UPLC-QTOF-MS Identification of the Chemical Constituents in Rat Plasma and Urine after Oral Administration of Rubia cordifolia L. Extract

An effective ultra-performance liquid chromatography coupled with the quadrupole time-of-flight tandem mass spectrometry (UPLC/Q-TOF/MS) method was developed for analysing the chemical constituents in rat plasma and urine after the oral administration of Rubia cordifolia L. extract. Under the optimized conditions, nine of 11 prototypes in rat plasma and four prototypes in urine were identified or characterized by comparing the retention time, accurate mass, fragmentation patterns, reference compounds, and literature data. In total, six metabolites, including alizarin-1-O-β-glucuronide, alizarin-2-O-β-glucuronide, alizarin-1-O-sulfation, alizarin-2-O-sulfation, purpurin-1-O-β-glucuronide, and purpurin-3-O-β-glucuronide, were identified in rat plasma, which were confirmed by lavaging standard solutions. Purpurin was found to be able to be transformed into alizarin based on the results in which alizarin was detected in rat plasma after the oral administration of a purpurin solution. In total, four metabolites were found in rat urine, but their chemical structures were not confirmed. The results indicate that the metabolic pathway of alizarin involves glucuronidation and sulfation, with the purpurins having undergone glucuronidation. The components absorbed into the blood, and the metabolites have the opportunity to become bioactive constituents. The experimental results would supply a helpful chemical basis for further research on the mechanism of actions of Rubia cordifolia L.


Introduction
Rubia cordifolia L. was officially listed in the 2015 edition of the Chinese Pharmacopoeia and is widely used as a traditional Chinese medicine for the treatment of tuberculosis, contusions, menoxenia, and rheumatism in China, Japan, Korea, and India [1]. Many constituents, including anthraquinones, naphthoquinones, naphthodroquinones, triterpenes [2,3], and iridoids [4,5] were isolated and identified from the genus Rubia. To the best of our knowledge, different parts of Rubia cordifolia L. have different chemical compositions. However, ony the bioactive components that are detected in the

Identification of Chemical Components in Rubia cordifolia L. Extract
The identification of the chemical constituents in the plasma and urine was based on the analysis of the components of the Rubia cordifolia L. extract. The information from the 40 compounds in Rubia cordifolia L. extract, including the chromatogram and MS 2 spectra, was recorded (Table 1). Compounds 23, 25, 26, 31, and 32 were confirmed to be 6-hydroxyrubiadin, alizarin, purpurin, physcion and rubiadin, respectively, by comparison with the reference compounds. A total of nine compounds were tentatively characterized based on the comparison of literature data and the analysis of the fragmentation regularities using the PeakView (Version 2.0, AB SCIEX) [15,21,22]. The remaining 26 ingredients were not further analyzed due to the lack of authentic compounds and reference data. The profiling of these compounds was shown by a total ion chromatogram (TIC) in Figure 1.    Figure 2). Thus, the compound 10 was identified as pseudopurpurin.

