Author's personal copy

Transcription

1 Analytica Chimica Acta 747 (2012) Contents lists available at SciVerse ScienceDirect Analytica Chimica Acta j ourna l ho me page: Tracing phenolic biosynthesis in Vitis vinifera via in situ C-13 labeling and liquid chromatography diode-array detector mass spectrometer/mass spectrometer detection Alexander W. Chassy a, Douglas O. Adams a, V. Felipe Laurie b, Andrew L. Waterhouse a, a Department of Viticulture and Enology, One Shields Avenue, University of California, Davis, CA, USA b School of Agricultural Sciences, Universidad de Talca, Talca, Chile h i g h l i g h t s g r a p h i c a l a b s t r a c t l-phenyl- 13 C 6 -alanine was incorporated into grape berries, intact on the vine, at two stages of maturity. Labeled anthocyanins were synthesized by the berry and quantified by LC UV/Vis MS/MS. Extremely tight regulation of phenylpropanoid pathways was observed. a r t i c l e i n f o Article history: Received 7 April 2012 Received in revised form 3 August 2012 Accepted 9 August 2012 Available online 21 August 2012 Keywords: Anthocyanin Stable-isotope Metabolism Flavonoid a b s t r a c t Phenolic compounds in Vitis vinifera contribute important flavor, functionality, and health qualities to both table and wine grapes. The plant phenolic metabolic pathway has been well characterized, however many important questions remain regarding the influence of environmental conditions on pathway regulation. As a diagnostic for this pathway s regulation, we present a technique to incorporate a stableisotopic tracer, l-phenyl- 13 C 6 -alanine (Phe 13 ), into grape berries in situ and the accompanying high throughput analytical method based on LC DAD MS/MS to quantify and track the label into phenylalanine metabolites. Clusters of V. vinifera cv. Cabernet Sauvignon, either near the onset of ripening or 4 weeks later, were exposed to Phe 13 in the vineyard. Phe 13 was present in berries 9 days afterwards as well as labeled flavonols and anthocyanins, all of which possessed a molecular ion shift of 6 amu. However, nearly all the label was found in anthocyanins, indicating tight regulation of phenolic biosynthesis at this stage of maturity. This method provides a framework for examining the regulation of phenolic metabolism at different stages of maturity or under different environmental conditions. Additionally, this technique could serve as a tool to further probe the metabolism/catabolism of grape phenolics Elsevier B.V. All rights reserved. 1. Introduction Phenylalanine metabolism, via the phenylpropanoid pathways, in the grape berries of V. vinifera yields a multitude of phenolic compounds; the major classes represented are anthocyanins, flavan-3-ols, hydroxycinnamates, flavonols, and stilbenes. Each Corresponding author. Tel.: ; fax: address: alwaterhouse@ucdavis.edu (A.L. Waterhouse). of these chemical classes contributes important characteristics to both table and wine grapes. In wine, color and mouthfeel from anthocyanins and proanthocyanidins, respectively, are perhaps the best understood contributions of phenolics. However, questions remain surrounding the environmental regulation and biosynthesis of grape phenolics. The biosynthetic pathways for hydroxycinnamic acids and flavonoids are, with a few exceptions, well characterized however research continues to evaluate the metabolic regulation and function of these important phenylalanine metabolites /$ see front matter 2012 Elsevier B.V. All rights reserved.

