Chemical Profiles and Simultaneous Quantification of Aurantii fructus by Use of HPLC-Q-TOF-MS Combined with GC-MS and HPLC Methods

Aurantii fructus (AF) is a traditional Chinese medicine that has been used to improve gastrointestinal motility disorders for over a thousand years, but there is no exhaustive identification of the basic chemical components and comprehensive quality control of this herb. In this study, high-performance liquid chromatography coupled with quadrupole time of flight mass spectrometry (HPLC-Q-TOF-MS) and gas chromatography coupled mass spectrometry (GC-MS) were employed to identify the basic chemical compounds, and high-performance liquid chromatography (HPLC) was developed to determine the major biochemical markers from AF extract. There were 104 compounds belonging to eight structure types, including 13 amino acids or peptides, seven alkaloids, 18 flavanones, 14 flavones, 15 polymethoxyflavonoids, six triterpenoids, nine coumarins, and 18 volatile oils, as well as four other compounds that were systematically identified as the basic components from AF, and among them, 41 compounds were reported for the first time. Twelve bioactive ingredients were chosen as the benchmark markers to evaluate the quality of AF. The analysis was completed with a gradient elution at a flow rate of 0.7 mL/min within 55 min. This efficient method was validated showing good linearity, precision, stability, repeatability and recovery. Furthermore, the method was successfully applied to the simultaneous determination of 12 chemical markers in different samples of AF. This study could be applied to the identification of multiple bioactive substances and improve the quality control of AF.


Introduction
Aurantii fructus (AF) is the dried and immature fruit of Citrus aurantium L. and its cultivars. It plays important roles in traditional Chinese medicine (TCM) and is a functional food that has been intensively applied to treat stagnation of dyspepsia, improve gastrointestinal motility dysfunction, and alleviate chest pain in traditional therapies [1,2]. There is a general consensus that the pharmacological effects of herbal medicines are significantly correlated with the chemical composition and the contents of bioactive compounds in herbs. The current research shows that the main bioactive constituents in AF are mainly flavonoids [2][3][4], alkaloids [1,4,5], coumarins [6,7], and volatile oils [1,8]. Modern pharmacological studies have shown these compounds to exhibit various activities, e.g., anti-inflammatory [9,10], anti-oxidation [10,11], regulation of gastrointestinal and prevention of cardiovascular disease [12,13], and they are considered to be the major bioactive constituents in AF. Plenty of clinical investigations have shown that the main pharmacological ingredients of AF are flavonoids, such as flavanones and polymethoxyflavonoids (PMFs), that mainly focus on regulating gastrointestinal dysmotility [12,14,15]. The chemical compounds from AF have been qualitatively or quantitatively analyzed via ultraviolet spectrophotometry (UV), thin layer chromatography (TLC), gas chromatography (GC), high performance liquid chromatography (HPLC), gas chromatograph-mass spectrometry (GC-MS), liquid chromatography mass spectrometry (LC-MS), infra-red spectrum (IR), and nuclear magnetic resonance (NMR) [1,2,7,16,17]. However, due to the complexity of the structural types of the basic chemical constituents of TCM, these methods have certain deficiencies in the analysis of the basic components, and there are incomplete reports for a comprehensive investigation on the constituents of AF. In particular, some ingredients are found in low sensitivity and trace amounts of the above methods often lead to insufficient and defective analysis. High-performance liquid chromatography quadrupole time-of-flight mass spectrometry (HPLC-Q-TOF-MS) is an efficient method with high sensitivity, high precision, high resolution, fast information acquisition, and has been employed to analyze complicated profiles of citrus plants in recent years [6,18,19]. Therefore, the detailed chemical constituents of AF extract can be identified systematically using the HPLC-Q-TOF-MS method. Alternatively, volatile components with low content can be identified by GC-MS technique.
TCM plays a clinical therapeutic role by applying multiple constituent works on multiple targets within the body [20], but only two flavanones (naringin, neohesperidin) are defined as the evaluated markers of quality control for AF in the Chinese Pharmacopoeia [21]. Due to the comprehensive pharmacological effects of TCM, the evaluated standards of quality control for TCM should be carefully considered and established based on multiple pharmacodynamic substances in the future.
The aim of this study is to establish a new comprehensive analysis method for gaining insight into the exhaustive chemical profiles of AF using HPLC-Q-TOF-MS combined with GC-MS. Then, in consideration of the major pharmacological effects of AF, the typical bioactive ingredients were successfully screened as the markers of quality control from the complicated compounds of AF based on the analytical data from mass spectrometry. Then, a simple and effective method for simultaneous determination of bioactive markers was established by high performance liquid chromatography (HPLC), and AF samples from different regions were evaluated. To the best of our knowledge, this is the first attempt to improve qualitative and quantitative methods for the analysis and determination of bioactive ingredients in AF with efficient procedures [22]. These results supply complete chemical components for further research and improve the practical application of the consistency evaluation and quality control of AF samples from different habitats. In addition, they provide comprehensive references for the research and development of AF to be used as a traditional medicine and functional food.

