Major bioactive phenolics in Bergenia species from the Indian Himalayan region: Method development, validation and quantitative estimation using UHPLC-QqQLIT-MS/MS

Bergenia species are important medicinal plants used in indigenous systems of medicine for their antilithiatic and diuretic properties. An ultra-high performance liquid chromatography coupled to hybrid linear ion trap triple quadrupole mass spectrometry (UHPLC-QqQLIT-MS/MS) method has been developed and validated for the estimation of quantitative variation of eight major bioactive phenolics in the rhizomes (150 samples) of four species of this herb, Bergenia (B. ciliata, B. ligulata, B. purpurascens and B. stracheyi). Chromatographic separation was obtained on a Waters ACQUITY UPLCTM BEH (ethylene bridged hybrid) C18 column with a mobile phase consisting of 0.1% (v/v) formic acid aqueous solution and acetonitrile under a gradient elution manner. A hybrid linear ion trap triple quadrupole mass spectrometer was operated in negative electrospray ionization mode with multiple reactions monitoring for detection and quantification of the eight compounds. The validated method demonstrated good linearity (r2 ≥ 0.9991), precision (RSD ≤ 1.87%) and accuracy (95.16–102.11%, RSD ≤ 1.83%) for all reference analytes. The quantitative results revealed that B. ligulata contains the highest amount of the major active marker-bergenin. The results also suggest that sensitive UHPLC-QqQLIT-MS/MS method, a sensitive, accurate and convenient one, could be helpful in identification of potential accession(s), rapid quality control and establishing authenticity of Bergenia species as raw material for pharmaceutical industries.


Introduction
The genus Bergenia Moench (Saxifragaceae) consists of about 10 species of perennial rhizomatous herbs distributed chiefly in the temperate and subtropical regions of Central and East Asia [1]. The four species of Bergenia viz., B. ciliata (Haw.) Sternb., B. ligulata (Wall.) Engl., B. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 analytical standards (purity ! 95%) of gallic acid, protocatechuic acid, chlorogenic acid and catechin were procured from Extrasynthese (Z.I. Lyon Nord, Genay Cedex, France). The analytical standards (purity ! 95%) of bergenin, arbutin, ferulic acid, syringic acid and vitexin were procured from Fluka, Sigma-Aldrich (St. Louis, MO, USA). The 150 rhizome samples of four Bergenia species were collected from various locations of India. All voucher specimens were deposited in the herbarium (LWG) of CSIR-National Botanical Research Institute, Lucknow, India. The sample details including sample code, voucher specimen number, collection location, altitude, latitude, longitude, date of collection, phenology and size of plant parts of each species are given in S1 Table. The species of Bergenia do not come under the threatened category; therefore, specific permission from the Forest Department to collect plant materials for research purpose was not required.

Extraction and sample preparation
The dried rhizomes of 150 samples of four Bergenia species (approximately 10 g) were milled into powder and sieved through a 40 mesh sieve and then extracted with methanol (200 mL × 4 times) in an extractor for 36 h by maceration. The extract was filtered through Whatman filter paper and evaporated to dryness using rotary evaporator (Buchi Rotavapor-R2, Flawil, Switzerland) under reduced pressure at 40˚C. To prepare 1000 μg/mL solution of each sample, dried residues (approximately 1 mg) were dissolved in appropriate amount of methanol and sonicated using ultrasonicator (Bandelin SONOREX, Berlin, Germany). The solutions were filtered through 0.22 μm syringe filter (Millex-GV, PVDF, Merck Millipore, Darmstadt, Germany). The filtrates were diluted with acetonitrile to the final working concentrations. The internal standard (vitexin) was spiked in each final working solution at a concentration of 20 ng/mL, vortexed for 30 s and 2 μL aliquot was injected into the LC-MS/MS system for analysis.

Preparation of standard solutions
A standard stock solution (1000 μg/mL) of each reference analyte was prepared in methanol. From these stock solutions, a mixed standard stock solution (10000 ng/mL) of eight analytes was prepared in methanol. The mixed standard solution was diluted with acetonitrile to a series of concentrations within the ranges from 0.1-500 ng/mL to prepare working standard solutions. The internal standard (vitexin) was spiked in each working solution at a concentration of 20 ng/mL. The standard stock and working solutions were stored at -20˚C until use and vortexed before injection.