2 52 A.W. Chassy et al. / Analytica Chimica Acta 747 (2012) One such area of research is on anthocyanin catabolism in vivo. In grapes, researchers have long known that warm climate causes lower anthocyanin content, however, the mechanism was poorly understood [1]. In wine grapes, anthocyanin content is considered highly reflective of quality [2]. For example, the central valley of California produces more red wine grapes than any other part of the country but the price for these grapes is significantly less than the price for grapes produced in cooler wine regions due to lower color and other factors. Scientists initially believed warm climates resulted in reduced enzyme expression and therefore less anthocyanin production until Shaked-Sachray et al. suggested that anthocyanin catabolism may play a role [3]. The concept of turnover among anthocyanins and flavonoids has only recently gained attention [4]. Anthocyanin turnover rates have been estimated to be as high as 70% per day in Brunfelsia calycina flowers to 3 6% per day in mustard seedlings [5,6]. In an attempt to understand the relationship between warm climates and reduced anthocyanin content in grapes, researchers fed isotope labeled Phe to grapes cultured in vitro. The labeled anthocyanins were then monitored under cool and warm conditions. In ripening grape berries cultured in vitro, turnover rates were less than 10% per day in 15 and 25 C conditions but up to 50% per day in 35 C conditions [7]. This study suggests that anthocyanin degradation may be a primary cause for reduced anthocyanin content in warm climate grapes. However, this data represents grape berries cultured in Petri dishes; grape anthocyanin turnover rates on the vine, in situ, are still unknown. Furthermore, diurnal temperature cycles further distance these estimates from actual turnover rates in situ. To further understand anthocyanin catabolism, a powerful method to track both anthocyanin production and degradation is needed. Analysis of grape phenolics is often performed using liquid chromatography coupled to UV/Vis spectral and/or mass spectrometer (MS) detectors. Methods have been developed that can detect a broad range phenolic constituents of grapes in one chromatographic run [8,9]. These methods are of great usefulness when profiling Phe metabolites, however not all methods are applicable to MS due to the presence of non-volatile salts and relatively poor quantitation by certain MS detection methods. This presents a conflict between the simplicity and precision of quantitation by UV/Vis and the crucial information provided by MS detection. Furthermore, when using an isotopic tracer it is critical to be able to detect and quantify isotopomers. Techniques such as metabolic flux analysis (MFA) use labeled substrates to provide an isotopomer traceable by NMR, mass spectrometry or liquid scintillation (when using radioisotopes) [10]. Since the isotopomer differs only by molecular mass from the native substrate, the biological system metabolizes it as if it were the endogenous compound. Some studies have utilized stable-isotope labeled Phe to probe the metabolic response to wounds in potatoes [11] or to identify alternate routes of phenolic biosynthesis in petunia flowers [12]. The incorporation of isotopically labeled Phe can also be used to study the nature of Phe metabolism to secondary products as well as subsequent metabolism. Isotopic tracers have long served scientists in understanding the biochemistry of plants and animals. Tracers have been incorporated into biological systems through various means, including through the root system, injection, as a gas or vapor, and through incisions into the stem or petiole [13]. A recently developed technique, while simple in nature, involves soaking plant material in a solution of isotopic tracer. This technique incorporated labeled guaiacol into grape leaves and berries which was then further conjugated with sugar moieties [14]. Here we present a technique to incorporate stable-isotope labeled phenylalanine into grape berries in situ and the accompanying high throughput analytical method based on LC/DAD/MS to quantify and track the label into phenylalanine metabolites. 