Simultaneous Determination of Eleven Compounds in the Chinese Herbal Medicine Flemingia prostrata by LC-MS

Abstract

Objectives

Flemingia prostrata is an important component of Chinese herbal medicine, renowned for its diverse therapeutic applications attributed to its abundant bioactive compounds. Accurate quantification of the bioactive compounds in this plant is crucial for ingredient quality control and pharmacological studies.

Methods

An LC-MS assay method was developed and validated for the simultaneous quantification based on external calibration of eleven bioactive compounds in F. prostrata. The target analytes comprised five isoflavones, three dihydroflavones, and three other characteristic constituents, including aureol, syringaresinol and 4′-O-methylgallocatechin. Chromatographic separation was achieved on a reversed-phase C18 column with gradient elution, followed by mass spectrometric detection under optimized parameters.

Results

The method was fully validated in terms of linearity, limits of detection, precision, accuracy, and recovery. All eleven compounds exhibited good linearity within their respective concentration ranges, with satisfactory intra- and inter-day precision and acceptable recoveries. The established LC-MS assay was successfully applied to the quantitative analysis of F. prostrata samples collected from six different geographical origins. Significant variations in the contents of individual compounds were observed among samples, indicating that geographical origin has a substantial influence on the chemical composition of F. prostrata.

Conclusion

The results demonstrate that the proposed method is suitable for comprehensive quality evaluation and comparative analysis of F. prostrata, providing a robust analytical tool for its standardization, quality control, and further pharmacological investigation.

Keywords

traditional Chinese medicine flavonoids isoflavones dihydroflavones coumarins lignans flavanols liquid chromatography mass spectrometry

1. Introduction

Flemingia prostrata, a perennial herb belonging to the family Fabaceae, is widely distributed in southern China and Southeast Asia and has long been used in traditional Chinese medicine.¹ Modern pharmacological studies have demonstrated that extracts of F. prostrata exhibit a variety of biological activities, including anti-inflammatory, antioxidant, and antibacterial and antifungal effects,² which are largely attributed to its abundant phenolic and flavonoid compounds. However, the chemical composition of F. prostrata is complex and may be influenced by factors such as geographical origins, environmental conditions, and harvesting practices, posing challenges for its quality evaluation and standardization.

Previous phytochemical investigations of the genus Flemingia, particularly Flemingia macrophylla, have revealed a diverse array of bioactive secondary metabolites, including isoflavones, dihydroflavones, lignans, and related phenolic compounds.³ Aureol belongs to the coumarin class, which possesses antioxidant and anti-inflammatory properties,⁴ while syringaresinol, a lignan, has been reported to possess antioxidant, anti-osteoporotic, and cytoprotective activities.⁵ 4′-O-methylgallocatechin, a methylated flavan-3-ol, is recognized for its potential free radical scavenging and anti-inflammatory effects.⁶ Although these compounds have been documented in F. macrophylla, they have also been identified in F. prostrata, suggesting close chemotaxonomic relationships between these species and highlighting the importance of comprehensive chemical profiling for F. prostrata.

Liquid chromatography-mass spectrometry (LC-MS) has emerged as one of the most powerful analytical platforms for the characterization and quantification of complex constituents in Chinese herbal medicines, owing to its high sensitivity, selectivity, and capability for multi-component analysis.⁷ Although LC-MS-based methods have been widely applied for analysis of flavonoid-rich herbal materials, most reported studies have focused on a limited number of marker compounds or a single class of constituents, often relying on relative quantification or semi-quantitative approaches.^8,9 For species of Flemingia, existing analytical reports have primarily concentrated on F. macrophylla or individual compounds, while comprehensive quantitative studies on F. prostrata remain scarce. Moreover, F. prostrata contains structurally diverse secondary metabolites, including isoflavones, dihydroflavones, coumarins, lignans and flavanols, which present considerable analytical challenges due to their similar chromatographic behaviors and varied ionization efficiencies. To date, no validated LC-MS assay has been reported for the external calibration-based quantification of multiple representative compounds spanning these different structural classes in F. prostrata. Therefore, the development of a single, robust LC-MS method capable of accurately and simultaneously quantifying chemically diverse constituents is highly desirable.

In this study, a sensitive and reliable LC-MS method was established and validated for the simultaneous determination of eleven characteristic compounds in F. prostrata. The developed assay was subsequently applied to the quantitative analysis of plant samples collected from six different geographical origins to evaluate variations in chemical composition. This method provides a comprehensive analytical approach for the quality assessment of F. prostrata and offers a valuable tool for future studies on its pharmacological activity, resource evaluation, and standardization. Moreover, the proposed LC-MS assay has broad application potential in chemotaxonomic research and quality control of related Flemingia species and other flavonoid-rich Chinese herbal medicines.¹⁰

2. Materials and Methods

2.1. Chemicals and Reagents

Reference standards of the eleven target compounds, including five isoflavones, three dihydroflavones, aureol, syringaresinol and 4′-O-methylgallocatechin as listed in Table 1, were purchased from MUST (Chengdu, China) Biotechnology Co., Ltd. With purity greater than 98%. Acetonitrile (LC-MS grade), methanol (LC-MS grade), and formic acid (LC-MS grade) were obtained from Fisher Scientific (Waltham, MA, USA). Ultrapure water used for mobile phase was prepared using a Milli-Q water purification system (Millipore, Billerica, MA, USA). All other chemical reagents were of analytical grade.

