Comparison of Chemical Compositions and Antioxidant Activities for the Immature Fruits of Citrus changshan-huyou Y.B. Chang and Citrus aurantium L.

Quzhou Aurantii Fructus (QAF), the dried immature fruit of Citrus changshan-huyou Y.B. Chang, is similar to Aurantii Fructus (AF), the dried immature fruit of Citrus aurantium L. or its cultivars, in terms of composition, pharmacological action, and appearance. However, potential chemical markers to distinguish QAF from AF remain unknown owing to the lack of a comprehensive systematic chemical comparison aligned with discriminant analysis. To achieve a better understanding of the differences in their composition, this study aimed to identify the basic chemical compounds in QAF (n = 42) and AF (n = 8) using ultra-performance liquid chromatography coupled with electron spray ionization and quadrupole time-of-flight mass spectrometry (UPLC−QTOF/MS) and gas chromatography coupled with mass spectrometry (GC−MS). Principal component analysis (PCA), orthogonal partial least squares-discriminant analysis (OPLS−DA), and hierarchical clustering analysis (HCA) were used to further analyze, screen, and verify potential chemical markers; the antioxidant capacity was assayed in vitro. A total of 108 compounds were found in QAF and AF, including 25 flavonoids, 8 limonoids, 2 coumarins, and 73 volatile components. The chemometric analysis indicated that the main components in QAF and AF were very similar. Trace differential components, including 9 flavonoids, 2 coumarins, 5 limonoids, and 26 volatile compounds, were screened as potential chemical markers to distinguish between QAF and AF. Additionally, the antioxidant capacity of QAF was found to be greater than that of AF. This research provides insights into the quality control and clinical application of QAF.


Introduction
Rutaceae is a significant plant family that has complex kinship and contributes to food and medicine, playing an indispensable role in daily life. Citrus changshan-huyou Y.B. Chang (HY), a cultivar belonging to the Rutaceae family, is a hybrid of Citrus aurantium L. (sour orange, SC) and Citrus grandis (L.) Osbeck (pummelo) [1]. It has been cultivated for approximately 600 years since the Ming Dynasty, mainly in Quzhou City, Zhejiang Province; its cultivation has now become one of the mainstay industries in the local agricultural economy, offering a great variety of agricultural and subsidiary products. Quzhou Aurantii Fructus (QAF), the dried immature fruit of HY, is gathered from the end of June to July. In 2018, it was selected as the new "Zhejiang Eight Flavors" cultivar of traditional Chinese medicine, with efficacy in regulating qi and relieving stagnation. The major bioactive constituents in QAF include volatile oils, flavonoids, terpenes (especially limonoids), of 50 samples are presented in Figure 1. The major components in QAF and AF were well separated and detected within 30 min. A total of 35 compounds were identified in both QAF and AF, based on the accurate mass determination of their precursor ions and MS E fragments in positive and negative ion modes. An overview of all major chromatographic peaks is presented, along with their retention times, MS data, and corresponding chemical compositions, in Table 1. The primary chemical components in QAF and AF were flavonoids, particularly flavanone, flavonoid glycosides, polymethoxyflavone, and flavone. However, other compounds were also present, such as coumarins and limonoids. The compounds are labelled in the representative chromatograms for the QAF and AF samples in positive and negative modes (Figure 2A-D).

UPLC−QTOF/MS Analysis and Identification
The chemical compositions of QAF and AF were successfully identified under optimal chromatographic and MS conditions; TIC obtained in the positive and negative ion modes of 50 samples are presented in Figure 1. The major components in QAF and AF were well separated and detected within 30 min. A total of 35 compounds were identified in both QAF and AF, based on the accurate mass determination of their precursor ions and MS E fragments in positive and negative ion modes. An overview of all major chromatographic peaks is presented, along with their retention times, MS data, and corresponding chemical compositions, in Table 1. The primary chemical components in QAF and AF were flavonoids, particularly flavanone, flavonoid glycosides, polymethoxyflavone, and flavone. However, other compounds were also present, such as coumarins and limonoids. The compounds are labelled in the representative chromatograms for the QAF and AF samples in positive and negative modes (Figure 2A-D).     Note: a These compounds were accurately identified with reference standards; b errors (ppm) were obtained by formula prediction software in the mass spectrometer; c these compounds were identified in QAF and AF for the first time; Arabic figures are the serial numbers of compounds according to RT; P and N represent the peaks under positive and negative ions separately in Peak No.; RT = retention time; and MF = molecular formula.

