Food Chemistry 136 (2013) 368–375
Contents lists available at SciVerse ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Analytical Methods
An improved mass spectrometric method for identification and quantification of phenolic compounds in apple fruits D. De Paepe a,b,⇑, K. Servaes a, B. Noten a, L. Diels c, M. De Loose b, B. Van Droogenbroeck b, S. Voorspoels a a
Flemish Institute for Technological Research (VITO), Environmental Analysis and Technology, Boeretang 200, 2400 Mol, Belgium Institute for Agricultural and Fisheries Research (ILVO), Technology and Food Science Unit (T&V), Burgemeester Van Gansberghelaan 115, 9820 Merelbeke, Belgium c Flemish Institute for Technological Research (VITO), Industrial Innovation, Boeretang 200, 2400 Mol, Belgium b
a r t i c l e
i n f o
Article history: Received 28 August 2011 Received in revised form 24 May 2012 Accepted 29 August 2012 Available online 7 September 2012 Keywords: Optimisation Validation Polyphenols Accurate mass Orbitrap UHPLC
a b s t r a c t Thirty-nine phenolic compounds were analysed using ultra high performance liquid chromatography (UHPLC) coupled with diode array and accurate mass spectrometry detection using electrospray ionisation (DAD/ESI-am-MS). Instrumental parameters such as scan speed, resolution, and mass accuracy were optimised to establish accurate mass measurements. The method was fully validated in terms of model deviation (r2 > 0.9990), range (typically 10–3500 ng g1), intra/inter-day precision (<6% and <8%, respectively) and accuracy (typically 100 ± 10%). The mass accuracy of each selected phenolic compound was below 1.5 ppm. The results confirmed that the UHPLC-DAD/ESI-am-MS method developed here was convenient and reliable for the determination of phenolic compounds in apple extracts. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Phenolic compounds have received increasing attention in recent years because of their bioactive functions and possible beneficial effects on human health. Epidemiological studies show relations between consumption of polyphenol-rich foods and prevention of diseases such as cancer, coronary heart disease and osteoporosis (Nováková, Spácil, Seifrtová, Opletal, & Solich, 2010; Sato et al., 2011). Ignat et al. recently reviewed the qualitative and quantitative analysis of phenolic compounds from fruits and vegetables (Ignat, Volf, & Popa, 2011). Despite the large number of investigations made, the separation and quantification of different phenolic compounds, especially the simultaneous determination of phenolic compounds belonging to several subclasses, remains an analytical challenge (Aldini et al., 2011). Method development is hampered by the wide variety of chemical and related physicochemical properties, great differences in concentration, and the lack of commercially-available standards (Vallverdú -Queralt, Jáuregui,
⇑ Corresponding author. Present address: Flemish Institute for Technological Research (VITO), Business Unit Environmental Analysis and Technology (MANT), Boeretang 200, 2400 Mol, Belgium. Tel.: +32 14 33 50 08, +32 14 31 94 72, +32 09 272 28 38; fax: +32 09 272 28 01. E-mail addresses:
[email protected], domien.depaepe@vito. be (D. De Paepe). 0308-8146/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2012.08.062
Medina-Remón, Andrés-Lacueva, & Lamuela-Raventós, 2010). The challenge is to develop an analytical method that is applicable on a large scale to separate and identify all phenolic compounds of interest (Abad-García, Berrueta, Garmon-Lobato, Gallo, & Vicente, 2009). High performance liquid chromatographic (HPLC) techniques are now widely used for quantification of phenolic compounds (Abad-García et al., 2009). Nevertheless, due to sensitivity disadvantages resulting sometimes in too high detection limits, HPLC methods present limitations for the analysis of complex matrices such as crude plant extracts (Kartsova & Alekseeva, 2008). These disadvantages make it necessary to perform an initial pre-concentration and purification step to remove potential interfering components prior to HPLC analysis (Ignat et al., 2011). Applicability calls for a compromise between speed and resolution, resulting in typical analysis times of 45 min or longer (Nováková et al., 2010; Spácil, Nováková, & Olich, 2008). These short comings can be dealt with by using state-of-the-art instruments such as ultra high performance liquid chromatography (UHPLC) systems (Spácil, Nováková, & Solich, 2010). UHPLC allows a higher separation efficiency on sub-2-lm particle sorbents and faster chromatographic separation while keeping the same resolution as HPLC sorbents with a conventional particle size (Guillarme, Nguyen, Rudaz, & Veuthey, 2007). This allows separation and detection of all the phenolic compounds in a single extract from plant material (Gómez-Romero, Segura-Carretero, & Fernández-Gutiérrez, 2010; Lin & Harnly, 2007).
