2.1 Chemicals and reagents

HPLC-grade of methanol, acetonitrile, ammonium acetate and formic acid were obtained from Merck (Darmstadt, Germany). Analytical grade of NaNO₂, Al(NO₃)₃·9H₂O and NaOH were purchased from Macklin (Shanghai, China). Analytical grade of acetic acid, potassium persulfate, 2,2-diphenyl-1-picrylhydrazyl (DPPH), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylicacid (Trolox), 2,2′-Azinobis-(3-ethylbenzthiazoline-6-sulphonate) (ABTS) and ferric chloride sodium acetate were supplied by J&K (Beijing, China). Sorbents PSA and C18 were provided by Agilent Technologies (Tianjin, China). g-MWCNTs (outer diameter: 10–20 nm, length: 5–30 μm), MWCNTs-COOH (outer diameter: 10–20 nm, length: 10–30 μm) and MWCNTs-OH (outer diameter: 10–20 nm, length: 5–30 μm) were acquired from XFNANO (Nanjing, China). Flavonoid standards including naringenin, apigenin, cynaroside, quercetin, eriodictyol, isorhamnetin, quercitrin, rutin, hesperetin, isoquercitrin, kaempferol, apiin, hyperoside, schaftoside, chrysoeriol, taxifolin and naringin were purchased from Anpu (Shanghai, China), with purities ranging from 97.1 % to 99.9 %. The structure of flavonoid compounds was listed in Table S1. Standard stock solutions of all flavonoid compounds (100 mg/L) were prepared by using methanol and stored at 4 °C until use.

2.2 Sample collection and pretreatment

Forty-seven samples of D. officinale were collected from production bases and markets. Detailed sample information including origin, cultivation method, growth years, collecting time, and processing was recorded in Table S2. Samples were cultivated under facility-aided cultivation mode or imitative wild cultivation mode. Fresh stems were processed into Fengdou, dry stems and slices. Generally, D. officinale of two-year-old was suggested to be harvested and sold in the market, and D. officinale of more than two-year-old were also collected to explore the effect of growth years on flavonoid contents. D. officinale was normally harvested during November to May of next year and D. officinale was also collected in non-harvest period to investigate the effect of collecting time on flavonoid contents. When fresh stems were collected, they were cut into small sections and then roasted using a DHG-9070A blast dryer (Shanghai Jinghong Experimental Equipment Co. LTD, China) for 24 h at 80 °C, following the same baking conditions utilized in production bases. The dry stems or slices were grounded into powders and stored at −20 °C until analysis.

2.3 Sample extraction

0.2 g of D. officinale sample was weighed and transferred into a 15 mL centrifuge tube. Subsequently, 8 mL of extraction solution (acetonitrile/water, volume ratio of 3:1) was added into the tube, and the mixture was vortexed for 5 min, followed by sonication for 30 min at room temperature. After sonication, the mixture was centrifuged at 4000 g for 5 min, and the supernatant was collected for the determination of flavonoid concentrations and the analysis of antioxidant activity (Fig. S1).

2.4 Determination of 17 flavonoid compounds by QuEChERS-HPLC-MS/MS

2.5 Method validation and uncertainty evaluation

The method used for analysis of 17 flavonoid compounds in D. officinale was validated according to the guidelines presented by European Commission (2021), and the International Union of Pure and Applied Chemistry (IUPAC, 2002). The linearity, sensitivity, accuracy and precision of the method were evaluated.

The linearity was evaluated by building external calibration curves for each compound. Each calibration curve was plotted with a minimum of five suitable concentrations within the range of 1 to 200 μg/L. The sensitivity of the method was determined by the limit of detection (LOD) and the limit of quantification (LOQ), which corresponded to the concentration of each analyte at a signal-to-noise ratio of 3 and 10, respectively. The accuracy was assessed by recoveries. Mixed standards containing 17 flavonoid components were spiked at three concentration levels, namely low, medium and high (80 %, 100 % and 120 % of the original amount) into the sample, and the recovery was calculated using the formula: Recovery (%) = (total mass in spiked sample − mass in non-spiked sample)/mass spiked × 100 (Xin et al., 2023). Both fortified and unfortified samples were treated identically. The precision of the method was evaluated by intra-day and inter-day variations.