Table 1.

Information and MS Parameters of the Eleven Compounds. Abbreviations: EP, Entrance Potential; CE, Collision Energy; CXP, Collision Cell Exit Potential

Name	CAS ID	Molecular weight	Molecular formula	Compound group	Q1 m/z	Q3 m/z	Dwell time (ms)	EP (V)	CE (V)	CXP (V)	Quantitative ion
genistein	446-72-0	270.24	C₁₅H₁₀O₅	isoflavone	271	153	20	10	41	14	Yes
genistein	446-72-0	270.24	C₁₅H₁₀O₅	isoflavone	271	215	10	10	35	12	Yes
genistin	529-59-9	432.38	C₂₁H₂₀O₁₀	isoflavone	433	271	20	10	20	18	Yes
genistin	529-59-9	432.38	C₂₁H₂₀O₁₀	isoflavone	433	215	10	10	61	12	Yes
pratensein-7-O-β-D-glucoside	36191-03-4	462.4	C₂₂H₂₂O₁₁	isoflavone	463	301	20	10	26	6	Yes
pratensein-7-O-β-D-glucoside	36191-03-4	462.4	C₂₂H₂₂O₁₁	isoflavone	463	286	10	10	54	20	Yes
prunetin	552-59-0	284.26	C₁₆H₁₂O₅	isoflavone	285	167	10	10	38	10	Yes
prunetin	552-59-0	284.26	C₁₆H₁₂O₅	isoflavone	285	242	20	10	45	14	Yes
sophoricoside	152-95-4	432.38	C₂₁H₂₀O₁₀	isoflavone	433	271	20	10	20	18	Yes
sophoricoside	152-95-4	432.38	C₂₁H₂₀O₁₀	isoflavone	433	215	10	10	61	12	Yes
flemichin D	57096-07-8	422.47	C₂₅H₂₆O₆	dihydroflavone	423	349	20	10	30	22	Yes
flemichin D	57096-07-8	422.47	C₂₅H₂₆O₆	dihydroflavone	423	231	10	10	39	8	Yes
kushenol E	99119-72-9	424.49	C₂₅H₂₈O₆	dihydroflavone	425	351	10	10	31	26	Yes
kushenol E	99119-72-9	424.49	C₂₅H₂₈O₆	dihydroflavone	425	232	20	10	40	16	Yes
lespedezaflavanone H	138822-59-0	492.6	C₃₀H₃₆O₆	dihydroflavone	493.3	419	10	10	28	24	Yes
lespedezaflavanone H	138822-59-0	492.6	C₃₀H₃₆O₆	dihydroflavone	493.3	363	20	10	39	20	Yes
aureol	88478-03-9	284.22	C₁₅H₈O₆	coumarin	285	242	20	10	44	16	Yes
aureol	88478-03-9	284.22	C₁₅H₈O₆	coumarin	285	167	10	10	37	8	Yes
syringaresinol	487-35-4	418.44	C₂₂H₂₆O₈	lignan	419	363	20	10	35	32	Yes
syringaresinol	487-35-4	418.44	C₂₂H₂₆O₈	lignan	419	307	10	10	47	22	Yes
4′-O-methylgallocatechin	PubChem CID:44257120	320.3	C₁₆H₁₆O₇	flavanol	321	153	20	10	19	14	Yes
4′-O-methylgallocatechin	PubChem CID:44257120	320.3	C₁₆H₁₆O₇	flavanol	321	139	10	10	36	12	Yes

2.2. Preparation of Standard Solutions

Individual stock solutions of the eleven reference standards were prepared by dissolving 2 mg of each compound in 1 mL methanol to a final concentration of 2 mg/mL and stored at 4 ^°C. A mixed standard working solution was prepared by serial dilutions of the stock solutions with methanol to obtain a series of calibration standards covering the expected concentration ranges in the samples. All standard solutions were centrifuged at 12,000 rpm for 10 min before injection.

2.3. Plant Materials and Sample Preparation

Six batches of Flemingia prostrata plant materials were collected from different geographical regions in China (Guangxi, Anhui, Henan, Guangdong, Jiangsu and Guizhou Provinces). The botanical identity of all samples was authenticated by a qualified taxonomist. The collected plant materials were air-dried at room temperature and ground into fine powder prior to analysis. Compound extraction procedure was performed following commonly used ultrasonic extraction procedures for flavonoids with modifications.^11,12 To be specific, accurately weighed powdered material (0.05 g) was transferred into a centrifuge tube and extracted with 2 mL of methanol by ultrasonic extraction for 60 min at room temperature. The extract was centrifuged at 12,000 rpm for 10 min, and the supernatant was collected. 1 μL of each sample was injected for LC-MS analysis. To achieve optimal chromatographic performance, a lower injection volume was chosen, provided the signal intensity was adequate. Three biological replicates were prepared for each sample. One LC-MS injection was performed for each biological replicate (No technical replicate was performed). The samples were subjected to LC-MS analysis in a random order.