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375
UV/VIS diode array (DAD) and/or mass spectrometry are the most common detection methods for phenolic compounds, but they share some weaknesses (Harnly, Bhagwat, & Lin, 2007; Spácil et al., 2010). They both lack of structural confirmation and specificity which could lead to possible sample matrix interference and misinterpretation of unknown compounds (Aldini et al., 2011). To identify the compounds, ion trap, single- and triple quadrupole mass spectrometers with electrospray ionisation (ESI) or atmospheric pressure chemical ionisation (APCI) are used (Lin & Harnly, 2007; Magiera, Baranowska, & Kusa, 2012). The main limitation of these technologies is that they can only identify and quantify a predefined list of target compounds. These techniques do not allow to perform a non-targeted screening analysis, thereby identifying unknown compounds present in the sample extract. (Abad-García et al., 2009). Furthermore, only a restricted number of target compounds can be simultaneously screened without loss of sensitivity by the aforementioned detection techniques (Moulard et al., 2011). To remedy these shortcomings, high-resolution mass spectrometers have recently increased in popularity because they can reveal the so-called accurate mass (am) of the analytes. The most common mass spectrometers of this type are time-of-flight (TOF), Fourier transform ion cyclotron resonance (FT ICR) and the Orbitrap detectors (Moulard et al., 2011). Accurate mass measurement coupled with sufficient resolution makes it possible to restrict the enormous number of possible molecular formulas corresponding with a particular molecular mass (Moulard et al., 2011). The fast elemental formula calculation of detected ions made possible by accurate mass measurement is the first step in the identification of unknown compounds and structure elucidation (VallverdúQueralt et al., 2010). In particular, the single stage Orbitrap (Exactive™, Thermo Fisher Scientific, Bremen, Germany) mass analyser provides high mass resolution, high mass accuracy and good sensitivity. In combination with retrospective analysis, this offers a new screening tool to identify phenolic compounds based on accurate mass and isotopic peak ratios (Makarov, 2000; Moulard et al., 2011). Furthermore, due to the sufficiently high scan rates, Orbitrap mass analysers provide sufficient points across narrow chromatographic peaks. This enables the coupling with UHPLC. The objective of the current study is to set up and validate an identification and quantification method for phenolic compounds (flavonoids, oligomeric flavonoids and phenolic acids) based on UHPLC-DAD/ESI-am-MS, that has the potential to be used as a generic screening method for phenolic compounds. For method development, the peel of apple fruit was the matrix of choice due to the high content of phenolic compounds from several phenolic subclasses.
2. Materials and methods 2.1. Chemicals and reagents UHPLC-grade methanol, acetonitrile, and water were purchased from Biosolve (Valkenswaard, The Netherlands). Formic acid, acetic acid, ammonium formate, ammonium acetate and (D-Ala)2-leucine enkephalin were supplied by Sigma–Aldrich (Bornem, Belgium). Commercially available mixtures to calibrate the mass spectrometer, i.e., MSCAL5–1EA (caffeine, tetrapeptide ‘‘Met-Arg-Phe-Ala’’, UltramarkÒ) for positive ion mode and MSCAL6–1EA (sodium dodecylsulfate, taurocholic acid sodium salt, UltramarkÒ) for negative ion mode, were purchased from SUPELCO (Bellefonte, PA, USA). A mixture of different compounds belonging to 7 flavonoid, 1 proanthocyanidin and 4 phenolic acid subclasses was chosen to develop the method. The following analytical reference standards
369
were purchased from Phytolab (Vestenbergsgreuth, Germany): flavones: apigenin, apigenin-7-O-glucoside (apigetrin), luteolin, luteolin-7-O-glucoside (cynaroside); flavonols: isorhamnetin, kaempferol, kaempferol-3-O-glucoside (astragalin), quercetin, quercetin-3-O-glucoside (isoquercitrin), quercetin-3-O-galactoside (hyperin), quercetin-3-O-rutinoside (rutin), quercetin-3-O-arabinoside (avicularin), quercetin-3-O-rhamnoside (quercitrin), galangin; dihydrochalcones: phloretin, phloretin-O-20 -glucoside (phloridzin); flavanones: naringenin, naringenin-7-O-neohesperidoside (naringin); flavanols: (+)-catechin, ()-epicatechin; flavanonols: (+)-dihydrokaempferol ((+)-aromadendrin), (+)dihydroquercetin ((+)-taxifolin); anthocyanidins: cyanidin chloride, cyanidin-3-O-glucoside chloride (kuromanin chloride), cyanidin-3O-galactoside chloride (ideain chloride), cyanidin-3-O-rutinoside chloride (keracyanin chloride) and procyanidins: procyanidin B2. Analytical reference standards of hydroxybenzoic acids: salicylic acid, protocatechuic acid, gallic acid, propyl gallate; hydroxycinnamic acids: p-coumaric acid, caffeic acid, ferulic acid, sinapinic acid, chlorogenic acid; hydroxyphenylpropanoic acids: dihydrocaffeic acid, dihydroferulic acid; hydroxyphenylacetic acids: 4-phydroxyphenyl acetic acid were obtained from Sigma–Aldrich (Bornem, Belgium). 2.2. Preperation of the stock and calibration solutions Standard stock solutions at a concentration of 1 mg mL1 were prepared in UHPLC-grade methanol for each analyte separately. From the stock solutions, a multi-compound standard solution was prepared in which each of the 39 individual components were present in a concentration of 25 000 ng mL-1. Twenty-three calibration solutions in a concentration range of 25 000–1 ng mL1 were made from the multi-compound stock solution. All solutions were stored at 4 °C in septum-capped amber-coloured vials to protect the compounds from light and moisture. Prior to analysis, each calibration solution was diluted 6:10 (v/v) in a microvial using a 40 mM ammonium formate buffer, resulting in a calibration series ranging from 15 000 to 0.6 ng mL1. 