Uncertainty of the measurement was assessed by the combination of individual uncertainty including purity of the standards, balance weighing (weighing of samples and standards), volume measuring by using a pipette (measuring of standard solutions and solvents), volumetric flask used for preparation of standard solution, and method repeatability obtained by the RSD of the measurements (ISO, 1993). The combined relative standard uncertainty (u) was calculated by Eq. (1). The expanded relative uncertainties (U) with a coverage factor of 2 (k = 2) was estimated by Eq. (2).

u_rel,p, u_rel,w, u_rel,m, u_rel,v and u_rel,rep indicated the relative uncertainty of purity of standards, balance weighing, volume measuring, volumetric flask in preparation of standard solution and the method repeatability, respectively.

2.6 Total contents of flavonoid determined by NaNO₂-Al(NO₃)₃ colorimetry

The total contents of flavonoid in D. officinale (C_t) were determined according to the method of Zhishen et al. (1999) with minor modifications. 2 mL of the sample extract was first mixed with 0.3 mL of 5 % NaNO₂ (w/w). Afterwards, 0.3 mL of 10 % Al(NO₃)₃ (w/w) and 4 mL of 4 % NaOH (w/w) were added and the sample was then thoroughly mixed. The absorbance was measured at 510 nm. Rutin was used as a standard, and the calculation of C_t was based on rutin equivalents (mg RE/g).

2.7 Antioxidant assays

The antioxidant activity of D. officinale was evaluated by using the DPPH radical scavenging capacity, ABTS assay and FRAP total antioxidant capacity according to available method (Jin et al., 2016; Zhang et al., 2017; Meng et al., 2019). Additionally, three specific methods for the determination of antioxidant activities were presented in Supplemental Materials.

2.8 Calculations and statistical analysis

To test the significance of the difference in C_t under different extraction conditions, t-test was performed using IBM SPSS Statistics 25 with a statistical significance level of 0.05. Spearman correlation was applied to determine the relationship between the antioxidant activity of 47 D. officinale samples and the total content of flavonoid (C_t) and the individual flavonoid concentration. Moreover, the relationship between flavonoid contents and factors including origin, cultivation method and growth years was explored by the Multiple Factor Analysis (MFA) performed on XLSTAT 2018 (Addinsoft, USA). In MFA analysis, flavonoid contents and antioxidant activity values were standardized by Z-score. Considering that D. officinale slices were obtained by cutting the dry stem into slices, and the mechanic cutting would not cause variations in flavonoid contents, the factor ‘processing’ was sorted as dry stem and Fengdou, and slices were merged with dry stem.

3.1 Method development for determination of 17 flavonoid compounds using QuEChERS-HPLC-MS/MS

3.1.1

3.1.1 Optimization of HPLC-MS/MS conditions

To achieve a satisfactory resolution, sensitivity and short sample running time, various mobile phases were tested. The pKa for flavonoid compounds is from 5.7 to 7.52, and LogP values vary from −3.06 to 3.19 (Table S4). Considering the acid-base property and wide polarity of flavonoid compounds, four mobile phases were examined. These were A (water − methanol), B (0.1 % formic acid − methanol), C (2 mmol/L ammonium acetate containing 0.1 % formic acid − acetonitrile) and D (2 mmol/L ammonium acetate containing 0.1 % formic acid water − methanol). A mixed standard solution of 17 flavonoid components with a concentration of 0.1 mg/L was injected, and the peak shape and peak area under the four mobile phases were compared. The peak shapes were comparable under all four mobile phases, appearing sharp and symmetrical. The peak areas of the 17 flavonoids under the four mobile phase conditions were shown in Fig. 1. Under each mobile phase, the intensity of chrysoeriol was the highest, which was nearly 64 times that of kaempferol. Meanwhile, the peak area of kaempferol was the highest under the mobile phase D. Herein, mobile phase D (2 mmol/L ammonium acetate containing 0.1 % formic acid water − methanol) was chosen for the analysis of flavonoid compounds. Furthermore, optimization of MS parameters is another essential step for obtaining the highest sensitivity. The precursor ion was determined in full scan with an m/z range of 100–1000 followed by the optimization of MRM transitions and the corresponding collision energy (CE). The optimized MRM transitions and CE for 17 flavonoid components were listed in Table S3.

Close

Fig. 1

Peak areas of flavonoid compounds under the four mobile phases. A (water − methanol), B (0.1 % formic acid − methanol), C (2 mmol/L ammonium acetate containing 0.1 % formic acid − acetonitrile) and D (2 mmol/L ammonium acetate containing 0.1 % formic acid water − methanol).