2. Materials and methods 2.1. In situ tracer incorporation on the vine The experiments were performed using grape vines from selfrooted Cabernet Sauvignon, Clone 8, planted in 1998 at the University of California, Davis, Hopkins experimental vineyard. During the summer of 2009, multiple methods were tested to evaluate the best technique to incorporate l-phenyl- 13 C 6 -alanine (Cambridge Isotope Labs, Cambridge, MA) into berries in the same vineyard. The best incorporation was achieved by soaking attached berry clusters in a solution of Phe 13. Clusters were wrapped in a small plastic bag containing 5 mm Phe 13. After 48 h, the bags were removed and the clusters were rinsed with distilled, deionized water. This procedure was conducted at two different timepoints during the 2010 growing season. The first incubation (July 29, 2010) was on green grapes, approximately 1 week before the onset of color development, and the second incubation (August 31, 2010) was on uniformly red berry clusters approximately 4 weeks after initial color development. For each incubation timepoint, 4 clusters were selected from Cabernet Sauvignon vines grown in a single row at the University of California, Davis experimental vineyard. From each cluster, 3 randomly selected berries were taken immediately prior to incubation and 9 days following the treatment, for a total of 12 biological replicates at each time point. These berries were frozen and stored at 80 C Extraction Frozen, de-seeded berry samples were weighed before and after freeze drying. Individual lyophilized berries, approximately 0.15 g, were extracted with 2 ml of 60% acetone. As a recovery check standard, 100 L of 1 mg ml 1 phlorizin dihydrate (Sigma, St. Louis, MO) in methanol was spiked into each sample to ensure sample was not lost in the extraction process. The mixture was vortexed and stored overnight at 4 C. The liquid extract was then transferred by Pasteur pipette to a microcentrifuge tube and placed in a Centrivap centrifugal vacuum evaporator (Labconco, Kansas City, MO) for 45 min at 25 C to evaporate the acetone. Once the acetone was removed, the resulting residue was transferred to a 2 ml volumetric flask and brought to volume with water. This extract was then filtered using 0.2 m PTFE syringe filters (Agilent Technologies, Santa Clara, CA) and analyzed by HPLC DAD MS/MS HPLC DAD Liquid chromatography was performed on an Agilent series 1200 instrument (Agilent Technologies, Santa Clara, CA). A solvent degasser, binary pump, refrigerated autosampler (4 C), temperature controlled column compartment, and diode-array detector (DAD) comprised the modular system. An Agilent Zorbax SB-C18 Rapid Resolution HT small particle column (4.6 mm 100 mm, 1.8 m) with a pre-column filter facilitated the separation. A flow rate of 0.75 ml min 1 at 40 C was used with a binary gradient of 10% formic acid (solvent A) and acetonitrile (solvent B) (Fisher Scientific, Fair Lawn, NJ), the same mobile phase as some previous studies [15,16]. The gradient began with 5% solvent B held for 2 min, then an increase to 13% B in 3 min, 15.3% B in 3 more min, and up to 20.5% in five additional minutes. From there a sharper increase in organic was applied: 2 min were given to increase solvent B to 31% and within 1 more min the gradient reached 50%. In another 1.5 min the gradient went to 65% where it was held for 3 more minutes. Afterwards, 3 min were given to drop solvent B back to the original 5% concentration which was held for an additional min. Also, one min of post-time was given to allow the column to

