Simultaneous Determination of Luteoloside, Apigetrin, And Hesperidin In Rat Plasma By UHPLC-MS/MS: Application to A Comparative Pharmacokinetic Investigation After Oral Administration Of Schizonepeta Tenuifolia Aqueous Extract With And Without Its Volatile Oil

Background: Schizonepeta tenuifolia Briq. (ST) has been used as an aromatic exterior-releasing medicine in clinical practice for thousands of years in China. Previous researches have revealed both volatile oil (STVO) and aqueous extract (STAE) from ST showed signicant pharmacological activities. However, the inuence between each other was still unknown. Methods: This study was designed to compare the pharmacokinetic proles of three main avonoids (luteoloside, apigetrin, and hesperidin) in STAE to illustrate the inuence of STVO. So, an ultra-ow liquid chromatography-tandem mass spectrometry (UHPLC-MS) method was established to quantitatively analyze the three absorbed ingredients in the plasma of healthy rats. Biological samples were analyzed on an Agilent Eclipse Plus C 18 column (3.0 mm × 150 mm, 3.5 μm) with gradient mobile phase (containing 0.2% formic acid and acetonitrile) at a ow rate of 0.8 mL/min. All the analytes and quercitrin (IS) were investigated with an electrospray ionization source (ESI) using multiple-reaction monitoring (MRM) in negative ionization mode. Results: This quantitative method showed good linearities (r ≥ 0.9995) and the lower limits of quantication were 0.590~1.19 ng/ml. The intra- and inter-day precisions ranged 3.47~10.45% and 4.29~11.28% for the three analytes. The mean extraction recoveries were in the range of 77.41~109.79% and the average matrix effects were within 83.41~112.67%. The validated method has been fully applied to compare the pharmacokinetic parameters of the three avonoid glycosides in rat plasma after oral administration of STAE and STAE+STVO. In comparison of luteoloside, apigetrin, and hesperidin in STAE group, it was found that different degree of increasing existed for the time to reach the maximum concentration (T max ), elimination half-life time (T 1/2 ), the area under the concentration curves (AUC of each analyte in the two groups differentiated to some extent. The results demonstrated that the mean T 1/2 values of luteoloside, apigetrin, and hesperidin in STAE+STVO group were found 1.12, 1.09, and 1.13 fold, respectively, compared to those in STAE group. there were no notable differences observed, indicating some weak inuence on elimination procedure of


Background
Schizonepeta tenuifolia (ST) is a common medicinal plant of Lamiaceae family widely spread in China, Korea and Japan. This medicinal plant has been widely used for treatment of common cold, sore throat, headache, allergic dermatitis, eczema, psoriasis and otitis media for a long time. Lines of evidence in modern pharmacological studies have demonstrated its various biological properties, such as immunomodulation [1,2], anti-in ammation [3,4], antioxidation [5], anti-diabetes [6]. Thus, this phytomedicine is still recorded in the present Chinese Pharmacopoeia (2020) [7].
For a long time, due to its fragrance, ST volatile oil (STVO) has been considered as the fundamental material basis of the phytomedicine and the bioactive volatile terpenoids have been listed as its quality markers, such as (+)-pulegone, (−)menthone, (−)-limonene and (+)-menthofuran, which was supported by many studies [8][9][10][11][12]. On the other hand, as the classical dosage form of traditional Chinese medicine (TCM) for thousands of years, aqueous extract (STAE) also exhibited similar effects with ST [4,[13][14][15]. What's more, there were many kinds of nonvolatile components from STAE, especially the avonoid glycosides with multiple bioactivities including hesperidin (hesperetin-7-O-rutinoside), luteoloside (luteolin 7-Oglucoside) and apigetrin (apigenin 7-O-glucoside) [16][17][18], possibly responsible for the e cacies of ST. For instance, hesperidin could ameliorate pancreatic beta-cell dysfunction and apoptosis in streptozotocin-induced diabetic rat model and attenuate LPS-induced acute lung injury in mice by inhibiting HMGB1 release [19,20]. Luteoloside showed its antiin ammatory and anti-allergic activity through reducing the number of T cells, mast cells and histiocytes in mouse skin with in ammation or atopic dermatitis [21]. Apigetrin was found to inhibit LPS-induced in ammation through the inactivation of AP-1 and MAPK signaling pathways in vivo and in vitro [22].
In addition, in the preparation of modern Chinese medicine, ST is often extracted to yield STVO by steam distillation method. At the same time, the decoction is also obtained and considered as STAE. Then, the concentrated STAE, STVO and some pharmaceutical excipients would be mixed and prepared to produce some pharmaceutics. Many similar examples containing ST could be found in the present Chinese Pharmacopoeia (2020), such as Chuanxiong Chatiao Pills, Xiaoer Jindan Pills, Xiaoer Chiqiao Qingre Granules, Niuhuang Shangqing Pills, and so on [23]. This pharmaceutics technology not only utilizes the phytomedicine completely, but also demonstrates the two parts' effects to the full extent. Then, besides the combined effects, what in uences would this combined application demonstrate on the active components in STVO and STAE as far as pharmacokinetics is concerned? To our knowledge, there have been some pharmacokinetic articles of luteoloside, apigetrin, and hesperidin in the administration form of monomeric compound, plant extract, and Chinese patent medicine [24][25][26][27][28]. The interaction studies of their combination with other drugs especially volatile oils were hardly involved. Therefore, when combination, whether the pharmacokinetic characteristics of the three active avonoids would change is an interesting issue worth investigating.
In this study, in view of the effect of STVO on bioavailability of luteoloside, apigetrin, and hesperidin in STAE, a simple, rapid and sensitive UHPLC-MS method was developed for the simultaneous determination of the three bioactive avonoids in rat plasma and applied to indicate their in uence between STAE and STVO in normal rat from the perspective of pharmacokinetics. Furthermore, we hope the obtained results would be helpful for safe utilization of ST and even for understanding of volatile oil's role in combined administration of drugs.

