Bioactive Flavonoid Glycosides and HPLC and UPLC Quantification of Commercial Astragali Complanati Semen

Eleven compounds, including nine known flavonoid glycosides (1–4, 6–8, and 10–11), one isoflavone glycoside (5), and a glansreginic acid (9), were isolated from the 80% ethanol extract of commercial Astragali Complanati Semen (ACS). All chemical structures were determined by spectroscopic analyses, including 1D and 2D NMR. Compounds 2, 4, 5, 6, 9, and 10 were isolated and identified from the title plant for the first time. Biological evaluation revealed that all the isolates showed promising anti-NO production, and 1, 2, 3, and 8 were more potent in antioxidant activity than vitamin E. The major peaks in the UPLC and HPLC profiles identified their chemical structures by comparing their retention time and UV spectra with those of the reference substances. Furthermore, nine of the eleven samples collected from North, Middle, and South regions of Taiwan possessed similar HPLC fingerprints and were identified as Astragali Complanati Semen, whereas the other two samples from southern Taiwan would be the adulterants due to the different fingerprinting patterns. In addition, an HPLC-UV method was employed to determine the content of target compound complanatuside (11) with good linear regression (R2 = 0.9998) for ACS in the Taiwanese market. Of the isolates, flavonol glycosides 1 and 3 were the major peaks in HPLC/UPLC, and showed more potent antioxidant and anti-NO production activities than that of 11, revealing that these compounds can be the available agents for the quality control of ACS.


Introduction
Astragali Complanati Semen (Sha Yuan Ji Zi/Flastem milkvetech seed; ACS) is a traditional Chinese medicine, made from the ripe seeds of the perennial herbaceous plant Astragalus complanatus in the Leguminosae (Fabaceae). ACS has been used traditionally to tonify yang, strengthen the kidney, and consolidate kidney qi (the essential elements and vital energy for the body in Traditional Chinese Medicine theory), in result of stabilizing the essence such as seminal emission, spermatorrhea, enuresis, frequent urination, leukorrhagia and lumbar pain [1][2][3]. In the clinical setting, ACS serves as the
Compound 4 presented as a yellowish powder, with a molecular formula of C26H28O16 determined from the ESIMS spectrum (m/z 619.1 [M + Na] + ). The 1 H and 13
Compound 5, obtained as a yellowish powder, was assigned with the molecular formula C 22 H 22 O 10 , determined from ESIMS, 13 C NMR, and DEPT spectroscopic data. The 1 H and 13 C NMR spectra (Table 1) showed resonance characteristics of isoflavone skeleton with a sugar moiety. The 1 H NMR spectrum showed the characteristic signals, δ H 8.33 (1H, s), for H-2 of the isoflavone framework, and two typical ABX system aromatic protons system for A-ring and B-ring in 5. In addition, the signals for one aromatic methoxy group at δ H 3.77 (3H, s) and an anomeric proton (δ H 5.07, 1H, d, J = 7.2 Hz) of a β-glucose were observed. The positions of methoxy and glucose groups were further determined at C-4 of B-ring and at C-7 of A-ring respectively, based on the HMBC spectrum ( Figure 2). Therefore, compound 5 was identified as 3 -hydroxy-4 -methoxyisoflavone-7-O-β-D-glucopyranoside [27]. 8. 4 Hz, H-5′), corresponding to a ABX system of aromatic protons, suggesting a quceretin skeleton for 4. The 13 C NMR spectrum (Table 2), together with two anomeric protons of β-glucose and xylose moieties in the 1 H NMR spectrum, confirmed that 4 contained a quceretin with a xylopyranose and a glucopyranose moieties. As shown in the HMBC spectrum ( Figure 2), a glucose and a xylose moieties were linked at C-3 and C-2′′, respectively. With the above evidence, 4 was identified as quercetin 3-O-β-D-xylopyranosyl(1→2)-β-D-glucopyranoside [22]. Compound 5, obtained as a yellowish powder, was assigned with the molecular formula C22H22O10, determined from ESIMS, 13 C NMR, and DEPT spectroscopic data. The 1 H and 13 C NMR spectra (Table 1) showed resonance characteristics of isoflavone skeleton with a sugar moiety. The 1 H NMR spectrum showed the characteristic signals, δH 8.33 (1H, s), for H-2 of the isoflavone framework, and two typical ABX system aromatic protons system for A-ring and B-ring in 5. In addition, the signals for one aromatic methoxy group at δH 3.77 (3H, s) and an anomeric proton (δH 5.07, 1H, d, J = 7.2 Hz) of a β-glucose were observed. The positions of methoxy and glucose groups were further determined at C-4′ of B-ring and at C-7 of A-ring respectively, based on the HMBC spectrum ( Figure 2). Therefore, compound 5 was identified as 3′-hydroxy-4′-methoxyisoflavone-7-Oβ-D-glucopyranoside [27].
Laricitin 3-O-β-D-glucopyranoside (6) was isolated as a yellowish powder, and its ESIMS gave a quasimolecular ion at m/z 517.1 [M + Na] + (C22H22O13Na). The 1 H and 13 C NMR spectra of 6, as well as the substituted pattern, were similar to those of 3 (Table 2), except for the presence of an aromatic methoxy signal and the absence of a rhamnose moiety in 6. The methoxy group was located at C-3′ based on HMBC correlations of H-2′/C-3′ and OCH3/C-3′ ( Figure 2). Thus, compound 6 was identified as laricitin 3-O-β-D-glucopyranoside [23].    (Table 2), except for the presence of an aromatic methoxy signal and the absence of a rhamnose moiety in 6. The methoxy group was located at C-3 based on HMBC correlations of H-2 /C-3 and OCH 3 /C-3 ( Figure 2). Thus, compound 6 was identified as laricitin 3-O-β-D-glucopyranoside [23].
The others, compounds 1, 3, 7, 10, and 11, were also isolated and identified by comparing their physical and spectroscopic data with those of authentic samples and references.