Export to PPT

3.1.2

3.1.2 Optimization of extraction conditions

Efficient and reproducible sample extraction is critical to the chemical analysis. Currently, flavonoids in herbs and foods are extracted using methods including hot water bath, Soxhlet extraction, ultrasonic assisted-extraction and microwave-assisted extraction (Søltoftet al., 2009; Wong et al., 2017; Moldovanet al., 2019; Manganget al., 2020). Ultrasound assisted-extraction shows advantages of simple operation, low cost and high yield since ultrasonic treatment could induce cavitation effects and destroy cell walls efficiently to promote solute transfer from the solid phase to the liquid phase (Cui et al., 2022; Chakraborty et al., 2020). Herein, the ultrasound-assisted extraction technology was selected to extract flavonoids from D. officinale. Further, the extraction procedure was optimized by investigating the influence of various factors such as types of organic solvent, water content, solid–liquid ratio and extraction time. The results were shown in Fig. 2.

Close

Fig. 2

Effects of extraction conditions on total content of flavonoid (C_t): (a) extraction solvents: MeOH (methanol), ACN (acetonitrile), ACTN (acetone) and EtOH (ethanol), (b) water content, (c) extraction time, and (d) solid–liquid ratio. The different letters in each subfigure indicated the significant difference among C_t values under various extraction conditions (p < 0.05).

Export to PPT

The extraction efficiencies of four organic solvents (methanol, acetonitrile, acetone and ethanol) were compared under constant parameters (water content of 25 %, solid–liquid ratio of 1:40 (g/mL) and an extraction time of 30 min). Acetone resulted in the highest C_t of 3.71 mg/g, followed by acetonitrile with a C_t of 3.60 mg/g (Fig. 2a). It was indicated that there was no significant differences in C_t values under these two extraction solvents (t-test, p > 0.05). Considering the volatility and toxicity of acetone, acetonitrile was selected as the extraction solvent. Subsequently, the effect of water content in the extraction solvent on extraction efficiency was investigated when acetonitrile was used as extraction solvent with a solid–liquid ratio of 1:40 (g/mL) and an extraction time of 30 min. The highest C_t of 3.77 mg/g was achieved when the water content was 25 % (Fig. 2b). Therefore, it was concluded that the optimal extractant was consisted of acetonitrile/H₂O (75/25, v/v). Extraction time was critical in achieving optimal ultrasound-assisted extraction efficiency. The C_t was measured at intervals between 15 and 75 min while keeping other parameters constant (acetonitrile/H₂O of 75/25, solid–liquid ratio of 1:40 (g/mL)). The flavonoid content increased initially and then stabilized after 30 min (Fig. 2c). It was appropriate to select 30 min as ultrasonic time. Finally, the effect of solid–liquid ratio on the C_t was assessed using the optimal conditions obtained above. When the solid–liquid ratio was 1:10 (g/mL), the sample matrix was not completely immersed by the extraction solvent, resulting in incomplete extraction and low flavonoid extraction efficiency. As shown in Fig. 2d, there was no significant difference (t-test, p > 0.05) between C_t values at ratios of 1:40 (g/mL) and 1:50 (g/mL). Therefore, a solid–liquid-ratio of 1:40 (g/mL) was selected for sample extraction to reduce solvent usage.

3.2 Method validation and comparison

The QuEChERS-HPLC-MS/MS method was validated to verify that its performance was qualified for the routine analysis of flavonoid components in D. officinale. Performance parameters including linearity, sensitivity, accuracy and precision were assessed.

All flavonoids showed a good linearity in the range of 1–200 μg/L, and the correlation coefficient was greater than 0.99 (Table 2). The LOD of 17 flavonoid components ranged from 2.1 to 49.55 μg/kg, and the LOQ ranged from 7.12 to 165.2 μg/kg (Table 2). The recoveries were examined at three levels, providing recoveries of 73 % to 112 %. The intra-day precision and inter-day precision were varied from 1.9 % to 13.3 % and 3.7 % to 15.1 %, respectively. Therefore, the validation results met the criteria of SANTE/11312/2021 (European Commission, 2021), demonstrating that the QuEChERS-HPLC-MS/MS method can provide good linearity, sensitivity, accuracy as well as precision for simultaneous analysis of 17 flavonoid components in D. officinale. Additionally, the relative uncertainties of 17 flavonoid components were in the range of 2.4 % to 15.4 % and the expanded relative uncertainties were in the range of 4.8 % to 30.8 % (Table S5).