2.4. LC-MS Instrumentation and Analytical Conditions

LC-MS conditions were established based on previously reported methods for flavonoids analysis with modifications.¹³ LC-MS analysis was performed using an ExionLC UHPLC (Shimadzu Corporation, Kyoto, Japan) coupled to an SCIEX 7500 QTRAP system (SCIEX, Framingham, MA, USA) equipped with an electrospray ionization (ESI) source. Separation was achieved on an XBridge Phenyl 2.1 mm x 10 cm, 3.5 μm column (Waters Corporation, Milford, MA, USA) at a temperature of 40 ^°C. The mobile phase consisted of solvent A (water containing 0.1% formic acid) and solvent B (acetonitrile containing 0.1% formic acid), delivered at a flow rate of 0.5 mL/min with gradient elution. The gradient program was optimized to achieve satisfactory separation of all analytes within a reasonable analysis time as follows: 0-0.6 min, 5% B; 0.6-9 min, 5-95% B; 9-12 min, 95% B; 12-12.5 min, 95-5% B; 12.5-15 min, 5% B. The mass spectrometer was operated in MRM mode targeting 22 transitions representing the eleven compounds (Table 1). The MRM assay was performed in positive ion mode without scheduling, as the target compounds exhibited stronger signals in this mode. The ion source parameters were set as follows: Curtain gas, 40 psi; ion source gas 1, 70 psi; ion source gas 2, 35 psi; CAD gas, 9 psi; source temperature, 350 ^°C. The ion spray voltage was 1.5 kV. Q0 Dissociation (Q0D) was set to simple mode at -10 V for all compounds.

2.5. Data Analysis

Raw MS data files were imported into the OS software for peak extraction (SCIEX). One transition per compound was selected for quantification, and the peak area corresponding to each transition was calculated by the software. Principal component analysis (PCA), hierarchical clustering analysis and one-way ANOVA with post-hoc test (Tukey’s HSD) for illustrating the results of compound profile in plant samples were performed using MetaboAnalyst web service.¹⁴

2.6. Method Validation

The developed LC-MS method was validated according to commonly accepted guidelines for analytical method validation (CLSI C62-A), including linearity (dynamic range), lower limit of quantification (LLOQ), limit of detection (LOD), precision, accuracy (recovery), and stability. Linearity was assessed by calibration curves constructed by plotting peak areas against corresponding concentrations of the analytes, and the correlation coefficients (R²) were calculated. The LLOQ was defined as the lowest concentration with a coefficient of variation (CV%) lower than 20% obtained from three replicates. LOD was defined as the lowest concentration that has a signal-to-noise ratio (S/N) greater than 3. Precision was evaluated by intra-day and inter-day repeatability, expressed as relative standard deviation (RSD, CV%). Intra-day precision was assessed using six replicates of one sample analyzed in one day. Inter-day precision was performed using three replicates on three consecutive days. Accuracy (recovery) was determined by a spike recovery experiment. Known amounts of standard solutions of three concentrations (low, medium and high) were spiked into a sample with known analyte concentration. Stability was evaluated by preparing a sample and storing it at room temperature for 1, 2, 4, 6, 8 and 12 hours before being analyzed, and a coefficient of variation (CV%) lower than 20% was considered acceptable.

2.7. Statistical Analysis

Coefficient of variation (CV%) and relative standard deviation (RSD) were calculated by dividing the standard deviation (SD) by the average. The calibration curves (y = ax + b) were constructed using the weighted regression method (1/x²) of the peak area of the analyte (y) versus actual concentrations (x). An acceptance criterion of the coefficient of determination (R²) value being greater than 0.99 and the accuracy at each concentration (accuracy is presented as a percentage of the calculated concentration over theoretical concentration) being between 85%-115% was employed to assess the linearity of the calibration curves. Recovery was calculated by comparing the actual concentration to theoretical concentration after spiking, and a recovery between 85% and 115% was considered acceptable. For quantification of the eleven target compounds in F. prostrata samples, the concentrations of the analytes were calculated using the corresponding calibration curves, and the results were expressed as mean ± standard deviation of three biological replicates. Final concentration results (ng/g) refer to dry weight of samples. For one-way ANOVA with post-hoc test (Tukey’s HSD), p < 0.05 was considered statistically significant.