The Optimization of Fingerprint Chromatographic Conditions
In order to obtain a good resolution within a short analysis time, the composition of mobile phase was optimized. Acidic mobile phase was used in order to suppress the ionization of phenolic hydroxyl groups of flavonoids. In this study, water and acetonitrile were chosen as the mobile phase for the low column pressure. We found that many interfering compounds were present in the plant material. Thus, the gradient elution program was carried out to separate these components in samples. Various acids, such as 0.1% ethylic acid, formic acid, and phosphoric acid in acetonitrile (B) and H 2 O (A) (v/v), were evaluated. Acetonitrile and water both containing 0.1% formic acid were chosen as the mobile phases because most components could be resolved under this condition. The elution program was as follows: 0-10 min, 15% B, 10-70 min, 15-28.8% B, 70-80 min, 28.8-38% B, 80-90 min, 38-100% B for HPLC, and 0-18 min, 15-21% B, 18-22 min, 21-32% B, 22-25 min, 32-100% for UPLC. Checking the UV spectra of the components recorded from 220 to 370 nm, 254 nm was finally selected for monitoring. As shown in Figure 3, 11 reference compounds in the real sample or standard mixture could be separated well within 20 min in UPLC, and good peak shapes were observed for all peaks.
Molecules 2020, 25, x FOR PEER REVIEW 6 of 17 hydroxyl groups of flavonoids. In this study, water and acetonitrile were chosen as the mobile phase for the low column pressure. We found that many interfering compounds were present in the plant material. Thus, the gradient elution program was carried out to separate these components in samples. Various acids, such as 0. Checking the UV spectra of the components recorded from 220 to 370 nm, 254 nm was finally selected for monitoring. As shown in Figure 3, 11 reference compounds in the real sample or standard mixture could be separated well within 20 min in UPLC, and good peak shapes were observed for all peaks. We studied the HPLC fingerprint profile (Figure 4) of the 80% EtOH crude extract of ACS by using the HPLC-UV method and identified 11 main peaks by comparing the retention times with the reference compounds isolated from the title plant. The mobile phase consisted of water (A) and acetonitrile (B) with 0.1% formic acid, using a gradient program of 15%-15%-28.8%-100% (B) in 0-10-70-75 min. In the fingerprint chromatograph, the retention times for compounds 1 to 11 were shown: 1: 8.  We studied the HPLC fingerprint profile (Figure 4) of the 80% EtOH crude extract of ACS by using the HPLC-UV method and identified 11 main peaks by comparing the retention times with the reference  We studied the HPLC fingerprint profile (Figure 4) of the 80% EtOH crude extract of ACS by using the HPLC-UV method and identified 11 main peaks by comparing the retention times with the reference compounds isolated from the title plant. The mobile phase consisted of water (A) and acetonitrile (B) with 0.1% formic acid, using a gradient program of 15%-15%-28.8%-100% (B) in 0-10-70-75 min. In the fingerprint chromatograph, the retention times for compounds 1 to 11 were shown: 1: 8