The quantitative analysis of 17 flavonoid components in this work applied QuEChERS method in sample preparation to reduce the interference in sample extracts. The study on the quantitative analysis of flavonoid components in D. officinale was relative rare and only one relevant publication was available. The method developed in this work was compared with those reported on D. catenatum flowers, D. chrysotoxum flowers and D. officinale, and sample extraction and instrumental analysis was summarized in Table S6. The method in this work cost less amount of solvent and less sample preparation time, and afforded higher sensitivity when compared with Zhang et al. (2019) and Hu et al. (2023). Meanwhile, the method in this work showed comparable sensitivity with that of Lv et al. (2017).

3.3 Total content of flavonoid in D. officinal

The total content of flavonoid (C_t) ranged from 1.57 to 9.36 mg/g, with an average value of 4.21 mg/g. The highest C_t of 9.36 mg/g was observed for sample 4 collected from Panan, Zhejiang Province, while the lowest C_t of 1.57 mg/g was measured in sample 44 taken from Taizhou, Zhejiang Province. Samples (1–4) were obtained from the same production base located in Jinhua, Zhejiang Province, being grown under imitative wild cultivation for 2, 3, 4 and 5 years, respectively. The C_t value of samples 1–4 indicated that the total content of flavonoid was enhanced with the increase of growth year (Fig. 3). The finding was similar with the study of Xin et al. (2015) who observed that flavonoid contents in Astragali Radix was increased with the extended growth time.

Close

Fig. 3

Total contents of flavonoid (C_t) in 47 of D. officinale samples.

Export to PPT

3.4 Contents of 17 flavonoid components

Flavonoids belong to a large group of secondary metabolites in D. officinal. The basic skeletons of flavonoids are apigenin, vitexin, quercetin and kaempferol. In this study, we employed the validated method to determine the content of 17 flavonoid components with available standards in D. officinal samples from Zhejiang, Yunnan, Anhui and Guangxi Provinces. The total concentration of 17 flavonoid components is denoted as C₁₇, ranging from 33.09 to 655.44 mg/kg. It was observed that naringenin, rutin and schaftoside were the three predominant flavonoid components in D. officinal (Fig. 4a), accounting for more than 80 % of the C₁₇. The concentration of naringenin, rutin and schaftoside in D. officinal were 5.19–57.18, 2.01–563.17 and 4.43–67.73 mg/kg, respectively. The highest concentration of naringin, rutin and schaftoside were observed in samples 4, 43 and 35, respectively. The cultivation methods for these three samples were imitative wild cultivation, while their origins were Jinhua, Quzhou and Taizhou, Zhejiang Province, respectively. The concentration of the remaining 14 flavonoid components in D. officinal was relatively low, accounting for only 1.05 %-16.83 % of the C₁₇ measured. The variation in individual flavonoid concentrations with extended growth years was examined in samples 1–4, which were grown under the same cultivation method yet for different years. It was noted that the variation of flavonoid components with growth years were compound-dependent. For example, the contents of naringenin, isorhamnetin, eriodictyol, kaempferol and taxifolin increased with the extension of growth year, while the contents of chrysoeriol, quercetin, and apiin remained relatively stable for 2–4 years before reaching their highest levels in the fifth year. This might be due to the different functions of flavonoids.

Close

Fig. 4

17 flavonoid components in 47 of D. officinale samples. (a) Contents of naringenin, rutin and schaftoside, and (b) Contents of the remaining 14 flavonoid components.

Export to PPT

3.5 Antioxidant capacity of D. officinale

D. officinale sample extracts were screened for antioxidant capacity by using DPPH, ABTS, and FRAP assays (Zhaoet al., 2021). DPPH radical scavenging capacity in D. officinale sample extracts ranged from 35.62 % to 89.64 %, a 2.5-fold variations (Table S7). ABTS free radical scavenging rate varied from 16.77 % to 96.02 %, a 5.7-fold differences (Table S7). FRAP values ranged from 277.51 to 1523.41, a 5.5-fold variations (Table S7). Sample 29 from Mangshi, Yunnan Province showed the lowest antioxidant capacity indicated by three antioxidant capacity assays, while sample 15 from Lishui, Zhejiang Province had the highest DPPH radical scavenging ability. Sample 4 from Jinhua, Zhejiang Province exhibited the highest ABTS and FRAP antioxidant activities. Correlation analysis was performed to investigate the relationship between the flavonoid concentrations and antioxidant activities of D. officinale sample extracts (Table S8). The correlation coefficients between C_t and DPPH radical scavenge, ABTS radical scavenge and total antioxidant capacity of FRAP exceeded 0.7, highlighting the important role of flavonoids in the antioxidant ability of D. officinale. As regard to flavonoid compounds, 12 out of 17 compounds, including naringenin, quercetin, apigenin and isorhamnetin, were significantly positively correlated with the antioxidant activity determined by the three assays. The results were comparable with Huang et al. (2014), who found the correlation between the content of flavonoids and the antioxidant activity of D. officinale extracts. Furthermore, quercetin, one of most well-known flavonoids, has observed to be a powerful scavenger of reactive oxygen species (Wang et al., 2016).

