Simultaneous determination of ten compounds in two main medicinal plant parts of Tibetan herb , Pterocephalus hookeri ( CB Clarke ) Höeck , by ultra-high performance liquid chromatography-photodiode array

Purpose: To develop an ultra-high performance liquid chromatography (UPLC) photodiode array (PDA) method to compare the chemical composition of two different medicinal components of Pterocephalus hookeri. Methods: Samples were chromatographically separated in succession using Waters Acquity UPLCR BEH C18 column (2.1 × 100 mm, 1.7 μm) and gradient elution (0.2 % phosphoric acid aqueous acetonitrile). Using partial least squares discriminant analysis and one-way analysis of variance, attempts were made to distinguish different medicinal parts of P. hookeri. Results: Regression equation for 10 compounds showed good linear regression (R > 0.9994). The relative standard deviations of precision, stability, repeatability and recovery were under 5 %. Compared with the aerial plant part, the root had significantly higher levels of sylvestroside I (p < 0.01), cantleyoside (p < 0.001), dipsanosides B (p < 0.01) and dipsanosides A (p < 0.01), but significantly lower levels of loganic acid (p < 0.001), chlorogenic acid (p < 0.01), and isochlorogenic acid (p < 0.01). There were no significant differences between loganin, sweroside and isochlorogenic acid C. Conclusion: The described method is simple, accurate and reproducible, and can be used for the simultaneous determination of 10 major compounds of P. hookeri. The results demonstrate that there is variation in the chemical composition of the aerialpart and root of P. hookeri and that loganic acid and cantleyoside are the primary chemical biomarkers.


INTRODUCTION
Pterocephali herba is the whole herb of the perennial plant Pterocephalus hookeri (C.B.Clarke) Höeck, a member of the Dipsacaceae family [1].In Tibetan, P. hookeri is called "Bangzi-du-wu," and it is recorded as a Tibetan medical classic named "rGyud-bZhi" (usually known as the Four Tantras) [2,3] in "Drug Standards of Tibetan Medicines" [4] and the 2015 edition of the Chinese Pharmacopoeia [5].As a traditional national medicine, P. hookeri has been widely used in many Tibetan herbal medicine formulations and is clinically used to treat diseases including common cold, rheumatoid arthritis, enteritis, dysentery, and other conditions [6,7].
Recent studies indicate that the main effective components of P. hookeri are glycosides, including iridoid glycosides (loganin and cantleyoside) and triterpenoid saponins (hookerosides A-D) [2,7,8].Zhang et al [9] demonstrated that P. hookeri exhibit antiinflammatory properties.To date, several analytical approaches including highperformance liquid chromatography (HPLC) [10], capillary zone electrophoresis [11], and ultra-fast liquid chromatography (UFLC) [12] have been used to quantify one compound or simultaneously quantify several compounds in P. hookeri.Partial least squares discriminant analysis (PLS-DA) is a supervised model based on dummy variable classification that can show the distribution and correlation between sample groups and has been used to distinguish between different species and identify potential biomarkers [13,14].As far as we know, there is no publication describing the use of PLS-DA and simultaneous quantitation of 10 compounds to distinguish between different medicinal parts of P. hookeri.
Here we exploit a quick, dependable, and reproducible UPLC-PDA method for the simultaneous quantitative assay of 10 compounds in two different medicinal plant parts of P. hookeri.

Materials and reagents
Acetonitrile (HPLC grade) was purchased from Merck (Darmstadt, Germany).Deionized water used for UPLC was obtained with the Milli-Q system from Millipore (Billerica, MA, USA).Other used reagents were of analytical grade.

Sample preparation
All samples were smashed, and passed a 50 mesh sieve.Powdered samples (0.5 g) were accurately weighed and dissolved in 70 % methyl alcohol (50 mL) by sonication for 30 min.Additional 70 % methanol was added to make up for lost weight.The extract was filtered through a 0.22 μm microfiltration membrane.

UPLC analysis
Samples were analyzed with a Waters Acquity UPLC system.PDA was utilized for the detection using 237 and 325 nm.All chromatographic parameters were optimized for better separation, including mobile phase composition, flow rate, gradient elution mode, and column temperature.

Validation of UPLC method
The UPLC-PDA method was validated in the terms of linearity, stability, recovery, repeatability, and accuracy [17].The storage solutions of 10 compounds were set up the regression equations, which were plotted after linear regression between the peak areas and concentrations of 10 compounds [18].The limit of detection (LOD) and quantification (LOQ) were determined at signal-to-noise ratios of 3 and 10, respectively [19,20].

Data analysis
The UPLC-PDA data were input to the SIMCA-P software package (version 12.0, Umetrics, Umeå, Sweden) for PLS-DA [21].GraphPad Prism software (version 5.0, GraphPad Software Inc, San Diego, CA, USA) was used to generate graphs and perform one-way analysis of variance (ANOVA) [22].Differences were considered significant at p < 0.05.

