Difference in volatile profile between pericarp tissue and locular gel in tomato fruit

Journal of Integrative Agriculture 2016, 15(12): 2911–2920 Available online at www.sciencedirect.com

ScienceDirect

RESEARCH ARTICLE

Difference in volatile profile between pericarp tissue and locular gel in tomato fruit WANG Li-bin1, 2, Jinhe Bai2, YU Zhi-fang1 1

College of Food Science and Technology, Nanjing Agricultural University, Nanjing 210095, P.R.China

2

U.S. Horticultural Research Laboratory, Agricultural Research Service (ARS), United States Department of Agriculture (USDA), Ft. Pierce 34945, USA

Abstract Aroma, a complex mixture of volatile compounds, plays an important role in the perception and acceptability of tomato products by consumers. Numerous studies have reported volatile profiles in tomatoes based on measurement of the whole fruit or pericarp tissue, however, little is understood regarding the volatile compositions in the inner tissues. The objective of this study was to investigate the differences in volatile profile between pericarp tissue and locular gel in tomato fruit. Based on HS-SPME-GC-MS analysis, totally 42 volatile compounds were detected in FL 47 and Tasti-Lee tomato fruits. Regardless of cultivars, a substantial higher concentration of total volatile compounds was observed in pericarp than that in locular gel, associated with higher levels of aldehydes, hydrocarbons, and nitrogen compounds. Pericarp tissue possessed higher levels of cis-3-hexenal, hexanal, heptanal, octanal, nonanal, cymene, terpinolene, undecane, dodecane, 2-phenylethanol, 6-methyl-5-hepten-2-one, 2-methylbutyl acetate, 1-nitro-pentane, and 1-nitro-2-phenylethane, while the abundances of 2-methylpropanal, butanal, 2-methylbutanal, 2-methyl-2-butenal, 2-methylpropanol, 3-methylbutanol, 2-methylbutanol, and 2-butanone were higher in locular gel. Principal component analysis (PCA) and cluster analysis using GC-MS and electronic nose (E-nose) data discriminated the two tissues. Keywords: Solanum lycopersicum, tomato fruit, volatile profile, pericarp, locular gel

(Fig. 1). Numerous studies have reported volatile profiles

1. Introduction

in tomatoes based on measurement of the whole fruit or

Tomato (Solanum lycopersicum L.) fruit is consisted of different tissues, including pericarp, septa, columella, placenta, seeds and locular gel according to van de Poel et al. (2014)

et al. 2011b; Wang et al. 2015a, b). Furthermore, the usage

pericarp tissue (Maul et al. 2000; Bai et al. 2011; Baldwin of pericarp for physiological/biochemical analysis has the advantage in uniform sample preparation (Maul and Sargent 1998; Moretti et al. 1998). However, information on the volatile profile in other tissues such as locular gel is not well known. Previously, Maul

Received 7 December, 2015 Accepted 7 March, 2016 WANG Li-bin, E-mail: [email protected]; Correspondence YU Zhi-fang, Tel: +86-25-84399098, E-mail: [email protected]

and Sargent (1998) reported that Solimar tomato pericarp

© 2016, CAAS. All rights reserved. Published by Elsevier Ltd. doi: 10.1016/S2095-3119(15)61324-7

chromatography (GC) when compared to locular gel (442

(including columnella) produced an average 219% concentration of the 16 volatile compounds quantified by gas and 203 μL L–1, respectively); meanwhile, the abundances of

2912

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

Pericarp Columella & Septa

Locular gel Seeds

Fig. 1 Schematic cross-section of a tomato fruit showing two locules and the different tissues according to van de Poel et al. (2014).

methanol, ethanol, 1-penten-3-one, cis-3-hexenal, hexanal, trans-2-hexenal, trans-2-heptenal, cis-3-hexenol, 6-methyl-5hepten-2-one, and geranyl acetone were higher in pericarp, while locular gel possessed high levels of acetaldehyde, acetone, and β-ionone (Maul and Sargent 1998). Recently, 3-methylbutanal and 2-methylbutanal were proposed by Klee (2010) to be the important contributors to tomato aroma, it is needed to determine the concentrations of these volatiles in the different tissues; also, only one variety used for study could not represent the situation in other cultivars. The objective of this study was to investigate the difference in volatile profile between pericarp and locular gel in tomato fruits. Two major Florida cultivars, FL 47 and Tasti-Lee, at full-ripe stage were used for the research. The results will provide researchers and general public a tool to evaluate and compare volatile data by different sampling methods.