3 A.W. Chassy et al. / Analytica Chimica Acta 747 (2012) Fig. 1. Separation of grape berry extract by LC UV/Vis detection at 316, 365, and 520 nm. Peak assignments are displayed in Table 1. equilibrate. The autosampler drew 8 L of sample. The DAD continuously monitored 4 wavelengths: 280, 316, 365, and 520 nm. Online spectra were recorded every 2 s from 190 to 600 nm to aid in identification MS/MS An Agilent 6430 Triple Quadrupole Mass Spectrometer with electrospray injection (ESI) was used in-line with the HPLC DAD system (Agilent Technologies, Santa Clara, CA). The ESI flowed 40 psi nitrogen through the needle while the capillary maintained 4 kv with 12 L min 1 drying gas at 325 C. Data acquisition was performed using dynamic MRM, a multiple reaction monitoring (MRM) method that only monitors specific mass transitions during preset retention times. Dynamic MRM also has the ability to simultaneously perform ESI in both positive and negative mode, allowing for the analysis of different phenolic classes within a single chromatographic run. The cycle time for each round of MRM was 550 ms. Individual compound fragmentation and collision energies as well as cell acceleration values can be seen in Supplementary materials (Supplemental 1) Quantitation An external standard representing each class of compounds was used for the quantification of total concentration by UV/Vis absorbance: caffeic acid for cinnamate-tartrate esters (Sigma, St. Louis, MO), rutin for flavonols (Sigma, St. Louis, MO), and malvidin- 3-glucoside for anthocyanins (Polyphenols, Sandnes, Norway). Additionally, MS/MS was acquired to identify the relative quantities of the natural molecular ion and the 6 amu shifted molecular ion. The concentration of labeled phenolics were then calculated using the following equation where [M+6] and [M] represent the labeled and natural molecular ions, respectively, of the same compound: [M + 6]fraction = [Total concentration by DAD] [M + 6]peak area [M + 6]peak area + [M]peak area 2.6. Field temperature measurements Daily temperature readings were accessed from the University of California Statewide Integrated Pest Management Program (UCDIPM Online). The data was collected at the Davis. A station, also known as CIMIS #6, Davis in Yolo County Results and discussion The phenylpropanoid pathway is responsible for a multitude of plant secondary metabolites. These compounds have diverse functions within plant systems and may also have significance in the human diet. The method presented in this paper describes an LC DAD MS/MS protocol that can accurately track the metabolism of labeled phenolic compounds in grapes over time. This analytical method supports a novel technique to incorporate a stable isotopic tracer into the phenylpropanoid pathway on intact grape berries in the vineyard. This study reveals a framework by which scientists may probe phenolic metabolic rates and products LC DAD MS/MS analysis Chromatography of the minimally processed grape sample achieved baseline separation of most phenolic analytes within 18 min, especially when monitored under corresponding DAD signals: 280 nm for proanthocyanins, 316 nm for hydroxycinnamates, 365 nm for flavonols, and 520 nm for anthocyanins (Fig. 1). Separation of anthocyanins is greatly aided by acidic conditions, as has been shown in previous studies [9,17]. Molecular features were tentatively identified by their chromatographic characteristics, their UV absorbance spectrum, and their MS/MS profile (Table 1) with help from previous studies [15,18]. To aid in identification, MS n was performed on an Agilent XCT Ultra Ion Trap MS (data not shown). For each class of phenolic, one external standard was used for quantitation by DAD: caffeic acid for hydroxycinnamates, rutin for flavonols, and malvidin-3-glucoside (M3G) for anthocyanins. For caffeic acid, rutin and M3G, the limits of detection were g ml 1 (0.331 nmol ml 1 ), g ml 1 (0.261 nmol ml 1 ), and g ml 1 (0.449 nmol ml 1 )

