, flowers, and fruits of D. nobile were obtained from Hejiang, Sichuan Province (28°49′N, 105°50′E). Roots, stems, leaves, and flowers were collected in May 2019 and May 2020, fruits were collected in November 2019 and November 2020. Tissue samples were obtained from more than 30 individual plants for each collection. e tissue samples were washed with pure water, dried at 40°C for a week, ground into powder, and screened with a 50 mesh sieve before extraction.

Chemical
Reagents. DPPH and ABTS were purchased from Beijing Zhongsheng Ruitai and Shanghai Macklin (China). e methanol, formic acid, potassium ferrocyanide, ferric chloride, and citric acid standard were purchased from Chengdu Kelon (China). e quantitative BCA protein kit and the bovine serum albumin (BSA) standard were purchased from Beijing Solarbio (China). e Rutin Standard was purchased from Chengdu Purechem-Standard (China).

Metabolites Extraction for Bioactivity
Analysis. Each 5 g fine powdered sample was immersed with 200 mL of solution (80% methanol contained 0.1% formic acid) at room temperature for 24 hrs, and then was paper filtered to remove the residues. Subsequently, the filtrates were condensed in a rotary evaporator at 40°C for 2 hrs, and then were evaporated under vacuum for final drying. e dry extracts were dissolved with 80% methanol that contained 0.1% formic acid at a final concentration of 100, 200, 500, 1000, and 2000 μg/mL for in vitro analysis.

DPPH Scavenging
Assay. Each 0.3 mL extract was mixed with 0.9 mL methanol containing 0.1 mM DPPH. e mixed solution was kept at room temperature in the dark for 30 min before measuring the absorbance at 517 nm. e DPPH scavenging activity was calculated as follows: DPPH scavenging activity (%) � (A 0 − A s )/A 0 * 100% (A 0 absorbance without sample; A s absorbance with sample). Vitamin C was used as a positive control [12].
2.5. ABTS Scavenging Assay. ABTS was dissolved in 0.01 M PBS at a final concentration of 7 mM (pH 7.4). e ABTS solution was reacted with 2.45 mM potassium persulfate at room temperature for 16 hrs without light to generate free radicals. Before use, the ABTS solution was diluted with 0.01 M PBS to an absorbance of 0.7 at 734 nm. Each 0.1 mL extract was mixed with 1 mL diluted ABTS solution. e mixed solution was kept at room temperature for 20 min before measuring absorbance at 734 nm. e ABTS scavenging activity was calculated as follows: ABTS scavenging activity (%) � (A 0 − A s )/A 0 * 100% (A 0 absorbance without sample; A s absorbance with sample). Vitamin C was used as a positive control [9].
2.6. Ferric Reducing Assay. Each 0.1 mL extract was mixed with 0.5 mL 0.2 M PBS (pH 6.6) and 0.5 mL potassium ferrocyanide (K 3 Fe(CN) 6 , 30 mM). e mixed solution was incubated at 50°C for 20 min before addition of 0.5 mL trichloroacetic acid (0.6 M). en, 0.5 mL mixed solution was further added with 0.5 mL deionized water and 0.1 mL ferric chloride (FeCl 3 , 6 mM). e absorbance was measured at 700 nm. e ferric reducing antioxidant power is calculated as follows: reduce capacity � A s − A 0 (A 0 absorbance without sample; A s absorbance with sample). Vitamin C was used as a positive control [13].

Metabolites Extraction for HPLC-MS/MS.
Each 100 mg fine powdered sample was suspended with a pre-chilled 500 μL solution (80% methanol contained 0.1% formic acid) by well vortex. e sample was incubated for 5 min and then centrifuged at 12, 000g for 10 min. e supernatant was diluted to a final concentration of 53% methanol by pure water. e sample was then transferred to a new tube and then centrifuged at 12, 000g for 20 min. e supernatant was used for chromatography. e mass spectrometer was operated in positive or negative polarity mode with curtain gas of 35 psi, medium collision gas, ion spray voltage of 5500 V or −4500 V, temperature of 500°C, ion source gas of 1 : 55, ion source gas of 2 : 55.

Identification and Quantification of Metabolic Molecules.
e data files generated by HPLC-MS/MS were processed using SCIEX OS version 1.4 to integrate and correct the peak.
e main parameters were set as minimum peak height of 500, signal/noise ratio of 5, and gaussian smooth width of 1. Each peak of the experimental samples was detected using multireaction monitoring (MRM) based on the Beijing Novogene internal database (China). e parent ion (Q1), the daughter ion (Q3), the retention times (RTs), the de-clustering potential (DP), the collision energy (CE), and the molecular weights (MWs) were used for the identification of the metabolites. e peak area of Q3 was used for relative quantification of the metabolites. ese metabolites were further annotated using the KEGG database (https://www.genome.jp/kegg/), the HMDB database (https: //www.hmdb.ca/), and the Lipidmaps database (https:// www.lipidmaps.org/).