The Optimization of Complanatuside Quantification Analysis Chromatographic Conditions
According to the China Pharmacopoeia, the chromatographic conditions of complanatuside quantification analysis were set as follows. The mobile phase system was 0.1% H 3 PO 4 aq. (A) and acetonitrile (B), with the gradient program of 0-17 min, 21% B, 17-17.1 min, 21-100% B. The UV detection wavelength was 254 nm. The column temperature was kept at 25 • C. The flow rate was 1.0 mL/min and the injection volume was 20 µL. These conditions assured that complanatuside (Rt: 13.9 min) in the real sample was baseline separated from the neighbor peaks.

Optimization of Extraction Conditions
The extraction solvent and extraction times were optimized in order to achieve satisfactory extraction efficiency. The extraction efficiency of 40%, 60%, 80%, 95% ethanol, 40%, 60%, 80% pure methanol, and H 2 O were evaluated. The 40% ethanol and 60% methanol extract had the same yield of complanatuside, but the sample dissolved in methanol seems unstable, especially the target compound complanatuside, which decreased after 3 weeks (Figure 4). Thus, the 40% ethanol was selected as the most suitable solvent for extraction. Extraction methods containing ultrasonic and reflux extraction were tested and compared. Although the extraction efficiency was approximated, the former was easier and simpler to perform than the latter. The extraction times were also compared, resulting in 97% and 3% yield of complanatuside for the first and second extract, respectively ( Figure 5). Therefore, ultrasonication for 30 min in 40% ethanol extracted one time was chosen as the optimal extraction condition for the following experiments.
selected as the most suitable solvent for extraction. Extraction methods containing ultrasonic and reflux extraction were tested and compared. Although the extraction efficiency was approximated, the former was easier and simpler to perform than the latter. The extraction times were also compared, resulting in 97% and 3% yield of complanatuside for the first and second extract, respectively ( Figure 5). Therefore, ultrasonication for 30 min in 40% ethanol extracted one time was chosen as the optimal extraction condition for the following experiments.

Chromatographic Analyses of Astragali Complanati Semen
We studied the HPLC profile of the 40% ethanol extracts of 11 samples collected from the North, Middle, and South Chinese herbal medicine companies of Taiwan. As shown in Figure 6, the 6 samples from the Northern and the Middle of Taiwan had similar secondary metabolite patterns but the contents depended on the source. In all samples collected from the Southern Taiwan, 3 samples (SE, SF, SH) showed similar secondary metabolite patterns, but samples SC and SG were different in in their HPLC profiles. The results identified nine samples collected as ACS, whereas two other samples from the Southern area would be the adulterants due to the different HPLC patterns.

Chromatographic Analyses of Astragali Complanati Semen
We studied the HPLC profile of the 40% ethanol extracts of 11 samples collected from the North, Middle, and South Chinese herbal medicine companies of Taiwan. As shown in Figure 6, the 6 samples from the Northern and the Middle of Taiwan had similar secondary metabolite patterns but the contents depended on the source. In all samples collected from the Southern Taiwan, 3 samples (SE, SF, SH) showed similar secondary metabolite patterns, but samples SC and SG were different in in their HPLC profiles. The results identified nine samples collected as ACS, whereas two other samples from the Southern area would be the adulterants due to the different HPLC patterns.