Optimization of extraction procedure
To achieve an efficient extraction of the 10 compounds from P. hookeri, key factors including the method, solvent type, time, solvent volume, and number of extractions were independently investigated.A comparison of different extraction methods (ultrasonic, refluxing, and standing extraction) revealed no significant differences in extraction rate; however, ultrasonic extraction was most convenient.Methanol was a more efficient solvent than ethanol.The concentrations of three different aqueous methanol solvents (50, 70 or 90 %) were compared, and 70 % was chosen.Furthermore, tests of the number of extractions (1, 2 or 3 times), solvent volume (30, 50 or 70 mL), and time (20, 30 or 40 min) were investigated.The sample processing method was finally established as follows: 0.5 g sample was extracted with 50 mL of 70 % methanol with ultrasonication for 30 min.

UPLC validation data
The method was validated by the guidelines for Validation of Quality Standard of Traditional Chinese Medicine [24].Table 1 lists detailed information.All regression equations showed good linear relationships (R 2 > 0.9994).LODs and LOQs for the 10 compounds were in the range of 0.12-3.35μg/mL and 0.45-11.20 μg/mL, respectively.Table 2 provides detailed data on the precision, repeatability, and stability for 10 compounds, relative standard deviations (RSDs) were in the range of 0.44-0.96,0.42-1.43,and 0.91-3.58,respectively.Table 3 shows the recovery rates for three various concentrations (low, middle, and high).The mean recovery rates were in the range of 96.62 to 100.76 %, and RSDs of the 10 compounds were less than 5 %.These findings demonstrate that the UPLC-PDA method is sufficiently precise and accurate for simultaneous quantification of 10 compounds in P. hookeri.

Method application to samples
Using the developed UPLC method, 10 compounds were determined simultaneously in 32 batches of different medicinal parts samples.Figure 1 shows representative the UPLC-PDA chromatograms of two different medicinal parts of P. hookeri.All samples were independently assayed three times to calculate the average content (mg/g), the results are shown in Table 4.Among the 10 compounds, dipsanosides A (0.30-9.52 mg/g) were more abundant than dipsanosides B (0.17-3.16 mg/g).In the root samples, the detected ranges were 1.60-9.13mg/g for loganic acid, 0.87-16.55mg/g for chlorogenic acid, 0.47-2.09mg/g for loganin,

DISCUSSION
PLS-DA was applied to achieve good sample separation.The resulting score plot (Figure 2A) demonstrated that the 32 batches of different medicinal parts from P. hookeri could be divided into two groups.In addition, the PLS-DA loading plot (Figure 2B) showed that loganic acid and cantleyoside were the primary chemical biomarkers.To better understand sample contents, graphs were produced from the data in Table 4. Figure 3 shows that compared with the aerial component, the root had significantly higher levels of sylvestroside I, cantleyoside, dipsanosides B, and dipsanosides A, but significantly lower levels of loganic acid, chlorogenic acid, and isochlorogenic acid.There were no significant differences in loganin, sweroside, and isochlorogenic acid C (p > 0.05).
Oleanolic acid and ursolic acid are the markers currently used to control P. hookeri quality.The 2015 edition of the Chinese Pharmacopoeia dictates that the total contents of oleanolic acid and ursolic acid in should not be less than 0.20 %.    4): sweroside, (5): isochlorogenic acid A, (6): sylvestroside I, (7): isochlorogenic acid C, (8): cantleyoside, ( 9): dipsanosides B. (10):dipsanosides A P. hookeri samples with lower levels are considered substandard and should not be used as a medicinal material for clinical application.In our previous study [25], we determined the total contents of oleanolic acid and ursolic acid in 32 batches of aerial part samples and found that the contents were all higher than 0.20 %.Interestingly, none of the root samples reached the standard of the 2015 edition of the Chinese Pharmacopoeia.
P. hookeri is a perennial herb that mainly grows at an altitude of 3200-5700 m on hillsides and in grasslands, meadows, and forests.Due to root excavation, it was listed as an endangered Tibetan medicine species in 2000.Understanding how to rationally use P. hookeri resources while preventing its depletion is an important issue.Therefore, the results of this study provide quality control references for P. hookeri.

CONCLUSION
The developed method is simple, accurate and reproducible, and can be used for the simultaneous determination of the contents of 10 major compounds of P. hookeri.The findings demonstrate that there is variation in the chemical composition of the two parts of P. hookeri, and that loganic acid and cantleyoside are the main chemical biomarkers.

Figure 2 :
Figure 2: Score (A) and loading (B) plots of PLS-DA of the 32 tested samples derived from UPLC quantitative data."a": the aerial part sample, "b": the root sample
total of 32 batches of different plant parts of P. hookeri were collected from different regions of Sichuan and Yunnan Provinces in China, consisting of 16 batches each of aerial part and root samples.as listed in 'Results' section.The plant samples were authenticated by Professor Yi Zhang of College of Ethnic Medicine, Chengdu University of Traditional Chinese Medicine, and voucher specimens kept in the herbarium of College of Ethnic Medicine, Chengdu University of Traditional Chinese Medicine. A

Table 1 :
Regression data, LOD and LOQ of 10 compounds

Table 2 :
Precision, repeatability and stability of the UPLC method for determination of the 10 compounds

Table 3 :
Recovery of the ten compounds