Materials and reagents
ST was purchased from Chinese herbal medicine market in Anguo City, Hebei Province in July 2019, and then was authenticated by Prof. Qinan Wu, a botanist of Nanjing University of Chinese Medicine. The voucher specimens were kept in the laboratory.
The reference standards of luteoloside, apigetrin, hesperidin, and quercitrin (IS) were bought from Nanjing Jin Yibai Biological Technology Co., Ltd. Their purities were determined as above 98% by HPLC method. They were stored in accordance with their instructions.
Methanol and acetonitrile were both of LC-MS grade and obtained from Merck (Merck, Darmstadt, Germany). Formic acid of HPLC grade was purchased from Anaqua Chemicals Supply (ACS, Houston, USA). Ultrapure water was puri ed by a Milli-Q water puri cation system (Millipore, Burlington, MA, USA). All other reagents were of analytical grade.

Preparation of STVO and STAE from ST
Using a set of standard apparatus including a round-bottom ask, a volatile oil extractor, a condenser pipe and a heating equipment, STVO with a yield of 1.08% was obtained by a steam distillation method recorded in the present Chinese Pharmacopoeia (2020). To remove the trace water, some ahydrous sodium sulfate was added into STVO, which was then stored in the dark glass bottle at 4 °C and ltered before use. At the same time, to remove some impurities, the decoction in the round-bottom ask was transferred into a separating funnel and added with three times the amount of 95% ethanol. After the mixture was kept at 4 °C over night, the supernatants were collected and evaporated by a vacuum-rotary unit and dried in a vacuum desiccator to prepare STAE with a yield of 10.22%.
An AB Sciex 4500 Qtrap™ mass spectrometer (AB Sciex, Foster City, CA, USA) with electron spray ionization (ESI) in the negative mode was utilized in this study. The precursor-to-product ion transitions in multiple-reaction monitoring (MRM) mode were used for mass analysis and quanti cation. Some MS setting conditions were de ned as follows: curtain gas pressure at 20 psi, ion spray voltage at 5000 V, turbo gas temperature at 300 °C, both nebulizer gas 1 and auxiliary gas 2 at 55 psi. Nitrogen of high purity was employed as both nebulizing gas and drying gas. All the data were obtained and analyzed by Analyst 1.6.3, a data acquisition and processing software.

Animal experiments
In the pharmacokinetic experiment, SPF male Sprague-Dawley rats (220-250 g) were supplied by Shanghai Jiesijie Experimental Animal Co., Ltd. (SCXK (Hu) 2018-0004, Shanghai, China). The rats were housed under ambient temperature of 23±2 °C with a 12 h light/dark cycle and 50% relative humidity in the laboratory. They were well looked after in the plastic cages with fresh water and free pellet food and were acclimatized for a week. Prior to the experiment, all the rats were fasted for 12 h with free access to water. To minimize the pain and suffering of the rats, the experimental procedures were performed in line with the guidelines of the Animal Care and Use Committee of Nanjing University of Chinese Medicine (Nanjing, China). The protocol was also approved by the animal experimental committee of Nanjing University of Chinese Medicine (201909A008).