3. Results

3.1. Optimization of LC-MS Conditions

A total of eleven compounds belonging to five compound classes were included in this assay method, namely isoflavones, dihydroflavones, coumarins, lignans and flavanols (Table 1). Chromatographic and mass spectrometric conditions were systematically optimized to achieve satisfactory separation and sensitive detection of the eleven target compounds. Different mobile phase systems, including methanol-water and acetonitrile-water combinations with various acidic modifiers, were evaluated. Different column types were also compared. As a result, the use of acetonitrile-water containing 0.1% formic acid combined with a Phenyl column provided overall balanced peak shape and ionization efficiency for all of the analytes. Gradient elution was optimized to resolve compounds with similar chemical structures, particularly the isobaric isomers that have the same molecular weight and fragment ions (such as genistin and sophoricoside), within a single analytical run. Under the optimized conditions, all eleven compounds were well separated with no significant interference from endogenous matrix components. Typical extracted ion chromatograms demonstrated clear baseline separation and stable retention times for all analytes (Figure 1). Mass spectrometric parameters were optimized individually for each compound to obtain maximum signal intensity. At least two transitions were selected for each compound, using one as quantitative transition and the other as qualitative transition (Table 1). The optimized LC-MS conditions allowed reliable detection and quantification of all target compounds in complex F. prostrata matrices.

Figure 1.

Extract ion chromatography of the eleven compounds. The exact retention time of each compound is marked above the peak

3.2. Method Validation

The method validation results of this developed LC-MS method were shown in Table 2. It showed good dynamic range for all eleven compounds within their respective concentration ranges basically ranging from 1 ng/mL to 1 μg/mL. Calibration curves exhibited correlation coefficients (R²) greater than 0.99, indicating excellent linearity within the dynamic range. The lower limit of quantification (LLOQ) and limits of detection (LOD), which ranged from 0.2-15.5 ng/mL and 0.05-0.35 ng/mL, respectively (Table 2), demonstrated the high sensitivity of the method, enabling the detection of trace-level constituents in F. prostrata samples. Intra-day and inter-day precision tests showed that the relative standard deviations (RSDs) were less than 10%, indicating good precision of the method. Accuracy was evaluated through recovery experiments at three spiking levels. The recoveries of all analytes were within acceptable ranges of 85% to 115%, indicating that the method is accurate. The stability of the analytes in the sample solution was confirmed for at least 12 hours at room temperature by CV% less than 10%. These results collectively demonstrate that the established LC-MS assay is reliable and suitable for simultaneous quantification of the eleven compounds.

Table 2.

Method Evaluation Result of the Eleven Compounds. Abbreviations: STD, Standard Deviation; CV%, Coefficient of Variation

		Genistein	Genistin	Pratensein-7-O-β-D-glucoside	Prunetin	Sophoricoside	Flemichin D	Kushenol E	Lespedezaflavanone H	Aureol	Syringaresinol	4′-O-methyl-gallocatechin
Precision	Average (ng/mL)	501.11	469.40	52.10	92.60	19.91	46.55	24.33	12.92	24.92	14.95	359.84
	STD	10.65	10.23	1.40	2.43	0.76	1.13	0.69	0.39	0.85	0.51	9.56
	CV% intraday	2.13%	2.18%	2.68%	2.62%	3.82%	2.43%	2.84%	3.02%	3.41%	3.41%	2.66%
	Average (ng/mL)	491.18	472.33	51.95	91.91	20.22	45.96	23.92	12.27	24.14	15.13	356.27
	STD	29.73	28.05	3.46	6.22	1.32	2.91	1.61	0.92	1.91	0.99	21.13
	CV% interday	6.05%	5.94%	6.66%	6.77%	6.53%	6.33%	6.73%	7.50%	7.91%	6.55%	5.93%
Recovery	Low concentration	107.25%	84.28%	84.99%	117.75%	86.16%	114.70%	117.87%	86.37%	91.13%	115.57%	112.12%
	Medium concentration	103.37%	115.24%	113.25%	89.78%	91.46%	88.80%	86.42%	97.55%	112.03%	92.94%	91.99%
	High concentration	98.01%	93.48%	111.25%	110.88%	114.63%	94.74%	91.99%	113.23%	108.97%	109.70%	106.11%
Stability	Average (ng/mL)	641.70	463.38	63.56	156.17	41.99	64.59	371.27	18.70	71.71	111.46	815.69
	STD	22.13	12.32	3.16	7.12	2.26	3.12	14.87	0.96	2.91	4.96	19.96
	CV%	3.45%	2.66%	4.98%	4.56%	5.39%	4.84%	4.00%	5.12%	4.06%	4.45%	2.45%
Linearity	Linear range (ng/mL)	3.5-1000	3.5-500	0.2-200	0.2-200	0.2-100	1.0-500	1.0-1000	0.4-100	0.4-200	0.4-100	15.5-2000
Linearity	r²	0.991	0.996	0.995	0.993	0.99	0.994	0.992	0.993	0.996	0.99	0.991
LOD	ng/mL	0.35	0.35	0.07	0.05	0.1	0.1	0.3	0.05	0.1	0.2	0.3