Validation of Quantitative Analysis Method
The complanatuside calibration curve linear equation was y = 18,221.2x − 3915.1 and showed good linear regression (R2 = 0.9998) within test ranges (1.9 to 150 μg/mL). The LOD (S/N = 3), LOQ (S/N = 10), precision (inter-day and intraday, 60.0, 15.0, 1.9 μg/mL), injection precision (15 μg/mL), reproducibility (n = 5), and stability (n = 5) for complanatuside are summarized in Table 2. The RSD values of the injection, intra-day, inter-day, and stability variations were less than 2.99%, except that of reproducibility, which was 4.25%. The results of the recovery test are summarized in Table 3, and the RSD was 3.30%. Therefore, the HPLC-UV method was precise, accurate, and sensitive enough for quantitative analysis of complanatuside in ACS. Table 2. Method validation of complanatuside quantification by HPLC.

Validation of Quantitative Analysis Method
The complanatuside calibration curve linear equation was y = 18,221.2x − 3915.1 and showed good linear regression (R 2 = 0.9998) within test ranges (1.9 to 150 µg/mL). The LOD (S/N = 3), LOQ (S/N = 10), precision (inter-day and intraday, 60.0, 15.0, 1.9 µg/mL), injection precision (15 µg/mL), reproducibility (n = 5), and stability (n = 5) for complanatuside are summarized in Table 2. The RSD values of the injection, intra-day, inter-day, and stability variations were less than 2.99%, except that of reproducibility, which was 4.25%. The results of the recovery test are summarized in Table 3, and the RSD was 3.30%. Therefore, the HPLC-UV method was precise, accurate, and sensitive enough for quantitative analysis of complanatuside in ACS.

Quantitative Determination of Astragali Complanati Semen
This HPLC assay method was subsequently applied to quantify complanatuside in ACS obtained from different Chinese herbal medicine stores distributed in the North, Middle, and South of Taiwan. Their contents are listed in Table 4. The contents varied from 0.062% to 0.134%, which were acceptable according to both of the 2010 China pharmacopoeia and Taiwan Herbal Pharmacopeia (0.06%), except the samples of SC and SG (Table 4). In the present study, a simple, accurate, and efficient HPLC method was developed to evaluate the quality of commercial ACS by establishing the complanatuside quantification method. The results demonstrate that this method is accurate, reproducible, and could be readily employed as a suitable quality control method for ACS.
We established the HPLC fingerprint of ACS, and identified 11 compounds in the profile. Ten of eleven compounds were flavonoids, suggesting that flavonoids were the main components in the ACS. The collected 11 commercial ACS samples showed similar morphological features. The HPLC/UPLC methods as described above can distinguish true ACS from the false ones among 11 commercial samples collected from three different areas of Taiwan (North area: NF, NH, NI, NJ; Middle area: CD, CR; South area: SC, SE, SF, SG, SH). The fingerprint indicated that samples SC and SG were pseudo, and the other nine samples were genuine ACS. Samples SC and SG were further identified as the seeds of A. adsurgens, based on the referenced HPLC profile. Therefore, the developed HPLC-UV method can be readily used in quality assurance, as well as inspection of adulteration of ACS.

The Isolates and Fractions of the Antioxidation and Anti-NO Production Activities
All isolated flavonoid glycosides and compound 9 were evaluated for antioxidant and anti-inflammatory activities. In the antioxidant assay, compounds 1, 2, 3, and 8 showed more potency than the positive control vitamins C and E (Table 5). Crude extract and partial purified fractions were evaluated for antioxidant activity, and 80ED25 fraction showed a scavenging DPPH radical effect (ED 50 = 37.39 ± 0.33 µg/mL). In this study, all tested compounds showed the promising potency on anti-NO production, and compound 8 was the most inhibitory ( Figure 7). Although the isolated compounds 1, 2, 3, and 8 all exhibited obvious inhibition both on antioxidant and anti-NO production, 1 and 3 were selected as the potential target compounds based on their much bigger peaks than those of 2 and 8 in HPLC/UPLC profiles of ACS. production, 1 and 3 were selected as the potential target compounds based on their much bigger peaks than those of 2 and 8 in HPLC/UPLC profiles of ACS.   In previous reports, ACS extracts exhibited potent activities on DPPH, superoxide radical, hydroxyl radical, and Fe 2+ -induced lipid peroxidation [19,29]. In an in vivo study, ACS extracts were reported to improve ROS effect by increasing the blood's important antioxidant enzymes, superoxide dismutase and glutathione peroxidase [30]. Another study revealed that the total flavonoids extracted from ACS had comprehensive effects on immune response, including anti-NO, anti-ROS, lymphocyte activation, and critical immune index, including thymus and spleen, in an aging mouse model [31]. Many flavonoids and their derivatives yielded from the other plants also possessed antioxidant and anti-NO production activities [32]. Our results suggested that the flavonoid glycosides of ACS played an important role in antioxidant and anti-NO production activities, especially compounds 1, 2, 3, and 8. In addition, herbal medicines used in treating premature ejaculation and sexual dysfunction activities have been shown to contain flavonoid derivatives as their major ingredients [33,34] which supports the therapeutic option for choosing ACS since ACS are also high in flavonoid content.