Flavonoids
Flavonoids are considered the most significant bioactive constituents obtained from QAF and AF, with the same nuclear 2-phenylchromone, and are classified into flavones, flavanones, flavonols, and isoflavones. They exhibit similar dissociation mechanisms in mass spectrometry after protonation, such as the retro-Diels-Alder reaction (RDA), and absorption wavelengths in the ranges of 300-400 nm and 240-280 nm.

Limonoids
Limonoids, highly oxygenated triterpenoid compounds, are the main cause of bitter taste in their aglycone and glucoside forms [30]. In this study, eight compounds were identified as limonoids through UPLC−QTOF/MS analysis, and their structures were either identified or provisionally assigned. Compounds 27 and 34 were identified as limonin and obacunone, respectively, based on a comparison with standard compounds. ] − ions were produced from decarboxylation in the seven-membered ring. Compound 15 was deduced to be a nomilin glucoside by comparison with data from a previous study [31]. Compounds 12, 16, 24, 28, and 31 were presumed to be obacunoic acid-17-β-D-glucoside, nomilinic acid 17-β-D-glucoside, obacunoic acid, nomilinic acid, and nomilin, respectively, using similar methods [18,26,27] were found in both compounds. Compound 23 was identified as isomeranzin, based on a comparison with the reference standard. Compound 11 was deduced to be meranzin through a comparison with compound 23 and information from previous studies [18,27].
Additionally, very similar chromatographic peaks appeared in the 42 batches of QAF from different plantings, based on both positive and negative ion modes of UPLC−QTOF/MS. These were also very similar to those in the eight batches of AF from different planting locations ( Figure 1). However, there were clear distinctions in the abundance of some chromatographic peaks between QAF and AF, indicating that the quantities of the corresponding compounds differed between QAF and AF.

Analysis of Volatile Compounds by GC−MS
All 50 batches of samples and n-alkane solution were analyzed, and TIC was acquired in the full scan mode for QAF and AF ( Figure 2C,D).
In total, 73 compounds were identified in QAF and AF ( Table 2). The relative content of the volatile compounds in QAF and AF was calculated using the area normalization method without a correction factor. The area percentages (%) of the volatile compounds in the 8 batches of AF and 42 batches of QAF are listed in Table 2. Limonene (48), a natural cyclic monoterpene, was the major compound in the volatile oils of both QAF and AF, with an average percentage of 65.76% (range 60.64-71.09% for 42 batches of QAF) and 85.93% (range 79.98-88.61% for 8 batches of AF), respectively. This result is consistent with those of previous studies [32,33]. The other high average percentages (≥1.00%) in QAF were γ-terpinene (51,8.86%), germacrene D (81, 7.99%), germacrene B (92, 1.91%), β-myrcene (43, 1.13%), α-cadinol (102, 1.02%), and dysoxylonene (88, 1.00%); while the other high average percentages (≥1.00%) in AF were γ-terpinene (51, 2.66%), linalool (53, 1.86%), β-myrcenein (68), and perillaldehyde (69) were found only in AF; the average relative content was below 0.05%. These results suggest a distinct difference in the chemical composition of the volatiles present in QAF and AF.  Note: a The retention index (RI) using DB-5MS column is calculated using closely eluted n-alkanes as the standard; b the retention times of these compounds were shorter than the n-alkanes that can be detected under these conditions; and ND: the target compounds were not detected or the contents of them were less than the quantitation limit in the samples.  Figure 3A,B). In addition, the samples in each group were essentially within a 95% confidence ellipse. These results demonstrate that the PCA model was reasonable and acceptable. Two distinct groups corresponding to QAF and AF were observed based on the UPLC−QTOF/MS and GC−MS data, indicating that the chemical compounds in QAF and AF were significantly different. As shown in Figure 3A,B, the data points for the QAF group were more closely clustered than those of the AF group, indicating that there were no clear differences in the chemical compositions in the QAF group even though these samples came from 14 different planting bases in four provinces. This was confirmed by the results of PCA based on UPLC−QTOF/MS and GC−MS data for the QAF group from 14 different planting bases ( Figure S1). As shown in Figure 3A,B, the data points for the QAF group were more closely clustered than those of the AF group, indicating that there were no clear differences in the chemical compositions in the QAF group even though these samples came from 14 different planting bases in four provinces. This was confirmed by the results of PCA based on UPLC−QTOF/MS and GC−MS data for the QAF group from 14 different planting bases ( Figure S1).