Identification and Analysis of Active Constituents in Ethanol Extract of AF
As many as 86 compounds were identified as the main constituents from AF by ESI-Q-TOF-MS in the positive and negative ion mode, and the total ion chromatogram was analyzed based on the standards, fragmentation patterns, literature and the ChemSpider database ( Figure 1). The retention time and mass spectrometry information of each chemical constituent of AF were detected as shown in Table 1. Eighty-six compounds in the ethanol extract of AF were accurately identified or preliminarily assigned. These compounds included 13 amino acids or peptides, seven alkaloids, 18 flavanones, 14 flavones, 15 polymethoxyflavonoids, six triterpenoids, nine coumarins, and four other compounds. The structures of these compounds were drawn and classified by ChemBio Draw Ultra 14.0, as shown in Figure S1. Thirty-seven of these compounds were reported in AF for the first time.
time and mass spectrometry information of each chemical constituent of AF were detected as shown in Table 1. Eighty-six compounds in the ethanol extract of AF were accurately identified or preliminarily assigned. These compounds included 13 amino acids or peptides, seven alkaloids, 18 flavanones, 14 flavones, 15 polymethoxyflavonoids, six triterpenoids, nine coumarins, and four other compounds. The structures of these compounds were drawn and classified by ChemBio Draw Ultra 14.0, as shown in Figure S1. Thirty-seven of these compounds were reported in AF for the first time.

Amino Acids and Polypeptides
Amino acids are important nutrients needed by the human body and are major primary metabolites of citrus plants [23], as well as the basic substrates in the biosynthesis of secondary metabolites of many plants. Although various free amino acids can be detected in different species of citrus, such as lemon, orange and other fruits [23,24], at present there is still no reports on the amino acid composition of AF. Owing to the low content and larger polarity of amino acids, it was difficult to obtain their MS 2 spectra. However, based on the high resolution and accuracy of Q-TOF-MS and the accuracy of the measured molecular weight in the MS data, multiple types of amino acids were found in AF.