Instrumentation and operating conditions
The Waters ACQUITY UPLC TM system (Waters, Milford, MA, USA) connected to a hybrid linear ion trap triple-quadrupole mass spectrometer (API 4000 QTRAP™ MS/MS system from AB Sciex, Concord, ON, Canada) via an electrospray (Turbo V) ion source was used in UHPLC-MS/MS analysis. The Waters ACQUITY UPLC TM system was equipped with a binary solvent manager, sample manager, column oven and photodiode array detector (PDA). The control of LC-MS/MS system, data acquisition and processing was done by Analyst software (version 1.5.1, AB Sciex). The Graph Pad Prism software version 5 was used for all statistical calculations related to quantitative analysis. UHPLC conditions. The chromatographic separation was carried out using an ACQUITY UPLC TM BEH C 18 column (50 mm × 2.1 mm id, 1.7 μm) maintained at 25˚C. The mobile phase consisted of 0.1% (v/v) formic acid aqueous solution (A) and acetonitrile (B). The gradient elution was performed as follows: 0-1.5 min, 5-8% B; 1.5-2.5 min, 8-28% B; 2.5-3 min, 28-35% B, 3-3.5 min, 35-50% B; 3.5-4 min, 50-70% B; 4-5 min, 70-5% B and finally, the initial conditions was held for 2 min for re-equilibration. The flow rate was kept at 0.3 mL/min throughout the analysis. The sample injection volume was 2 μL.
MS conditions. For the quantitative determination of the target analytes, the MS instrument was operated in negative electrospray ionization mode with MRM acquisition at the unit resolution for Q1 and Q3. The optimized conditions for the electrospray source were as follows: ion spray voltage, -4200 V; curtain (CUR) gas, 20 psi; nebulizer gas (GS1) and heater gas (GS2), 50 psi; ion source temperature, 400˚C; collision activated dissociation (CAD) gas, medium and the interface heater was on. The compound dependent MRM parameters including declustering potential (DP), entrance potential (EP), collision energy (CE) and cell exit potential (CXP) for precursor-to-product ion transition of each analyte were optimized by direct infusion and listed in Table 1. The 200 ms dwell time was used for each MRM transition.

Statistical analysis
Principal component analysis (PCA) was carried out to interpret the differences between 150 samples of four Bergenia species based on the contents of eight bioactive constituents using STATISTICA 7.0 software [33].

Optimization of LC conditions
For achieving optimal separation of all analytes in a short analysis time, UHPLC conditions such as the column, mobile phase, gradient elution program, flow rate, injection volume and column temperature were optimized. Two different short length columns, ACQUITY UPLC TM BEH C 18 column (50 mm × 2.1 mm id, 1.7μm) and Thermo Scientific Hypersil GOLD C 18 column (50 mm × 2.1 mm id, 1.9μm) were examined for rapid separation. An ACQUITY UPLC TM BEH C 18 column (50 mm × 2.1 mm id, 1.7 μm) was ultimately chosen for comparatively good separation efficiency and better peak shapes. Further, different gradient and mobile phase systems (water-methanol, water-acetonitrile, 0.1% (v/v) formic acid aqueous solution-methanol and 0.1% (v/v) formic acid aqueous solution-acetonitrile) were compared at different column temperatures (20˚C, 25˚C, 35˚C and 45˚C) and flow rates (0.2, 0.25, 0.3, 0.35 and 0.4 mL/min). Optimization results showed that mobile phase system composed of 0.1% (v/v) formic acid aqueous solution and acetonitrile afforded optimum separation and ionization at a flow rate of 0.3 mL/min and 25˚C column temperature within 5 min. The developed method is rapid as compared to previously reported HPLC, UHPLC-QqQ-MS and HPLC-QTOF-MS methods [23,27,32] for screening and quantitation of phenolics, having chromatographic run time of 8.5-50 min.

Optimization of MS conditions
The mass spectrometric conditions were optimized by direct infusion of 50 ng/mL solution of each targeted analyte into mass spectrometer at a flow rate of 10 μL/min using a Harvard '22' syringe pump (Harvard Apparatus, South Natick, MA, USA). MS spectra were recorded in both positive and negative ionization mode. Finally, negative ionization mode was selected due to the high signal sensitivity of all target analytes in that mode. In order to achieve most abundant, specific and stable MRM transition for each analyte, the compounds dependent MRM parameters (DP, EP, CE and CXP) and source parameters (curtain gas, GS1, GS2 and ion source temperature) were optimized ( Table 1) [34][35][36][37][38]. The developed UHPLC-MS/MS method is more specific than the earlier reported ones [12,23,27,30] as the targeted active phenolics were confirmed by their [M-H]ions and diagnostic fragment ions in the samples investigated during the present study.