3.3. Quantitative Analysis and Comparison of F. prostrata Samples

The validated LC-MS method was successfully applied to the quantitative determination of the eleven target compounds in batches of F. prostrata collected from six different geographical origins. All analytes were detected in the tested samples, although their contents varied considerably among different batches (Table 3). Isoflavones of genistein and genistin were basically the predominant constituents in most samples, accounting for a major proportion of the total quantified flavonoids. Aureol, syringaresinol and 4′-O-methylgallocatechin were also consistently detected. The simultaneous determination of structurally diverse compounds highlights the applicability of the developed method for comprehensive chemical profiling of F. prostrata. Marked differences in the contents of individual compounds and total quantified constituents were observed among F. prostrata samples from different geographical origins, as shown in the PCA score plot in Figure 2A. Some compounds contributed more to the separation of geographical origins in PCA score plot, such as genistein, sophoricoside, kushenol E and 4′-O-methylgallocatechin according to the PCA loading plot shown in Figure 2B. A heatmap of hierarchical clustering analysis was used to show the expression patterns of individual compounds (Figure 2C). Certain samples (from Henan and Jiangsu) exhibited significantly higher levels of dihydroflavones (kushenol E, lespedezaflavanone H and flemichin D), lignans (syringaresinol) and flavanols (4′-O-methylgallocatechin), while others were characterized by elevated contents of isoflavone (genistein, genistin, pratensein-7-O-β-D-glucoside, prunetin and sophoricoside) or coumarins (aureol). The sample of one geographical origin (Guangxi) exhibited significantly low contents in all kinds of compounds compared to others. Figure 2D showed the statistical results of the differences of each compound among samples.

Table 3.

Concentrations of the Eleven Compounds in the F. prostrata Samples Collected From Six Geographical Origins. Concentrations are Expressed as Mean ± SD of Three Biological Replicates

Concentration (ng/g)
	Guangxi	Anhui	Henan	Guangdong	Jiangsu	Guizhou
genistein	501.11±44.91	973.58±22.04	786.13±49.61	704.05±26.29	307.25±28.71	578.08±62.62
genistin	469.4±37.45	442.82±22.06	437.67±8.52	642.18±16.43	423.41±41.42	364.8±17.57
pratensein-7-O-β-D-glucoside	52.1±4.55	47.19±2.82	38.52±1.11	46.87±3.11	56.92±11.24	139.72±6.48
prunetin	92.59±4.93	475.36±28.62	160.58±4.36	66.66±7.31	114.31±2.79	27.48±1.15
sophoricoside	19.91±1.64	78.97±3.18	11.66±0.48	95.67±2.53	21.26±2.77	24.43±2.47
flemichin D	46.55±2.96	53.24±1.68	109.65±2.67	57.75±2.72	114.61±1.75	5.69±0.25
kushenol E	24.32±1.02	283.27±11.48	709.27±22.96	365.19±17.17	814.56±22.5	30.99±1.31
Lespedezaflavanone H	12.91±1.24	18.4±1.8	30.89±0.7	13.18±1.43	34.23±0.95	2.56±0.14
Aureol	24.91±1.72	87.26±2.9	11.94±0.27	139.33±2.22	17.23±0.25	149.58±2.76
syringaresinol	14.95±1.16	82.07±2.15	194.57±4.91	143.7±4.12	145.02±3.98	88.43±0.63
4′-O-methylgallocatechin	359.84±6.48	181.7±4.07	1865.77±48.59	664.26±31.38	1798.4±13.25	24.12±0.36

Figure 2.

Compound profiles of the F. prostrata samples. (A) PCA score plot of the samples with three biological replicates. (B) PCA loading plot showing the variables that contribute to sample separation in PCA score plot. (C) The heatmap of hierarchical clustering analysis showing the compound expression patterns of the samples with three biological replicates. The suffix of R1 to R3 represents the first to third biological replicates. (D) Content variation of the eleven compounds across different groups. Box plots illustrate the interquartile range (25th to 75th percentiles), with whiskers extending to the minimum and maximum values. The central horizontal line and yellow diamond represent the median and mean, respectively. Statistical analysis was performed using one-way ANOVA followed by Tukey’s HSD post-hoc test. Pairs without statistically significant differences are labeled as “ns” (not significant, p ≥ 0.05); all unlabeled pairs indicate statistical significance (p < 0.05)

4. Discussion

The chemical composition of F. prostrata is characterized by structural diversity and complexity, which poses significant challenges for comprehensive quality evaluation using conventional analytical approaches. In the present study, a validated LC-MS method was established for the simultaneous quantification of eleven representative compounds belonging to different compound classes, including isoflavones, dihydroflavones, coumarins, lignans, and flavanols. Compared with previously reported methods that focused on individual compounds or a single class of constituents, the developed assay provides a more comprehensive and efficient strategy for chemical characterization of F. prostrata.

The optimized LC-MS conditions enabled satisfactory chromatographic separation and sensitive detection of all analytes despite their structural similarities and differences (Figure 1). The use of gradient elution combined with mass spectrometric detection effectively minimized matrix interference and allowed reliable quantification in complex herbal matrices. The favorable validation results, including good linearity, precision, accuracy, and sensitivity, further confirm the robustness and reliability of the proposed method. These characteristics are essential for multi-component analysis in traditional Chinese medicine, where batch-to-batch variability is often substantial.