Preparation of standard solutions and quality control samples
The reference standards of the three target avonoids were separately and accurately weighed and dissolved in methanol to prepare the stock standard solutions with the concentrations of 0.238 mg/mL for luteoloside, 0.193 mg/mL for apigetrin and 0.118 mg/mL for hesperidin. Then, the calibration standard solutions were obtained by mixing and serially diluting above three stock standard solutions with methanol. Additionally, IS stock solution of quercitrin was prepared with the concentration of 200 ng/mL in methanol. All these solutions were enclosed and stored at 4 °C until analysis.
The plasma samples of calibration standards were prepared by spiking 10 μL of appropriate standard solution into 90 μL of blank plasma to obtain the nal concentrations of 1.19~238 ng/mL for luteoloside, 0.965~193 ng/mL for apigetrin, 0.590~118 ng/mL for hesperidin, respectively. Quality control (QC) samples at low, middle and high concentrations (1.19, 119 and 190 ng/ml for luteoloside; 0.965, 96.5 and 154 ng/ml for apigetrin; 0.590, 59.0 and 94.4 ng/ml for hesperidin) were also prepared according to the similar procedures.
Biological sample preparation 100 µL of plasma samples and 10 µL of IS solution was added to a centrifuge tube and mixed with 500 μL acetonitrile for protein precipitation. The mixture was vortexed for 3 min and then centrifuged at 13000 rpm for 10 min at 4 °C. The supernatant was transferred to another centrifuge tube and then desiccated with an Eppendorf Concentrator Plus (Hamburg, Germany). Finally, the residue was redissolved and vortex-mixed with 120 μL of acetonitrile/water (50:50, v/v) for 5 min and centrifuged at 13000 rpm for 10 min. The supernatant was then injected into the UHPLC-MS/MS system for analysis.

Method validation
Based on the bioanalytical guidance of FDA, the method validation was carried out in terms of its speci city, linearity, sensitivity, accuracy, precision, matrix effect, recovery and stability.

Speci city
The chromatograms of some biological samples were compared to investigate the method speci city, including blank plasma samples, plasma samples spiked with luteoloside, apigetrin, hesperidin at lower limits of quanti cation (LLOQ) and IS, plasma samples after oral administration of STAE+STVO and STAE.

Linearity and LLOQ
The calibration curve of each analyte was obtained by plotting its peak area ratio (y) to IS versus its nominal concentration (x) with weighted (1/x 2 ) least-squares linear regression. The method sensitivity was evaluated by LLOQ of each analyte with its signal-to-noise ratio of 10:1 in the matrix-analyte samples, which was conformed to the lowest concentration of the standard curve with an acceptable accuracy (RE≤±20%) and precision (RSD≤20%).

Precision and accuracy
Six QC samples at each level (low, medium, high) were used to in the method precision and accuracy test. The intra-day precision test was performed by repeatedly analyzing each QC sample in a day while the inter-day precision test was carried out by analyzing each QC sample on three consecutive days. The analytical precision was expressed by relative standard deviation (RSD) and relative error (RE), respectively. RE was calculated with this formula: RE (%) = (measuring valuenominal value) × 100%/nominal value. The pass criterion for RSD was required to be less than 15%, and that for RE was required to be within ±15%.

Extraction recovery and matrix effect
The extraction recovery and the matrix effect were evaluated with six replicates by the formulas as follows: (A 1 /A 2 ) × 100% and (A 2 /A 3 ) × 100%, respectively. A 1 : the peak areas obtained from the post-treated plasma samples spiked with the target components at the three QC levels; A 2 : the peak areas obtained from the pre-treated plasma samples spiked with the target components at the three QC levels; A 3 : the peak areas obtained from the standard solutions at the equivalent concentrations. The extraction recovery and matrix effect of IS were also measured by the same approach at the identical concentration.

Stability
The stability of each analyte in rat plasma was evaluated by analyzing six replicates of the QC samples at low, middle, and high concentrations in the following practical experimental conditions: storage at 4 °C for 12 h, three freeze-thaw cycles (freezing at -80 °C for 24 h and thawing at room temperature) and long-term storage at -80 °C for 2 weeks. The acceptable RE was within ±15.0% of the nominal concentration.