2.3. Sample preparation Apples (Malus domestica Barkh cv. KanziÒ) were collected during the commercial harvest on September 17th, 2010, at the experimental agricultural station PCFruit (Velm, Belgium). Apples were cooled and stored at 2 °C prior to sample preparation. Prior to freeze-drying, the apples were peeled using a semi-automatic device for a reproducible peel thickness of 3 mm. The samples were immediately frozen in liquid nitrogen to avoid enzymatic browning. Directly from the liquid nitrogen, the samples were transferred into a freeze dryer with heated shelves at 25 °C (GAMMA 1–16 LSC Martin Christ, Osterode am Harz, Germany). Following the freeze-drying process, the apple peel was grounded in a commercial blender (DP705 LA Moulinette, Group SEB, Fleurus, Belgium) and consequently stored under N2 atmosphere in an amber-coloured flask at 25 °C. The inert atmosphere of N2 gas avoided rehydration, biological contamination, and compound degradation. 2.4. Instrumental method Identification and quantification of the selected phenolic compounds were performed via an UHPLC-DAD/ESI-am-MS configuration. The LC system consisted of an Accela™ quaternary solvent manager, a ‘Hot Pocket’ column oven (Thermo Fisher Scientific, Bremen, Germany) and a CTC PAL™ autosampler (CTC Analytics, Zwingen, Switzerland). A reversed phase separation was performed on a Waters Acquity UPLCÒ BEH SHIELD RP18 column, with
370
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375
dimensions 3.0 150 mm, 1.7 lm (Waters, Milford, MA). To protect the UHPLC column, a Acquity BEH RP18 VanGuard pre-column, with dimensions 1.7 lm, 2.1 5 mm (Waters, Milford, MA) was coupled with the analytical column. The mobile phase consisted of water + 0.1% formic acid (solvent A) and acetonitrile + 0.1% formic acid (solvent B). The gradient was varied linearly from 0% to 26% B (v/v) in 9.91 min, to 65 % B at 18.51 min, and finally to 100% B at 18.76 min and held at 100% B to 20.76 min. Afterwards, the initial conditions of 100% A were re-equilibrated from 20.88 min to 23 min prior to the next injection. The flow rate was 500 lL min1 and the column temperature was set at 40 °C. Aliquots of 5 lL of the sample extract were injected into the chromatographic system. The UV spectra of all selected phenolic compounds were recorded in the range of 220–400 nm for tentative identification using an Accela™ photo diode array (PDA) detector. The UHPLC system was coupled to an Orbitrap mass spectrometer (Exactive™, Thermo Fisher Scientific, Bremen, Germany) operating with an Ion Max™ ESI source (Thermo Fisher Scientific, Bremen, Germany) in negative ionisation mode (ESI-) using the following operation parameters: spray voltage 2.5 kV; sheath gas (N2, >99.99%) 47 (adimensional); auxiliary gas (N2, >99.99%) 15 (adimensional); skimmer voltage 25 V; tube lens voltage 110 V; and capillary temperature 350 °C. The mass spectra were acquired using an acquisition function as follows: resolution, high (equivalent to a mass resolving power of 50,000 FWHM at m/z
200); automatic gain control (AGC), balance (target value of 1 106), and scan speed, 2 Hz. Mass range in the full scan experiments was set at m/z 90–1800. To guarantee high mass accuracy during run-time, the Orbitrap™ was externally calibrated in both positive and negative ionisation mode prior to each measurement. All the analyses were performed using a lock spray with internal lock mass of a solution of (D-Ala)2-leucine enkephalin (5000 ng mL1, 12C[M–H], m/z 568.27767) delivered to the ESI source at 5 lL min1 by using an additional LC pump (Hewlett Packard HP/Agilent 1100 HPLC Pump, Santa Clara, CA, USA). Detection of the targeted phenolic compounds was based on theoretical exact mass and on retention time (Table 1). Data were evaluated by Xcalibur 2.2.1 (Thermo Fisher Scientific, Bremen, Germany). 2.5. Extraction and cleanup The extraction system consists of a consecutive extraction of 0.5 g dry apple peel weighted in a BD Falcon™ conical tube (BD, Sunderland, United Kingdom) with MeOH:water (20/80, v/v) in a first step and 100% MeOH in a second step. Each extraction was performed by ultrasound-assisted solid–liquid extraction with 5 mL of the appropriate solvent by using a 2200 R-4 Ultrasonic sonicator (40 kHz, 100 W) (Branson Ultrasonic Corporation, Danbury, USA) for 60 min at room temperature. After adding the solvent to the extraction tube and after 30 min of extraction, the
Table 1 Basic parameters of the analytes including calibration range, equation, regression coefficient (r2), number of levels (N), instrumental limit of detection (IDL), instrumental quantification limit (IQL), limit of detection (LOD) and limit of quantification (LOQ). Phenolic compound
Calibration range (ng mL1)
r2
N
IDL(ng mL1)
IQL (ng mL1)
Working range (ng mL1)
LOD (ng g1)
LOQ (ng g1)
Salicylic acid Protocatechuic acid Gallic acid Propyl gallate p-Coumaric acid Caffeic acid Ferulic acid Sinapinic acid Chlorogenic acid Dihydrocaffeic acid Dihydroferulic acid 4-p-Hydroxyphenyl acetic acid Apigenin Apigetrin Luteolin Cynaroside Isorhamnetin Kaempferol Astragalin Quercetin Isoquercitrin Hyperin Rutin Avicularin Quercitrin Galangin Phloretin Phlorizin Naringenin Naringin (+)-Catechin ()-Epicatechin (+)-Aromadendrin (+)-Taxifolin Cyanidin chloride Kuromanin chloride Ideain chloride Keracyanin chloride Procyanidin B2
9–3577 12–3779 20–3796 0.5–3744 0.5–3775 0.5–3836 9–3800 9–3666 34–3891 0.5–3618 43–3926 36–3940
0.9999 0.9995 0,9998 0.9988 0.9999 0.9999 0.9996 0.9995 0.9997 0.9998 0.9997 0.9998
11 11 10 15 16 15 11 11 9 15 8 9
10 10 35 0.5 1 3 35 20 35 1 45 35
20 20 70 1 2 6 70 40 70 2 90 70
20–3577 20–3779 70–3796 1–3744 2–3775 6–3836 70–3800 40–3666 70–3891 2–3618 90–3926 70–3940
200 200 700 10 20 60 700 400 700 20 900 700
400 400 1400 20 40 120 1400 800 1400 40 1800 1400