Amino Acids and Polypeptides
Amino acids are important nutrients needed by the human body and are major primary metabolites of citrus plants [23], as well as the basic substrates in the biosynthesis of secondary metabolites of many plants. Although various free amino acids can be detected in different species of citrus, such as lemon, orange and other fruits [23,24], at present there is still no reports on the amino acid composition of AF. Owing to the low content and larger polarity of amino acids, it was difficult to obtain their MS 2 spectra. However, based on the high resolution and accuracy of Q-TOF-MS and the accuracy of the measured molecular weight in the MS data, multiple types of amino acids were found in AF.
Compound 1 displayed a protonated molecule at m/z 147.1130 [M + H] + with the molecular formula C 6 H 14 N 2 O 2 . From the MS data (Table 1), compound 1 corresponds to lysine, which was isolated previously from citrus fruits [23]. Compound 2 showed a protonated molecule at m/z 147.0748 [M + H] + with the molecular formula C 5 H 10 N 2 O 3 . Compound 2 was presumed tentatively as glutamine [23]. By use of the same method and mass spectrometer formula prediction software (absolute error <5 ppm), combined with literature references, compounds 3, 4, 5, 6, 7, 8, 9, 10 and 11 can be identified tentatively as histidine [23], arginine [23,24], proline [23,24], valine [23], isoleucine [23], leucine [23], tyrosine [23], phenylalanine, and tryptophan [23], respectively (Table 1). In addition, two polypeptides were found in AF. Compound 12 displayed a protonated molecule at m/z 728.3966 [M + H] + with the molecular formula C 36 H 53 N 7 O 9 , and corresponded to citrusin III [25][26][27], isolated previously from citrus plants. Compound 13 showed a protonated molecule at m/z 704.3978 [M + H] + with the molecular formula C 34 H 53 N 7 O 9 , which is consistent with the literature, and presumed as citrusin I [26,27]. These 13 compounds were reported in AF for the first time.  [28,29], and it was unambiguously identified by comparison with the standard. Compound 18 showed a protonated molecule at m/z 152.1071 [M + H] + with the molecular formula C 9 H 13 NO, which is consistent with the literature, and presumed to be N-methyltyramine [28,29]. Compound 19 displayed a protonated molecule at m/z 268.1027 [M + H] + with the molecular formula C 10 H 13 N 5 O 4 , which is consistent with the literature, and presumed to be adenosine [30]. Compound 20 showed a protonated molecule at m/z 196.0970 [M + H] + with the molecular formula C 10 H 13 NO 3 , which is consistent with the literature, and presumed to be N-acetylnorsynephrine [29]. In addition, other alkaloids (absolute error <5 ppm) were analyzed in this experiment. Due to the high solution of Q-TOF-MS, these compounds were preliminarily assigned to aminobenzoic acid, pipecolic acid, and piperidine acetic acid, respectively. These three compounds were reported in AF for the first time.

Flavones
A total of 14 flavones were identified or preliminarily assigned in the study. Compound 39 was accurately identified as quercetin-3-O-rutinoside (rutin) [ [35] as a nucleus structure according to the literature data. In positive ion mode, it was presumed that the two compounds contained rutinoside or neohesperidoside because of the low weight of 308 Da, and the abundance of the protonated molecule at m/z 271 [M + H] + in compound 46 was lower than compound 50. Based on the above information and literature reports [2,27], it was speculated that compounds 46 and 50 were apigenin-7-O-rutinoside (isorhoifolin) [28] and apigenin-7-O-neohesperidoside (rhoifolin) [28], and compound 43 was preliminarily assigned to apigenin  Table 1). The compounds 42, 43, 44, 45, 49, 51 were reported in AF for the first time.

Polymethoxyflavonoids
Polymethoxyflavonoids (PMFs) are based on the core aglycone structure that is modified by different numbers and/or positions of methoxyl/hydroxyl substituents [18]. The molecular weights of PMFs can be calculated in advance by adding n × 30 and/or n × 16 according to the basic flavone structure at 222 Da. Then, the chemical formula and the corresponding molecular weight of every possible PMF isomer can be designated, which is very helpful in speculating the structures of PMFs in the complex extracts of AF.
In the study, PMFs were analyzed in the positive ion mode due to the ideal protonated effect for these compounds.

Triterpenoids
Triterpenoids, especially limonoids, are often reported in citrus plants. Studies have shown that they have good anti-tumor activity, and have antifeedant activity to insects [40]. A total of six triterpenoids were identified or speculated in AF by Q-TOF-MS analysis. Compound 68 was identified certainly as limonin [41] by comparing with a reference substance, the retention time and the precise molecular weight. Compound 73 was speculated to be an isomer of limonin [41] O 15 , and it can be presumed to be nominin-17-β-D-glucoside [42]. Compounds 71 and 72 were assigned tentatively to obacunoic acid-17-β-D-glucoside [42] and nomilinic acid-17-β-D-glucoside [41,42], respectively. Compounds 70, 71 and 73 were reported in AF for the first time.

