A comprehensive highresolution mass ... - Wiley Online Library

Research Article Received: 10 April 2014

Revised: 4 June 2014

Accepted: 5 June 2014

Published online in Wiley Online Library

Rapid Commun. Mass Spectrom. 2014, 28, 1779–1791 (wileyonlinelibrary.com) DOI: 10.1002/rcm.6960

A comprehensive high-resolution mass spectrometry approach for characterization of metabolites by combination of ambient ionization, chromatography and imaging methods Arton Berisha1, Sebastian Dold1, Sabine Guenther1, Nicolas Desbenoit1†, Zoltan Takats1,2, Bernhard Spengler1 and Andreas Römpp1* 1

Institute of Inorganic and Analytical Chemistry, Justus Liebig University Giessen, Schubertstrasse 60, 35392 Giessen, Germany Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, London SW7 2AZ, UK

2

RATIONALE: An ideal method for bioanalytical applications would deliver spatially resolved quantitative information in real time and without sample preparation. In reality these requirements can typically not be met by a single analytical technique. Therefore, we combine different mass spectrometry approaches: chromatographic separation, ambient ionization and imaging techniques, in order to obtain comprehensive information about metabolites in complex biological samples. METHODS: Samples were analyzed by laser desorption followed by electrospray ionization (LD-ESI) as an ambient ionization technique, by matrix-assisted laser desorption/ionization (MALDI) mass spectrometry imaging for spatial distribution analysis and by high-performance liquid chromatography/electrospray ionization mass spectrometry (HPLC/ESI-MS) for quantitation and validation of compound identification. All MS data were acquired with high mass resolution and accurate mass (using orbital trapping and ion cyclotron resonance mass spectrometers). Grape berries were analyzed and evaluated in detail, whereas wheat seeds and mouse brain tissue were analyzed in proof-of-concept experiments. RESULTS: In situ measurements by LD-ESI without any sample preparation allowed for fast screening of plant metabolites on the grape surface. MALDI imaging of grape cross sections at 20 μm pixel size revealed the detailed distribution of metabolites which were in accordance with their biological function. HPLC/ESI-MS was used to quantify 13 anthocyanin species as well as to separate and identify isomeric compounds. A total of 41 metabolites (amino acids, carbohydrates, anthocyanins) were identified with all three approaches. Mass accuracy for all MS measurements was better than 2 ppm (root mean square error). CONCLUSIONS: The combined approach provides fast screening capabilities, spatial distribution information and the possibility to quantify metabolites. Accurate mass measurements proved to be critical in order to reliably combine data from different MS techniques. Initial results on the mycotoxin deoxynivalenol (DON) in wheat seed and phospholipids in mouse brain as a model for mammalian tissue indicate a broad applicability of the presented workflow. Copyright © 2014 John Wiley & Sons, Ltd.

Biological samples consist of several thousand compounds that belong to different compounds classes and span a wide range of concentrations. An ideal method for bioanalytical applications would deliver spatially resolved quantitative information in real time and without sample preparation. In reality these requirements can typically not be met by a single analytical technique.

* Correspondence to: A. Römpp, Institute of Inorganic and Analytical Chemistry, Justus Liebig University Giessen, Schubertstrasse 60, 35392 Giessen, Germany. E-mail: [email protected] †

Rapid Commun. Mass Spectrom. 2014, 28, 1779–1791

Copyright © 2014 John Wiley & Sons, Ltd.

1779

Present address: Centre de Recherche Public, Gabriel Lippmann, 41 rue du Brill, L-4422 Belvaux, Luxembourg

A frequently used method for analysis of complex samples is liquid chromatography (LC) coupled to mass spectrometry (MS) by electrospray ionization (ESI). The chromatographic separation reduces the complexity of the sample and results in reduction of ion suppression effects. Retention times can be used for identification if reference compounds are available. Fragmentation experiments (MS/MS) can provide information about the molecular structure of an analyte. LC/MS/MS is now the primary analytical tool for many research areas including metabolomics,[1,2] proteomics,[3–5] and environmental chemistry.[6,7] It is also a powerful analytical tool for sensitive analysis of plant metabolites.[8–10] Significant progress has been made in speeding up the analysis and thus increasing sample throughput with the advent of ultrahigh-performance liquid chromatography (UPLC).[11] However, sample preparation is still the time-limiting factor in many applications.

A. Berisha et al.

1780

These limitations are overcome when ambient ionization techniques are used that allow fast and direct analysis with no (or minimal) sample preparation. The field has attracted increased interest ever since the invention of desorption electrospray ionization (DESI) in 2004.[12] Nowadays numerous ambient ionization techniques are available which also include direct analysis in real time (DART),[13] plasma-based ambient ionization methods (such as low temperature plasma, LTP),[14] rapid evaporative ionization mass spectrometry (REIMS)[15] and laser desorption ionization (LDI).[16] Application fields for these methods are as numerous and diverse as the methods themselves. Ambient ionization methods have been used for the rapid analysis of agrochemicals,[17,18] explosives,[14] drug compounds,[19] crude oil,[20] tissue,[21,22] and in clinical applications.[15,16] In this work we utilize a laser-based ambient ionization method. The potential of laser desorption ionization (LDI)[23] for analysis of organic substances was recognized early on,[24] but the adaptation to ambient conditions has broadened its application most recently.[16] However, LDI generally exhibits low ionization efficiencies (typically around 1%), especially for biological samples. Therefore, modifications of the method such as desorption from activated surfaces (desorption ionization on silicon, DIOS)[25,26] or post-ionization of the desorbed molecules by a second ionization method have been applied to increase the ion yield in laser-based ionization methods. Electrospray ionization is the most commonly used post-ionization method,[27] but laser ionization[28] and corona discharge ionization[29] have also shown to improve ion signal intensity. Laser desorption followed by electrospray post-ionization has been performed with ultraviolet (UV) lasers (electrospray-assisted laser desorption ionization, ELDI[30]) and infrared (IR) lasers (laser ablation electrospray ionization, LAESI[31]) in the past. Applications of these methods were the direct analysis of drugs, peptides, and proteins.[30,32] Additionally, it has been demonstrated to be well suited for primary and secondary plant metabolites.[33,34] In contrast to ELDI and LAESI, desorption and post-ionization in our setup are spatially separated by an ion transfer system, hence it is called laser desorptionelectrospray ionization (LD-ESI). In this setup the sample does not have to be placed directly in front of the mass spectrometer inlet, which provides more flexibility in analyzing (larger) samples. The employed ion source is a combination of two previously published systems. The post-ionization and the ion transfer system were described in a study on surgical aersols.[27] In analogy to a previous study on LDI,[16] a UV laser (355 nm) was used for desorption in our experiments. In many cases the spatial distribution of an analyte can be linked to functional processes. This information is however lost in homogenized bulk samples and punctual measurements which are typically used in chromatographybased and in situ techniques, respectively. For this reason mass spectrometry imaging, that provides information about the spatial distribution of an analyte, has gained significant attention in recent years.[35,36] Matrix-assisted laser desorption/ionization (MALDI) mass spectrometry imaging (MSI) was introduced 20 years ago[37] and has been applied extensively for the analysis of tissue in the following years.[38,39] Major application areas nowadays include