Flavonoids, an important group of bioactive compounds widely distributed in Chinese herbal medicine, have exhibited diverse pharmacological activities.^10,15,16 In our previously reported assay method,¹³ two flavonoid subclasses—isoflavones and dihydroflavones—were included. Isoflavones occur in many plant species, and are especially high in soybeans. Studies have found their potential ability to prevent cancer or other chronic diseases.^17,18 Dihydroflavones are various aromatic, colorless ketones derived from flavone that often occur in plants as glycosides. They have shown potential therapeutic anti-inflammatory and immunomodulatory effects.¹⁹ In the present study, the scope of the aforementioned assay method was further expanded to achieve a more comprehensive compound coverage by incorporating three additional classes of bioactive constituents, namely coumarins, lignans, and flavanols. Coumarins represent a class of naturally occurring phenolic compounds commonly found in medicinal plants and are well recognized for their broad spectrum of biological activities, such as antioxidant,²⁰ anti-inflammatory²¹ and anticoagulant effects.²² These compounds have been reported to contribute significantly to the therapeutic efficacy of various traditional herbal medicines. Lignans are another important group of plant-derived phenolic compounds, widely distributed in medicinal herbs and dietary plants. They have been suggested to play a role in the prevention of chronic diseases through modulation of oxidative stress and inflammatory pathways due to their diverse bioactivities, including antioxidant, anti-inflammatory and antitumor effects.²³ Flavanols, a subclass of flavonoids characterized by the flavan-3-ol skeleton, are abundant in many herbal medicines and plant-derived foods. These compounds are known for their ability to scavenge free radicals and modulate cellular signaling pathways, contributing to their wide range of biological functions such as strong antioxidant capacity,²⁴ cardioprotective effect²⁵ as well as osteoprotective properties.²⁶ Although only one representative compound from each class of coumarin, lignan, and flavanol was included in the present study, the structural similarity within each class suggests that the assay could be readily expanded to incorporate additional target compounds as required by specific analytical objectives. This flexibility enhances the adaptability of the method and supports its broader application in natural product research and comprehensive chemical profiling.

Application of the validated method to samples of F. prostrata from six different geographical origins revealed pronounced variations in the contents of individual compounds and total quantified constituents (Figure 2A). The observed chemical variability may be attributed to a combination of environmental and biological factors, including climate, soil composition, altitude, and ecological conditions, as well as genetic diversity within F. prostrata populations. More importantly, the results indicate that no single geographical origin was able to provide consistently high levels of all quantified compounds. For instance, as shown in Figure 2C and D and Table 3, samples from Henan and Jiangsu exhibited significantly higher contents of dihydroflavones (kushenol E, lespedezaflavanone H and flemichin D), lignan (syringaresinol), and flavanol (4′-O-methylgallocatechin), but showed relatively lower levels of coumarin (aureol) compared with samples from Guangdong and Guizhou, and lower levels of isoflavones (genistein, genistin, pratensein-7-O-β-D-glucoside, prunetin or sophoricoside) compared with samples from Anhui, Guangdong and Guizhou. In contrast to the absence of an “ideal” sample with uniformly high contents of all compounds, one sample (from Guangxi) was characterized by uniformly low levels across all compound classes. Taken together, these findings highlight the necessity of including multiple representative compounds, rather than relying on a limited number of marker constituents, for accurate and reliable quality assessment of herbal materials. From a practical perspective, while quality evaluation is useful for identifying and excluding low-quality samples, achieving a more balanced and comprehensive chemical profile may be better accomplished by combining materials from multiple origins or batches rather than relying on a single source.

In terms of application, the established LC-MS assay offers significant advantages for quality control and standardization of F. prostrata. By simultaneously quantifying multiple bioactive constituents, the method provides a more comprehensive evaluation of chemical consistency than single-marker approaches. Moreover, the assay may serve as a valuable tool for chemotaxonomic studies, enabling comparison of F. prostrata with closely related Flemingia species, such as F. macrophylla. In addition, the quantitative data generated by this method may facilitate future pharmacological and pharmacokinetic studies by providing reliable information on the distribution and abundance of bioactive constituents in F. prostrata and in other biological samples such as cultured cells or mouse models.

While flavonoids are the primary bioactive constituents of Flemingia, other chemical classes also contribute to its multifaceted therapeutic effects in Traditional Chinese Medicine. Therefore, in this present study, in addition to isoflavones and dihydroflavones, we included representatives from three additional categories (coumarins, lignans, and flavanols) to develop a robust LC-MS method capable of simultaneous analysis in a single injection. This panel serves as a representative cross-section of the Flemingia chemical profile, but these compounds do not encompass the entirety of the plant’s bioactivity; however, the current assay is designed for extensibility, allowing for the future inclusion of additional targets or even application to other similar biological samples. What’s more, sample preparation procedure could be further optimized to increase the extraction efficiency for target compounds depending on sample types. For example, homogenization procedure is different between fresh herbal samples and dried ingredients such as the roots we used in this study. Additional soaking or ultrasonic extraction in solvents would be necessary for samples that are dry and tough. Our present study is a proof-of-concept study which only included three kinds of herbs for flavonoid extraction, so if this assay method would be used for other types of samples, the sample preparation procedure should be optimized accordingly.