Analysis of the Constituents of Volatile Oils Obtained from AF
The volatile components of AF were extracted by supercritical fluid extraction technology (SFE), which is widely used in the extraction of low-pole components with the advantages of low energy consumption, no pollution, high extraction rate and good product purity. Separation of particular volatile components was carried out by means of the optimal GC-MS method, as shown in Figure 2. Then, the AF volatiles were identified by the comparison of their mass spectra with those recorded in the National Institute of Standards and Technology mass spectral library (NIST05a.L). A total of 18 volatile compounds [1,8] were identified, as shown in Table 2. The structures of these compounds were drawn as shown in Figure S2.
Molecules 2018, 23, x FOR PEER REVIEW 12 of 21 (NIST05a.L). A total of 18 volatile compounds [1,8] were identified, as shown in Table 2. The structures of these compounds were drawn as shown in Figure S2.

Selection of the Quantitative Chemical Markers from AF
The effective constituents obtained from AF are complex and various, according to a previous qualitative study. For the selection of quantitative chemical markers, four main principles are followed [45]. First, they are components absorbed in vivo. Generally, the components absorbed in vivo are considered as potential directly effective materials with therapeutic effects. Second, the quantitative markers should also exhibit the same or similar activity as indicative of the TCM. Third, in consideration of the quality control of TCM, quantitative markers should be found from different individual samples in the TCM. Another important point is that the quantitative markers of different samples should be taken into account to ensure the improvement of practicality, based on the Chinese Pharmacopoeia (2015 edition).
In previous studies, flavanones such as naringenin, naringin, narirutin, neohesperidin, hesperidin, neoeriocitrin and poncirin were identified as the main bioactive components absorbed in rat plasma after oral administration of AF extract [5,46,47]. Moreover, PMFs, such as nobiletin and tangeretin, were also absorbed as constituents, because they were found in rat plasma [16,46]. Many investigations have explored that the main pharmacological effects of these flavonoids, and mainly focus on regulating gastrointestinal dysmotility [12,14,15], which was consistent with the traditional clinic application of AF. These compounds were easily detected from different samples of AF, but only naringin and neohesperidin are defined as the chemical markers in the Chinese Pharmacopoeia (2015 edition). Hence, the quality control for AF should be comprehensively considered and established based on multiple pharmacodynamic substances. Except for naringin and neohesperidin, ten other components (eriocitrin, neoeriocitrin, narirutin, hesperidin, poncirin, naringenin, hesperetin, nobiletin, tangeretin, auraptene) were selected as quantitative chemical markers in order to establish a comprehensive quality control for AF.

Optimization of Chromatographic Conditions
Due to the variety and complexity of the constituents in AF, it is necessary to establish a chromatographic separation method with specificity for the analysis of all the chemical markers. Therefore, several experimental parameters that influence the separation performance of the proposed HPLC method were investigated and optimized in the study.
For example, acetonitrile-water was selected as the mobile phase, due to its better separation of peaks compared with methanol-water in this study, and formic acid was added in the mobile phase to improve the tail behavior of the flavonoids. Full-wavelength scanning was performed during the analysis of AF because of the difference in maximum absorption wavelength of these compounds. The main absorption wavelengths of these markers are 273 nm, 284 nm, and 325 nm, respectively. According to the analysis of preliminary experiments, the final wavelength was set to 284 nm, which exhibited good detection of these chemical markers. Moreover, the flow rate of the mobile phase, column temperature and injection volume were also optimized via pre-experience at an efficient level, and the solvents saved. These markers were well separated and directly determined, as shown in Figure 3, which was conducive to the HPLC quantitative analysis of AF.

