Development and validation of a simple, sensitive and reproducible method for simultaneous determination of six polyphenolic bioactive markers in Dendrobium plants

2.1 Sample collection

D. stricklandianum was collected and authenticated by Dr. Boonchoo Sritularak and Mr. Yanyong Punpreuk. D. crumenatum, D. friedericksianum and D. officinale were purchased from local markets in Bangkok, Thailand. The voucher specimens were deposited at the Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Chulalongkorn University, Bangkok, Thailand.

2.2 Material and reagents

(2S)-Eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin were provided from the Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Chulalongkorn University, Bangkok, Thailand. They were isolated from orchid plants and confirmed their structures using infrared spectrophotometry (IR), nuclear magnetic resonance spectroscopy, and high-resolution mass spectrometry (Chanvorachote et al., 2013; Tanagornmetar et al., 2014; Klongkumnuankarn et al., 2015). Chromatographic purity of the isolated compounds was determined by UPLC coupled with a photodiode array (PDA) detector to confirm their purity of > 98.0% and used as standards in the validation and assay. Chromatographic purities of (2S)-Eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin are 99.8, 98.3, 98.2, 99.3, 100.0, and 98.6 respectively. Reagent grade trifluoroacetic acid was purchased from Sigma-Aldrich (St. Louis, Missouri, USA). HPLC gradient grade acetonitrile and methanol were obtained from Fisher (Seoul, Korea). High purity water was prepared in-house using Ultra-high quality (UHQ) water purification system (Milli-Q, Millipore Corporation, Bedford, MA, USA).

2.3 Instruments and chromatographic conditions

The Waters ACQUITY UPLC^TM H-Class system (Waters Corporation, Milford, MA, USA) consisted of a quaternary pump, a column oven, an autosampler and a PDA detector was used for method development and validation. The Empower 3 software was used for instrument control and data acquisition process. Chromatographic separation was achieved on an ACQUITY UPLC^TM BEH C18 1.7 μm, 2.1 × 50 mm column (Waters Chromatography Ireland Limited, Dublin, Ireland). The mobile phase consisting of water (A) and acetonitrile (B), containing 1 %v/v TFA each, was delivered at a flow rate of 0.2 mL/min. The gradient elution program was optimized as follows: initial isocratic A-B of 68:32 from 0.0 to 1.0 min; linear gradient A-B of 35:65 from 1.0 to 7.0 min; linear gradient A-B of 68:32 from 7.0 to 7.3 min; isocratic A-B of 68:32 from 7.3 to 11.0 min. All analytes and interferences were achieved with a complete chromatographic separation under the total chromatographic run time of 11 min. The column oven temperature and detection wavelength were set at 30 °C and 280 nm, respectively. The injection volume was 2 μL.

2.4 Preparation of a system suitability solution

Six standard stock solutions of (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin were separately prepared by dissolving 5 mg of each compound with methanol in 25-mL volumetric flasks (200 μg/mL). A 3 mL of the standard stock solution was transferred to a 10-mL volumetric flask and diluted to volume with methanol to obtain a working standard solution of each compound at the concentration of 60 μg/mL.

A system suitability solution was then prepared by co-spiking 1 mL of (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin working standard solutions to obtain a solution mixture at a concentration of 10 μg/mL. The system suitability solution was filtered through a 0.22 μm nylon membrane filter prior to analysis.

2.5 Calibration curve preparation

A five-point calibration curve of each compound was constructed in the range of 8–80, 1–40, 5–100, 0.5–50, 1–50 and 1–40 μg/mL for (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin, respectively. Calibration samples were prepared in a dilution of the standard stock solution of each compound. The solution was filtered through a 0.22 μm nylon membrane filter before analysis.

2.6 Matrix and sample preparation

2.7 Method validation

The method was validated according to the ICH Q2 (R1) guidance for the validation of analytical procedures in the aspect of assay procedures (The International Conference on Harmonization, 2005). The validation parameters including system suitability, specificity, linearity, accuracy, precision, LOD, LOQ and robustness, were assessed under the appropriately designed validation experiments.

3.1 Method development

The analytical method was developed for simultaneous determination of six polyphenolic markers in orchid plants. The compounds contain phenolic groups that can ionize in the mobile phase. Separation of polyphenolic analytes on reverse-phase chromatography can be achieved by an ion-suppression technique using organic acids such as formic acid, acetic acid, or TFA (Lian et al., 2005). Formic acid is most often chosen as a modifier for ion suppression due to its eco-friendly with no pungent smell like glacial acetic acid. However, formic acid is insufficient to completely suppress the ionization of polyphenolic functional groups. A stronger acid such as TFA at 1 %v/v is usually preferred because of its ability to control pH of the mobile phase and suppress ionization of phenolic groups. Buffers such as acetate and phosphate at higher pH can cause peak splitting, especially long-eluting peaks. An ACQUITY UPLC^TM BEH C18 column was chosen because it can tolerate the low pH of the mobile phase and provide symmetrical peaks. In contrast, some phenolic compounds cannot retain on the C8 column and co-elute with matrix interferences. Alternatively, a phenyl column can make much more retention of phenolic analytes, resulting in a longer analysis time. Chromatographic separation of the six polyphenolic compounds was achieved at a flow rate of 0.2 mL/min with the gradient elution program varying the ratio of aqueous (1% v/v TFA in water) and organic (1% v/v TFA in acetonitrile) phases. Initially, the column condition was equilibrated at a 32% organic composition for 1 min. Then, the organic phase was slowly gradient up to 65 % to achieve the complete separation of all analytes within 7 min. It was noted that too fast ramping of an organic composition might make the peak coelution. After that, the mobile phase composition was gradient to the initial composition within 0.3 min. Conditioning the column should be performed with sufficient time to ensure column re-equilibration. An attempt to use methanol was unsuccessful due to peak broadening and long analytical run time. As the UV spectra of six detected polyphenols shown in Fig. 2, the detection wavelength was selected at 280 nm representing the maximum absorption wavelength of the analytes (λ_{(2S)-Eriodictyol} = 287; λ_{(2S)-Homoeriodictyol} = 287; λ_Moscatilin = 280; λ_Gigantol = 280; λ_Chrysotoxine = 279 and λ_Crepidatin = 279). The developed chromatographic condition provided sufficient separation among all analytes and other matrix interferences.

Close

Fig. 2

The UV spectra of six detected polyphenols: (2S)-eriodictyol, (2S)-hyomoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin.

Export to PPT

3.2 System suitability

Suitability was verified during method validation and sample analysis to ensure the validity of the chromatographic condition. The system suitability solution containing (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin at a concentration of 10 μg/mL for each compound was prepared by mixing 1 mL of each working standard solution (60 μg/mL). Fig. 3A and B represent the overlaid chromatograms of each working standard solution and the typical chromatogram of the system suitability solution. The retention times of (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin were 1.35, 2.21, 2.98, 3.39, 4.63 and 4.95, respectively. System performance was achieved with sufficient resolution between the adjacent analytes (R > 2.2), tailing factor (T < 1.1) and theoretical plate (N > 2,104). Five-replicate injection of the system suitability solution demonstrated that the system was reproducible with low variation of peak area, retention time, resolution, tailing factor and theoretical plate. The overall system suitability results summarized in Table 1 suggest that the developed chromatographic condition is suitable for further validation of simultaneous analysis of the six phenolic compounds.

Close

Fig. 3

Representative chromatograms of (A) the overlaid working standard solutions of each polyphenol standard at a concentration of 20 μg/mL, (B) the system suitability solution containing six polyphenols at a concentration of 10 μg/mL, (C) the diluent (methanol), and (D) the matrix extract of D. stricklandianum.

Export to PPT

3.3 Specificity

Specificity is crucial in method development to separate analytes from solvent, matrices, and potential impurity. Fig. 3C and D represent blank chromatograms of methanol and matrix extract of D. stricklandianum, respectively. Compared with the chromatograms of polyphenolic analytes as shown in Fig. 3A and B, the chromatograms of methanol and the matrix extract of D. stricklandianum revealed no co-eluting peaks at the retention times of the polyphenolic analytes. The specificity result suggests that the chromatographic system is sufficient to revolve the analyte peaks from other interference peaks.

3.4 Linearity and range

The linear relationships between analyte concentrations and signal responses were evaluated at the concentration range of 8–80, 1–40, 5–100, 0.5–50, 1–50 and 1–40 μg/mL for (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin, respectively. The calibration curves were constructed by plotting between the concentrations versus the peak responses of analytes. Using the linear least square model, the calibration curve was linear over the tested concentration range (r² > 0.995). The percent deviations of the mean back-calculated concentrations and actual concentrations of (2S)-eriodictyol, (2S)-homoeriodictyol, moscatilin, gigantol, chrysotoxine and crepidatin were in the ranges of −2.3–6.9, −9.0–4.1, −1.1–6.0, −2.3–2.4, −3.6–1.6, and −1.7–9.1, respectively. The %CV (n = 3) of the back-calculated concentrations was<6.5%. The calibration curve results summarized in Table 2 indicate that the ranges of the analytical procedure possess suitable levels of precision, accuracy and linearity.