There are several limitations in the present study. First, internal standards were not included because of the inaccessibility of isotope-labeled counterparts, thus this assay method is only an external calibration-based quantification. The drawback of not involving internal standards may fail to eliminate the error existing in compound extraction and mass spectrometry analysis. As a suboptimal alternative, structural analogues could be used as internal standards. However, their performance must be thoroughly evaluated in terms of chromatographic and mass spectrometric performance. Therefore, the selection and evaluation of internal standards were not included in this study but will be considered in future work. The second limitation is that sample preparation recovery rates and matrix effects during LC-MS analysis were not assessed, primarily due to the lack of an ideal blank matrix—i.e., a sample free of all target compounds. A suboptimal approach to evaluate these parameters would involve using structural analogues for which blank matrices are available, and this will be explored in future studies. Lastly, due to the limited sample size, our findings and interpretations regarding differences among samples’ geographic origins are preliminary. Additional experiments with larger sample sets are necessary before drawing definitive conclusions.

5. Conclusion

In this study, a sensitive and reliable LC-MS method was successfully developed and validated for the simultaneous quantification of eleven representative analytes belonging to five bioactive compound groups (isoflavones, dihydroflavones, coumarins, lignans and flavanols) in F. prostrata. The established assay demonstrated satisfactory performance in terms of linearity, precision, accuracy and sensitivity, enabling comprehensive analysis of structurally diverse constituents within a single run. Application of the method to samples collected from six different geographical origins revealed significant variations in chemical composition, highlighting the influence of origin on the quality of F. prostrata. The proposed LC-MS approach provides a robust and efficient tool for quality evaluation, standardization, and comparative studies of F. prostrata, and offers broad potential for future pharmacological research of Flemingia species and related Chinese herbal medicines.

Footnotes

Acknowledgement

The Authors would like to thank the staff at the mass spectrometry facility at the Center of Medical and Health Analysis in Peking University Health Science Center for maintenance of LC-MS instruments and technical support.

ORCID iD

Jiao Liu

Ethical Consideration

Ethical Approval is not applicable for this article.

Consent to Participate

There are no human subjects in this article and informed consent is not applicable.

Author contributions

BZ, JL and FW conceived and designed the experiments. BZ, ZZ YL and YZ conducted the sample collection and preparation. DW, DX, JAL and JZ developed the assay method. JZ, JL and FW wrote the manuscript. JL and FW oversaw the completion of this study and edited the manuscript. All authors read and approved the final manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Natural Science Foundation of Hebei Province (H2023209038 and H2023209072), Major Science and Technology Support Programme Projects in Hebei Province (24297701Z), Hebei Provincial Administration of Traditional Chinese Medicine Research Plan Project (2026104 and 2020222), National Natural Science Foundation of China (NSFC) project (82174212 and 31700898), Graduate Innovation Project of North China University of Science and Technology (2026B19), and Scientific Research Project of Colleges and Universities of Hebei Province (No.QN2025501).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data used to support the findings of this study are available from the corresponding author upon request.*

Statement of Human and Animal Rights

This article does not contain any studies with human or animal subjects.

References

Wei

Lin

Tang

Liang

Qin

. Comparative analysis of complete chloroplast genomes of Flemingia prostrata and Flemingia macrophylla, two commonly used medicinal plants in southern China. Frontiers in plant science. 2025;16:1591427. doi:10.3389/fpls.2025.1591427.

Huang

, et al. Flemiphilippinin A induces paraptosis in lung cancer cells via c-Myc-driven endoplasmic reticulum stress and CHOP-mediated mitochondrial dysfunction. Phytomedicine. 2025;146:157098. doi:10.1016/j.phymed.2025.157098.

Qin

Cui

, et al. The complete chloroplast genome of Flemingia macrophylla (Willd.) Prain (Fabaceae) from Guangxi, China. Mitochondrial DNA Part B, Resources. 2021;6(12):3378-3380. doi:10.1080/23802359.2021.1997124.

Di Stasi

. Coumarin Derivatives in Inflammatory Bowel Disease. Molecules (Basel, Switzerland). 2021;26(2):422. doi:10.3390/molecules26020422.

Mssillou

Bakour

Amrati

Oraibi

Elmobark

Khalid

. Syringaresinol: Exploring Its Natural Sources, Bioactivity, and Health Benefits. Chemistry & biodiversity. 2025;22(10):e00352. doi:10.1002/cbdv.202500352.

Shi

Yang

You

, et al. Updates on the chemistry, processing characteristics, and utilization of tea flavonoids in last two decades (2001-2021). Crit Rev Food Sci Nutr. 2023;63(20):4757-4784. doi:10.1080/10408398.2021.2007353.

Wang

Zhang

, et al. The Current Application of LC-MS/MS in Pharmacokinetics of Traditional Chinese Medicines (Recent Three Years): A Systematic Review. Current drug metabolism. 2020;21(12):969-978. doi:10.2174/1389200221666201009142418.