3–3718 0.5–3385 0.5–3563 0.5–3749 0.5–3810 0.5–3814 0.5–3712 0.5–3906 0.5–3606 0.5–3711 0.5–3671 0.5–3463 0.5–3672 0.5–3810 0.5–3695 0.5–3925 0.5–3871 0.5–3840 0.5–3731 0.5–3660 0.5–3857 0.5–4073 37–4297 13–4499 4–4355 2–4272 2–5230
0.9995 0,9997 0.9995 0.9998 0.9995 0.9997 0,9993 0.9994 0.9991 0.9991 0.9995 0.9991 0.9992 0.9996 0.9991 0.9992 0.9993 0.9995 0.9992 0.9994 0,9992 0.9991 0.9997 0.9998 0.9991 0.9993 0.9993
12 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 16 16 15 15 9 11 12 14 14
4 1 1 2 2 3 1 2 1 1 1 1 1 2 2 1 1 1 2 2 3 1 40 10 10 5 5
8 2 2 4 4 6 2 4 2 2 2 2 2 4 4 2 2 2 4 4 6 2 80 20 20 10 10
8–3718 2–3385 2–3563 4–3749 4–3810 6–3814 2–3712 4–3906 2–3606 2–3711 2–3671 2–3463 2–3672 4–3810 4–3695 2–3925 2–3871 2–3840 4–3731 4–3660 6–3857 2–4073 80–4297 20–4499 20–4355 10–4272 10–5230
80 20 20 40 40 60 20 40 20 20 20 20 20 40 40 20 20 20 40 40 60 20 800 200 200 100 100
160 40 40 80 80 120 40 80 40 40 40 40 40 80 80 40 40 40 80 80 120 40 1600 400 400 200 200
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375
solutions were stirred with a IKA MS2 Minishaker (IKAÒWerke GmbH & Co. KG, Staufen, Germany) for 15 min. During sonication, the temperature was kept below 40 °C. Following extraction, a separation between the solid particles and the liquid phase was obtained by centrifuging at 3000 rpm using a Allegra™ Centrifuge (Beckman Coulter Inc., CA, USA). Subsequently, the supernatans was collected and stored in a capped vial at 4 °C. When the two consecutive extraction cycles were performed, the two supernatants were combined in a microtube (50/50, v/v) and centrifuged using a Galaxy 16DH ultracentrifuge (VWR, Leuven, Belgium). Finally, the obtained supernatant was diluted (dilution factor 1/5 (v/v)) in a microvial by adding MeOH:water (60/40, v/v) and stored at 4 °C prior to injection into the UHPLC-DAD/ESI-am-MS system. 2.6. Method validation The developed UHPLC-DAD/ESI-am-MS approach was validated using MS detection in negative ionisation mode, since this was the main detector used to identify and quantify the peak signals. The performance characteristics taken into account for the validation of the measurement method were curve fit, range, sensitivity (instrumental detection limit (IDL), instrumental quantification limit (IQL), method limits of detection and quantification (MDL and MQL, respectively), precision and trueness, as well as specificity. In the validation study, precision (repeatability, intermediate precision) was expressed by the relative standard deviation (RSD %) between the replicate measurements. As a measure for the trueness, percentage recovery was calculated using the formula REC % = (measured value/theoretical value) 100. The mass accuracy was calculated with the formula D (ppm) = [(theoretical mass measured mass)/theoretical mass] 10,00 000. 2.6.1. Calibration range Calibration curves of all 39 phenolics were measured in a concentration range 0.6–15 000 ng mL1 corresponding to the expected levels in matrix. The curves were built by plotting the areas for each concentration level versus the nominal concentration of each calibration standard. The best curve fit was obtained with a non-linear curve with a 1/x statistical weight. The acceptance criteria for the calibration curves were a correlation coefficient (r2) of 0.9990 or better and each back-calculated standard concentration being within 15% deviation from the nominal value. 2.6.2. Sensitivity Because the noise level in the Orbitrap mass spectrometer, especially at m/z values >200 is virtually absent due to high mass accuracy, a standard signal-to-noise approach to determine the instrumental detection limits (IDL) is not feasible. Therefore, the IDLs were defined as 3 times the standard deviation at the lowest concentration on the calibration curve that can be measured with a precision lower than e.g., 10% and an accuracy higher than e.g., 90%. They were obtained by the injection of the calibration solutions at the 8 lowest concentration levels of the calibration range (from 250 to 0.6 ng mL-1) in 5 replicates. 2.6.3. Trueness and precision To evaluate the trueness and the repeatability (intra-day precision) of the analytical method, triplicate sets of 0.5 g dry apple peel were analysed in fivefold. In each triplicate set, one sample was left blank, one was fortified with 0.35 mL of a 25000 ng mL1 standard solution corresponding to a concentration level of 875 ng mL1 in final extract, and one was fortified with 0.025 mL of the same standard solution, corresponding to a concentration level of 65 ng mL1 in final extract. Each set was matured over night. The level of extraction was based on the recovery of the spiked concentrations after blank correction. To obtain intermediate precision (inter-day
371
precision), this procedure was repeated once in a 14-day period (n = 2). 2.6.4. Specificity Retention time stability (intra-day, inter-day) and mass accuracy (two days) were determined by extracting the retention times and accurate masses from the raw files generated during the above-described trueness and precision tests. This was performed by using ToxID software (version 2.1.1, Thermo Fisher Scientific, Bremen, Germany). 3. Results and discussion 3.1. Compound selection The challenge was to develop a simple and reproducible analytical method with high selectivity that has the potential to be used as a generic screening tool for phenolic compounds. In order to achieve this goal, 39 phenolic compounds from 7 flavonoid, 1 proanthocyanidin and 4 phenolic acid subclasses were selected based on an extensive literature review (Hoffmann-Ribani, Huber, & Rodriguez-Amaya, 2009; Lata, Trampczynska, & Paczesna, 2009; Salta et al., 2010; Vieira et al., 2009; Wu et al., 2007). The selected phenolic compounds include both non-esterfied and esterfied phenolic acids as well as glycosylated and non-glycosylated flavonoids (Table 1). 