4 54 A.W. Chassy et al. / Analytica Chimica Acta 747 (2012) Table 1 Peak annotation of tentatively identified phenolics by LC DAD MS/MS separation of grape berry. Peak Annotated compound RT (min) Monitored UV/Vis (nm) Transition native (M+6) +/ ESI Compound class 1 Caftaric acid ( ) 2 cis-coutaric acid ( ) 3 trans-coutaric acid ( ) 4 Delphinidin-3-O-glucoside ( ) 5 Cyanidin-3-O-glucoside ( ) 6 Myricetin-3-O-glucuronide ( ) 7 Petunidin-3-O-glucoside ( ) 8 Myricetin-3-O-glucoside ( ) 9 Peonidin-3-O-glucoside ( ) 10 Malvidin-3-O-glucoside ( ) 11 Delphinidin-3-O-(6-O-acetyl)glucoside ( ) 12 Quercetin-3-O-glucuronide ( ) 13 Quercetin-3-O-glucoside ( ) 14 Laricitrin-3-O-glucoside ( ) 15 Cyanidin-3-O-(6-O-acetyl)glucoside ( ) 16 Kaempferol-3-O-galactoside ( ) 17 Petunidin-3-O-(6-O-acetyl)glucoside ( ) 18 Kaempferol-3-O-glucoside ( ) 19 Isorhamnetin-3-O-glucoside ( ) 20 Peonidin-3-O-(6-O-acetyl)glucoside ( ) 21 Malvidin-3-O-(6-O-acetyl)glucoside ( ) 22 Malvidin-3-O-(6-O-p-trans-caffoyl)glucoside ( ) ( ) ( ) 23 Petunidin-3-O-(6-O-p-coumaryl)glucoside ( ) ( ) ( ) 24 Peonidin-3-O-(6-O-p-cis-coumaryl)glucoside ( ) ( ) ( ) 25 Malvidin-3-O-(6-O-p-cis-coumaryl)glucoside ( ) ( ) ( ) 26 Peonidin-3-O-(6-O-p-trans-coumaryl)glucoside ( ) ( ) ( ) 27 Malvidin-3-O-(6-O-p-trans-coumaryl)glucoside ( ) ( ) ( ) Hydroxycinnamoyltartaric acid Hydroxycinnamoyltartaric acid Hydroxycinnamoyltartaric acid Flavonol glucuronide Flavonol glucuronide + Anthocyanin caffoyl glycoside

5 Table 2 Concentrations of grape berry phenolics (± standard error) at two stages of maturity (green and red) before and after incubation with Phe 13. Hydroxycinnamates are expressed as nmols caffeic acid equivalents per berry, flavonols as nmols rutin equivalents per berry, and anthocyanins as nmols M-3-G equivalents per berry. Compounds Green incubation Red incubation Day 0 Day 9 Day 0 Day 9 7/27/2010 8/5/2010 8/31/2010 9/9/2010 Native Labeled Native Labeled Native Labeled Native Labeled Phenylalanine (nmol berry 1 ) ± ± ± ± ± ± 0.94 Hydroxycinnamic tartrate esters Caftaric acid ± ± ± ± 2.96 c-coutaric acid ± ± ± ± 0.31 t-coutaric acid ± ± ± ± 0.86 Flavonols Myricetin-3-O-glucuronide 1.11 ± ± ± ± 1.32 Myricetin-3-O-glucoside 0.34 ± ± ± ± 8.24 Quercetin-3-O-glucuronide ± ± ± ± Quercetin-3-O-glucoside ± ± ± ± ± Laricitrin-3-O-glucoside 0.17 ± ± ± ± 1.65 Kaempferol-3-O-galactoside 0.18 ± ± ± ± 1.12 Kaempferol-3-O-glucoside 0.61 ± ± ± ± 3.60 Isorhamnetin-3-O-glucoside nd 0.73 ± ± ± 3.63 Anthocyanins Delphinidin-3-O-glucoside nd ± ± ± ± ± 0.03 Cyanidin-3-O-glucoside nd ± ± ± ± ± Petunidin-3-O-glucoside nd ± ± ± ± ± 0.03 Peonidin-3-O-glucoside nd ± ± ± ± ± 0.03 Malvidin-3-O-glucoside nd ± ± ± ± ± 0.33 Delphinidin-3-O-(6-O-acetyl)glucoside nd ± ± ± ± ± Cyanidin-3-O-(6-O-acetyl)glucoside nd 8.07 ± ± ± ± 0.59 nd Petunidin-3-O-(6-O-acetyl)glucoside nd ± ± ± ± ± Peonidin-3-O-(6-O-acetyl)glucoside nd ± ± ± ± ± Malvidin-3-O-(6-O-acetyl)glucoside nd ± ± ± ± ± 0.23 Malvidin-3-O-(6-O-p-caffoyl)glucoside nd nd nd ± ± ± Petunidin-3-O-(6-O-p-coumaryl)glucoside nd 6.93 ± ± ± ± ± Peonidin-3-O-(6-O-p-cis-coumaryl)glucoside nd 6.46 ± ± ± ± 0.13 nd Malvidin-3-O-(6-O-p-cis-coumaryl)glucoside nd 6.61 ± ± ± ± ± Peonidin-3-O-(6-O-p-trans-coumaryl)glucoside nd 9.77 ± ± ± ± ± Malvidin-3-O-(6-O-p-trans-coumaryl)glucoside nd ± ± ± ± ± 0.09 A.W. Chassy et al. / Analytica Chimica Acta 747 (2012)