wileyonlinelibrary.com/journal/rcm

biological applications,[40] pharmaceutical studies,[41] and clinical research.[42] MSI studies of plants have been reported for wheat,[43,44] Arabidopsis thaliana,[45,46] and Oryza sativa rice.[47] MALDI imaging of secondary plant metabolites, such as anthocyanins, was reported for black rice[48] and rabbiteye blueberry,[49] where samples were measured at a pixel size of 100 μm. Recently, a MSI method was introduced that combines high mass accuracy and high spatial resolution at the cellular level (5 μm pixel size)[50] for analysis of drug compounds,[51] neuropeptides,[52] tryptic peptides,[53] and single HeLa cells,[54] thus providing in-depth information about biological samples.[36] Impressive progress has been made in recent years in each of the three discussed mass spectrometric methods. While LC/MS/MS is one of the most widely used mass spectrometric techniques, ambient ionization techniques and imaging applications are currently among the most active areas in terms of instrumental and methodological development. Each of these methods has significant advantages, but in most studies only one of these three methods was applied. In this work we present an approach to combine in situ/ambient, chromatography-based and MSI approaches in order to obtain comprehensive information about complex biological samples. For an initial screening we employed LD-ESI as an in situ technique followed by a compound distribution analysis using MALDI imaging. In a final step, high-performance liquid chromatography coupled to electrospray ionization mass spectrometry (HPLC/ESI-MS) was used for quantification and confirmation of identified compounds. The advantages and drawbacks of each method are presented in detail for the example of plant metabolites. A grape sample was used as an example because it contains a high number of metabolites. A special focus is placed on anthocyanins as a representative for a complex compound class which is present over a wide concentration range. These flavonoid compounds are also known for their health-promoting effects due to anti-oxidative properties[55] and are therefore of interest for health-care and nutritional science. Another example on plant samples demonstrates the direct detection of mycotoxins in a fungus-infected wheat seed and confirmation by LC/MS. The combination of all three methods is also demonstrated for mammalian tissue (mouse brain) in order to show the general applicability of the presented approach. High mass resolution and high mass accuracy measurements and MS/MS experiments were used throughout all methods and experiments allowing reliable interpretation and combination of the different approaches.

EXPERIMENTAL Samples The grape variety Accent[56] was provided by the Geisenheim Research Center, Department for Wine Analysis and Beverage Research, Geisenheim, Germany. This type is a cross between Kolor (Blauer Spätburgunder x Teinturier) and Chancellor (Seibel 7053). Grape samples were stored at 22°C until measurement.



Comprehensive MS approach for complex samples LD-ESI MS

HPLC/ESI-MS

For the measurements with the LD-ESI source samples were brought to room temperature and no further sample preparation was needed. A pulsed (20 Hz) Brilliant Nd:YAG laser (Quantel, Les Ulis Cedex, France) at 355 nm wavelength (frequency tripled) was used with an energy of 10.5 mJ. The laser beam was directed to the sample surface via an optical mirror and a lens (focal length of f = 10 cm). The sample was placed 12.5 ± 0.3 cm below the lens. Laserdesorbed material was transferred to the mass spectrometer via a 0.75 m long (1/8" o.d., 1/16" i.d.) PTFE tubing (Fluidflon PTFE tubing; LIQUID-scan GmbH Co. KG, Überlingen, Germany) and a venturi pump (VAC-100, Veriflo, UK) in analogy to previously described instrumentation.[16,27] For electrospray post-ionization of desorbed material a capillary was integrated into the venturi pump as described in a previous study.[27] A solvent mixture of methanol/water/formic acid (50:50:0.1, v/v/v) was provided at a flow rate of 5 μL/min by the built-in syringe pump of the mass spectrometer. The spray voltage was set at 3.6 kV. Measurements were performed in positive ion mode on a LTQ FT Ultra mass spectrometer (Thermo Fisher Scientific GmbH, Bremen, Germany) using automatic gain control with a target value of 1E6 and mass resolution was set to 100 000 at m/z 400.

Grape peel extracts of Accent grapes were obtained using an extraction method described in the literature.[9,60] HPLC separation of grape peel extracts was carried out using a gradient of 0.7% formic acid in water (Solvent A) and methanol (Solvent B) at a flow rate of 400 μL/min. Gradient conditions were: 0 min, 2% B; 1 min, 7% B; 5 min 15% B; 10 min 40% B; 13 min 70% B; 16 min 70% B and 18 min, 2% B. The HPLC system consisted out of a Kinetex C18 (100 mm × 2.1 mm, 2.6 μm packing; Phenomenex, Aschaffenburg, Germany) and an UltiMate 3000 Rapid Separation LC system with a HPG-3200RS pump and a WPS-3000 automatic sampler (Thermo Fisher Scientific GmbH, Bremen, Germany). Mass spectrometry measurements in positive ion mode were carried out on a LTQ FT Ultra mass spectrometer equipped with an IonMax electrospray source (both Thermo Fisher Scientific GmbH, Bremen, Germany) with resolving power 100 000 at m/z 400. Quantitation of all anthocyanins was based on the most abundant species (malvidinmonoglucoside), which is a common approach for the analysis of anthocyanins. Additional details on sample preparation and calibration can be found in section 4 of the Supporting Information.

MALDI MSI


Peaks in the mass spectrometric data derived from LD-ESI, MALDI imaging and HPLC/ESI were assigned to compounds using information from collaboration partners from the Geisenheim Research Center and literature.[8,56,61,62] This exact mass list contained 20 amino acids, several carbohydrates (both with sodium, potassium and ammonium adducts) and anthocyanins in various modifications. Measurements of the different approaches were compared to this list using a maximum mass tolerance of 2 ppm. Selected compounds were confirmed by MS/MS experiments.