3.6 Relationship among flavonoid components, antioxidant activity and influencing factors

Flavonoid contents in D. officinale may be influenced by various factors such as origin, cultivation method, growth years, collecting time and processing techniques. MFA was applied to study the relationship among the origin, processing, growth years, collection period, cultivation method, total flavonoid content (C_t), 17 flavonoid component contents and antioxidant activity of D. officinale (Fig. 5). The first two factors exhibited a variability of 26.20 % and 14.05 %, respectively. 12 out of 17 flavonoid compounds, including quercetin, isorhamnetin, eriodictyol, schaftoside, luteoloside, apiin, hesperetin, taxifolin, apigenin, kaempferol, hyperoside and isoquercitrin were clustered together in the first quadrant. On the other hand, rutin, naringin, and quercetin were separated from these 12 flavonoid compounds. The C_t and antioxidant activity were found to be closely related to the 12 flavonoid components since they were clustered in the near proximity, and it suggested that those 12 flavonoids contributed significantly to the antioxidant activity. This observation was consistent with the results obtained from correlation analysis. Eight different origins of D. officinale were distributed across four quadrants. Contents of 12 flavonoid components were relatively high in D. officinale sampled from Quzhou, Guangxi and Anhui compared to samples from Yunnan, Wenzhou and Taizhou, and it was consistent with the fact that samples obtained from Quzhou, Guangxi and Anhui were found to be closely associated with the 12 flavonoid components. A positive correlation was found between the D. officinale collected from Taizhou and quercetin. Accordingly, the contents of 12 flavonoid components in samples grown under imitative wild cultivation were observed to be relatively high compared to those cultivated in green-house. Moreover, long-growth-year (more than 2 years) was favored for accumulating these 12 flavonoid components. Fig. 5b plotted the coordination of all classes and indicated that the origin and cultivation methods contributed more to the flavonoid content and antioxidant activity because they have a more proximate relationship. By differentiating the color of sample dot according to cultivation method, D. officinale samples cultivated under facility-aided and imitative wild methods were mainly scattered in two separate areas on MFA score plot (Fig. 5c). Nevertheless, when the colour of sample dots was sorted according to the origin, samples from Jinhua, Quzhou and Lishui had large overlapping, suggesting that multiple factors other than the origin account for the accumulation of flavonoid in D. officinale.

Close

Fig. 5

The results of MFA analysis (1-Naringenin, 2-Quercetin, 3-Rutin, 4-Isorhamnetin, 5-Apigenin, 6-Eriodictyol, 7-Chrysoeriol, 8-Hesperetin, 9-Isoquercitrin, 10-Quercitrin, 11-Luteoloside, 12-Kaempferol, 13-Schaftoside, 14-Apiin, 15-Hyperoside, 16-Naringin and 17-Taxifolin). (a) Correlation map of all variables, (b) coordinates of the eight variable classes, (c) observation scores plot of MFA (cultivation method) and (d) observation scores plot of MFA (origin).

Export to PPT

Close

Fig. 5

The results of MFA analysis (1-Naringenin, 2-Quercetin, 3-Rutin, 4-Isorhamnetin, 5-Apigenin, 6-Eriodictyol, 7-Chrysoeriol, 8-Hesperetin, 9-Isoquercitrin, 10-Quercitrin, 11-Luteoloside, 12-Kaempferol, 13-Schaftoside, 14-Apiin, 15-Hyperoside, 16-Naringin and 17-Taxifolin). (a) Correlation map of all variables, (b) coordinates of the eight variable classes, (c) observation scores plot of MFA (cultivation method) and (d) observation scores plot of MFA (origin).

Export to PPT