Methodological Verification
Calibration and verification of analytical methods are very important for quality assurance (QA) and quality control (QC) in practical applications. The feasibility of the chromatographic method for determining these bioactive markers should be evaluated via methodological verification including linearity, the limits of detection (LOD) and quantification (LOQ), precision, stability, repeatability and recovery. The analysis was performed by HPLC, and the linearity, precision, accuracy, recovery of each bioactive marker, and the stability of this method were analyzed as listed in Table 3.
The high-correlation coefficient values (r > 0.9995) obtained indicated that there were good linear correlations between the concentrations of the relative standard deviations (RSDs) of investigated compounds and their peak areas within the test concentrations, and the obtained LODs and LOQs were less than 0.21 μg/mL and 0.67 μg/mL, respectively. The precision, repeatability, and stability for each compound was less than 1.97%, 3.50%, and 3.38%, respectively. In addition, the developed method showed good recoveries at a range of 98.35%-102.95%. The results indicated that the HPLC method was efficient, accurate, and sensitive for quantitative determination of the major chemical markers in AF.

Methodological Verification
Calibration and verification of analytical methods are very important for quality assurance (QA) and quality control (QC) in practical applications. The feasibility of the chromatographic method for determining these bioactive markers should be evaluated via methodological verification including linearity, the limits of detection (LOD) and quantification (LOQ), precision, stability, repeatability and recovery. The analysis was performed by HPLC, and the linearity, precision, accuracy, recovery of each bioactive marker, and the stability of this method were analyzed as listed in Table 3.
The high-correlation coefficient values (r > 0.9995) obtained indicated that there were good linear correlations between the concentrations of the relative standard deviations (RSDs) of investigated compounds and their peak areas within the test concentrations, and the obtained LODs and LOQs were less than 0.21 µg/mL and 0.67 µg/mL, respectively. The precision, repeatability, and stability for each compound was less than 1.97%, 3.50%, and 3.38%, respectively. In addition, the developed method showed good recoveries at a range of 98.35%-102.95%. The results indicated that the HPLC method was efficient, accurate, and sensitive for quantitative determination of the major chemical markers in AF.  (Table S1) were analyzed by the proposed method to determine the 12 chemical markers by HPLC. The quantitative results are presented in Figure 4A. Naringin and neohesperidin were the main constituents, the contents of which varied from 80.40-106.81 mg/g, and 26.97-102.3 mg/g in 11 samples, respectively. All the samples were in accordance with the standards of the Chinese Pharmacopoeia except for batch 11, because it might be a variety. In addition, it was found that narirutin, hesperidin, poncirin, and neoeriocitrin were also considerably varied from 2.87-9.15 mg/g, 5.10-38.14 mg/g, 1.80-9.40 mg/g, and 1.27-6.48 mg/g, respectively (Table S2). Meanwhile, the content of 12 markers in 11 samples had a relatively large variation. Therefore, it was recommendable to evaluate the quality in different samples of AF.
Hierarchical cluster analysis was performed based on the quantitative determination results from the different samples of AF in order to finding relatively homogeneous clusters. The contents of the 12 components in AF samples were defined as the variables in the analysis to differentiate and classify the 11 samples. In the dendrogram, the quantitative relationship reflected the degree of similarity between different AF samples. As shown in Figure 4B, 11 samples could be divided into three main clusters. Generally, samples with common contents of the compounds, such as samples 1, 2, 3, 4, 5, and 10, were arranged to be a clusters (I) in normal content range of the 12 chemical markers. Higher contents of the markers in the samples 6, 7, 8 and 9 were gathered to clusters II with high quality. Sample 11 was a single, generated as clusters III. Because of the huge difference content of the main chemical markers, naringin and neohesperidin in the sample were lower than other samples, but with a higher content of hesperidin ( Figure 4A). The different samples of AF were discriminated by hierarchical cluster analysis for quality evaluation.