Detection of Total
Flavonoids. Each 0.25 g fine powdered sample was added with 4 mL of 80% methanol in a 10 mL centrifuge tube. After ultrasonic extraction for 30 min and centrifugation at 9, 000g for 10 min (4°C), the supernatant was collected. e residue was extracted with 4 mL of 80% methanol once-more. e combined supernatant was fixed to 10 mL with methanol. Each of 0.5 mL sample solution was mixed with 0.15 mL 5% sodium nitrite solution for 6 min. ey were mixed with 0.15 mL 10% aluminum nitrate solution for 6 min. ey were further mixed with 2 mL 4% sodium hydroxide solution and 2.2 mL distilled water for 3 min. e absorbance was determined at 508 nm and the total flavonoid content was calculated with the rutin standard.
2.11. Detection of Total Proteins. Each 0.1 g fine powdered sample was extracted with 1 mL 0.05 mM PBS (pH 7.8) by shaking for 2 hrs at room temperature. en, each 20 μL of the extracted filtrate was added with 200 μL of BCA working solution (50 : 1 of bicinchoninic acid and Cu reagent). After mixing well, they were placed at 37°C for 30 min. e absorbance at 562 nm was used for calculation of total proteins with BSA standard.

Detection of Total Organic
Acids. Each 0.25 g fine powdered sample was extracted with 100 mL of distilled water by shaking for 3 hrs at room temperature. Accurately take 50 mL of the extracted filtrate into a 250 mL beaker. en, basic burette filled with sodium hydroxide solution was used for titration. e end point of the titration was pH 7.0. Citric acid standard was used for calculation. Organic acid content � (C * V * M)/(3 * m) * 100% (C concentration of sodium hydroxide solution; V volume of sodium hydroxide solution consumed by titration; M mass of citric acid; m mass of sample).
2.13. Statistics Analysis. All measurements and experiments were repeated at least three times. Quantitative data were presented as mean ± standard deviation (SD). e correlation analysis was performed using PASW statistics 18.0 (IBM, USA). Pearson correlation coefficients and p value were used for evaluating the correlations. Student's t-test was used for comparison between two groups. Log 2 (fold change) was used for comparison of relative quantification. CSCF/ TCCF (ratio of the contents of one specific component in flower to the total contents of all components in flower) and CSCF/CSCA (ratio of the contents of one specific component in flower to the contents of this component in all of root, stem, leaf, flower, and fruit) were used for comparison of different distributions.

Relatively Higher Antioxidant Activities Showed by Extracts of D. Nobile Flower.
e ABTS and DPPH scavenging rates and ferric-reducing power of the extracts from root, stem, leaf, flower, and fruit increased in a concentrationdependent manner (Figure 1). Under the concentration of 100, 200, and 500 μg/mL, the ABTS scavenging rates of flower extracts were significantly higher than those of extracts from root, stem, leaf, and fruit (p < 0.01). Under the concentration of 100, 200, 500, and 1000 μg/mL, the DPPH scavenging rates of flower extracts were significantly higher than those of extracts from root, stem, leaf, and fruit (p < 0.05). At a concentration of 500, 1000, and 2000 μg/ mL, the ferric-reducing power of flower extracts was significantly higher than those of extracts from root, stem, leaf, and fruit (p < 0.05). When the concentration was greater than 500 μg/mL, the ABTS and DPPH scavenging rates of flower extracts were close to those of vitamin C. In summary, flower extracts showed higher ABTS and DPPH scavenging rates and much more ferric-reducing power than those extracts from other tissues. ese results revealed relatively higher antioxidant activities in vitro in the D. nobile flower.