Analytical method validation
The developed UHPLC-MS/MS method was validated in terms of linearity, limit of detection (LOD), limit of quantitation (LOQ), precision, stability and accuracy in accordance with the International Conference on Harmonization (ICH, Q2R1) guidelines. Linearity, LOD and LOQ. The linearity of the developed method was assessed at seven concentration levels from 0.1-500 ng/mL. The calibration curves were prepared by plotting the analytes-to-IS peak area ratios against the corresponding concentrations. The LOD and LOQ were determined by calibration curve method using following equations: LOD = (3.3×S xy )/S a and LOQ = (10×S xy )/S a , where S xy is the residual standard deviation of the regression line and S a is the slope of a calibration curve ( Table 2). The calibration curves of all analytes exhibited good linearity with correlation coefficients (r 2 ) ranging from 0.9991-1.0000 within the test ranges. The LOD for eight analytes was ranged from 0.05-0.78 ng/mL and LOQ from 0.14-2.36 ng/mL. The previously reported LOD and LOQ values for these phenolics ranged from 0.04 to 10.7 μg/mL by HPLC and HPTLC methods. However, in the present study the LOD and LOQ values for all analytes are 2.36 ng/mL, thus showing higher sensitivity than the earlier reports [8, 19, 21, 23-24, 27-28, 39]. Precision, stability and accuracy. The intra-day and inter-day variations were selected to determine the precision of the established method. For intra-day and inter-day variability test, three concentration levels of working standard solutions were analyzed in six replicates within a single day and by repeating the experiments on three consecutive days. Variations of the peak area were expressed by percentage relative standard deviations (RSD) and taken as the measures of precision. The overall intra-day and inter-day precisions were not more than 1.87%. For stability test, sample solutions stored at room temperature were analyzed by replicate injections at 0, 2, 4, 8, 12 and 24 h. The RSD values of stability of the eight analytes were 2.55% ( Table 2).
The recovery experiment was used to assess the accuracy of the developed method. It was performed by adding known amounts of eight analytical standards at low, medium and high levels into samples. The spiked samples were then analyzed at each level in triplicate. The average recovery of each analyte was estimated using the following equation: Recovery ð%Þ ¼ ðobserved amount À original amountÞ=spiked amount Â 100%: The average recoveries of all analytes were ranged from 95.16-102.11% with RSD 1.83%, demonstrated that the developed method is accurate ( Table 2).
The results of method validation assay showed that the developed method fulfilled all criteria of a validated method as per ICH guidelines with the acceptable ranges of linearity (r 2 0.9991), LOD and LOQ ( 2.36 ng/mL), precision and accuracy (RSD 1.87%) [40].