3.2. Chromatographic separation In developing the UHPLC method, several chromatographic conditions were investigated to obtain the separation of the selected phenolic compounds. Although under the established conditions baseline separation was not obtained for all the compounds (e.g., ideain and kuromanin), an acceptable separation of structurally similar compounds, including several structural isomers (e.g., isoquercitrin and hyperin) and epimers ((+)-catechin and ()-epicatechin) was achieved. The general group elution order is shown Fig. 1. Due to their high water solubility, the free sugars and cyanidins eluted at the beginning of the gradient. Next, relatively small molecules with polar groups that are not strongly retained on C18 eluted, i.e., hydroxybenzoic acids, hydroxycinnamic acids, hydroxyphenyl propionic acids. Subsequently, flavonoid glycosides eluted, as they contain many polar groups including sugars and other hydroxy groups, but they are rather large molecules. Finally, flavonoid aglycones eluted as non-polar compounds according to the number of hydroxy groups and their position. An exception to this were the non-glycosylated flavanols (catechines) and the procyanidins, which occured considerably earlier in the chromatogram. This is in agreement with previously published literature (Gómez-Romero et al., 2010; Nováková et al., 2010). 3.3. Mass spectrometric conditions Despite the superior chromatographic resolution of UHPLC compared to HPLC, co-elution of the analytes of interest with matrix compounds remains a common issue in multi-compound analysis in complex biological matrices. This problem can be solved by using an Orbitrap™ mass analyzer which makes it possible to measure at an extremely high mass accuracy (<1 ppm), which compensates the lack of chromatographic resolution by an increased mass spectrometric resolution. Furthermore, the high mass resolution of the Orbitrap instrumentation prevents inaccurate mass measurements caused by unresolved background matrix interferences. These advantages in relation to other MS technology permits the
372
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375
Fig. 1. Chromatogram of KanziÒ apple peel obtained by UPLC-DAD/ESI-am-MS in negative ion mode. The elution areas are indicated. Elution % gradient is also shown.
definition of a very narrow mass window, providing to possibility to cut off disturbing interferences which increases the selectivity of the measurement method. Furthermore, accurate mass measurements make it possible to limit the enormous number of possible molecular formulas corresponding with a particular molecular mass which is a first step in the identification of unknown compounds. However, selecting the highest possible mass accuracy and mass resolution of the instrument is not an option. Several compromises must be made to obtain a selective, sensitive and repeatable measurement method. A complementary set of scan settings has to be carefully defined, before proceeding to the method validation. Firstly, a compromise must be made between resolution and scan speed. High resolution mass scans (high selectivity) result in an increased scan time, decreasing the number of chromatographic data points over a peak. Insufficient chromatographic data points can dramatically affect the repeatability of the measurement method and can lead to increased LODs. In this study, a resolution of 50,000 (FWHM at m/z of 200) that corresponds to a scan speed of 2 Hz was selected. This value delivers sufficient mass separation to exclude mass interferences from matrix as well as sufficient data points per peak (minimum 15) permitting acceptable integration and quantification performance. Secondly, mass accuracy and dynamic range are correlating factors. In Orbitrap instrumentation, mass accuracy is not only established by mass calibration (external, internal), but also by the system parameters such as maximum injection time (trap filling time) and the automatic gain control (C-trap target capacity, AGC). These two parameters control the ion population in the Orbitrap. A high ion population in the Orbitrap mass analyser results in a high dynamic range (high sensitivity), but overloading of the trap
can result in mass shifts. Conversely, high mass accuracy (high selectivity) can lead to false negative findings. Both a C-trap target capacity of 106 (balanced) and a maximum injection time of 250 ms reflect the compromise that was made between sensitivity and selectivity. Finally, dynamic range is also affected by the scan range. Low scan starting ranges cause filling of the C-trap with ions originating from the mobile phase which reduce the dynamic range and consequently the sensitivity of the instrument. In function of the development of a generic screening method for phenolic compounds, a large scan range stretching from m/z 90 to 1800 was taken without loss of sensitivity. This complementary set of scan settings permits the application of an extraction mass window of 2 ppm for all selected phenolic compounds, without affecting the sensitivity or repeatability of the method, which was reflected in the method validation. 3.4. Method validation 3.4.1. Working range Phenolic compounds occur in a wide range of concentration levels in apple fruit. From this fact, a calibration range as large as possible was investigated during method validation. To avoid large deviations of the back-calculated standard concentration at the lowest concentration levels, the calibration was based on non-linear curves for all phenolic compounds. Basic parameters of the analytes including calibration range, regression coefficient (r2), and number of levels (N) are summarised in Table 1. Non-linear calibration curves with a minimum of 9 levels were plotted through the concentration range of three orders of magnitude obtaining an excellent fit to the non-linear model (r2 > 0.9990) for all analytes.