RESULTS AND DISCUSSION In situ measurement The first step in the developed approach was the application of an in situ ionization method for a fast screening of sample composition. In this work, laser desorption-electrospray ionization (LD-ESI) was applied. The employed setup provides the possibility to post-ionize desorbed material by electrospray ionization in order to increase sensitivity. The schematic of the LD-ESI source and the measured grape sample are shown in Figs. 1(a)–1(c). The mass spectrum from the analysis of an Accent grape is shown in Fig. 1(d). The mass spectrum shows a broad range of compound classes that can be detected directly from the surface of the sample. This includes amino acids such as glutamine and arginine as well as carbohydrates such as mono- and di-hexoses. Amino acids were detected as protonated species, whereas the carbohydrates occurred as sodium or potassium adducts. In addition, a variety of anthocyanins were detected with different aglycones such as cyanidin, peonidin, delphinidin, petunidin and malvidin. These were detected as naturally occurring flavylium cations [M]+ (see section 1 of the Supporting Information for general structure) and were also



1781

For mass spectrometry imaging (MSI), 60 μm thick grape sections were cut using a cryotome (HM 525 cryostat, Thermo Fisher Scientific GmbH, Dreieich, Germany) at 26 °C. The sections were mounted on glass slides and stored in a desiccator for 1 h to avoid condensation of humidity on the sample surface. An Olympus BX-40 (Olympus Europa GmbH, Hamburg, Germany) microscope was used to capture optical images of the sections before sample preparation. A solution of 500 μL 2,5-dihydroxybenzoic acid (DHB; 98% purity, Sigma Aldrich, Germany), 30 mg/mL in acetone/ water/trifluoroacetic acid (50:50:0.1 %, v/v/v), was applied with a home-built pneumatic sprayer.[57] Experiments were performed using a AP-SMALDI10 high-resolution MALDI imaging ion source (TransMIT GmbH, Giessen, Germany), which was operated at atmospheric pressure and coupled to an Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific GmbH, Bremen, Germany).[50,58] The ion source includes a nitrogen laser (λ = 337 nm) operating at a repetition rate of 60 Hz. The laser beam was focused by a centrally bored objective lens to an optical diameter of 8.4 μm (1/e2 definition).[58] DHB signals were used as lock masses (m/z 273.0394, 444.0925 and 716.1246). MS data was acquired in positive ion mode. The mass resolution of the measurement was set to 100 000 at m/z 200. On-tissue MS/MS measurements were carried out with the AP-SMALDI10 ion source attached to a LTQ Orbitrap Discovery mass spectrometer (Thermo Fisher Scientific GmbH, Bremen, Germany). Mass spectrometry images were generated using the in-house developed software package MIRION.[59] Ion images of selected mass-to-charge values were generated with a bin width of Δm/z = ±0.005.

Compound identification

A. Berisha et al.

Figure 1. (a) Schematics of the LD-ESI measurement setup. (b, c) Optical images of Accent grape sample before and after measurement (15 laser shots on the exocarp). (d) LD-ESI high-resolution mass spectrum of Accent grape. The compound classes or modification groups can be found in the heading. The mass deviation of an identified peak is given in ppm. Arg = arginine, Gln = glutamine, Hex = hexose, Cy = cyanidin, Mal = malvidin, Del = delphinidin, Peo = peonidin, and Pet = petunidin.

1782

present in different modifications such as mono-glucosides, acetylmonoglucosides, coumaroyl-monoglucosides, diglucosides, or coumaroyl-diglucosides. Thus compounds with a range of biological functions could be detected with high mass accuracy in this sample. Carbohydrates are an important energy source and metabolic intermediates. Anthocyanins are vacuolar pigments and protect plants from UV radiation thus preventing damages to the plant cells. They play a role due to their intense colors in the attraction of animals for pollination and seed dispersal.[63] High-resolution mass spectrometry is especially important for screening of complex samples as compounds with very slight mass differences can be detected. The mass difference of the two signals shown in the inset in Fig. 1(d) is only Δm/z = ±0.021. Low-resolution instruments such as ion traps would have led to misinterpretation as only one peak would have been detected. Here, the ion signal m/z 625.1767 was identified as peonidin-diglucoside. The ion signal m/z 625.1559 could correspond to the isomeric compounds petunidin-coumaroyl-monoglucoside or peonidincaffeoyl-monoglucoside. These two compounds cannot be


distinguished in in situ measurements by accurate mass alone, but additional approaches are required as discussed below. Nevertheless, the used LD-ESI method provides substantial chemical information about the sample in a very short time. One high-resolution mass spectrum was acquired in less than 2 s. A typical measurement was performed within 1 min per sample. Within this period a blank signal and 20 to 30 spectra of the sample were acquired. One application where LD-ESI can be used for fast screening purposes is the differentiation between various types of grapes or maturation states (monitor change of metabolites during maturation). An example for characterization of grape types is shown in section 2 of the Supporting Information. The grape types Accent, Dunkelfelder and Dakapo were differentiated based on their LD-ESI mass spectra using principal component analysis (PCA). Here it was possible to rapidly distinguish between the various types of grapes due to their metabolite profiles. This example demonstrates that LD-ESI provides reproducible results (n = 10 biological replicates were analyzed for each group) and that semiquantitative data was obtained which can be used for statistical analysis.



Comprehensive MS approach for complex samples




1783

Figure 2. (a) Optical image of a 60 μm thick Accent grape section. (b) Overlay of selected ion images; red: delphinidin-monoglucoside [M]+ = 465.103 ± 0.005 exocarp; green: phosphatidylcholine [PC(36:4) + K]+ = 820.525 ± 0.005 endosperm; blue: arginine [M + H]+ = 175.115 ± 0.005 mesocarp. (c) Overlay of selected ion images: red: peonidin-monoglucoside [M]+ = 463.124 ± 0.005 exocarp and pulp; green: triglyceride [TG(54:6) + Na]+ = 901.725 ± 0.005 endosperm; blue: di-hexose [M + Na]+ = 365.105 ± 0.005 mesocarp. MS images were generated within an area of 150 × 268 pixels with a pixel size of 20 μm. (d, e) Mass spectra acquired from individual 20 μm pixels of the exocarp and endosperm, respectively.