Method application
The newly developed UHPLC-MS/MS method was applied to estimate the quantitative variation of eight bioactive phenolics in 150 samples of four Bergenia species that included 103 samples of B. ciliata, 24 samples of B. ligulata, 7 samples of B. purpurascens and 16 samples of B. stracheyi. The content of eight analytes was calculated with internal standard method based on their corresponding calibration curves ( Table 3). UHPLC-MRM extracted ion chromatograms of investigated Bergenia species are shown in Fig 2. Variation of bioactive phenolics in Bergenia species. Comparative quantitative analysis of four Bergenia species from different geographical regions of Himalaya showed that the content of investigated phenolics varied significantly among species. The quantity of bergenin with the content range of 28.50-243.00 mg/g was comparatively higher in all 24 samples of B. ligulata than other species and detected highest (243.00 mg/g) in BS-114 collected from Rajouri (Jammu Kashmir). Similarly, arbutin with the content range of 6.77-77.70 mg/g was comparatively higher in all 16 samples of B. stracheyi than other species and detected highest (77.00 mg/g) in BS-135 collected from Kullu (Himachal Pradesh). The highest quantity of gallic acid (23.5 mg/g) and ferulic acid (0.53 mg/g) was found in B. ligulata, whereas protocatechuic acid (0.56 mg/g), chlorogenic acid (23.6 mg/g), catechin (38.30 mg/g) and syringic acid (0.80 mg/g) were found highest in B. ciliata. The total content of eight bioactive phenolics was detected highest (244.61 mg/g) in B. ligulata (BS-114) and lowest (8.34 mg/g) in B. ciliata (BS-31).
The comprehensive quantitative analysis indicated significant intra and interspecies quantitative variation of major bioactive phenolics among four Bergenia species. Phenolic compounds reported from medicinal plants possess a wide range of bioactivities. Bergenin Estimation of quantitative variation in major bioactive phenolics in Bergenia species Estimation of quantitative variation in major bioactive phenolics in Bergenia species (Trihydroxy benzoic acid glycoside) is found as a principle constituent of B. ligulata and B. ciliata. It has been reported to have antioxidant, anti-inflammatory, antiviral, antihyperglycemic, immunostimulant, and antipyretic potential [40]. Catechin, which is also present in a significant amount in Bergenia, is a flavanol found in a variety of foods and drinks such as fruits, chocolates, wine and tea. It has mainly antioxidant properties [41]. Phenolic acids like gallic acid, protocatechuic acid, chlorogenic acid, and ferulic acid are known to inhibit tumor cells and induce apoptosis [42]. Chlorogenic acid, present in the highest quantity in B. cilliata, is a is reported to have natural antioxidant as it prevents the dinitrogen trioxide formation by scavenging nitrogen dioxide generated in the human oral cavity. According to an in-vitro experiment chlorogenic acid inhibits the mutagenic and carcinogenic N-nitroso compounds and DNA formation in single strand breaks [43]. Thus the present estimates of major bioactive phenolics demonstrated that B. ligulata, which is one of the key ingredients in various Ayurvedic and Unani herbal formulations, is the best among all species having the highest total content of eight bioactive phenolics (244.61 mg/g) and the highest quantity of bergenin.  The chemical pattern of BS-58 was entirely different from rest of the samples as all the three markers viz., catechin (32.4 mg/g), bergenin (39.2 mg/g) and arbutin (20.1 mg/g) were dominant. The characteristic pattern of BS-42, BS-67 and BS-71 was due to the comparatively higher quantity of catechin than the bergenin, which is otherwise the major marker amongst a large number of samples.
The first two PCs explained 47.5% variation in the 24 samples of B. ligulata (Fig 3B). The characteristic variation was observed in 8 samples viz., BS-105, BS-108, BS-113, BS-114, BS-115, BS-123, BS-124 and BS-127, whereas rest of the 16 samples showed a similar pattern. The two samples BS-105 (225 mg/g) and BS-114 (243 mg/g) showed close similarity due to the comparatively high amount of bergenin than other samples. BS-123 and BS-124 were distinguished due to catechin content which was 7.7 mg/g and 24.4 mg/g, respectively. BS-127 was unique from the rest of the samples because of a high quantity of gallic acid (23.5 mg/g) which was less than 4 mg/g in other samples.
The PC1 and PC2 explained 49.06% variation in 16 samples of B. stracheyi. The PC1 score has dominance of bergenin, arbutin and ferulic acid while PC2 score has dominance of syringic acid, chlorogenic acid and arbutin. The factor loading of syringic acid was high, or absent or in negligible quantity in all samples of B. stracheyi. The BS-135, BS-136, BS-145 and BS-147 have characteristic features. The BS-145 and BS-147 showed distinct features because of the high quantity of catechin in BS-147 and almost 10% of this in BS-145. The BS-135 and BS-136 have characteristic features due to higher quantity of bergenin and arbutin in BS-135 as compared to BS-136, and the highest quantity of chlorogenic acid (4.1 mg/g) in BS-136 than other samples (Fig 3D).  Table. https://doi.org/10.1371/journal.pone.0180950.g003 Estimation of quantitative variation in major bioactive phenolics in Bergenia species

Conclusions
The newly standardized method we developed and applied in this study has not only facilitated rapid and accurate determination of target analytes but also fulfilled all the criteria of a validated method as per ICH guidelines in terms of sensitivity, precision and accuracy. Results indicated that B. ligulata collected from Jammu and Kashmir contained highest total content of eight phenolics and the maximum amount of major active marker (Bergenin) which might be medicinally potent regarding cytotoxic and antioxidant activity, whereas B. stracheyi collected from Kullu (HP) contained the highest quantity of arbutin. However, direct activity relation of phenolics to the geographic location in a large scale sample is a subject of future study. The present study could be helpful in identifying potential accession(s) as a raw material of Bergenia species for pharmaceutical industries for preparation of herbal formulations and sustainable utilization of natural resources. The established method also provided a sensitive, accurate and convenient approach for rapid quality control and the establishment of authenticity of Bergenia species.