OPLS−DA
OPLS−DA is a relational model between omics data and a set of samples. In this study, a supervised OPLS−DA method was adopted to further identify differences in chemical composition between QAF and AF and to screen potential chemical markers.
In the OPLS−DA model, Q 2 and R 2 are vital parameters for evaluating the rationality of the model. In this study, Q 2 and R 2 (Q 2 = 0.961, R 2 X = 0.702, R 2 Y = 0.983) were both greater than 0.5 in UPLC−QTOF/MS according to Simca 14.1. A permutation test for OPLS−DA was performed 200 times to assess the predictability of the model ( Figure S2). All test boxes were lower than the original boxes, and the intersections of curvilinear regressions and coordinate axes were in the negative semi-axis, indicating that the models were acceptable. A three-dimensional (3D) score scatterplot derived from the UPLC−QTOF/MS data is shown in Figure 3C. The data points for the 50 samples were classified into two groups corresponding to QAF and AF. These results were consistent with the results of PCA.
However, the major components, including narirutin (6), naringin (8), hesperidin (9), neohesperidin (10), and naringenin (19), showed no difference between QAF and AF. Previous research has shown that these components were the principal biologically active ingredients assimilated in rat plasma after the oral ingestion of AF and QAF extracts [35]. The pharmacological functions of these flavonoids have been shown to mainly regulate gastrointestinal dysmotility [19], which is in accordance with the conventional clinical applications of QAF and AF. It is well-known that all the pharmacological activities of herbal medicines are significantly related to the composition of their bioactive compounds, implying that QAF and AF have very similar pharmacological effects.
The relevant R 2 X = 0.766, R 2 Y = 0.990, and Q 2 = 0.981 in the OPLS−DA model from the GC−MS data indicated that the model had good prediction and goodness-of-fit. The 3D score scatterplot of OPLS−DA is displayed in Figure 3E and suggests two separate groups, corresponding to QAF and AF. Based on the permutation test results, shown in Figure S3, the models were appropriate. A total of 26 volatile compounds with VIP values > 1 and p < 0.05 ( Figure S4) were selected as chemical markers for discriminating between QAF and AF. Previous reports have observed differences between the volatile components of QAF and AF using HS-GC-IMS [2]; however, the total average relative content of the 26 differential components in the volatile oils of QAF and AF was 13.20%.

HCA
To further validate the results of OPLS−DA, area data of all differential peaks with VIP > 1 and p < 0.05 in GC−MS and UPLC−QTOF/MS were imported into Origin 2023 software for HCA.
The data was analyzed using the group average as a clustering method and similarity as a distance type. In the cluster dendrogram, the samples were divided into two groups: one group comprised only the species QAF and the other group was composed of the species AF. Using a similarity > 70% as the standard, all 42 batches of QAF and 8 batches of AF were correctly classified by HCA ( Figure 4A). The outcome of HCA was in agreement with that of the PCA and OPLS−DA, indicating that the selected differential components were valid and could discriminate between QAF and AF.  A heatmap was employed to visualize the differences between QAF and AF, which included all the peaks of the differential components ( Figure 4B). The quantities of eriocitrin (2), neoeriocitrin (3), nomilinic acid (28) (103), and juniper camphor (106) were higher in QAF; the contents of the other differential components were higher in AF.
In total, all components, including nine flavonoids, two coumarins, five limonoids, and twenty-six volatile compounds, could serve as biological markers to distinguish between QAF and AF and help to verify the botanical origin of crude drugs in an application.
Rutaceae is an important source of food and medicine and has played an important role in the history of traditional Chinese medicine, mostly in qi-regulating drugs, and has a good influence on the digestive and respiratory systems on the basis of abundant flavonoids [7,36]. Flavonoids are the main components of QAF and AF that exert the pharmacological effect of regulating gastrointestinal motility [37,38], and they are also the main indicators in the comprehensive quality evaluation model of medicinal herbs [39]. As shown in the TIC (Figure 1), the main components of QAF and AF were similar in composition, but there existed differences in trace components. Chemometric analysis validated the differences presented by the TIC and presented them in a more visual way, while specific differential components were screened out, which provided a basis for the identification of herbs. Fingerprint analysis combined with clustering analysis could dis- A heatmap was employed to visualize the differences between QAF and AF, which included all the peaks of the differential components ( Figure 4B). The quantities of eriocitrin (2), neoeriocitrin (3), nomilinic acid (28) (98), T-muurolol (100), cadin-4-en-10-ol (101), α-cadinol (102), neointermedeol (103), and juniper camphor (106) were higher in QAF; the contents of the other differential components were higher in AF.
In total, all components, including nine flavonoids, two coumarins, five limonoids, and twenty-six volatile compounds, could serve as biological markers to distinguish between QAF and AF and help to verify the botanical origin of crude drugs in an application.
Rutaceae is an important source of food and medicine and has played an important role in the history of traditional Chinese medicine, mostly in qi-regulating drugs, and has a good influence on the digestive and respiratory systems on the basis of abundant flavonoids [7,36]. Flavonoids are the main components of QAF and AF that exert the pharmacological effect of regulating gastrointestinal motility [37,38], and they are also the main indicators in the comprehensive quality evaluation model of medicinal herbs [39]. As shown in the TIC (Figure 1), the main components of QAF and AF were similar in composition, but there existed differences in trace components. Chemometric analysis validated the differences presented by the TIC and presented them in a more visual way, while specific differential components were screened out, which provided a basis for the identification of herbs. Fingerprint analysis combined with clustering analysis could distinguish QAF from other Rutaceae herbs in a holistic perspective [40], but the analysis of differential components was missing. Flavonoids and volatile oils could also be used as signature components to distinguish herbs of the Rutaceae family [41,42], which was consistent with our results. In addition, the result of OPLS−DA showed that coumarins and limonoids could also be used as markers to distinguish QAF from AF.