Detection of the Prototype Components and Metabolites in Rat Plasma
After eliminating the interference of endogenous substances, 21 compounds (including 11 prototype components and 10 metabolites) were found, but only 13 of these components were tentatively identified ( Figure 3).  As shown in Table 1 and in the rat plasma, respectively. It was concluded that these compounds can be regarded as prototype components.
According to Table 2, compound 2 shows a parent ion at a m/z of 459.0562 and product ions at the m/z of 283, 239, 211, and 195. The product ion at a m/z of 283 was 176 Da less than the deprotonated molecule. Based on the chemical constituents of Rubia cordifolia L., the fragment at the m/z of 283 represents the structure of munjistin. Thus, it was concluded that compound 2 was generated through the glucuronide conjugation of munjistin. Therefore, compounds 4 and 5 were identified as glucuronide conjugations of alizarin, while compounds 3 and 7 were confirmed to be the glucuronide conjugation of purpurin. The MS 2 spectrum of compound 2 is shown in Figure 2. The parent ions of compounds 8 and 13 at the m/z of 319 are 80 Da larger than alizarin. Thus, they were identified as sulfation products of alizarin.
The metabolites of alizarin and purpurin metabolites were observed in rat plasma after the administration of an alizarin solution and a purpurin solution, respectively. Figure 4 and Figure 5 obtained using Metabolite Pilot 1.5 software show the metabolite chromatograms of alizarin and purpurin, respectively. The information for the metabolites are shown in Table 3 and Table 4. M1 and M3 were identified as alizarin-1-O-β-glucuronide and alizarin-2-O-β-glucuronide, based on their same molecular weight, different retention times, and different peak areas (the peak area of M3 > M1), because the β-OH of antraquinones is more active and the result is consistent with the study of emodin [23]. Therefore, M2 and M4 were similarly confirmed to be alizarin-1-O-sulfation and alizarin-2-O-sulfation, respectively. According to the Table 4 and the MS 2 spectra (shown in Figure  5), M1 and M2 in the metabolite chromatogram of purpurin were identified as purpurin-1-O-βglucuronide and purpurin-3-O-β-glucuronide, respectively, while M4 was regarded as alizarin. The interference of alizarin in purpurin standard can be excluded due to the purity of purpurin standard and the peak area of alizarin (shown in Table 4). The MS 2 spectrum of M4 can be found in Figure 5. Therefore, it is concluded that purpurin could transform into alizarin in the rat body. The result is consistent with the study of emodin that stated that emodin could transform into chrysophanol [24]. The M3 was not determined, but it was deduced to be a metabolite of M2 due to their same fragment ions at the m/z of 431, 255, and 227. The metabolic pathways of alizarin and purpurin are shown in Figure 6 and Figure 7, respectively. The chemical structures of the metabolites and the identified compounds are summarized in Table 5. The glucuronic acid (GlcA) and sulfuric acid ester group binding sites in the anthraquinones are α-OH and β-OH. As shown in Table 2, the compounds 1, 2, 6, and 12 were found as metabolites. However, further studies are needed to reveal the structures of these compounds. As shown in Tables 1 and 2, the compounds 10, 11, 12, 15, 17, 19, 21, 25, 26, 27, and 32 in the Rubia cordifolia L. extract correspond to the compounds 9, 10, 11, 14, 15, 16, 17, 18, 19, 20, and 21 in the rat plasma, respectively. It was concluded that these compounds can be regarded as prototype components.
According to Table 2, compound 2 shows a parent ion at a m/z of 459.0562 and product ions at the m/z of 283, 239, 211, and 195. The product ion at a m/z of 283 was 176 Da less than the deprotonated molecule. Based on the chemical constituents of Rubia cordifolia L., the fragment at the m/z of 283 represents the structure of munjistin. Thus, it was concluded that compound 2 was generated through the glucuronide conjugation of munjistin. Therefore, compounds 4 and 5 were identified as glucuronide conjugations of alizarin, while compounds 3 and 7 were confirmed to be the glucuronide conjugation of purpurin. The MS 2 spectrum of compound 2 is shown in Figure 2. The parent ions of compounds 8 and 13 at the m/z of 319 are 80 Da larger than alizarin. Thus, they were identified as sulfation products of alizarin.
The metabolites of alizarin and purpurin metabolites were observed in rat plasma after the administration of an alizarin solution and a purpurin solution, respectively.  Tables 3 and 4. M1 and M3 were identified as alizarin-1-O-β-glucuronide and alizarin-2-O-β-glucuronide, based on their same molecular weight, different retention times, and different peak areas (the peak area of M3 > M1), because the β-OH of antraquinones is more active and the result is consistent with the study of emodin [23]. Therefore, M2 and M4 were similarly confirmed to be alizarin-1-O-sulfation and alizarin-2-O-sulfation, respectively. According to the Table 4 and the MS 2 spectra (shown in Figure 5), M1 and M2 in the metabolite chromatogram of purpurin were identified as purpurin-1-O-β-glucuronide and purpurin-3-O-β-glucuronide, respectively, while M4 was regarded as alizarin. The interference of alizarin in purpurin standard can be excluded due to the purity of purpurin standard and the peak area of alizarin (shown in Table 4). The MS 2 spectrum of M4 can be found in Figure 5. Therefore, it is concluded that purpurin could transform into alizarin in the rat body. The result is consistent with the study of emodin that stated that emodin could transform into chrysophanol [24]. The M3 was not determined, but it was deduced to be a metabolite of M2 due to their same fragment ions at the m/z of 431, 255, and 227. The metabolic pathways of alizarin and purpurin are shown in Figures 6 and 7, respectively. The chemical structures of the metabolites and the identified compounds are summarized in Table 5. The glucuronic acid (GlcA) and sulfuric acid ester group binding sites in the anthraquinones are α-OH and β-OH. As shown in Table 2, the compounds 1, 2, 6, and 12 were found as metabolites. However, further studies are needed to reveal the structures of these compounds.               Table 5. Chemical structures of some of the chemical constituents. Anthraquinone structure β-Dglucopyranosyl (Glc) Glucuronic acid (GlcA).
regarded as prototypes. However, most of the chemical constituents were not identified. The TIC was shown in Figure 8.