Al-Dalahmeh

Al-Bataineh

Al-Balawi

, et al. LC-MS/MS Screening, Total Phenolic, Flavonoid and Antioxidant Contents of Crude Extracts from Three Asclepiadaceae Species Growing in Jordan. Molecules. 2022;27(3):859. doi:10.3390/molecules27030859.

David

Norazhar

Che Zain

. Integrating LC-MS/MS and Molecular Networking for Advance Analysis and Comprehensive Flavonoids Annotation. J Sep Sci. 2025;48(7):e70230. doi:10.1002/jssc.70230.

10.

Yunusoğlu

Ayaz

Dovankaya

. Pharmacological, medicinal and biological properties of flavonoids: A comprehensive review. Journal of Research in Pharmacy. 2025;29(2):561-584. doi:10.12991/jrespharm.1661054.

11.

Jiang

Zeng

Liu

. Optimization of ultrasonic-assisted extraction of total flavonoids from Oxalis corniculata by a hybrid response surface methodology-artificial neural network-genetic algorithm (RSM-ANN-GA) approach, coupled with an assessment of antioxidant activities. RSC Adv. 2024;14(52):39069-39080. doi:10.1039/d4ra05077k.

12.

Zhang

Yin

Guo

Chen

Zhang

. Ionic liquid-based ultrasonic-assisted extraction coupled with HPLC to analyze isoquercitrin, trifolin and afzelin in Amygdalus persica L. flowers. BMC Chem. 2023;17(1):102. doi:10.1186/s13065-023-01018-w.

13.

Zhang

Zhou

Liu

, et al. Development of an LC-MS Based Assay for Quantifying Eight Flavonoid Compounds and Its Application in Chinese Herbal Medicine Analysis. Natural Product Communications. 2025;20(8):1934578X251372414. doi:10.1177/1934578X251372414.

14.

Pang

Zhou

, et al. MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Res. 2024;52(W1):W398-w406. doi:10.1093/nar/gkae253.

15.

Dias

Pinto

Silva

AMS

. Plant Flavonoids: Chemical Characteristics and Biological Activity. Molecules. 2021;26(17):5377. doi:10.3390/molecules26175377.

16.

Cheng

Bell

Lamport

Williams

. Dietary Flavonoids and Human Cognition: A Meta-Analysis. Mol Nutr Food Res. 2022;66(21):e2100976. doi:10.1002/mnfr.202100976.

17.

Messina

Barnes

Setchell

. Perspective: Isoflavones-Intriguing Molecules but Much Remains to Be Learned about These Soybean Constituents. Adv Nutr. 2025;16(5):100418. doi:10.1016/j.advnut.2025.100418.

18.

Chen

. Utilization of Isoflavones in Soybeans for Women with Menopausal Syndrome: An Overview. Int J Mol Sci. 2021;22(6):3212. doi:10.3390/ijms22063212.

19.

Zaragozá

Villaescusa

Monserrat

Zaragozá

Álvarez-Mon

. Potential Therapeutic Anti-Inflammatory and Immunomodulatory Effects of Dihydroflavones, Flavones, and Flavonols. Molecules (Basel, Switzerland). 2020;25(4). doi:10.3390/molecules25041017.

20.

Pouladvand

Azarnia

Fathi

Zeinali

Vesali

Tavana

. Coumarin attenuates cyclophosphamide-induced premature ovarian failure in mice by suppressing oxidative stress and apoptosis. Biomed Pharmacother. 2025;191:118528. doi:10.1016/j.biopha.2025.118528.

21.

Nayeli

Maribel

Enrique

, et al. Anti-inflammatory activity of coumarins isolated from Tagetes lucida Cav. Nat Prod Res. 2020;34(22):3244-3248. doi:10.1080/14786419.2018.1553172.

22.

Liao

Wang

, et al. Graveoumarins A-C: chiral resolution, absolute configuration, and anticoagulant/anti-inflammatory activities of 3'-methyl-3'-butenyl coumarins from ruta graveolens L. Pharm Biol. 2026;64(1):17-26. doi:10.1080/13880209.2025.2599599.

23.

Jang

Kim

Cho

. Antioxidant, Anti-Inflammatory, Anti-Menopausal, and Anti-Cancer Effects of Lignans and Their. Metabolites. 2022;23(24):15482. doi:10.3390/ijms232415482.

24.

Pérez-Torres

Castrejón-Téllez

Soto

Rubio-Ruiz

Manzano-Pech

Guarner-Lans

. Oxidative Stress, Plant Natural Antioxidants, and Obesity. International journal of molecular sciences. 2021;22(4):1786. doi:10.3390/ijms22041786.

25.

Luo

Jian

Liu

Jiang

Muhammad

Wang

. Flavanols from Nature: A Phytochemistry and Biological Activity Review. Molecules. 2022;27(3):719. doi:10.3390/molecules27030719.

26.

Chen

Huang

Zhang

, et al. The effect of flavan-3-ols on ovariectomy-induced bone loss in mice and the potential mechanisms. The Journal of nutritional biochemistry. 2025;145:110001. doi:10.1016/j.jnutbio.2025.110001.