Antioxidant Capacity
To explore the impact of the differences in composition between QAF and AF in terms of efficacy, DPPH, ABTS, and FRAP methods were used to determine the total antioxidant activity of the extract solutions of QAF and AF ( Figure 5).
Molecules 2023, 28, x FOR PEER REVIEW 14 of 21 consistent with our results. In addition, the result of OPLS−DA showed that coumarins and limonoids could also be used as markers to distinguish QAF from AF.

Antioxidant Capacity
To explore the impact of the differences in composition between QAF and AF in terms of efficacy, DPPH, ABTS, and FRAP methods were used to determine the total antioxidant activity of the extract solutions of QAF and AF ( Figure 5).

Antioxidant Potency Composite (APC)
APC was selected to characterize the total antioxidant capacities of the samples. The APC index was computed using the method described by Seeram et al. [43]. Briefly, an identical weight coefficient was allocated to three tests, the best point in each test was set to an index value of 100, and the index points for all other samples were computed using the following equation: antioxidant index point = [(sample point/best point) × 100]. The APC values for QAF ranged from 56.81 to 94.82, while those for AF varied from 49.48 to 82.28 ( Figure 5D). The APC of QAF was significantly greater than that of AF (** p < 0.01), suggesting that QAF has greater antioxidant ability than AF.
Oxidative stress is a predominant factor in the development of various diseases, including liver, cardiovascular, neurodegenerative, and digestive system diseases as well as psychiatric disorders, and is a potential therapeutic target [44][45][46][47][48]. Studies have shown that QAF can suppress radical production and scavenge radicals to achieve a hepatoprotective effect in vivo and in vitro [11], as well as hypolipidemic effect in hamsters with hyperlipidemia by alleviating oxidative stress [49]. The differences in antioxidant capacity between QAF and AF may impact their ability to treat certain diseases, which needs to be further verified due to different antioxidant mechanisms in vivo and in vitro. In this study, QAF and AF were compared based on their total antioxidant capacity as determined by the DPPH, ABTS, and FRAP methods. Our results suggest that the total antioxidant capacity of QAF was significantly better than that of AF, indicating that QAF has better antioxidant ability than AF, which was consistent with the previous findings [50]; however, the antioxidant capacity of QAF may change with different processing methods [51].

Chemicals and Reagents
All standards (purity ≥ 98.0%) were obtained from Chengdu Push Bio-technology Co., Ltd.

Plant Materials
To eliminate the impact of harvest time on the constituent compounds [52,53], 42 batches of QAF and 8 batches of AF were collected from June 25th to July 11th. The AF samples were obtained from Jiangxi Province and identified as Citrus aurantium L. In contrast, 34 batches of QAF were harvested from 11 different planting bases in Zhejiang Province, and the remaining 8 batches of QAF were provided from Hubei (n = 3), Hunan (n = 3), and Jiangxi (n = 2) Provinces. All the batches were identified as Citrus changshan-huyou Y.B. Chang, and their voucher specimens were preserved at Hangzhou Medical College. The sample IDs, time of gathering, origin, and other pertinent information are listed in Table S1.