Detection of the Chemical Constituents in Rat Urine
It could be observed from Table 6 that eight compounds were detected in rat urine. Among them, the compounds 1, 2, 3, and 8 were regarded as metabolites. In comparison, the other compounds were regarded as prototypes. However, most of the chemical constituents were not identified. The TIC was shown in Figure 8.  The parent ion of Compound 1 at a m/z of 349.0031 was 80 Da more than the precursor ion of compound 4 (6-Hydroxyrubiadin). It was regarded as the sulfation product of 6-Hydroxyrubiadin, based on the MS 2 spectra (Figure 2). The 6-hydroxyrubiadin was not detected in the rat plasma but it was found in the urine, suggesting that it may not be a bioactive constituent.

Chemcials, Reagents, and Meterials
Rubia cordifolia L. were purchased from Guangdong Medicinal Materials and Yin Pian Company (Guangzhou, China), and were further identified by Professor Ruo-Ting Zhan (Guangzhou University of Traditional Chinese Medicine, Guangzhou, China). The reference standards of alizarin, purpurin, and physcion (purity of >98%) were provided by the China Institute of Pharmaceutical and Miological products, while the 6-pydroxyrubiadin and rubiadin (purity of >98%) were obtained from BioBioPha. HPLC-grade methanol, acetonitrile, and formic acid were purchased from Merck (Merck, Darmstadt, Germany). Purified water was prepared from a Milli-Q system (Millipore Billerica, MA, USA). The parent ion of Compound 1 at a m/z of 349.0031 was 80 Da more than the precursor ion of compound 4 (6-Hydroxyrubiadin). It was regarded as the sulfation product of 6-Hydroxyrubiadin, based on the MS 2 spectra (Figure 2). The 6-hydroxyrubiadin was not detected in the rat plasma but it was found in the urine, suggesting that it may not be a bioactive constituent.

Chemcials, Reagents, and Meterials
Rubia cordifolia L. were purchased from Guangdong Medicinal Materials and Yin Pian Company (Guangzhou, China), and were further identified by Professor Ruo-Ting Zhan (Guangzhou University of Traditional Chinese Medicine, Guangzhou, China). The reference standards of alizarin, purpurin, and physcion (purity of >98%) were provided by the China Institute of Pharmaceutical and Miological products, while the 6-pydroxyrubiadin and rubiadin (purity of >98%) were obtained from BioBioPha. HPLC-grade methanol, acetonitrile, and formic acid were purchased from Merck (Merck, Darmstadt, Germany). Purified water was prepared from a Milli-Q system (Millipore Billerica, MA, USA).
The MS data were acquired on an AB SCIEX Triple TOF 5600 (AB sciex Pte. Ltd., Singapore). The Mass spectrometric parameters were as follows: interface of negative electrospray ionization (ESI); gas 1 and 2 being nitrogen 55 psi; curtain gas being nitrogen 40 psi; source temperature of 400 • C; ion spray voltage of 5500 V; de-clustering potential of 100 V and collision energy of 45 eV. The Peakview (Version 2.0, AB SCIEX) and MetabolitePilot TM (Version 1.5, AB SCIEX) were employed for the analyses.