Plant Material
The dried seed of Astragalus Complanatus (Astragali Complanati Semen; ACS) for the isolation was purchased from Tianshun Chinese herbal medicine pharmacy (Taipei, Taiwan

Sample Preparation for HPLC and UPLC Assays
All raw samples of ACS were ground into powder and filtered through a 65-mesh sieve (0.25 mm). An accurately weighed 0.5 g powder sample was placed into a 50 mL Centrifuge tube with stopper, and 25 mL 40% ethanol was added exactly to the centrifuge tube, and then ultrasonic extracted for 30 min (500 W, 40 kHz). The resultant mixture was adjusted to the original weight with extraction solvent, and the supernatant was removed through a 0.45 or a 0.22 µm membrane filter before HPLC and/or UPLC injection, respectively.

Complanatuside Quantification Analysis
Standard stock solution of complanatuside (11) was prepared with 40% ethanol. From the stock solution, a series of working standard solutions were prepared by dilution with 40% ethanol to the proper concentrations and stored at 4 • C. The mobile phase system for 11 quantification analysis was used as 0.1% H 3 PO 4 aq. (A) and acetonitrile (B), with the gradient program of 0-17 min, 21% B; 17-17.1 min, 21-100% B. The UV detection wavelength was 254 nm. The column temperature was kept at 25 • C. The flow rate was 1.0 mL/min and the injection volume was 20 µL. The UPLC fingerprint was performed on an Agilent 1200 Infinity Series liquid chromatography system (Santa Clara, CA, USA) equipped with a binary solvent delivery system (1290 Bin Pump), a diode array detector (1290 DAD), a column temperature controller (1290 TCC), and an autosampler (1290 Sampler). Chromatographic data were recorded and processed using Agilent chromatographic work station software. The UPLC experiment was carried out on an Agilent Poroshell 120, EC-C 18 Column (150 × 3.0 mm, 2.7 µm), which was protected by an EC-C 18 guard column (5.0 × 3.0 mm, 2.7 µm). The mobile phase condition was similar to that of HPLC, the binary gradient elution system consisted of 0.1% formic acid aq. (A) and 0.1% formic acid acetonitrile (B), and the UPLC fingerprint separation was achieved using the following gradient: 0-18 min, 15-21% B; 18-22 min, 21-32% B; 22-25 min, 32-100% B. The UV detection wavelength was 254 nm, the injection volume was 2 µL, and the flow rate was 0.6 mL/min.

Validation Procedure
The calibration curve was established by plotting the peak area (y) versus the concentration (x) of each analysis, using working standard solutions of complanatuside, and contained five different concentrations performed in triplicate. The stock solution was further diluted to a series of concentrations with 40% EtOH to determine the limit of detection (LOD) and limit of quantitation (LOQ) on the basis of the signal-to-noise ratio (S/N) of 3 and 10, respectively. The injection precision was determined consequently by replicate injection of the same standard solution (15.0 µg/mL), five times. The intra-day precision was determined by analyzing three replicates of each standard solution (60.0, 15.0, and 1.9 µg/mL) within one day, while inter-day variation was determined on three consecutive days. In order to evaluate the repeatability of the developed method, sample solutions prepared from five parallel samples (0.5 g, SF sample) were determined within one day. To choose one from the sample solutions, the complanatuside concentrations were analyzed at 0, 3, 6, 12, and 24 h to evaluate the stability of the solution at 25 • C. The recovery test was carried out by the standard addition method. The standard stock solution (150 µg/mL, 2.5 mL, accurately) was added into the sample (0.5 g, SF sample), and then extracted, processed, and quantified, and five parallel samples were prepared. The average recovery was calculated by the following equation: recovery (%) = (amount found − amount contained)/amount added × 100%, and RSD (%) = (SD/mean) × 100%. Samples from different regions of Taiwan were prepared as described above. An aliquot (20 µL) of the filtrate was directly subjected to HPLC analysis. Each sample was determined in triplicate. Chromatographic peaks of the samples were quantified by the external standard method.