Pharmacokinetic study
For pharmacokinetic study, all the rats were divided into two groups (n=6 per group) randomly. The given dosages of two test extracts were equivalent to about 5g ST/kg. STAE group: the rats were administrated orally with STAE (500mg/kg), dissolved in 0.5% carboxy methyl cellulose sodium (CMC-Na) solution. STAE+STVO group: the rats were administrated orally with STAE (500mg/kg) and STVO (50mg/kg), dissolved in 0.5% CMC-Na solution. Blood samples were collected from the venous plexus of the eye socket and transferred in a heparinized centrifuge tube (1.5 mL) after oral administration (0, 5, 10, 20, 30, 45, 60, 120, 240, 480, and 720 min). Each sample was immediately centrifuged at 13000 rpm for 10 min at 4 °C to acquire the plasma. At last, the plasma was immediately stored at -80 °C until analysis.

Data analysis
The plasma concentrations of the three avonoids during the experiment were calculated according to the daily calibration curve. With Drug And Statistics Version 3.2.8 (DAS 3.2.8) software (Mathematical Pharmacology Professional Committee of China, Shanghai, China), a non-compartmental method was employed to analyze some representative pharmacokinetic parameters, such as the time to reach maximum concentration (T max ), elimination half-life time (T 1/2 ), the area under the concentration curves (AUC 0→t and AUC 0→∞ ) and the maximum concentration (C max ). Data were demonstrated as the mean ± standard deviation (SD) with triplicate. The differences of pharmacokinetic parameters from two groups were tested by Student's t-test. Statistical analysis was operated by Graphpad Prism 7.0 software.

Results And Methods
Optimization of chromatographic and MS conditions As far as peak shape and analysis time were concerned, it was revealed upon comparison that an Agilent Eclipse Plus C 18 column (3.0 mm × 150 mm, 3.5 μm) showed better performance than a Phenomenon Luna C 18 column (4.6mm × 250mm, 5 μm), a Hanbon Megres C 18 column (4.6mm × 250mm, 5 μm). The mobile phase system consisting of 0.2 % formic acid and acetonitrile was employed for gradient elution with the ow rate of 0.8 mL/min.  Table 1.

Sample preparation
It is well known that protein precipitation, liquid-liquid extraction and solid phase extraction are the common methods for plasma sample preparation. In our trial test, low extraction recoveries of the analytes were found in liquid-liquid extraction and could not meet the speci cations of pharmaceutical analysis. Too many cartridges would be used and signi cant time would be required when solid phase extraction was employed. Thus, protein precipitation with acetonitrile was selected to prepare plasma samples owing to its easy operation, high e ciency and low cost.

Method validation
Speci city The representative chromatograms acquired from blank plasma, blank plasma added with the three target avonoids at LLOQ and IS, and plasma sample at 1 h after oral administration of STAE and STAE+STVO are shown in Figure 2. Due to the e cient sample preparation and high selectivity of MRM, there was no obvious endogenous interference for the separation and analysis of the four avonoids in the plasma.

Linearity and LLOQ
The linear regression equations, correlation coe cients (r) and LLOQs of luteoloside, apigetrin and hesperidin were summarized in Table 2. The correlation coe cients (≥ 0.9995) indicated their good linearities and the LLOQs (0.590 1.19 ng/mL) were appropriate for quantitative detection.

Precision and accuracy
As shown in Table 3, the results of the intra-and inter-day precision and accuracy of the analytes in QC samples were exhibited. For the three QC levels of the analytes, in the intra-day test, RSD values varied from 3.47% to 10.45%, while RE values ranged between -3.82% and 5.43%. Similarly, in the inter-day test, RSD values varied from 4.29% to 11.28%, while RE values ranged between -7.09% and 9.24%. All these precision and accuracy results were within the acceptable ranges for analysis in biological media, indicating the reliability and reproducibility of the method.

Extraction recovery and matrix effect
As for the three QC levels, the average extraction recoveries of luteoloside, apigetrin and hesperidin ranged from 77.41% to 109.79% while their average matrix effects were between 83.41% and 112.67% (Table 4). Besides, all of the RSD values were below 15.00%. Thus, the results within the acceptable ranges demonstrated that the analytes could be extracted from the plasma e ciently and matrix effects were seldom found for this method.