6 56 A.W. Chassy et al. / Analytica Chimica Acta 747 (2012) Table 3 Quantity of native and labeled malvidin moieties 9 days after in situ addition of Phe 13 to green grapes. Native M+6 M+6 M+12 Malvidin-3-O-glucoside Transition (m/z) nmol berry ± ± 0.09 % Total 1.13 malvidin-3-o-(6-o-acetyl)glucoside Transition nmol M-3-G equiv. berry ± ± 0.06 % Total 1.12 malvidin-3-o-(6-o-p-trans-coumaryl)glucoside Transition nmol M-3-G equiv. berry ± ± ± ± % Total respectively. Total concentrations of individual phenolics based on equivalence with standards were calculated based on UV absorbance, and MS/MS was performed to tentatively identify and quantitate the ratio of labeled to unlabeled phenolics. For each compound, the ratio of percent labeled was then multiplied by the total concentration to yield a quantity for each the native compound and the labeled compound (Table 2) Tracer incorporation l-phenyl- 13 C 6 -alanine was selected as a tracer due to its unique isotopic characteristics. The phenyl ring is completely labeled with 13 C, which imparts labeled phenolics with an unusual shift of 6 amu, providing a signal with little interference. The addition of 6 extra mass units provides a baseline separation from the natural isotopic profile of a given compound, therefore allowing more accurate quantification of labeled metabolites. Grape clusters were soaked in l-phenyl- 13 C 6 -alanine for 48 h and phenolics were monitored immediately before and 9 days after the treatment. This treatment was performed on grape clusters at 2 separate stages of maturity. The first treatment was performed on green berries on the cusp of ripening and the second treatment was around 4 weeks after the onset of ripening. While the first treatment was performed on berries with no perceptible color, in some berry samples trace levels of anthocyanins were detectable by MS, but since they were not detectable by DAD 520 nm, the values were below the limit of quantification. Labeled anthocyanins were readily found after both treatment timepoints. Malvidin is the predominant anthocyanin moiety in Cabernet Sauvignon grapes; 13 C label was found in all of the abundant species of malvidin (Table 3). The soaking technique yielded labeled anthocyanin up to 1% of the total pool in grapes entering the ripening stage. The yield of labeled anthocyanin was much less when maturing red grapes were soaked in Phe 13. This is to be expected as red grapes had already accumulated anthocyanins for almost 4 weeks prior to treatment. A soaking technique has previously been shown to be effective for incorporating stable-isotope labeled guaiacol into grape berries [14]. Guaiacol is lipophilic and would be expected to assimilate into the waxy grape skin. Terpenes from eucalyptus are also observed to accumulate in berries near such trees [19]. However the incorporation of hydrophilic, zwitterionic phenylalanine represents a novel technique, as grape berries possess a waxy cuticle surrounding the epidermis, so the admission of substantial amounts of Phe 13 into the berry was somewhat surprising. Recent work shows water can enter at the stem berry junction [20] suggesting an intriguing possibility, but the process by which Phe 13 enters the grape needs to be clarified. Whatever the mechanism, Phe 13 clearly enters the berry and was found to be 5 7% of the total Phe content 9 days after treatment (Table 2). Additionally, labeled anthocyanins were present 9 days after treatment, suggesting that the total quantity of Phe 13 that entered the berry is even higher since a significant portion was already metabolized. In both green and red grape incubation treatments, labeled acylated and p-coumaroylated anthocyanins were observed. On the coumaroylated molecules, there are 3 M+6 isotopomers possible: a labeled anthocyanidin, a labeled p-coumaroyl group, and a molecule which contains both the labeled anthocyanidin and p-coumaroyl group. The proportions of different malvidin moieties in green grapes are illustrated in Table 3. Phe 13 was incorporated equally into all of the malvidin moieties (1% of total), with the exception of the M+12 moiety which demonstrated that its concentration