A. Berisha et al. MALDI imaging for distribution analysis

1784

Information about the spatial distribution of compounds within the sample were obtained by matrix-assisted laser desorption/ionization (MALDI) mass spectrometry imaging (MSI). A tissue section of Accent grape (optical image shown in Fig. 2(a)) was imaged with 20 μm pixel size (150 × 268 pixels, 3.0 × 5.4 mm). Figure 2(b) shows the spatial distribution of an anthocyanin (delphinidin-monoglucoside, red), a phospholipid (phosphatidylcholine PC(36:4), green) and an amino acid (arginine, blue) as determined by MALDI imaging. The overlay of selected ion images of a second anthocyanin (peonidinmonoglucoside, red), a triglyceride (TG(54:6), green) and a carbohydrate (di-hexose, blue) is shown in Fig. 2(c). All ion images were generated with a bin size of Δm/z = ±0.005. The imaged compounds show a distinct distribution within the grape section with high localization in exocarp, endocarp and endosperm, respectively. As expected, anthocyanins were detected predominantly in the exocarp. During ripening anthocyanins are known to accumulate in the berry skin due to light activation.[64] Peonidin-monoglucoside (Fig. 2(c), red) shows a distinctly different distribution with relatively high signal intensities in the mesocarp. As described in previous work,[36] MS images are not overlaid with optical images and do not include any spatial smoothing or pixel-wise normalization. Gray scale images of ion intensities for individual anthocyanin compounds are provided in section 3 of the Supporting Information. The distributions found in this approach are in accordance with a previous study on blueberry conducted at a lower spatial resolution of 100 μm pixel size.[49] The different localization can be explained by different activity of enzymes which are responsible for modification of anthocyanin compounds. Similar effects (different distribution patterns of anthocyanin species) were found in the pericarp of black rice seed.[48] A large number of additional compounds were detected as demonstrated by the mass spectra shown in Figs. 2(d) and 2 (e), which correspond to individual 20 μm pixels acquired from the exocarp and of the endosperm (seed), respectively. For example, the amino acid glutamine and carbohydrate monohexose were detected in the exocarp. Anthocyanins were detected modified as monoglucosides, acetyl-monoglucosides, coumaroyl-monoglucosides, caffeoyl-monoglucosides, diglucosides and coumaroyl-diglucosides. In accordance with the LD-ESI MS measurement (Fig. 1) the inset in Fig. 2(d) shows two signals at m/z 625.1766 corresponding to peonidindiglucoside and a second peak at m/z 625.1560 which can be assigned to the isomeric compounds peonidin-caffeoylmonoglucoside or petunidin-coumaroyl-monoglucoside. Compounds in the mass spectrum originating from the endosperm (Fig. 2(e)) were predominantly lipids such as phosphatidylcholines detected as protonated species or potassium adducts and triglycerides detected as sodium and potassium adducts. Lipid compounds are important for energy storage and are therefore expected to be localized in the endosperm. These measurements demonstrate that MALDI imaging with 20 μm pixel size is possible for plant tissue, which is the highest spatial resolution that has been reported for MSI of plant so far. A range of compound classes can be detected. The obtained detailed distribution of plant metabolites is in accordance with their biological function. Mass spectra were


acquired with high mass resolution and mass accuracy which proved to be essential for reliable identification of imaged compounds. This is an improvement compared to previous studies of blueberry, which concluded that low mass resolution analysis is not sufficient to resolve the high complexity of these samples.[49] Sample preparation of plant material is more demanding than for typical mammalian tissue samples. Especially plant organs with high water content such as fruit bodies (berry in this study) are difficult to section. However, with some practice and by using slightly thicker sections (60 μm), these samples can be analyzed reproducibly. This is also demonstrated by MALDI imaging measurements of additional grape species (Dunkelfelder and Dakapo) which are shown in section 3 of the Supporting Information. These experiments confirmed the distinct distribution patterns of anthocyanins and lipids. HPLC/ESI-MS measurements The screening by LD-ESI and distribution analysis using MALDI imaging provided detailed information about the identity of a range of plant metabolites, but these measurements were not able to fully resolve the complexity of the sample and provided only semiquantitative information. Therefore we employed a chromatographybased mass spectrometry method in order to separate isomeric compounds and to obtain absolute quantitation. Table 1 shows the quantification results for 13 anthocyanin species in an extract of the grape exocarp (peel). Concentrations were in the range of several nmol per mg (dry weight). The anthocyanin monoglucosides were determined to have the highest concentration. Delphinidin-monoglucoside with a concentration of 4.3 (±1%) nmol/mg was the most abundant anthocyanin species. The concentrations of other anthocyanins modified as acetyl-monoglucosides, coumaroyl-monoglucosides and di-glucosides were lower. Table 1. Quantification of anthocyanin compounds in Accent grape exocarp extract using HPLC/ESI-MS. Concentrations are given in nmol per mg peel (dry weight). Quantitation of all anthocyanins was based on calibration with malvidin-monoglucoside Retention Concentration* time [min] [nmol/mg]

Anthocyanin Delphinidin-diglucoside Petunidin-diglucoside Peonidin-diglucoside Malvidin-diglucoside Delphinidin-monoglucoside Petunidin-monoglucoside Peonidin-monoglucoside Malvidin-monoglucoside Petunidin-acetylglucoside Peonidin-acetylglucoside Malvidin-acetylglucoside Peonidin-coumaroylglucoside Malvidin-coumaroylglucoside

6.6 7.7 8.2 8.4 8.5 9.4 9.8 10.0 11.5 11.9 11.9 12.7 12.7

0.3 (±5%) 0.6 (±3%) 0.7 (±1%) 1.1 (±3%) 4.3 (±1%) 3.1 (±1%) 2.1 (±2%) 3.3 (±1%) 0.9 (±6%) 0.4 (±13%) 0.8 (±7%) 0.3 (±5%) 0.3 (±6%)

*Average values of three technical replicates. The standard deviation of these three measurements is given in parentheses.



Comprehensive MS approach for complex samples LD-ESI and MALDI-MS measurements of the Accent grape showed already that there are more than one compound with nominal mass m/z 625, which could not be unambiguously identified. The selected ion chromatogram of the analyzed grape extract revealed three compounds in the mass range m/z 625.14–625.19 (Fig. 3(a)). Chromatographic peak I (m/z 625.176) corresponds to peonidindiglucoside which was also detected in MALDI and LD-ESI. The identification was confirmed by MS/MS measurements which showed the neutral loss of two hexose residues (Δm/z 162). The resulting fragment ion at m/z 301 corresponds to the peonidin aglycon (Fig. 3(b)). Chromatographic separation reveals that two compounds (peaks II and III) with m/z 625.155 are present in the sample. This mass peak was already detected by MALDI and LD-ESI, but could not be assigned unambiguously. The MS/MS experiment of the precursor ion m/z 625.2 of peak II (RT 12.16 min) results in a fragment ion at m/z 301 which corresponds to the peonidin aglycon (Fig. 3(b)). The neutral loss of Δm/z 324 results from loss of a hexose and caffeic acid. Consequently, peak II was assigned to peonidin-caffeoyl-monoglucoside. The second signal at m/z 625.155, peak III, can be assigned to petunidin-coumaroyl-monoglucoside. The fragment ion at m/z 317 in the MS/MS spectrum (Fig. 3(d)) corresponds to the petunidin aglycone which results from neutral loss of a