Preparation of Standard Solutions and Sample Solutions
To prepare the standard stock solutions, all the reference standards were weighed and dissolved in 50% methanol and stored at 4 • C until use.
The optimal extraction method was determined by the comparison of different methanol concentrations. The samples were crushed into a fine powder and passed through a 100 mesh (150 µm) sieve. The powder (0.25 g) was accurately weighed into a 25 mL brown glass volumetric flask with an appropriate amount of 50% methanol (v/v). The total weight was recorded, and the mixture was subjected to 20 min of ultrasonic extraction at 40 kHz. After cooling to 25 • C, the flask was reweighed and a further amount of 50% methanol (v/v) was added to offset any loss. The extracts were then filtered through 0.22 µm membranes (JLSP042201) obtained from Tianjin Keyilong Lab Equipment Co., Ltd. (Tianjin, China) and stored at 4 • C until UPLC−QTOF/MS analysis. The sample solutions for the antioxidant assay were extracted using the same method, but at different concentrations: 0.5, 1, and 16 mg of crude drug /mL for the ABTS, DPPH, and FRAP assays, respectively. Possible molecular formulae were inferred based on the parent and fragment ion information using the self-built database of compounds, previously reported compounds, and MassLynx V4.2 software (Waters Corporation, Milford, MA, USA), with a mass error of less than 5 ppm between the theoretical and measured mass values. To identify the compounds, the target compound information, key fragment ions, and fragmentation pathways were compared with those of standard compounds or those in the literature.

Extraction of Volatile Oil and GC−MS Analysis
Volatile oils were extracted from QAF and AF via steam distillation, according to the procedure outlined in ChP (2020 edition); dried using anhydrous sodium sulfate; and analyzed using optimal GC−MS analysis conditions after dissolving in ethyl acetate.
GC−MS analyses of the volatile oil and n-alkane solution were performed using an Agilent 7890B gas chromatograph coupled to an Agilent 5975C mass spectrometer with a triple-axis detector (TAD), equipped with an Agilent DB-5MS capillary column (30 m × 0.25 mm, 0.25 µm) and an Agilent 7693 automatic sampler (Agilent Technologies, Santa Clara, CA, USA).
The temperatures of the injector, ion source, and detector were 200 • C, 230 • C, and 270 • C, respectively. The GC oven temperature was initially held at 70 • C for 2 min, then raised to 90 • C at a rate of 5 • C/min, held for 1 min and then raised to 100 • C at 3 • C/min, then to 135 • C at 10 • C/min, then to 185 • C at a rate of 2 • C/min, held for 1 min, and finally increased to 280 • C at 20 • C/min and held for 5 min; an electron impact ionization (EI) of 70 eV was used. Data in the range of 40-400 atom mass units (amu) were collected and analyzed in SCAN mode, with a solvent delay time of 3 min.
Masshunter GC/MS acquisition B.07.06.2704 and Workflows B.08.00 (Agilent Technologies) were used for data acquisition and processing, respectively. The retention index (RI) using a DB-5MS column was calculated using closely eluted n-alkanes as the standard. The volatile compounds were identified by a comparison of the fragmentation patterns and RI with the mass spectral library in the National Institute of Standards and Technology (NIST, version NIST 17), using a standard of MS matching similarity ≥90%, and those in the literature [25,32,54].

Chemometric Analysis
A total of 50 samples were analyzed and the total ion chromatograms (TIC) were obtained. Automatic integration, including automatic noise measurement and smoothing, was performed for TIC of UPLC−QTOF/MS in positive and negative ions with MassLynx software to obtain the peak area of each compound. The peaks of each component in the volatile oil were also extracted by automatic integration of Workflows software to obtain a reasonable background deduction. The areas of their common peaks from UPLC−QTOF/MS in the positive and negative ion modes and GC−MS were set as x variables and normalized to perform chemometric analysis. PCA was used to determine whether there were differences between QAF and AF. OPLS−DA was then used to explore the potential differential components; the results were validated using HCA. Chemometric analyses were performed using Simca 14.1 (MKS Umetrics, Umea, Sweden) and Origin 2023 (OriginLab, Northampton, MA, USA).