373
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375
3.4.2. Sensitivity High resolution MS technology permits to extract the so-called accurate mass (down to several decimals of a mass unit) from the ion chromatogram in a very narrow mass window which often devoid of any noise. Therefore, the calculation of the IDLs based on signal-to-noise delivers only arbitrary results. More realistic values of the IDLs are obtained when calculations are based on the relative standard deviation of the replicate measurement of the lowest concentration standard solution of the calibration curve that can be measured with a precision lower than e.g., 10% and an accuracy higher than e.g., 90%. The instrumental quantification limits (IQL) were defined as 2 times the corresponding IDLs. The limits of detection (LOD, method detection limits) were reassessed based on the sample intake (0.5 g apple peel). The limits of quantification (LOQ, method quantification limits) were defined as 2 times the LODs. The LODs and LOQs ranged from 200 to 900 ng g1 dw (dry weight) and from 400 to 1800 ng g1 dw, respectively. Except for p-coumaric acid and propyl gallate, the flavonoid LODs are about 10 times lower than those of the phenolic acids. This is consistent with the findings of Gómez-Romero et al. (2010). The values of the LOQs are comparable (Wen, Li, Di, Liao, & Liu, 2005) or lower (Chinnici, Gaiani, Natali, Riponi, & Galassi, 2003; Gómez-Romero et al.,
2010; Harnly et al., 2007; Nováková et al., 2010; Spácil et al., 2010; Tsao & Yang, 2003) than other quantification methods for a wide range of phenolic compounds. An overview of all LODs and LOQs is given in Table 1. 3.4.3. Trueness and precision Table 2 summarises the results of the trueness (average, n = 5), repeatability (intraday precision, n = 5) and intermediate precision (interday precision, n = 2). The trueness at five replicates ranged from 74% (cyanidin chloride) to 100% (kuromanin chloride) for the higher spike level and 72% (chlorogenic acid) and 99% (protocatechuic acid) for the lower spike level. The repeatability ranged from 1% to 10% and 1–11% for the high and low spike level respectively. The values of the trueness, repeatability and intermediate precision are comparable with those of other quantification methods for phenolic compounds in apple (Spácil et al., 2010), tea (Chinnici et al., 2003) and herbal medicines (Wen et al., 2005). 3.4.4. Specificity The intra- and interday precision and accuracy in the retention time, as well as the mass accuracy, were listed in Table 3. Consecutive injections (n = 5) within 1 day demonstrate exceptional intraday retention time repeatability and mass accuracy, ranging from
Table 2 Validation parameters trueness, repeatability (intra-day precision, n = 5) and intermediate precision (inter-day precision, n = 2). Concentration level: 875 ng mL1 Trueness REC %
a
a
Concentration level: 65 ng mL1
Repeatability (n = 5) RSD %
Intermediate precision (n = 2) RSD %
Truenessa REC %
Repeatability (n = 5) RSD %
Intermediate precision (n = 2) RSD %
Salicylic acid Protocatechuic acid Gallic acid Propyl gallate p-Coumaric acid Caffeic acid Ferulic acid Sinapinic acid Chlorogenic acid Dihydrocaffeic acid Dihydroferulic acid 4-p-Hydroxyphenyl acetic acid Apigenin Apigetrin Luteolin Cynaroside Isorhamnetin Kaempferol Astragalin Quercetin Isoquercitrin Hyperin Rutin Avicularin Quercitrin Galangin Phloretin Phlorizin Naringenin Naringin (+)-Catechin ()-Epicatechin (+)-Aromadendrin (+)-Taxifolin Cyanidin chloride Kuromanin chloride Ideain chloride Keracyanin chloride Procyanidin B2
90 96 94 83 78 76 86 95 94 97 94 89 82 93 96 90 81 92 84 78 92 86 83 91 89 84 77 78 82 93 91 88 78 92 74 100 99 98 89
3 5 3 3 6 5 9 7 3 6 10 8 1 1 6 3 3 2 7 1 3 7 4 4 3 6 3 5 5 6 3 2 8 3 3 5 10 9 6
5.6 2.7 3.9 3.2 1.8 1.7 3.1 4.5 3.6 3.1 2.4 1.1 2.4 1.5 1.5 2 2.3 1.6 1.8 3.2 1.4 2 1.4 2.9 2 2.2 5 3.8 1.3 1.1 3.1 2.8 3.1 4 2.3 2 3.5 2.9 1.8
98 99 83 98 94 78 95 94 72 98 92 94 74 93 80 97 87 88 96 82 83 81 89 81 84 85 73 89 98 86 84 82 89 81 78 73 74 84 74
4 5 10 6 4 4 5 6 3 3 4 1 3 6 3 4 6 8 3 11 11 4 3 4 3 9 6 10 3 5 3 6 3 4 3 9 5 5 4
4.4 5.5 3.3 6.3 4.8 6.3 5.5 3.8 2.9 3 4.5 4.7 5.7 3.3 6 2.9 2.7 2.1 3.3 5.6 5.5 6.2 4.2 4.4 6.6 2 3.2 3.4 5.9 5.6 4.1 4.9 2.9 6.3 2.2 6.3 3.7 5.3 4.8
Min Max Average
74 100 88
1 10 5
1 6 3
72 99 86
1 11 5
2 7 4
average, n = 5.