Stability
The stability of the three analytes in rat plasma during the sample storing and processing procedures was fully evaluated. The results (Table 5) showed that luteoloside, apigetrin and hesperidin were stable in rat plasma after three freeze-thaw cycles, for 12 h at 4 °C and for 2 weeks at -80 °C after plasma sample preparation. The RE values were calculated in the range from -7.13% to 11.54% and the RSD values were not more than 9.38%, which demonstrated that the sample solutions were stable under above various conditions.

Pharmacokinetic study
The developed assay method was sensitive and accurate enough to be applied in the pharmacokinetic study of luteoloside, apigetrin, and hesperidin in rat plasma following oral administration of STAE and STAE+STVO. The plasma concentrationtime curves of the three analytes are present in Figure 3. The pharmacokinetic parameters are summarized in Table 6, including T max , T 1/2 , C max , AUC 0→t and AUC 0→∞ .
In Figure 3, it can be seen intuitively that the concentration-time curve of each avonoid glycoside in rat plasma after oral administration of STAE+STVO is similar as that of STAE. Acturally, the pharmacokinetic pro les of each analyte in the two groups differentiated to some extent. The results demonstrated that the mean T 1/2 values of luteoloside, apigetrin, and hesperidin in STAE+STVO group were found 1.12, 1.09, and 1.13 fold, respectively, compared to those in STAE group. However, there were no notable differences observed, indicating some weak in uence on elimination procedure of the three analytes. In addition, for the three avonoids, after co-administration of STAE and STVO, the values of AUC 0→t and AUC 0→∞ were investigated signi cantly larger than administration of STAE alone (p<0.01, p<0.05), except AUC 0→∞ value of apigetrin.
The data reminded us promoting effect of STVO on absorption of the three analytes. For hesperidin and apigetrin, T max values in STAE+STVO group were markedly longer than those in STAE group (p<0.01) while C max values were increased without notable differences. On the contrary, C max value of luteoloside was elevated notably (p<0.05) while no signi cant difference was found for its T max value as far as the same comparison was concerned. It was concluded that STVO might ameliorate absorption and improve bioavailability of the three analytes.

Discussion
In the past decades, it has become a consensus that the plant-originated volatile oils are investigated as the favourable transdermal absorption enhancing agents, such as the volatile oils from Rosmarinus o cinalis [29], Magnolia fargesii [30], Eugenia caryophyllata [31] and Alpiniae O cinarum [32]. Moreover, it has been documented that the volatile oil from Acori Tatarinowii might have promoting effect on pharmacokinetic fates of three coumarins including byakangelicin, xanthotoxol and oxypeucedanin hydrate from Angelicae Dahuricae when co-administration in rat [33]. In another study, the volatile oil from Xiangfu Siwu Decoction (a famous formula of traditional Chinese medicine) has been found to improve the absorption and bioavailability of ten non-volatile components (caffeic acid, ferulic acid, gallic acid, vanillic acid, albi orin, paeoni orin, berberine, protopine, tetrahydrocolumbamine and tetrahydropalmatine) in the same formula [34]. Taken together, these ndings indicated that the volatile oils could enhance the absorption and bioavailability of some drugs not only through transdermal absorption, but also through intestinal absorption, which was consistent with the results of our present study.
Our investigation veri ed the rationality of oral co-administration of STAE and STVO in terms of pharmacokinetics, which would be helpful for the pharmaceutical technology study of the Chinese herbal preparation containing ST and for the novel application exploration of STVO. However, there still existed some interesting issues. Except for the pharmacokinetic parameters, would the metabolites of the three analytes in STAE be affected by STVO? Which volatile component(s) showed absorption-promoting effect? What was the absorption-promoting mechanism of STVO? Was P-glycoprotein involved? Which effects would STAE show on the pharmacokinetic pro les of the components in STVO after oral coadministration? These would be listed as the research priorities in the future.

Conclusions
In the present study, a sensitive and rapid UHPLC-MS/MS analytical method was rstly established for simultaneous quantitative analysis of luteoloside, apigetrin, and hesperidin in rat plasma. This method provided eligible recovery and stability with ne precision and accuracy and then was successfully applied to the pharmacokinetic study of three bioactive avonoid glycisodes after oral administration of STAE and STAE + STVO. The results obtained from this study demonstrated that STVO could promote the absorption and bioavailability of the three avonoid glycosides in vivo. Therefore, we hope that these ndings would help to broaden the applied range of STVO and to provide some scienti c basis and guidance for the safe clinical use of ST.