hexose and coumaric acid (Δm/z 308). Similar results were obtained for additional anthocyanin compounds. For example, three compounds with nominal mass 611 were separated by HPLC and identified by MS/MS experiments (data shown in section 4 of the Supporting Information). In addition to separating analytes of very similar mass, HPLC/ESI-MS measurements also revealed compounds which were not detected in LD-ESI or MALDI such as pelargonidin-coumaroyl-monoglucoside (m/z 579.1497). The concentration of this anthocyanin can only be estimated as the signal intensity was below the range of the calibration series. The detected signal was 360-fold lower than that of the most abundant anthocyanin delphinidin-monoglucoside. This higher sensitivity of HPLC/ESI-MS/MS is a result of the reduced ion suppression due to chromatographic separation of a complex biological mixture. A critical step in the HPLC/ESI-MS approach is sample preparation. Typically, different extraction protocols have to be used for different compound classes. Here, an extraction protocol was used that was optimized for anthocyanins and flavonols, and discussion is thus focused on these compounds. In contrast, amino acids, carbohydrates and anthocyanins were detected by the LD-ESI and MALDI approach in a single MS measurement. (Lipids were observed in the MALDI measurements in the same MS measurement, see Fig. 2(e).)




1785

Figure 3. HPLC/ESI-MS and -MS/MS measurements of Accent grape extract. (a) Extracted ion chromatogram (XIC) of mass range m/z 625.14–625.19. The peaks are assigned due to their MS/MS spectra (ion trap data) to peonidin-diglucoside (b), peonidin-caffeoyl-monoglucoside (c), and petunidin-coumaroyl-monoglucoside (d). All MS/MS spectra were carried out using collision-induced dissociation (’normalized collision energy’ was set to 20). * The fragment ion m/z 317 in (c) belongs to the petunidin-aglycone and results from the chromatographic shoulder of the petunidin-coumaroyl-monoglucoside peak.

A. Berisha et al. Comparison of the three mass spectrometric approaches The results of the three different mass spectrometric approaches are compared in the following. Mass spectra of Accent grape exocarp acquired by LD-ESI, MALDI and HPLC/ESI-MS, respectively, are shown in Figs. 4(a)–4(c). MALDI and LD-ESI measurements were performed directly on tissue. All mass spectra contain various compound classes such as amino acids, carbohydrates and anthocyanins. High mass resolution and accurate mass measurements for all three approaches provide for reliable compound identification. The maximum mass tolerance for compounds identification was 2 ppm. The overlap of different methods is confirmed by MS/MS measurements for the example of precursor ion m/z 655.19 shown in Figs. 4(d)–4(f). The fragment ion signals confirm the assignment to a diglucoside. All MS/MS spectra contain the fragment ions m/z 493 after neutral loss of one hexose residue and m/z 331 after neutral loss of a second hexose residue. The detected fragment m/z 331 is the characteristic ion signal of a malvidin aglycone. Consequently, the assignment of the ion signal m/z 655.19 to malvidin-

diglucoside can be confirmed for all three ionization methods. The use of accurate mass significantly facilitates the comparison between the different MS approaches and in addition MS/MS measurements can be used for confirmation. A Venn diagram (Fig. 5(a)) was used to illustrate the number of identified compounds detected by the different mass spectrometry approaches (the list of identified compounds is shown in section 5 of the Supporting Information). Interestingly, most of the identified compounds were detected throughout all three mass spectrometric approaches, in total 41 compounds. There is a slightly higher overlap between HPLC/ESI and LD-ESI (rather than between MALDI and HPLC/ ESI or MALDI and LD-ESI). This is most likely due to the use of electrospray both for LD-ESI and HPLC/ESI. The elevated number of detected compounds with the HPLC/ESI approach is due to the higher sensitivity of this method, as discussed above. Figs. 5(b)–5(d) show an example for the differentiation of compounds with small differences in m/z using accurate mass and highresolution measurements.

1786

Figure 4. Comparison of the different MS approaches: (a) LD-ESI, (b) MALDI imaging spectrum from exocarp, and (c) HPLC/ESI (summed spectrum from retention time 0.5–14.0 min). All spectra show metabolites of the measured grape sample Accent with a mass deviation below 2 ppm. High-resolution MS/MS experiments were performed for malvidin-diglucoside (precursor mass m/z 655.19) by LD-ESI (d), MALDI (e), and HPLC/ESI (f). For LD-ESI and MALDI (imaging) the MS/MS measurements were performed on tissue.




Comprehensive MS approach for complex samples

Figure 5. (a) Venn diagram comparing the identified compounds in grape exocarp. Most of the compounds were identified with all three methods within a mass deviation of maximal 2 ppm. In this manner, compounds with small differences in m/z could be differentiated using accurate mass and high resolution. (b–d) Zoomed-in mass spectra showing cyanidin-diglucoside and the isomeric compounds delphinidincoumaroylglucoside and cyanidin-caffeoylglucoside detected by LD-ESI, MALDI and HPLC/ESI, respectively.




1787

Figure 6. (a) Extracted ion chromatograms of deoxynivalenol (DON) for the LD-ESI-MS measurement of infected wheat seed with Fusarium graminearum and (b) healthy wheat seed as control. (*Laser was switched off during blank measurement.) (c) The mycotoxin DON was detected as [M + H]+ in the infected sample. (d) Averaged intensity of DON (arbitrary units) in the infected and control wheat.

A. Berisha et al. Application to mycotoxins in crop plants and phospholipids in mammalian tissue The results and discussion presented above are based on (secondary) plant metabolites in order to demonstrate the benefits and capabilities of the combined approach. However, the method can also be used for other application areas such as fungus infection of crop plants. The detection of the mycotoxin deoxynivalenol (DON) from the surface of a seed from a wheat plant that was infected with Fusarium graminearum is shown in Fig. 6. This measurement was obtained without any sample preparation. Significantly higher signal intensities of the mycotoxin were detected in the infected sample compared to the uninfected control material (healthy wheat seed) (Fig. 6(d)). DON was identified by accurate mass (mass deviation 0.1 ppm) (Fig. 6(c)) and MS/MS measurements directly from the plant surface (section 6 in the Supporting Information). HPLC/ ESI-MS/MS measurements of the infected seed extract are also presented in section 6 in the Supporting Information. Initial MALDI imaging measurements of DON have not yielded satisfactory result so far, but the method is being further optimized. In addition to DON a number of fungus-related compounds were detected which are currently investigated in more detail. Wheat seed was chosen as an example because it has a significantly lower water content than grape berries and thus demonstrates that our method can be used for a wide range of sample properties. A potential scenario to apply our approach to