374
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375
Table 3 Stability of retention time (RT) and mass accuracy (d, day 1 and 2) of the selected phenolic compounds. Phenolic compound
Retention time
Repeatability (n = 5) RSD %
Intermediate precision (n = 2) RSD %
14.48 7.05 4.79 13.76 12.67 10.34 12.85 12.66 8.3 8.65 11.28 8.61
0.1 0.3 0.4 0.1 0.2 0.2 0.2 0.2 0.3 0.6 0.2 0.3
18.02 13.69 16.57 12.72 18.15 18.36 13.75 16.73 13.07 13.01 12.58 13.68 13.97 20.26 17 13.68 17.12 13 9.33 10.16 14.79 13.27 8.52 8.53 8.3 8.73 10.2
Average RT (min) Salicylic acid Protocatechuic acid Gallic acid Propyl gallate p-Coumaric acid Caffeic acid Ferulic acid Sinapinic acid Chlorogenic acid Dihydrocaffeic acid Dihydroferulic acid 4-p-hydroxyphenyl acetic acid Apigenin Apigetrin Luteolin Cynaroside Isorhamnetin Kaempferol Astragalin Quercetin Isoquercitrin Hyperin Rutin Avicularin Quercitrin Galangin Phloretin Phlorizin Naringenin Naringin (+)-Catechin ()-Epicatechin (+)-Aromadendrin (+)-Taxifolin Cyanidin chloridea Kuromanin chloridea Ideain chloridea Keracyanin chloridea Procyanidin B2 Min Max Average a
Mass accuracy
Day 1 (n = 5) d (ppm)
Day 2 (n = 5) d (ppm)
Elemental composition
Exact mass (amu)
1.3 1.9 0.1 1 0.4 1.6 0.2 1.2 1.5 1.5 0.4 0.3
C7H6O3 C7H6O4 C7H6O5 C10H12O5 C9H8O3 C9H8O4 C10H10O4 C11H12O5 C16H18O9 C9H6O4 C10H12O4 C8H8O3
137.02 153.02 169.01 211.06 163.04 179.03 193.05 223.06 353.09 177.02 195.07 151.04
0.8 0.1 0.4 0.3 0.6 0.6 1.3 1 0.1 0.2 1 1.8
0.9 0 0.5 0.4 0.7 0.7 1.4 1.1 0 0.3 1.1 1.7
0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.4 0.2 0.1 0.2 0.3 0.3 0.4 0.2 0.2
0.2 0.6 0.1 0.3 0.1 0.4 0.1 0.1 0.1 0.3 0.5 0.5 0.3 0.1 0.2 0.7 0.2 0.2 0.9 1 0.3 0.6 0.2 1.1 0.1 2.1 0.6
C15H10O5 C21H20O10 C15H10O6 C21H20O11 C16H12O7 C15H10O6 C21H20O11 C15H10O7 C21H20O12 C21H20O12 C27H30O16 C20H18O11 C21H20O11 C15H10O5 C15H14O5 C21H24O10 C15H12O5 C27H32O14 C15H14O6 C15H14O6 C15H12O6 C15H12O7 C15H12O7 C21H22O12 C21H22O13 C27H32O16 C30H26O12
269.05 431.1 285.04 447.09 315.05 285.04 447.09 301.04 463.09 463.09 609.15 433.08 447.09 269.05 273.08 435.13 271.06 579.17 289.07 289.07 287.06 303.05 303.05 465.1 481.1 611.16 577.14
0.8 0.6 0.8 0.3 0.5 0.6 0.2 0.9 0.3 0.3 0.4 0.9 0.2 0.5 0.3 0.8 0.6 0.8 0.1 0.2 0.5 0.8 0.3 0.1 0.1 0.3 0.1
0.6 0.2 0.2 0.1 0.1 0.4 0.2 0.3 0.5 0.5 0.6 0.1 0.2 0.6 0.1 0.3 0.3 0.2 0.2 0.2 0.3 0.2 0.4 0.2 0.2 0.1 0.1
0.1 0.6 0.2
0.4 2.1 0.6
0.1 1.8 0.5
0 1.7 0.4
Anthocyanidins are measured in the carbinol pseudobase form.
0.1% to 0.6% and 0.3% to 3.3%, respectively. In addition, the intermediate precision on retention time was less than 2% for each of the standard compounds. The mass errors of each of the standard compounds was less than 2 ppm. The smallest average mass error was 0.1 ± 0.3 ppm for chlorogenic acid and the largest average mass error was 1.8 ± 0.7 for 4-p-hydroxyphenyl acetic acid. These mass errors are generally 5 times lower than those obtained with high mass accuracy time-of-flight systems used for the quantification of phenolic compounds in biological extracts (Verardo et al., 2010). The data clearly demonstrate that the instrumentation used in this study is routinely capable of sub-ppm mass accuracy measurements.