Antioxidant Capacity Assays
The antioxidant capacity of the sample extract solution was determined using DPPH, ABTS, and FRAP antioxidant assay kits following the manufacturer's instructions. The absorbance of samples and standards was measured on a BioTek Cytation 1 Cell Imaging Multimode Reader (Agilent Technologies), and data collection was performed using Gen 5 3.08 (Agilent Technologies).
Sample solutions were prepared according to the method described in Section 3.3 and diluted with 50% methanol. Appropriate concentrations were chosen to determine the absorbance within a rational scope to obtain accurate data. All results were converted to a potency at 1 mg/mL drug concentration. The Trolox equivalent antioxidant capacity was calculated for each sample, with units of µmol Trolox/mL for ABTS and FRAP and µg Trolox/mL for DPPH; a higher value indicated a stronger potency.

DPPH Radical Scavenging Assay
DPPH was weighed and dissolved in absolute ethanol. A standard curve was generated using 0, 10, 20, 30, 40, and 60 µg/mL Trolox standard solutions, with the concentration of Trolox and the radical scavenging activity (RSA) set as the x and y variables, respectively. A 50 µL sample extract solution was mixed with 150 µL of DPPH solution in a 96-well plate; the absorbance was determined at 517 nm after being placed at 25 • C for 30 min in the dark. RSA was calculated using the following equation: RSA (%) = (1 − A/A 0 ) × 100 (where A 0 is the absorbance of the control and A is the absorbance of the sample). Finally, the antioxidant capacity was calculated using the standard curve.

ABTS Radical Scavenging Assay
The working solution was prepared according to the manufacturer's instructions. A standard curve was generated using 0, 0.04, 0.08, 0.16, and 0.20 µmol/mL Trolox standard solutions with the concentration of Trolox and the difference of absorbance (∆A; ∆A = A 0 − A, where A 0 is the absorbance of the control and A is the absorbance of the sample) set as the x and y variables, respectively. The absorbance was measured at 419 nm after a 96-well plate containing 10 µL of sample extract solution and 190 µL of the working solution was left at 25 • C for 6 min in the dark. The ∆A value was calculated and the antioxidant capacity was determined using the standard curve.

FRAP Assay
The working solution was prepared according to the manufacturer's instructions. A standard curve was generated using 0, 0.4, 1.2, 2.0, 2.8, and 3.6 µmol/mL Trolox standard solutions, with the concentration of Trolox and the difference of absorbance at 590 nm set as the x and y variables, respectively. To assess the antioxidant capacity, 5 µL of sample extract solution, 25 µL of distilled water, and 170 µL of working solution were mixed in a 96-well plate at 25 • C for 10 min in the dark. The ∆A was calculated and the antioxidant capacity was determined using the standard curve.

Statistical Analysis of Antioxidant Capacity
Three replicates were performed for each sample and data were expressed as the means ± standard deviation. Comparison between the QAF and AF groups was performed by unpaired t-test or Wilcoxon rank-sum test using SPSS software (version 23.0; IBM Corp., Armonk, NY, USA). Results with a p value p < 0.05 were considered statistically significant.

Conclusions
This study presents a systematic comparison of the total chemical components and antioxidant capacity of QAF and AF, using UPLC−QTOF/MS and GC−MS for the first time. A total of 108 compounds, 25 flavonoids, 8 limonoids, 2 coumarins, and 73 volatile compounds, were systemically identified as the foundational components of QAF and AF. Four of these compounds were identified in QAF and AF for the first time. The results of the chemometric analysis indicated that the main components in QAF and AF were very similar. The trace differential components, 26 volatile compounds, 9 flavonoids, 2 coumarins, and 5 limonoids, were screened as potential metabolic markers for discriminating decoctions of QAF and AF to determine their origins. Furthermore, a comparison of the total antioxidant capacity revealed that QAF had a greater antioxidant capacity than AF. As an AF cultivar, QAF can be used as a source of AF, but further investigation is required to understand its properties and applications.
These findings suggest the chemical composition characterization combined with chemometric analysis is an effective approach to distinguish the origin and determine the authenticity of Rutaceae herbs to ensure clinical efficacy and regulate the production of preparation.
Author Contributions: Q.Z.: conceptualization, data curation, formal analysis, methodology, software, writing-original draft, and visualization. W.S.: data curation, methodology, investigation, and formal analysis. G.T.: conceptualization, methodology, software, and data curation. Q.L.: software, investigation, and writing-review and editing. L.W.: resources, investigation, and visualization. W.H.: methodology, project administration, supervision, validation, and writing-review and editing. L.G.: supervision, visualization, and writing-review and editing. L.Y.: conceptualization, resources, supervision, and funding acquisition. Y.Y.: conceptualization, methodology, data curation, visualization, supervision, funding acquisition, and writing-review and editing. All authors have read and agreed to the published version of the manuscript.