fungus infection could be to screen a larger number of plants by LD-ESI in order to detect infected specimen, analyze the distribution of fungus-related signals in these samples with MALDI imaging and quantify mycotoxin levels by LC/MS/MS subsequently. The presented workflow is not restricted to plant material, but can be used for a range of sample types. The same approach is currently being applied for the analysis of phospholipids and other metabolites in mammalian tissue (mouse model and human tumor samples). Figure 7 shows the analysis of mouse brain tissue by all three methods. In contrast to plant material mouse brain tissue is characterized by a high content of lipids. More details on the mouse brain samples can be found in section 7 of the Supporting Information. Figure 7(a) shows a mass spectrum acquired from mouse brain tissue by LD-ESI. A large number of phospholipids were detected in this experiment. Several lipid species are labeled and demonstrate that significant overlap can be achieved with MALDI and HPLC/ESI (Figs. 7(b) and 7(c)) in analogy to the plant metabolite study discussed above. The detection of phospholipids by LD-ESI-MS is in accordance with previous studies of mammalian tissue by in situ analysis.[27] Likewise, previous studies by MALDI imaging[36] have shown that phospholipids are highly specific for different cell types and can be used for detailed characterization of tissue, e.g. for the investigation of tumor heterogeneity. HPLC/ESI-MS is routinely used for the analysis of phospholipids and delivers quantitative

1788

Figure 7. Comparison of the different MS approaches: (a) LD-ESI, (b) MALDI imaging, and (c) HPLC/ESI spectra from mouse brain tissue.




Comprehensive MS approach for complex samples information.[65] Therefore, we expect that a similar gain in information can be obtained for mammalian tissue samples as for the presented plant metabolite example.

CONCLUSIONS


The authors would like to thank the Geisenheim Research Center, Department for Wine Analysis and Beverage Research, Geisenheim, Germany, for providing the grape samples and the Institute of Plant Breeding, especially Sven Gottwald (JLU Giessen, Germany), for providing the wheat samples and for helpful discussions. This work was funded by the Hessian Ministry of Science and Arts (HMWK) through LOEWE focus "Ambiprobe", by the Federal Ministry of Education and Research (BMBF 0315379A), by the European Research Council ERC-STG 210356, and by the Deutsche Forschungsgemeinschaft DFG Sp314/12-1, Sp314/13-1.

REFERENCES [1] N. Vinayavekhin, E. A. Homan, A. Saghatelian. Exploring disease through metabolomics. ACS Chem. Biol. 2010, 5, 91. [2] S. Becker, L. Kortz, C. Helmschrodt, J. Thiery, U. Ceglarek. LCMS-based metabolomics in the clinical laboratory. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2012, 883/884, 68. [3] R. Aebersold, M. Mann. Mass spectrometry-based proteomics. Nature 2003, 422, 198. [4] J. Cox, M. Mann. Annu. Rev. Biochem. 2011, 80, 273. [5] B. Domon, R. Aebersold. Review – Mass spectrometry and protein analysis. Science 2006, 312, 212. [6] M. Petrovic, M. Farre, M. L. de Alda, S. Perez, C. Postigo, M. Kock, J. Radjenovic, M. Gros, D. Barcelo. Recent trends in the liquid chromatography-mass spectrometry analysis of organic contaminants in environmental samples. J. Chromatogr. A 2010, 1217, 4004. [7] S. D. Richardson. Environmental mass spectrometry: Emerging contaminants and current issues. Anal. Chem. 2010, 82, 4742. [8] H. B. Wang, E. J. Race, A. J. Shrikhande. Characterization of anthocyanins in grape juices by ion trap liquid chromatography-mass spectrometry. J. Agric. Food Chem. 2003, 51, 1839. [9] M. Downey, S. Rochfort. Simultaneous separation by reversed-phase high-performance liquid chromatography and mass spectral identification of anthocyanins and flavonols in Shiraz grape skin. J. Chromatogr. A 2008, 1201, 43. [10] W. Mullen, S. Larcombe, K. Arnold, H. Welchman, A. Crozier. Use of accurate mass full scan mass spectrometry for the analysis of anthocyanins in berries and berry-fed tissues. J. Agric. Food Chem. 2010, 58, 3910. [11] M. I. Churchwell, N. C. Twaddle, L. R. Meeker, D. R. Doerge. Improving LC–MS sensitivity through increases in chromatographic performance: Comparisons of UPLC–ES/MS/MS to HPLC–ES/MS/MS. J. Chromatogr. B 2005, 825, 134. [12] Z. Takats, J. M. Wiseman, B. Gologan, R. G. Cooks. Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science 2004, 306, 471. [13] R. B. Cody, J. A. Laramee, H. D. Durst. Versatile new ion source for the analysis of materials in open air under ambient conditions. Anal. Chem. 2005, 77, 2297. [14] J. D. Harper, N. A. Charipar, C. C. Mulligan, X. Zhang, R. G. Cooks, Z. Ouyang. Low-temperature plasma probe for ambient desorption ionization. Anal. Chem. 2008, 80, 9097. [15] K.-C. Schaefer, J. Denes, K. Albrecht, T. Szaniszlo, J. Balog, R. Skoumal, M. Katona, M. Toth, L. Balogh, Z. Takats. In vivo, in situ tissue analysis using rapid evaporative ionization mass spectrometry. Angew. Chem. Int. Ed. 2009, 48, 8240. [16] K.-C. Schaefer, T. Szaniszlo, S. Guenther, J. Balog, J. Denes, M. Keseru, B. Dezso, M. Toth, B. Spengler, Z. Takats. In Situ,