correlated with the probability of a labeled anthocyanin being further modified with a labeled p-coumaroyl group (1% 1% = 0.01%). While labeled p-coumaroyl groups were detected on anthocyanins, no labeled p-coumaroyl tartrate esters were found. This may result from Phe 13 not entering the mesocarp, or pulp, of the berry. Since almost all hydroxycinnamate tartrate esters are produced in the mesocarp, if Phe 13 were not present, labeled caftaric and coutaric acids would not be expected [21]. An alternate explanation may be the level of grape maturity; hydroxycinnamate tartrate esters are produced early in the development of grape berries, however accumulation ends at the onset of ripening and levels gradually decrease to a constant amount [21,22]. It is possible that the metabolic pathways for tartrate esterification were not as active as those for esterification onto anthocyanins. Similarly, with the exception of quercetin-3-glucoside (Q3G) in green berries, labeled flavonols were not detected after Phe 13 treatments. This again may be explained by the level of maturity. While levels of Q3G, the most abundant flavonol in ripe grape berries, have been shown to steadily increase after the onset of ripening, anthocyanin production is simultaneously times greater [21,23]. This indicates that most of a ripening grape s phenylalanine pool is used for anthocyanin production, so the relatively small pool (5 7%) of Phe 13 does not produce flavonols in high enough concentrations for detection. In green grapes however there was sufficient Q3G produced for quantification. This suggests that the phenylpropanoid pathway is exceptionally selective at this point in the season with anthocyanin metabolism dominating all other sinks for Phe. Only one non-anthocyanin was observed in the green berries; Q3G represented about 15% of all labeled phenolics. An interesting feature of the data is that the levels of almost all phenolics were lower in red, maturing grapes, 9 days after the treatment as compared to day 0. This would not be expected, especially in anthocyanins where one would expect continued accumulation at this stage of development. This could be explained by natural variation, the treatment method, or perhaps by in vivo catabolism due to climate. Natural variation between grape berries within a single cluster is well documented [24,25], and it is certainly possible that the 12 berries selected on day 9 were simply lower that those selected on day 0 due to that variation. Since the values are listed on a per berry basis, any dilution effects would not be influential. A

7 A.W. Chassy et al. / Analytica Chimica Acta 747 (2012) remaining explanation is that there was degradation of the phenolics due to climate. Between August 31, 2010 and September 9, 2010, seven of the nine days had temperatures above 30 C ( 90 F), with a maximum temperature of 37.8 C on September 2. High temperatures have long been associated with low anthocyanin content [1] in grapes and recent research has suggested that anthocyanin degradation may be a contributing cause [7]. Anthocyanin catabolism has been established as a critical factor for color levels in vivo, but the mechanism of this degradation is still unknown [7,26,27]. Future studies using these methods will examine the rates of both biosynthesis and catabolism of in situ, labeled anthocyanins. Additionally, labeled anthocyanins may serve as tracers in the search for degradation products, which might aid in the identification of the anthocyanin catabolism pathway. 