GC-MS Conditions
GC-MS analysis of volatile oils was performed on an Agilent 6850 Network GC System coupled to an Agilent 5975C VL Mass Selective Detector (MSD). An HP-5 MS capillary column (30.0 m × 250 µm × 0.25 µm) was used for the separation. Helium (purity 99.99%) was employed as the carrier gas, with a flow rate of 3.0 mL/min. The splitting ratio was set at 10:1. The injection volume was 1.0 µL and the interface temperature to 250 • C. The MS source temperature was set to 230 • C and the MS quadrupole temperature to 150 • C. The mass spectra plot was acquired using the full scan monitoring mode with a mass scan range of m/z 45-450. The column temperature was initially set at 50 • C (3 min held) and programmed to rise at 5 • C/min to 220 • C (3 min held), 5 • C/min to 250 • C (3 min held), and 5 • C/min to 310 • C (3 min held).

Preparation of AF Extract (Ethanol Solution)
The conditions were optimized in our previous study via response surface methodology. In short, 50 mg of each AF sample was added to 20 mL of 58% ethanol, and extracted for 17 min in a 70 • C water bath by use of an ultrasonic instrument with a constant power (200 W, 40 Hz). After the process, the solution weight was re-adjusted and complement, then the extract solution was filtered through a 0.22 µm membrane for further direct Q-TOF-MS analysis and HPLC determination. This extraction procedure was fast, low energy, low solvent consumption, and low toxicity, and could be considered to be an economical preparation method.

Preparation of AF Extract (Volatile Oils)
A 500 g AF sample was added into the supercritical fluid extraction device. The conditions of the device included temperature (50 • C), pressure (20 Mpa), time (2 h) and flow of CO 2 (2 L/h), and were improved. Then, the volatile oils were sealed immediately for direct GC-MS analysis. This extraction method was a friendly procedure with low volatility and no toxicity.

Validation of Quantitative Analysis HPLC
To determine the limits of detection and quantification, methanol stock solution containing the 12 reference compounds was diluted into a series of appropriate concentrations with the same solvent, and aliquots of the diluted solutions were injected into HPLC for analysis. The limits of detection (LOD) and quantification (LOQ) under the present chromatographic conditions were determined at S/N (the ratios of signal to noise) of three and 10, respectively. Calibration curves were fitted by HPLC response for at least seven appropriate concentrations in triplicate of each markers. The precision was investigated by analyzing a proper calibration sample with the chromatographic system and repeated six times. For repeatability, the mixed solution was examined for six replicates within one day. To investigate the stability of the sample, the sample solution was analyzed at different time points (0, 1, 2, 4, 6, 12, 18, and 24 h). To evaluate the accuracy of the analytical method, the recovery study was measured by analyzing spiked samples. A certain amount of reference substances (low, medium, and high concentrations) were added into a certain amount of AF samples, respectively, and then were extracted and analyzed under the proposed method.

Conclusions
A sum of 104 compounds were identified or tentatively assigned in this study from ethanol extraction and volatile oils of AF by HPLC-Q-TOF-MS and GC-MS methods. The detailed chemical composition of AF was systematically illustrated and classified by comparing retention times, MS spectra of exact mass, fragmentation behaviors and data previously reported on this issue. Among them, 41 compounds were reported in AF for the first time. This provides the basic chemical data for further pharmacological and clinical research. Then, the direct quantitative determination of 12 biochemical markers for AF was efficiently established by the HPLC method that could be suitable for the quality control of AF. The main advantage of the proposed method is the low required amount of the sample, low organic solvents, and low energy consumption, from sample preparation to analytical procedures. Thus, it could be considered an efficient method for analysis and determination of AF samples [22,48]. We anticipate that this research will become an important basis for better comprehension of the pharmacological profiles and for practical application of the QA and QC of AF, which should be developed into a high quality traditional Chinese medicine and health functional foods in the future.