4. Conclusion A novel UHPLC-DAD/ESI-am-MS method for the quantification of 39 phenolic compounds in extracts of apple peel and flesh was developed and validated. The developed method was validated in
terms of model deviation (r2 > 0.9990 for all 39 compounds), IDL, IQL, LOD, LOQ, trueness (>72%), repeatability and intermediate precision (<10% and <11% respectively). The mass accuracy of each selected phenolic compound was below 2 ppm. The described UHPLC-DAD/ESI-am-MS method represents a valuable tool and a good alternative for simultaneous characterisation of phenolic components in biological extracts. Knowledge of the phenolic content in apples, including phenolic acids, flavonoids and their derivatives, will help to define their potential as a dietary source of phenolic antioxidants. Acknowledgements The authors thanks Tom Deckers from the Research Station for Fruit Culture (PCfruit, Velm) for delivery of the KanziÒ apple fruit. References Abad-García, B., Berrueta, L. A., Garmon-Lobato, S., Gallo, B., & Vicente, F. (2009). A general analytical strategy for the characterization of phenolic compounds in
D. De Paepe et al. / Food Chemistry 136 (2013) 368–375 fruit juices by high-performance liquid chromatography with diode array detection coupled to electrospray ionization and triple quadrupole mass spectrometry. Journal of Chromatography A, 1216, 5398–5415. Aldini, G., Regazzoni, L., Pedretti, A., Carini, M., Cho, S. M., Park, K. M., et al. (2011). An integrated high resolution mass spectrometric and informatics approach for the rapid identification of phenolics in plant extract. Journal of Chromatography A, 1218, 2856–2864. Chinnici, F., Gaiani, A., Natali, N., Riponi, C., & Galassi, S. (2003). Improved HPLC determination of phenolic compounds in Cv. Golden Delicious apples using a monolithic column. Journal of Agricultural and Food Chemistry, 52, 3–7. Gómez-Romero, M., Segura-Carretero, A., & Fernández-Gutiérrez, A. (2010). Metabolite profiling and quantification of phenolic compounds in methanol extracts of tomato fruit. Phytochemistry, 71, 1848–1864. Guillarme, D., Nguyen, D. T. T., Rudaz, S., & Veuthey, J. L. (2007). Recent developments in liquid chromatography–Impact on qualitative and quantitative performance. Journal of Chromatography A, 1149, 20–29. Harnly, J., Bhagwat, S., & Lin, L. Z. (2007). Profiling methods for the determination of phenolic compounds in foods and dietary supplements. Analytical and Bioanalytical Chemistry, 389, 47–61. Hoffmann-Ribani, R., Huber, L. S., & Rodriguez-Amaya, D. B. (2009). Flavonols in fresh and processed Brazilian fruits. Journal of Food Composition and Analysis, 22, 263–268. Ignat, I., Volf, I., & Popa, V. I. (2011). A critical review of methods for characterisation of polyphenolic compounds in fruits and vegetables. Food Chemistry, 126, 1821–1835. Kartsova, L., & Alekseeva, A. (2008). Effect of milk caseins on the concentration of polyphenolic compounds in tea. Journal of Analytical Chemistry, 63, 1107–1111. Lata, B., Trampczynska, A., & Paczesna, J. (2009). Cultivar variation in apple peel and whole fruit phenolic composition. Scientia Horticulturae, 121, 176–181. Lin, L. Z., & Harnly, J. M. (2007). A screening method for the identification of glycosylated flavonoids and other phenolic compounds using a standard analytical approach for all plant materials. Journal of Agricultural and Food Chemistry, 55, 1084–1096. Magiera, S., Baranowska, I., & Kusa, J. (2012). Development and validation of UHPLCESI–MS/MS method for the determination of selected cardiovascular drugs, polyphenols and their metabolites in human urine. Talanta, 89, 47–56. Makarov, A. (2000). Electrostatic axially harmonic orbital trapping: a highperformance technique of mass analysis. Analytical Chemistry, 72, 1156–1162. Moulard, Y., Bailly-Chouriberry, L., Boyer, S., Garcia, P., Popot, M. A., & Bonnaire, Y. (2011). Use of benchtopexactive high resolution and high mass accuracy orbitrap mass spectrometer for screening in horse doping control. Analytica Chimica Acta, 700, 126–136.
375
Nováková, L., Spácil, Z., Seifrtová, M., Opletal, L., & Solich, P. (2010). Rapid qualitative and quantitative ultra high performance liquid chromatography method for simultaneous analysis of twenty nine common phenolic compounds of various structures. Talanta, 80, 1970–1979. Salta, J., Martins, A., Santos, R. G., Neng, N. R., Nogueira, J. M. F., Justino, J., et al. (2010). Phenolic composition and antioxidant activity of Rocha pear and other pear cultivars – A comparative study. Journal of Functional Foods, 2, 153–157. Sato, Y., Itagaki, S., Kurokawa, T., Ogura, J., Kobayashi, M., Hirano, T., et al. (2011). In vitro and in vivo antioxidant properties of chlorogenic acid and caffeic acid. International Journal of Pharmaceutics, 403, 136–138. Spácil, Z., Nováková, L., & Olich, P. (2008). Analysis of phenolic compounds by high performance liquid chromatography and ultra performance liquid chromatography. Talanta, 76, 189–199. Spácil, Z., Nováková, L., & Solich, P. (2010). Comparison of positive and negative ion detection of tea catechins using tandem mass spectrometry and ultra high performance liquid chromatography. Food Chemistry, 123, 535–541. Tsao, R., & Yang, R. (2003). Optimization of a new mobile phase to know the complex and real polyphenolic composition: Towards a total phenolic index using high-performance liquid chromatography. Journal of Chromatography A, 1018, 29–40. Vallverdú -Queralt, A., Jáuregui, O., Medina-Remón, A., Andrés-Lacueva, C., & Lamuela-Raventós, R. M. (2010). Improved characterization of tomato polyphenols using liquid chromatography/electrospray ionization linear ion trap quadrupole orbitrap mass spectrometry and liquid chromatography/ electrospray ionization tandem mass spectrometry. Rapid Communications in Mass Spectrometry, 24, 2986–2992. Verardo, V., Arráez-Román, D., Segura-Carretero, A., Marconi, E., FernándezGutiérrez, A., & Caboni, M. F. (2010). Identification of buckwheat phenolic compounds by reverse phase high performance liquid chromatographyelectrospray ionization-time of flight-mass spectrometry (RP-HPLC-ESI-TOFMS). Journal of Cereal Science, 52, 170–176. Vieira, F. G. K., Borges, G. D. S. C., Copetti, C., Amboni, R. D. D. M. C., Denardi, F., & Fett, R. (2009). Physico-chemical and antioxidant properties of six apple cultivars (Malus domestica Borkh) grown in southern Brazil. Scientia Horticulturae, 122, 421–425. Wen, D., Li, C., Di, H., Liao, Y., & Liu, H. (2005). A universal HPLC method for the determination of phenolic acids in compound herbal medicines. Journal of Agricultural and Food Chemistry, 53, 6624–6629. Wu, J., Gao, H., Zhao, L., Liao, X., Chen, F., Wang, Z., et al. (2007). Chemical compositional characterization of some apple cultivars. Food Chemistry, 103, 88–93.