1789

In this study we present a method that combines three mass spectrometric approaches in order to obtain comprehensive information about complex biological samples. This combined method starts with an in situ technique, LD-ESI, which allows rapid identification of a range of plant metabolites. Results were acquired directly from the surface without any sample preparation and allow for semiquantitative interpretation as demonstrated by PCA. The detailed distribution of the compounds within the grape sample was visualized with 20 μm pixel size using MALDI MSI, which constitutes the most detailed MSI analysis of plant tissue so far. Metabolites showed different distributions which were in accordance with their biological function. Finally HPLC/ESI-MS was used within this workflow for the quantification of 13 anthocyanin species. These chromatography-based measurements also provided higher sensitivity due to reduced ion suppression effects. High mass resolution and high mass accuracy were used for all three approaches in order to increase the reliability of analysis. In this manner anthocyanin diglucosides (e.g. peonidindiglucoside) could be easily differentiated from the coumaroyl/ caffeoyl compounds which differ by only Δm/z = 0.02 in LD-ESI and MALDI-MS experiments. HPLC/ESI-MS/MS confirmed these findings and was used for the separation of isobaric compounds, e.g. petunidin-coumaroyl-monoglucoside and peonidin-caffeoyl-monoglucoside. The comparison of the mass spectra showed that correlation with regard to identified signals was quite high although different ionization methods were applied. In total, 41 metabolites were identified with all three approaches. Initial measurements were performed in order to detect the mycotoxin deoxynivalenol in fungus-infected wheat seeds as well as phospholipids for characterization of tissue types in mouse brain sections as a model system for mammalian tissue. These examples span a wide range of sample properties (in terms of water and lipid content) and thus demonstrate that the presented workflow can be used as a general approach for the detailed characterization of complex (biological) samples. The proposed strategy is also applicable to other combinations of MS methods, e.g. when LD-ESI is replaced by DESI or another ambient ionization technique. MSI studies would benefit from more frequent use of complementary techniques to validate results. HPLC/ESI and ambient ionization can also be used separately in order to back up quantitative data or to identify compounds. We expect that our strategy can be applied for a wide range of sample types, as ambient ionization, chromatography-based and imaging methods have been used individually for numerous applications, but not in the proposed combination yet. It should be noted again that, in any case, high resolution and high accuracy mass spectrometry data for each experiment is a key requirement in order to reliably combine data from different MS techniques.

Acknowledgements

A. Berisha et al.

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

1790

real-time identification of biological tissues by ultraviolet and infrared laser desorption ionization mass spectrometry. Anal. Chem. 2011, 83, 1632. J. S. Wiley, J. F. Garcia-Reyes, J. D. Harper, N. A. Charipar, Z. Ouyang, R. G. Cooks. Screening of agrochemicals in foodstuffs using low-temperature plasma (LTP) ambient ionization mass spectrometry. Analyst 2010, 135, 971. J. F. Garcia-Reyes, A. U. Jackson, A. Molina-Diaz, R. G. Cooks. Desorption electrospray ionization mass spectrometry for trace analysis of agrochemicals in food. Anal. Chem. 2009, 81, 820. F. M. Fernandez, R. B. Cody, M. D. Green, C. Y. Hampton, R. McGready, S. Sengaloundeth, N. J. White, P. N. Newton. Characterization of solid counterfeit drug samples by desorption electrospray ionization and direct-analysisin-real-time coupled to time-of-flight mass spectrometry. Chemmedchem 2006, 1, 702. M. Benassi, A. Berisha, W. Romão, E. Babayev, A. Römpp, B. Spengler. Petroleum crude oil analysis using lowtemperature plasma mass spectrometry. Rapid Commun. Mass Spectrom. 2013, 27, 825. J. M. Wiseman, S. M. Puolitaival, Z. Takats, R. G. Cooks, R. M. Caprioli. Mass spectrometric profiling of intact biological tissue by using desorption electrospray ionization. Angew. Chem. Int. Ed. 2005, 44, 7094. Z. Takats, V. Kobliha, K. Sevcik, P. Novak, G. Kruppa, K. Lemr, V. Havlicek. Characterization of DESI-FTICR mass spectrometry – from ECD to accurate mass tissue analysis. J. Mass Spectrom. 2008, 43, 196. M. Karas, F. Hillenkamp. Laser desorption ionization of proteins with molecular masses exceeding 10000 daltons. Anal. Chem. 1988, 60, 2299. B. Spengler, M. Karas, U. Bahr, F. Hillenkamp. Laser mass analysis in biology. Ber. Bunsen-Gesellschaft-Phys. Chem. Chem. Phys. 1989, 93, 396. J. Wei, J. M. Buriak, G. Siuzdak. Desorption-ionization mass spectrometry on porous silicon. Nature 1999, 399, 243. K. Huikko, P. Ostman, C. Sauber, F. Mandel, K. Grigoras, S. Franssila, T. Kotiaho, R. Kostiainen. Feasibility of atmospheric pressure desorption/ionization on silicon mass spectrometry in analysis of drugs. Rapid Commun. Mass Spectrom. 2003, 17, 1339. S. Guenther, K. C. Schafer, J. Balog, J. Denes, T. Majoros, K. Albrecht, M. Toth, B. Spengler, Z. Takats. Electrospray post-ionization mass spectrometry of electrosurgical aerosols. J. Am. Soc. Mass Spectrom. 2011, 22, 2082. B. Spengler, U. Bahr, M. Karas, F. Hillenkamp. Postionization of laser-desorbed organic and inorganic compounds in a time of flight mass spectrometer. Anal. Instrum. 1988, 17, 173. J. J. Coon, K. J. McHale, W. W. Harrison. Atmospheric pressure laser desorption/chemical ionization mass spectrometry: a new ionization method based on existing themes. Rapid Commun. Mass Spectrom. 2002, 16, 681. J. Shiea, M. Z. Huang, H. J. Hsu, C. Y. Lee, C. H. Yuan, I. Beech, J. Sunner. Electrospray-assisted laser desorption/ ionization mass spectrometry for direct ambient analysis of solids. Rapid Commun. Mass Spectrom. 2005, 19, 3701. P. Nemes, A. Vertes. Laser ablation electrospray ionization for atmospheric pressure, in vivo, and imaging mass spectrometry. Anal. Chem. 2007, 79, 8098. I. X. Peng, J. Shiea, R. R. O. Loo, J. A. Loo. Electrosprayassisted laser desorption/ionization and tandem mass spectrometry of peptides and proteins. Rapid Commun. Mass Spectrom. 2007, 21, 2541. B. Shrestha, A. Vertes. In situ metabolic profiling of single cells by laser ablation electrospray ionization mass spectrometry. Anal. Chem. 2009, 81, 8265.