Metabolites Associated with Antioxidant Activities in
Flower of D. nobile. After correlation analysis, there were 72 metabolites showed significant association (coefficient >0.8, p < 0.05) with the ABTS and DPPH scavenging rates and ferric-reducing power ( Table 1). As shown in Figure 5, the 72 metabolites were mainly belongs to three classes of amino acid and its derivatives (13, 60.45% of CSCF/TCCF), organic acid and its derivatives (11,19.05% of CSCF/TCCF), flavonoids (20, 17.05% of CSCF/TCCF). e average CSCF/ CSCA of amino acid and its derivatives, organic acid and its derivatives, and flavonoids were 55.05%, 67.42%, and 81.15%. Antioxidant activities associated with amino acids and their derivatives showed a higher distribution in the flower itself, but antioxidant activities associated with flavonoids showed a higher distribution in the flower compared to the root, stem, leaf, and fruit. But none of them showed more than 80% of CSCF/CSCA (Figure 7(a)). Among the 11 antioxidant activities associated with organic acid and its derivatives, pipecolinic acid (50.43%), caffeic acid (20.79%), pipecolic acid (10.74%), p-coumaric acid (9.37%), and caffeic acid O-glucoside (5.11%) showed a relatively high distribution in the flower itself ( Figure 6(b)). But only caffeic Acid, p-coumaric acid, and caffeic acid O-glucoside showed more than 80% of CSCF/CSCA (Figure 7 (Figure 9(a)). e BCA method showed that total protein concentrations in root, stem, leaf, flower, and fruit were 32.24 mg/g, 24.29 mg/g, 253.59 mg/g, 288.18 mg/g, and 92.09 mg/g, respectively (Figure 9(d)). Relative quantification by HPLC-MS/MS showed that the distributions of organic acid and its derivatives in root, stem, leaf, flower, and fruit were 13.81%, 10.54%, 21.56%, 25.02%, and 29.06%,       respectively (Figure 9(b)). e titration method showed the concentrations of total organic acids in root, stem, leaf, flower, and fruit were 1.17 mg/g, 0.68 mg/g, 1.66 mg/g, 1.73 mg/g, and 2.33 mg/g, respectively (Figure 9(e)). Relative quantification by HPLC-MS/MS showed that the distributions of flavonoids in root, stem, leaf, flower, and fruit were 0.65%, 4.55%, 27.25%, 53.62%, and 13.94%, respectively (Figure 9(c)). e colorimetric method showed that total flavonoids concentrations in root, stem, leaf, flower, and fruit were 8.72 mg/g, 9.23 mg/g, 12.49 mg/g, 31.30 mg/g, and 11.91 mg/g, respectively (Figure 9(f)). ese results indicate that the relative quantification by HPLC-MS/MS was consistent with absolute quantification by the corresponding common methods.  throughput property of HPLC-MS/MS makes it capable of analyzing hundreds of metabolites simultaneously. Recently, some reports revealed the attempts to use it for metabolic identification and quantification related to some specific biofunctions [19][20][21]. HPLC-MS/MS was used for the analysis of bioactive ingredients responding to UV-B radiation in D. officinale [19]. HPLC-MS/MS was used for co-analysis between metabolites and anti-inflammatory activities in D. chrysanthum [25]. HPLC-MS/MS was used for the identification of polysaccharides that prevent ethanol-induced liver injury in D. huoshanense [26]. HPLC-MS/MS was used for co-analysis between polysaccharides and polycystic ovary syndrome in D. nobile [27]. HPLC-MS/MS was used for co-analysis between metabolites and diabetic myocardial fibrosis in D. officinale [28]. HPLC-MS/MS was used for co-analysis between metabolites and suppression rates in A549 lung cancer cells in D. nobile [29]. HPLC-MS/ MS was used for the comparison of chemicals related to antioxidant activities between D. huoshanense and D. officinale [20]. HPLC-MS/MS was used for identification of antioxidant compounds in D. catenatum flower [12].

HPLC-MS/MS was Suitable for Metabolic Identification and Quantification in
Here, HPLC-MS/MS was used for identification of the

Some Enriched Flavonoids and Organic Acids Formed the
Main Antioxidant Basis of the D. nobile Flower. ABTS scavenging, DPPH scavenging, and ferric reduction were generally used to evaluate in-vitro antioxidant activities [9,12]. e extracts from flower of D. nobile showed significant higher ABTS scavenging rates, DPPH scavenging rates, and ferric-reducing power than those from root, stem, leaf, and fruit in this paper. e flower extracts of D. officinale, D. sabin, D. devonianum, and D. catenatum had also been reported to possess relatively high antioxidant activities [9,[11][12][13]. But the antioxidant activities related chemical basis was poorly studied in Dendrobium flower. Polysaccharides in the flowers of D. devonianum have been reported to be correlated with its antioxidant activities [11]. Phenolic glycosides in the methanolic extract of the flower were identified as antioxidant components in D. catenatum [12]. Here, 72 compounds mainly belong to three classes of metabolites amino acid and its derivatives, organic acid and its derivatives, and flavonoids were correlated to the higher antioxidant activities of flower in D. nobile. Furthermore, eight components of rutin, astragalin, isomucronulatol-7-Oglucoside, quercetin 4′-O-glucoside, methylquercetin O-hexoside, p-coumaric acid, caffeic acid and caffeic acid O-glucoside were identified to play a key contribution to antioxidant activities in vitro. Quercetin extracted from D. officinale showed antioxidant effect to UV-B exposure [19,21]. e major compounds contributed to the antioxidative activities were identified as 1-O-caffeoyl-β-Dglucoside, rutin, and isoquercitrin in D. catenatum [12]. e antioxidant activities of D. huoshanense were also mainly attributed to its high content of flavonoids [20]. e novel finding of antioxidative flavonoids and organic acids further enriched acknowledge about the antioxidant basis of Dendrobium flower. is will be helpful in the production of related healthy or beauty products, such as flower-tea, flower-wine, flower-biscuits, flower-mask, flower-cream, flower-toothpaste, and flower-capsules [1,16,21].

Conclusions
is paper firstly confirmed the best in-vitro antioxidant activities of D. noblie flower. A total of seventy-two metabolites were identified to be corresponded to antioxidant activities in vitro. Eight flavonoids and organic acids formed the key antioxidant basis of D. nobile flower. e quantification results of HPLC-MS/MS were also verified by the common methods. ese results suggest that HPLC-MS/MS is suitable for quantitative chemical-function analysis in D. nobile.

Data Availability
All related data are included within the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest.