Rapid HPLC-MS Method for the Simultaneous Determination of Tea

An effective and rapid HPLC-MS method for the simultaneous separation of the eight most abundant tea catechins, gallic acid, and caffeine was develope...
2 downloads 12 Views 2MB Size
Download PDF
Article pubs.acs.org/JAFC

Rapid HPLC-MS Method for the Simultaneous Determination of Tea Catechins and Folates Monica Araya-Farias,†,‡,§ Alain Gaudreau,†,‡ Elodie Rozoy,†,‡ and Laurent Bazinet*,†,‡ †

Institute of Nutrition and Functional Foods (INAF) and ‡Department of Food Sciences and Nutrition, Laval University, Quebec, QC, Canada G1V 0A6 ABSTRACT: An effective and rapid HPLC-MS method for the simultaneous separation of the eight most abundant tea catechins, gallic acid, and caffeine was developed. These compounds were rapidly separated within 9 min by a linear gradient elution using a Zorbax SB-C18 packed with sub 2 μm particles. This methodology did not require preparative and semipreparative HPLC steps. In fact, diluted tea samples can be easily analyzed using HPLC-MS as described in this study. The use of mass spectrometry detection for quantification of catechins ensured a higher specificity of the method. The percent relative standard deviation was generally lower than 4 and 7% for most of the compounds tested in tea drinks and tea extracts, respectively. Furthermore, the method provided excellent resolution for folate determination alone or in combination with catechins. To date, no HPLC method able to discriminate catechins and folates in a quick analysis has been reported in the literature. KEYWORDS: green tea, catechins, folates, HPLC, mass spectrometry

■

INTRODUCTION Green tea, a nonfermented tea, has become one of the most widely consumed beverages in the world. The increasing consumption of green tea can be explained by the health effects of some of its constituents. Indeed, green tea is rich in flavonoids (300−400 mg/g) and an excellent source of catechins, up to 20−30% of the green tea dry weight.1,2 The eight most abundant catechins in tea are (−)-gallocatechin, (−)-epigallocatechin, (+)-catechin, (−)-epicatechin, (−)-epigallocatechin gallate, (−)-gallocatechin gallate, (−)-epicatechin gallate, and (−)-catechin gallate. The principal tea catechins are known for their strong antioxidant activity and their health benefits, including antiviral, antiallergic, anti-inflammatory, and anticariogenic properties.3,4 In addition to tea catechins, green tea contains other polyphenols such as gallic acid, the main tea polyphenolic acid, and certain amounts of alkaloids such as caffeine. Although caffeine is well-known for its stimulatory effect, it has been reported that caffeine also exerts beneficial effects on human health including cardiovascular, gastrointestinal, and respiratory effects.5 Caffeine affects also the taste of tea.6 The composition of green tea catechins, phenolic acids, and caffeine varies according to variety, climate, horticultural conditions, and mainly in technologies applied during extraction, concentration, and storage. Folates play a critical role in the reduction of neural tube defects in newborns and the risk of cardiovascular diseases.7 Folic acid (FA) is not a significant natural form of folate. However, FA is synthetically produced and used as a dietary supplement because it is the most stable form of this vitamin. L5-Methyltetrahydrofolic acid (L-5-MTHF) is a natural form of folate, and it is more bioactive than folic acid. However, it has limited stability especially when exposed to oxygen, light, or high temperatures.8 Recently, research was focused on finding ways to stabilize this molecule in food matrices to reduce or replace the use of synthetic FA in fortified foods.9,10 Rozoy et al.10 showed that degradation of L-5-MTHF could be © 2014 American Chemical Society

controlled by the addition of ascorbic acid. Some studies have proven that the antioxidant capacity of (−)-epigallocatechin gallate exceeds the effect of ascorbic acid at least 100 times.11 Thus, the use of molecules with strong antioxidant activity such as catechins could be an interesting alternative to protect L-5-MTHF against oxidation. To date, high-performance liquid chromatography (HPLC) remains the most used technique for the determination of catechins. Traditional HPLC coupled with ultraviolet (UV), photodiode array (DAD), electrochemical detection, and mass spectrometry (MS) are among the most employed methods in the identification of catechins.12−20 In general, the analysis time is relatively long, 40−105 min, and between 5 and 10 compounds of interest can be simultaneously separated. From a quantitative point of view, fast HPLC approaches have not been extensively investigated, and only a few works have reported a simultaneous separation of all tea catechins in a relatively short time (within 25 min).21−24 Most recently, ultrahigh-pressure liquid chromatography (UHPLC) has shown a considerable potential in terms of speed and separation efficiency.25,26 However, UHPLC is still relatively new and requires specialized equipment able to withstand pressures of 1000 bar. Up to now, only a few applications of UHPLC have been reported for the analysis of catechins in tea samples.27−30 To enhance chromatographic performances in terms of efficiency and rapidity, several strategies can be implemented. An interesting approach would be to evaluate the use of columns packed with sub 2 μm particles. As described 30 years ago by Knox, Martin, and other authors, the use of short columns packed with small particles induces an increase in efficiency, optimal velocity, and mass transfer.31−34 Received: Revised: Accepted: Published: 4241

November 28, 2013 March 31, 2014 April 15, 2014 April 15, 2014 dx.doi.org/10.1021/jf4053258 | J. Agric. Food Chem. 2014, 62, 4241−4250

Journal of Agricultural and Food Chemistry

Article

according to the procedure developed by Bazinet et al.35 The green tea used in this experiment was a commercially available and Japanese green tea obtained from local retailer. Before being used, the green tea was stored in vacuum bags at room temperature in a dark and dry space. The EGCG-enriched tea drink was produced at a volume of 100 L, in conditions and with unit operations similar to the ones met in the food industry for the production of low-acid beverages. Dry leaves (3.5 kg) were brewed twice in 100 L of water. The first brewing step was carried out at 30 ± 2 °C for 10 min using a stainless steel double-jacket heated reservoir. The infusion was filtered to separate the leaves and to collect the filtrate. The filtered tea leaves were brewed again in 100 L of water at 80 ± 2 °C for1 h. The green tea leaves were squeezed and discarded. The second filtrate was adjusted to a pH value of 3.5 by lemon juice addition. The ECGC-enriched green tea was then pasteurized and aseptically bottled and stored at 4 °C. Enriched Green Tea Extract. An enriched freeze-dried tea extract was prepared in our laboratory using the second filtrate, which was prepared according to the procedure previously cited. This extract was not adjusted to pH 3.5. The extracts were frozen at −30 °C for 48 h and lyophilized at 25 °C for 72 h (Virtis, Gardiner, NY, USA). For the analyses, 400 mg of extract was diluted to 25 mL with HPLC grade water. All prepared tea samples were filtered through 0.2 μm PVDF syringe filters. A 10 μL aliquot of this tea sample was injected onto HPLC. All tea samples were prepared and analyzed in triplicate for a total of nine analyses per sample. Preparation of Folates. Solutions of folic acid and L-5-MTHF were prepared separately in CHES−HEPES buffer at a concentration of 227 μM. On the other hand, solutions of 1 mM L-5-MTHF with an enriched green extract (at 400 mg/L) were prepared in Britton− Robinson buffer10 and analyzed by HPLC-MS. Data Analysis. Calibration curves were done using standard solutions, and linearity was evaluated over six calibration points with three measurements for each calibration point. Detection limits (LOD) and quantitation limits (LOQ) were determined according to the ICH Harmonized Tripartite Guideline (Validation of Analytical Procedures: Text and Methodology Q2 (R1)).36 Repeatability of the method was calculated in terms of percent relative standard deviation (% RSD) by performing HPLC analysis on aliquots of each sample in triplicate on three consecutive days. Relative standard deviation (RSD) was defined as sample standard deviation divided by sample mean, multiplied by 100%.

The aim of this study was to develop an efficient and rapid HPLC method for the simultaneous separation of catechins, gallic acid, and caffeine in tea samples. To evaluate the resolution in other complex matrices, the method was applied to the determination of folates. Chromatographic analyses of L5-MTHF in combination with green tea enriched extracts were also performed to evaluate the analytic performance of the method proposed in a complex solution. To perform faster analysis, a conventional HPLC system was equipped with a Zorbax Stable Bond (SB) C18 analytical column (4.6 × 50 mm) packed with 1.8 μm particles. HPLC was coupled with MS to attain sufficient sensitivity and selectivity for the identification of catechins, folates, and other constituents of tea samples. To the best of our knowledge, there is no method published yet on the simultaneous separation of catechins and folates.

■

MATERIALS AND METHODS

Chemical and Standards Preparation. Standards were purchased from Sigma Chemical Co. (St. Louis, MO, USA): gallic acid (GA, 98%), (−)-gallocatechin (GC, 98%), (−)-epigallocatechin (EGC, 98%), (+)-catechin (C, 98%), (−)-epicatechin (EC, 98%), (−)-epigallocatechin-3-gallate (EGCG, 95%), (−)-gallocatechin gallate (GCG, 98%), (−)-epicatechin gallate (ECG, 98%), (−)-catechin gallate (CG, 98%), and caffeine (CAF, 99%). HPLC grade water was obtained using a Siemens Purelab Ultra Water-Purification System (Siemens, USA). HPLC grade acetonitrile and formic acid were purchased from VWR International (Mississauga, ON, Canada) and Sigma-Aldrich (St. Louis, MO, USA) respectively. The stock solutions were prepared by dissolving 5−10 mg of each standard compound in 10 mL of HPLC grade water/formic acid (99.9:0.1 v/v). Each stock solution was then used for the preparation of a mix of standards. Serial dilutions were then prepared with formic acid solution to produce calibration curves at different concentrations ranging between 3 and 50 μg/mL. All standard solutions were filtered through 0.2 μm PVDF syringe filters. A 10 μL aliquot of solutions was injected onto HPLC. HPLC-MS Analysis. The HPLC UV-MS analyses were performed on an Agilent 1100 series (Santa Clara, CA, USA) HPLC system equipped with an Agilent diode array UV−vis detector (Agilent G1315 DAD), a vacuum degasser, a binary pump delivery (Agilent G1312A), a refrigerated autosampler (Agilent G1329A), and an HP mass spectrometer detector (MSD, model G1948B). Integration and data processing were performed by Chemstation software (LC/MSD Chemstation B.01.03 SR1). An Agilent Zorbax SB-C18 (4.6 mm id × 50 mm, 1.8 μm) column with a precolumn filter was used. HPLC UVMS analyses were carried out using mobile phases consisting of 0.2% formic acid in water (A) and 0.1% formic acid in acetonitrile (B) (HPLC grade, EMD Chemicals Inc., Gibbstown, NJ, USA). The linear gradient elution system was as follows: 0−8 min, 0−5% (v/v) of B; 8− 10 min, 5−25% (v/v) of B; 10−17 min, 25−100% (v/v) of B; and 17−23 min, 100−5% (v/v) of B, at a flow rate of 1 mL/min. The quantification of catechins by UV-DAD was performed at 230 nm. The column temperature was maintained at 25 °C, and the injection volume was 10 μL. The same column of HPLC-DAD was used for analyses by MS. Mass spectra of catechins were recorded in the positive ionization mode using an electrospray (API-ES) ionizing source with nitrogen as drying gas. Spray chamber parameters were as follows: capillary potential, 3000 V; gas temperature, 350 °C; drying gas flow, 13 L/min; nebulizer pressure, 60 psi. Quantitative analysis by MS was carried out in the scanning mode, scan range being m/z 150−1000, setting the fragmentor value at 70 V with a molecular ion [M+H]+ for each catechin chosen as the most abundant and representative signal. Identification of compounds by HPLC-MS analysis was carried out by comparing retention times and mass spectra of the unknown peaks to those of the standards. Preparation of Green Tea Samples. Enriched Green Tea Drink. An EGCG-enriched tea drink was prepared in our laboratory

■

RESULTS AND DISCUSSION Description of Developing Steps To Improve the Methodology. A reversed-phase HPLC method coupled with mass spectrometry was developed. Two mobile phases, one containing HPLC grade water and the other containing acetonitrile, were tested. The formic acid was added to both mobile phases. The presence of acid in the mobile phases is essential to attain a complete separation of catechins.17 Moreover, it was reported that catechins are most stable in acidic conditions.37 Two different concentrations of formic acid (0.1 and 0.2%) were tested in the present study. Although a good separation was achieved with 0.1% formic acid in both mobile phases, improved peak shapes were obtained with the inclusion of 2% formic acid in the water mobile phase. Another reason for the use of the formic acid was its compatibility with mass spectrometry. The elution gradient program was described earlier. Different absorption wavelengths have been reported in the literature for the determination of tea catechins, varying from 210 to 280 nm.12−20 In our study, all catechins, gallic acid, and caffeine exhibited maximun absorbance at 210 and 230 nm. The absorbances at 210 nm were slightly larger than those at 230 nm, but the baseline shifts of chromatograms were also increased. Folates were found to have maximum absorbances at 280−290 nm, as was already mentioned in the literature.38,39 4242

dx.doi.org/10.1021/jf4053258 | J. Agric. Food Chem. 2014, 62, 4241−4250

Journal of Agricultural and Food Chemistry

Article

Figure 1. Chromatogram of mixed standards of GA, catechins, and CAF recorded by UV-DAD at 230 nm. GA, gallic acid; GC, gallocatechin; EGC, epigallocatechin; C, catechin; CAF, caffeine; EC, epicatechin; EGCG, epigallocatechin gallate; GCG, gallocatechin gallate; ECG, epicatechin gallate; CG, catechin gallate.

column was better when the trifluoroacetic acid was added to the mobile phases as reported in our previous work.19 The critical step in the quantification of catechins, phenolic acids, and caffeine in tea by HPLC methods is the sample extraction. The extraction method must enable complete extraction of the compounds of interest without chemical modification or degradation. At present, a single extraction step with acetonitrile and methanol has been widely used to quantify polyphenols in tea.14,16,24,40 However, our method allows the quantification of catechins in tea without any extraction with solvents. In fact, after the preparation of tea samples in boiling water, the samples needed only to be diluted and filtered prior to the HPLC analysis. Hu et al.24 also analyzed catechins in tea after hot water brewing, but their method needed a semipreparative/preparative HPLC step prior to the identification of catechins by analytical HPLC. Mass spectra of tea catechins was recorded in the 40−120 V fragmentor range to determine the highest sensitivity for each molecule. A fragmentor value of 70 V was chosen as the best compromise among all ions detected and applied to the tea samples. The method was tested in negative and positive ionization modes, but a better sensitivity and a significant improvement in signal-to-noise response were achieved when the mass spectrometer was operated in positive ionization mode. Pelillo et al.22 also reported a better signal response when HPLC-MS analysis of green tea extracts was obtained with positive ion detection. HPLC-MS Analysis and Peak Identification. The typical chromatogram of mixed standards of GA, catechins, and CAF is shown in Figure 1. For this separation, the generated back

However, a strong absorption was also observed at the shortest wavelength (230 nm). The detector wavelength was then set at 230 nm as the best compromise to reach maximun absorbance for all compounds of interest. To increase separation efficiency and speed of analysis, the HPLC system was equipped with a Zorbax SB-C18 column (4.6 mm id × 50 mm) packed with 1.8 μm silica-based particles. The use of small particles is one of the best solutions to improve chromatographic performances.31−34 Moreover, good efficiency can be maintained with a short column and a high flow rate. On the other hand, the special and ultrapure silica support of the Zorbax SB-C18 column is designed to reduce or eliminate undesirable interactions between analytes and the silica surface, thereby providing best separation of all compounds. It is especially suited in applications that use high-sensitivity detectors such as mass spectrometers. That was another reason for the use of this specific column. The performance of the Zorbax SB-C18 1.8 μm column was compared to that of a YMC-Pack ODS-AM C18 5 μm column, which has already been used for the determination of catechins. Both columns were tested under the same conditions. The Zorbax column gave good separation for all of the analytes within 9 min (Figure 1). However, the YMC-Pack ODS-AM column did not achieve a satisfactory resolution of EC, EGCG, ECG, and CG in the water/acetonitrile/formic acid system, and the separation time was generally long (47 min) (data not shown). In this context, the analysis time and solvent consumption were decreased by 5.2 times using the Zorbax SB-C18 column. The performance of the YMC-Pack ODS-AM 4243

dx.doi.org/10.1021/jf4053258 | J. Agric. Food Chem. 2014, 62, 4241−4250

Journal of Agricultural and Food Chemistry

Article

Table 1. Retention Times and MS Spectral Data of Gallic Acid, Caffeine, and Catechin Standards retention time (min) compda GA GC EGC C CAF EC EGCG GCG ECG CG

regression eq y y y y y y y y y y

= = = = = = = = = =

40.302x 19.293x 27.426x 44.752x 13.534x 33.639x 33.756x 16.302x 23.987x 43.903x

+ 45.137 − 13.746 − 20.407 + 15.259 − 3.347 + 7.819 − 11.891 − 8.936 − 1.002 + 15.146

R2

LOD (μg)

LOQ (μg)

UVb

MS

MW (g/mol)

M [+H]+ (m/z)

0.9998 0.9999 0.9999 1.0000 0.9999 1.0000 0.9998 0.9999 0.9999 0.9998

1.279 0.509 0.972 0.450 1.014 0.495 1.328 0.717 0.515 1.342

4.265 1.696 3.241 1.501 3.380 1.652 4.427 2.392 1.716 4.475

1.275 2.690 5.966 6.324 6.534 7.353 7.518 7.819 8.770 8.853

1.290 2.704 6.001 6.338 6.548 7.365 7.534 7.830 8.863 8.863

170.12 306.27 306.03 290.03 194.14 290.20 458.40 458.40 442.40 442.37

171.03 307.08 307.08 291.10 195.10 291.12 459.10 459.07 443.08 443.08

a

GA, gallic acid; GC, gallocatechin; EGC, epigallocatechin; C, catechin; CAF, caffeine; EC, epicatechin; EGCG, epigallocatechin gallate; GCG, gallocatechin gallate; ECG, epicatechin gallate; CG, catechin gallate. bUV, detection at 230 nm.

Figure 2. Chromatograms of catechin standards recorded by both UV-DAD (blue line) and MSD (red line) detectors at 230 nm. GA, gallic acid; GC, gallocatechin; EGC, epigallocatechin; C, catechin; CAF, caffeine; EC, epicatechin; EGCG, epigallocatechin gallate; GCG, gallocatechin gallate; ECG, epicatechin gallate; CG, catechin gallate.

40−105 min.12−20 Using a Zorbax SB-C18 sub 2 μm column, the HPLC analyses could be performed 4.4−12 times faster than those reported for a C18 5 μm column, and because the pressure was only 220 MPa and solvent consumption 9 mL by analysis, this method is an excellent option for the determination of catechins. Catechin detection was linear over the ranges tested (3−50 μg/mL) with coefficients of determination (R2) between 0.9998 and 1.0000. LODs and LOQs for most standards were in the range of 0.5−1.0 and 1.0−4.0 μg, respectively. CG had the highest LOD at 1.342 μg, and C had the lowest LOD at 0.450 μg. The results are summarized in Table 1.

pressure was around 220 MPa, which is very compatible with conventional HPLC system (HPLC standard pressure, up to 400 bar, 40 MPa). The principal advantage of the here-developed HPLC-MS method is the high baseline resolution of 10 compounds in a very short time with an excellent separation quality for GA and CAF. The elution order was GA, GC, ECG, C, CAF, EC, EGCG, CGC, ECG, and CG. These 10 compounds were successfully separated within 9 min by linear gradient elution of formic acid solution and acetonitrile (Figure 1). Most HPLC methods for the separation of tea catechins are carried out on C18 5 μm columns (4.6 mm id × 250 mm or 4.6 mm id × 150 mm), and 5−10 compounds can be simultaneously separated in 4244

dx.doi.org/10.1021/jf4053258 | J. Agric. Food Chem. 2014, 62, 4241−4250

Journal of Agricultural and Food Chemistry

Article

GA, and CAF were separated within 15 min by a linear gradient elution of formic acid and methanol.24 However, the method was focused principally on the preparation and purification of methylated catechins; thus, several steps of isolation and semipreparative HPLC were required. Furthermore, the time for cleaning and column conditioning was not mentioned. Although those methods are rapid, the preparation of samples may become a difficult task due to the numerous steps. To our knowledge, only one chromatographic method based on the utilization of a Zorbax SB-C18 sub 2 μm column was developed for the analysis of phenolic compounds including tea catechins.30 The analysis time was 12 min, and 6 standards (GC, EGCG, EC, GC, EC, CG) were separated with satisfactory resolution. However, there was no resolution for standards EGC and C. In the matrix of green tea infusion, the resolution was more critical, and only the peaks of GCG and ECG had good response and were sufficiently separated. Separation and Identification of Catechins in Tea Samples. Identification of peaks of tea samples was carried out by comparing retention times and mass spectra of the standards and data in the literature (Tables 2 and 3). These results are in good agreement with catechin standards. The typical HPLCUV profiles of enriched tea drink and green tea extracts are shown in Figures 4 and 5, respectively. In both tea samples we also observed a peak eluting at 8.4 min (not identified on the chromatogram) that could correspond to a second peak of GC. The catechin contents in enriched green tea drink are shown in Table 2. The results obtained indicate that EGCG was in highest concentration (660.9 mg/L) in green tea drinks, whereas GA (3.0 mg/L) was the lowest. CG was not quantified due to its low level. As expected, calculated amounts of EGCG (the most interesting catechin compound) were higher than those of similar tea infusions or beverages15 because we produced an EGCG-enriched green tea drink. Chen et al.44 reported that one canned or bottled tea drink (250 mL) contained less green tea catechins (3−60 mg) than one cup of conventional brewed green tea drink (400−500 mg). The content of catechins in tea samples can vary depending on the conditions and technology used for the extraction and storage.22 In terms of quantification, % RSD values obtained were quite low, ranging from 0.8 to 3.5%. The relatively low % RSD for GA, most catechins, and CAF indicated that this

The retention time and MS spectral data of standards are shown in Table 1. The results are in agreement with the values reported in the literature.22−24,40 In Figure 2 chromatograms of standards recorded by both UV and MSD detectors are compared. As shown in Figure 2, EGC and CG exhibited different retention times, and two peaks can be easily differentiated in UV spectra. However, EGC and CG had identical [M+H]+ ions at 443.08, and only one peak (with a shoulder) can be observed in MS spectra. A typical mass spectrum of EGCG is shown in Figure 3. The mass spectra

Figure 3. Mass spectrum of EGCG recorded by API-ES+ with the fragmentor set at 70 V.

showed a characteristic product ion of EGCG (m/z 459.1 is the [M+H]+ion). Similar mass spectra were reported in the literature.22 The mass spectrum contains several extra signals at m/z 151.1, 289.1, 622, 0 and 921.1. The m/z signal at 289.1 would be a typical fragment ion detected in the mass spectrum of catechins.40 To summarize the information given above, separation and identification of green tea compounds by HPLC are typically slow, requiring complex and time-consuming gradients.12−20,41−,43 To date, only a few methods have allowed the simultaneous determination of the main catechins, gallic acid, and caffeine in