[34] B. Shrestha, J. M. Patt, A. Vertes. In situ cell-by-cell imaging and analysis of small cell populations by mass spectrometry. Anal. Chem. 2011, 83, 2947. [35] K. Chughtai, R. M. A. Heeren. Mass spectrometric imaging for biomedical tissue analysis. Chem. Rev. 2010, 110, 3237. [36] A. Römpp, B. Spengler. Mass spectrometry imaging with high resolution in mass and space. Histochem. Cell Biol. 2013, 139, 759. [37] B. Spengler, M. Hubert, R. Kaufmann, in Proc. 42nd Conf. Mass Spectrometry and Allied Topics, Chicago, Illinois, May 19–June 3, 1994, p. 1041. [38] R. M. Caprioli, T. B. Farmer, J. Gile. Molecular imaging of biological samples: Localization of peptides and proteins using MALDI-TOF MS. Anal. Chem. 1997, 69, 4751. [39] M. Stoeckli, P. Chaurand, D. E. Hallahan, R. M. Caprioli. Imaging mass spectrometry: A new technology for the analysis of protein expression in mammalian tissues. Nat. Med. 2001, 7, 493. [40] J. D. Watrous, T. Alexandrov, P. C. Dorrestein. The evolving field of imaging mass spectrometry and its impact on future biological research. J. Mass Spectrom. 2011, 46, 209. [41] B. Prideaux, M. Stoeckli. Mass spectrometry imaging for drug distribution studies. J. Proteomics 2012, 75, 4999. [42] P. Neubert, A. Walch. Current frontiers in clinical research application of MALDI imaging mass spectrometry. Exp. Rev. Proteomics 2013, 10, 259. [43] M. M. Burrell, C. J. Earnshaw, M. R. Clench. Imaging matrix assisted laser desorption ionization mass spectrometry: a technique to map plant metabolites within tissues at high spatial resolution. J. Exp. Bot. 2007, 58, 757. [44] S. Robinson, K. Warburton, M. Seymour, M. Clench, J. ThomasOates. Localization of water-soluble carbohydrates in wheat stems using imaging matrix-assisted laser desorption ionization mass spectrometry. New Phytologist 2007, 173, 438. [45] J. H. Jun, Z. Song, Z. Liu, B. J. Nikolau, E. S. Yeung, Y. J. Lee. Highspatial and high-mass resolution imaging of surface metabolites of Arabidopsis thaliana by laser desorption-ionization mass spectrometry using colloidal silver. Anal. Chem. 2010, 82, 3255. [46] D. C. Perdian, Y. J. Lee. Imaging MS methodology for more chemical information in less data acquisition time utilizing a hybrid linear ion trap-orbitrap mass spectrometer. Anal. Chem. 2010, 82, 9393. [47] N. Zaima, N. Goto-Inoue, T. Hayasaka, M. Setou. Application of imaging mass spectrometry for the analysis of Oryza sativa rice. Rapid Commun. Mass Spectrom. 2010, 24, 2723. [48] Y. Yoshimura, N. Zaima, T. Moriyama, Y. Kawamura. Different localization patterns of anthocyanin species in the pericarp of black rice revealed by imaging mass spectrometry. PloS One 2012, 7, e31285. [49] Y. Yoshimura, H. Enomoto, T. Moriyama, Y. Kawamura, M. Setou, N. Zaima. Visualization of anthocyanin species in rabbiteye blueberry Vaccinium ashei by matrix-assisted laser desorption/ionization imaging mass spectrometry. Anal. Bioanal. Chem. 2012, 403, 1885. [50] A. Roempp, S. Guenther, Y. Schober, O. Schulz, Z. Takats, W. Kummer, B. Spengler. Histology by mass spectrometry: Label-free tissue characterization obtained from high-accuracy bioanalytical imaging. Angew. Chem. Int. Ed. 2010, 49, 3834. [51] A. Roempp, S. Guenther, Z. Takats, B. Spengler. Mass spectrometry imaging with high resolution in mass and space (HR(2) MSI) for reliable investigation of drug compound distributions on the cellular level. Anal. Bioanal. Chem. 2011, 401, 65. [52] S. Guenther, A. Roempp, W. Kummer, B. Spengler. APMALDI imaging of neuropeptides in mouse pituitary gland with 5 μm spatial resolution and high mass accuracy. Int. J. Mass Spectrom. 2011, 305, 228. [53] Y. Schober, S. Guenther, B. Spengler, A. Römpp. Highresolution matrix-assisted laser desorption/ionization imaging of tryptic peptides from tissue. Rapid Commun. Mass Spectrom. 2012, 26, 1141.



Comprehensive MS approach for complex samples [54] Y. Schober, S. Guenther, B. Spengler, A. Roempp. Single cell matrix-assisted laser desorption/ionization mass spectrometry imaging. Anal. Chem. 2012, 84, 6293. [55] S. J. Duthie, P. T. Gardner, P. C. Morrice, S. G. Wood, L. Pirie, C. C. Bestwick, L. Milne, G. G. Duthie. DNA stability and lipid peroxidation in vitamin E-deficient rats in vivo and colon cells in vitro - Modulation by the dietary anthocyanin, cyanidin-3glycoside. Eur. J. Nutr. 2005, 44, 195. [56] H. Dietrich, B. Fröhling, D. Hofmann, E. Rühl, F. Will. Deutsche Lebensmittel Rundschau 2009, 105, 695. [57] W. Bouschen, O. Schulz, D. Eikel, B. Spengler. Matrix vapor deposition/recrystallization and dedicated spray preparation for high-resolution scanning microprobe matrix-assisted laser desorption/ionization imaging mass spectrometry (SMALDI-MS) of tissue and single cells. Rapid Commun. Mass Spectrom. 2010, 24, 355. [58] M. Koestler, D. Kirsch, A. Hester, A. Leisner, S. Guenther, B. Spengler. A high-resolution scanning microprobe matrixassisted laser desorption/ionization ion source for imaging analysis on an ion trap/Fourier transform ion cyclotron resonance mass spectrometer. Rapid Commun. Mass Spectrom. 2008, 22, 3275. [59] C. Paschke, A. Leisner, A. Hester, K. Maass, S. Guenther, W. Bouschen, B. Spengler. MIRION – a software package for automatic processing of mass spectrometric images. J. Am. Soc. Mass Spectrom. 2013, 24, 1296.

[60] M. O. Downey, M. Mazza, M. P. Krstic. Development of a stable extract for anthocyanins and flavonols from grape skin. Am. J. Enol. Viticulture 2007, 58, 358. [61] J. Wang, P. Sporns. Analysis of anthocyanins in red wine and fruit juice using MALDI-MS. J. Agric. Food Chem. 1999, 47, 2009. [62] W. Mullen, S. C. Marks, A. Crozier. Evaluation of phenolic compounds in commercial fruit juices and fruit drinks. J. Agric. Food Chem. 2007, 55, 3148. [63] J. M. Kong, L. S. Chia, N. K. Goh, T. F. Chia, R. Brouillard. Analysis and biological activities of anthocyanins. Phytochemistry 2003, 64, 923. [64] A. Fujita, N. Goto-Yamamoto, I. Aramaki, K. Hashizume. Organ-specific transcription of putative flavonol synthase genes of grapevine and effects of plant hormones and shading on flavonol biosynthesis in grape berry skins. Biosci. Biotechnol. Biochem. 2006, 70, 632. [65] C. Wolf, P. J. Quinn. Lipidomics: Practical aspects and applications. Prog. Lipid Res. 2008, 47, 15.

SUPPORTING INFORMATION Additional supporting information may be found in the online version of this article at the publisher’s website.

1791