Scavenging Activity of DPPH Radical Assay
The antioxidative activity of the isolates on DPPH (1,1-diphenyl-2-picrylhydrazyl)-free radical was measured using the method of Rangkadilok et al., with minor modifications [35]. An aliquot of each sample (120 µL, 200-5 µg/mL) or vitamin E (40-5 µg/mL) was mixed with 30 µL of 0.75 mM of DPPH methanol solution in a 96-well microplate. The mixture was shaken vigorously with an orbital shaker in the dark at room temperature for 30 min and the absorbance was measured at 517 nm with an ELISA reader. Methanol was used as the negative control by replacing the sample in the reaction solution. The DPPH radical scavenging activities of the test samples were compared to the negative control and vitamins C and E as the positive controls. The final results were performed as the concentrations of ED 50 , which is the concentration of each sample required to give 50% of the absorbance shown by the negative control.

LPS-Induced NO Production of Anti-Inflammatory Assays
The macrophage cell line RAW 264.7 was obtained from ATCC (Rockville, MD, USA) and cultured in DMEM containing 10% heat-inactivated fetal calf serum, 100 U/mL penicillin, and 100 µg/mL streptomycin, and grown at 37°C with 5% CO 2 in fully humidified air. Cells were plated at a density of 2 × 10 5 cells/well in a 96-well culture plate and stimulated with lipopolysaccharide (LPS, 1 µg/mL) in the presence or absence of different concentrations of tested compounds (20 µg/mL) for 18 h, simultaneously. All compounds were dissolved in DMSO and further diluted with sterile PBS and sterilized via a 0.2 µm filter. Nitrite (NO 2 − ) accumulation in the medium was used as an indicator of NO production which was measured by adding Griess reagent (1% sulfanilamide and 0.1% naphthylenediamine in 5% phosphoric acid). NaNO 2 was used to generate a standard curve, and nitrite production was determined by measuring optical density at 550 nm [36]. All experiments were performed in triplicate. NO production by LPS stimulation was designated as 100% for each experiment. A well-known iNOS inhibitor, quercetin, was employed as a positive control (IC 50 = 7.87 ± 0.14 µM).

Conclusions
In our study, flavonoid glycosides were the main components of Astragali Complanati Semen (ACS) and proved to have potent antioxidation and anti-NO effects. Our established HPLC-UV method can be used to determine the content of flavonol glycoside complanatuside (11) with good linear regression (R 2 = 0.9998) for commercial ACS in the Taiwanese market. Using our established HPLC/UPLC methods, nine out of eleven commercial samples collected from three areas of Taiwan were reliably identified to be genuine ACS.
Biological evaluation revealed that flavonol glycosides 1, 2, 3, and 8 isolated from ACS had more potent antioxidant activity (by DPPH) than the positive control vitamins C and E, and all the isolated flavonol derivatives showed promising effects on anti-NO production. By comparing all the isolates, both myricetin 3-O-β-D-xylopyranosyl(1→2)-β-D-glucopyranoside (1) and myricetin 3-O-β-D-glucopyranoside (3) showed the most obvious peaks in HPLC/UPLC. The results, together with 1 and 3 exhibiting more potent antioxidant and anti-NO production activities than that of 11, the quality control component for ACS in China and Taiwan pharmacopoeia, suggested that flavonol glycosides 1 and 3 can be the available compounds for the quality control of ACS.