Animals, Dosage, and Biological Sample Collection
Eighteen male Sprague-Dawley rats (weight of 180-220 g) were provided by the Experimental Animal Center of the Guangzhou University of Chinese Medicine and randomly divided into three groups (Group I, Group II, and Group III) of six rats each. These animals were housed in a breeding room at a controlled temperature (20-24 • C) and humidity (40%-60%) in a 12 h light/dark cycle with free access to food and water for three days. All rats were fasted for 12 h with free access to water prior to the experiment.
Rubia cordifolia L. was heating by being extracted with 70% ethanol for 1 h and then filtered. The moisture in the filtrate was evaporated and the residue was dissolved in methanol to a concentration equivalent to 1 mg/mL of the Rubia cordifolia L. for analysis.
Rubia cordifolia L. was immersed in 70% ethanol (1:8, w/v) and extracted three times (1 h each time). The extracted solutions were combined and concentrated under a reduced pressure to a density of 1 g/mL for oral administration.
The Rubia cordifolia L. solution (1 g/mL), the alizarin solution (0.63 mg/mL), and the purpurin solution (3.25 mg/mL) were orally administered to group I (10 g/kg body weight), group II (6.3 mg/kg body weight), and group III (32.5 mg/kg body weight), respectively. Blank blood samples and medicated blood samples were collected from the suborbital vein before administration and 2 h after administration, respectively. Following this, these samples were and then immediately centrifuged for 5 min at 14,000 rpm at 4 • C.
All urine samples from group I were collected for 12 h post-dosing and combined into one sample for the purpose of eliminating the individual variability.

Biological Sample Preparation
The supernatants of the two types of blood samples from the same group were mixed into a single sample to eliminate individual interference. A total of 200 µL aliquots of mixed plasma samples were precipitated with 800 µL methanol and vortexed for 5 min. The sample was centrifuged at 14,000 rpm for 5 min, and beforehand the supernatant was separated and dried under a stream of nitrogen gas at 30 • C. The residue was dissolved in 200 µL of methanol, before 10 µL of this mixture was injected into the UPLC/Q-TOF/ MS for analysis.
The mixed urine sample was homogenized with methanol at a ratio of 1:4, before being vortexed and centrifuged at 14,000 rpm for 5 min. Following this, the supernatant was removed and evaporated to dryness. The residue was dissolved in 200 µL of methanol for UPLC/Q-TOF MS analysis.

Conclusions
In this study, a UPLC/Q-TOF-MS method was established for studying the metabolism of Rubia cordifolia L. Eventually, nine prototype components and six metabolites including alizarin-1-O-β-glucuronide, alizarin-2-O-β-glucuronide, alizarin-1-O-sulfation, alizarin-2-O-sulfation, purpurin-1-O-β-glucuronide, and purpurin-3-O-β-glucuronide were identified in the plasma. Also, this indicates that the metabolic pathway of alizarin involves glucuronidation and sulfation, with the purpurins having undergone glucuronidation. Thus, Rubia cordifolia L. possibly express its effects through their metabolites. Additionally, it is interesting that purpurin can transform into alizarin in rat body. Research has made it clear that alizarin and purpurin both can induce gene mutations that contribute to cases of nephrotoxicity [25,26]. Thus, it is hypothesized that alizarin, instead of purpurin, is associated with the noxious effects to kidney. The analysis of the changes in purpurin content in the body may be the best way to understand the metabolic pathways of purpurin. Furthermore, four prototype components were identified in urine. The 6-pydroxyrubiadin was only detected in the urine, suggesting that it may not be the active substance of Rubia cordifolia L. However, further studies based on the nuclear magnetic resonance technology (NMR) are needed to identify the unidentified compounds. This experiment might provide a basis for further pharmacological and pharmacokinetic research on Rubia cordifolia L.