The residual plots and regression generated by one-way analysis of variance demonstrated that the F values (F_ANOVA) of all regression lines were significantly less than the calculated F values (F_cal), indicating an excellent linear relationship between the peak response (y) and the concentration of analyte (x). The p-value is the regression parameter used to indicate whether the slope and y-intercept are significantly different from zero at a 95% confidence interval. The p-values of slope and y-intercept are summarized in Table 2. The p-values of slopes were<0.05, indicating a significant difference from zero. The p-values of intercepts exceeded 0.05, indicating that the intercepts of all regression lines were insignificantly different from zero. The linearity results suggest that the calibration curve within the specified range can be applied for the routine analysis of the six polyphenolic compounds, and the single-point standard can also be used for the single-point assay.

3.5 LOD and LOQ

LOD and LOQ of six polyphenolic analytes were experimentally determined, and the results are summarized in Table 3. The LOD of analytes showed the S/N ratio ranging from 4 to 6 with a %CV of six replicate injections of ≤ 1.9. The LOQ of analytes showed the S/N ratio ranging from 15 to 45. The accuracy and precision at the LOQ of all analytes were achieved with percent recovery ranging from 88.3 to 104.2 % and precision (%CV) of ≤ 6.5.

3.6 Accuracy and precision

The accuracy and precision of the matrix extract spiked with individual polyphenolic standards were determined at three concentration levels. The accuracy and precision results are summarized in Table 4. The intra-day accuracy and precision exhibited percent recoveries ranging from 86.0 to 105.7% with relative standard deviations of %CV between 0.1 and 8.4. The inter-day accuracy and precision showed percent recoveries ranging from 86.1 to 104.2% with relative standard deviations of %CV between 0.9 and 6.5. The results demonstrated satisfactory accuracy and precision of the proposed method.

3.7 Robustness

The system suitability solution was used to assess the robustness of the method under slight variations of a column temperature and a TFA concentration in the mobile phase. The robustness results are summarized in Table 5. The retention times of six polyphenolic analytes were insignificantly different among the modified chromatographic conditions. In addition to the retention time, the slight variations in the column temperature and TFA concentration had no impact on the reproducibility of injections. The %CV values of the retention time and peak response were within the limit criteria with % CV < 2. The variations in the column temperature and TFA concentration had a subtle effect on the peak separation but the chromatographic resolutions between two adjacent analytes remained>2. The slight changes in tested chromatographic factors did not affect the system’s performance as measured by the tailing factor (T < 1.1) and the number of theoretical plates (N > 2,041).

3.8 Method applications for determination of six polyphenolic bioactive markers in Dendrobium spp.

The content of six phenolic compounds (moscatilinm gigantol, crepidatin, chrysotoxine, eriodictyol and homoeriodictyol) in Dendrobium spp. was quantified using the developed and validated UPLC-UV method. Typical UPLC chromatograms of methanolic extracts of D. crumenatum, D. friedericksianum and D. officinale are depicted in Fig. 4. The content of six phenolic compounds in stems from the three Dendrobium plants is summarized in Table 6. The results show that only a few markers were found in each Dendrobium spp. with different amounts. These observations are in agreement with previous reports (Yang et al., 2006; Yang et al. 2007; Xu et al., 2009). For example, the HPLC method developed by Yang et al. (2006) for determination of 11 compounds in Dendrobium spp. demonstrated that only three primary detected markers were found including gigantol, moscatilin, and moscatin. In addition, Yang et al. (2007) reported that only 4 phenolic compounds including moscatilin, gigantol, moscatin, and coumarin were detected in which gigantol was observed as the primary bioactive marker (Yang et al., 2007). Xu et al. (2009) also found only gigantol and moscatilin in most orchid hybrids. In our study, moscatilin, gigantol and crepidatin were mainly found in D. crumenatum. However, eriodictyol, homoeriodictyol and chysotoxine were primarily located in D. friedericksianum and D. officinale.

Close

Fig. 4

Typical UPLC chromatograms of methanolic extracts of (A) D. crumenatum, (B) D. friedericksianum and (C) D. officinale.

Export to PPT