4. Conclusion This new method of introducing a labeled polar metabolic precursor in V. vinifera berries opens the door for investigating many related phenylpropanoid and perhaps unrelated pathways with polar precursors that may benefit from additional scrutiny. The simplicity and efficiency, combined with the sensitivity and precision of MS measurements provide a powerful tool for exploring specific systems, and this tool can be a strong complement to investigations of metabolic networks using genomic and related evolving technologies. Future studies will characterize the turnover of grape phenolics and identify novel or interesting sinks of phenolic metabolism. This technique provides a framework for further inquiries into the biosynthesis and catabolism of phenolics as well as a diagnostic tool for probing seasonal and environmental regulation of the phenylpropanoid pathway and comparisons with regulatory mechanism, i.e. RNA and expression of pertinent enzymes. Acknowledgments AWC was supported by graduate student fellowships including the Mario P. Tribuno Memorial Research Fellowship, Wine Spectator Scholarship, Horace O. Lanza Scholarship, Adolf L. & Richie C. Heck Research Scholarship, Leon D. Adams Research Scholarship, and the Curtis J. Alley Memorial Research Scholarship. Travel to Chile to aid in method development was funded by the Wine Spectator International Research Award. This research was not possible without the LC DAD MS/MS instrument time provided by Dr. Anita Oberholster. Undergraduate student researchers Christopher Barberi, Monthakan Boonpermpol, Aaron Garlits, Sarah Suharwardy, and Nina Vuoso helped with sample extraction. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at References [1] A. Winkler, J. Cook, W. Kliewer, L. Lider, General Viticulture, vol. 4, University of California Press, Berkeley, [2] T.C. Somers, M.E. Evans, K.M. Cellier, Vitis 22 (1983) 348. [3] L. Shaked-Sachray, D. Weiss, M. Reuveni, A. Nissim-Levi, M. Oren-Shamir, Physiol. Plantar. 114 (2002) 559. [4] M. Oren-Shamir, Plant Sci. 177 (2009) 310. [5] H. Vaknin, A. Bar-Akiva, R. Ovadia, A. Nissim-Levi, I. Forer, D. Weiss, M. Oren-Shamir, Planta 222 (2005) 19. [6] K. Zenner, M. Bopp, J. Plant Physiol. 126 (1987) 475. [7] K. Mori, N. Goto-Yamamoto, M. Kitayama, K. Hashizume, J. Exp. Bot. 58 (2007) [8] S. Gomez-Alonso, E. Garcia-Romero, I. Hermosin-Gutierrez, J. Food Compos. Anal. 20 (2007) 618. [9] R.M. Lamuela-Raventos, A.L. Waterhouse, Am. J. Enol. Vitic. 45 (1994) 1. [10] R.G. Ratcliffe, Y. Shachar-Hill, Plant J. 45 (2006) 490. [11] Y.N. Yang, J.S. Zhang, J. Hoh, F. Matsuda, P. Xu, M. Lathrop, J. Ott, Proc. Natl. Acad. Sci. U.S.A. 100 (2003) [12] J. Boatright, F. Negre, X. Chen, C. Kish, B. Wood, G. Peel, I. Orlova, D. Gang, D. Rhodes, N. Dudareva, Plant Physiol. 135 (2004) [13] R.H. Burris, Bot. Rev. 16 (1950) 150. [14] Y. Hayasaka, G.A. Baldock, K.H. Pardon, D.W. Jeffery, M.J. Herderich, J. Agric. Food Chem. 58 (2010) [15] S. Pati, M. Liberatore, G. Gambacorta, D. Antonacci, La NotteF E., J. Chromatogr. A 1216 (2009) [16] Q.G. Tian, M.M. Giusti, G.D. Stoner, S.J. Schwartz, J. Chromatogr. A 1091 (2005) 72. [17] A. Heier, W. Blaas, A. Dross, R. Wittkowski, Am. J. Enol. Vitic. 53 (2002) 78. [18] N. Castillo-Munoz, S. Gomez-Alonso, E. Garcia-Romero, M.V. Gomez, A.H. Velders, I. Hermosin-Gutierrez, J. Agric. Food Chem. 57 (2009) 209. [19] L. Farina, E. Boido, F. Carrau, G. Versini, E. Dellacassa, J. Agric. Food Chem. 53 (2005) [20] T. Becker, E. Grimm, M. Knoche, Aust. J. Grape Wine Res. 18 (2012) 109. [21] D.O. Adams, Am. J. Enol. Vitic. 57 (2006) 249. [22] B.F. Desimon, T. Hernandez, I. Estrella, Food Chem. 47 (1993) 47. [23] H.S. Brar, Z. Singh, E. Swinny, Sci. Hortic. Amsterdam 117 (2008) 349. [24] B.G. Coombe, Am. J. Enol. Vitic. 43 (1992) 101. [25] S.T. Lund, F.Y. Peng, T. Nayar, K.E. Reid, J. Schlosser, Plant Mol. Biol. 68 (2008) 301. [26] D. Rowan, M. Cao, K. Lin-Wang, J. Cooney, D. Jensen, P. Austin, M. Hunt, C. Norling, R. Hellens, R. Schaffer, New Phytol. 182 (2009) 102. [27] A. Bar-Akiva, R. Ovadia, I. Rogachev, C. Bar-Or, E. Bar, Z. Freiman, A. Nissim-Levi, N. Gollop, E. Lewinsohn, A. Aharoni, J. Exp. Bot. 61 (2010) 1393.