2. Materials and methods 2.1. Plant materials Uniform and defect-free FL 47 and Tasti-Lee tomato fruits at full-ripe stage (red color on entire fruit surface) (USDA 1997), 20 fruits per cultivar, with average a* value of 17.57 and 20.61, respectively, were purchased from a local grocery store on Feb 14, 2015. CIELAB was used for determining the color coordinates by using a colorimeter (model CR300, Minolta, Tokyo, Japan). The instrument was calibrated using a white tile and color a* values were recorded where the positive a* value indicates red color while the negative a* value represents green color. Fifteen fruits per cultivar were then selected and further divided into five biological replicates with three fruits per replicate. For sampling, pericarp and locular gel including seeds, were taken with a sharp stainless steel knife, immersed in liquid N2, fractured, and then stored at –80°C until analyzed. Other inner tissues were discarded.

2.2. Volatile analysis Volatile analysis was conducted by headspace, solid phase

micro-extraction, and gas chromatography-mass spectrometry system (HS-SPME-GC-MS), following the method of Wang et al. (2015c). Frozen pericarp tissue was ground to powder under liquid nitrogen and 4.3 g of powder, together with 1.7 mL of saturated CaCl2 solution were transferred to a 20-mL vial and sealed with Teflon-lined septa. For headspace analysis, the homogenized samples were incubated for 30 min at 40°C in a incubator, then they were exposed to the headspace for another 30 min at 40° after 2-cm SPME fiber (50/30 μm DVB/Carboxen/PDMS; Supelco, Bellefonte, PA) was inserted into the vial. After exposure, the SPME fiber was inserted into the injector of a GC-MS (Model 6890, Agilent, Santa Clara, CA) to desorb the extract for 15 min at 250°C. The GC-MS equipment and settings were: DB-5 (60 m length, 0.25 mm i.d., 1.00 μm film thickness; J&W Scientific, Folsom, CA) columns, coupled with a 5973 N MS detector (Agilent). The column oven was programmed to increase at 4°C min−1 from the initial 40 to 230°C, then ramped up at 100°C min−1 to 260°C and held for 11.70 min for a total run time of 60 min. Helium was used as carrier gas at flow rate of 1.5 mL min−1. Inlet, ionizing source and transfer line were kept at 250, 230, and 280°C, respectively. Mass units were monitored from 30 to 250 m/z and ionized at 70 eV. Data were collected using the ChemStation G1701 AA data system (Hewlett-Packard, Palo Alto, CA). A mixture of C-5 to C-18 n-alkanes was run at the beginning of each day to calculate retention indices (RIs). Volatile compounds were identified by comparison of their mass spectra with library entries (NIST/EPA/NIH Mass Spectral Library, ver. 2.0d; National Institute of Standards and Technology, Gaithersburg, MA), as well as by comparing RI with authentic standard aroma compounds purchased from Sigma-Aldrich (St. Louis, MO) or Fluka Chemical Corporation (Buchs, Switzerland). Quantification was conducted by using a peak size vs. concentration curve built by serially diluted five point standard solutions (Baldwin et al. 2009). Briefly, a standard compound was dissolved in pure methanol and the mixture was then introduced into a deodorized tomato homogenate. The range of concentrations in the standard curve for each compound covers the concentrations found in the samples.

2.3. Headspace electronic nose analysis Headspace electronic nose (E-nose) analysis was conducted according to the method of Wang et al. (2015b). For sample preparation, 2.15 g frozen pericarp tissue ground to powder under liquid nitrogen, together with 0.85 mL of saturated CaCl2 solution were transferred to a 10-mL vial and sealed with Teflon-lined septa before analysis. For E-nose analysis, a FOX 4000 system (Alpha MOS, Toulouse, France) was used, fitted with 18 metal oxide gas sensors (LY2/LG, LY2/G, LY2/AA, LY2/GH, LY2/gCTl, LY2/

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

gCT, T30/1, P10/1, P10/2, P40/1, T70/2, PA/2, P30/1, P40/2, P30/2, T40/2, T40/1, and TA/2), some with coated surfaces (Baldwin et al. 2012). The electrical output from the sensors was measured at 0.5 s intervals. Samples were incubated in an agitator at 500 r min–1 and 40°C for 2 min before the headspace sample (500 µL) was taken from the vial and injected into the E-nose. The carrier gas was pure air with a flow rate of 150 mL min−1. The E-nose data acquisition program was a 2-min sampling time followed by an 18-min delay between samples for sensor recovery.

2.4. Statistical analysis Data presented were the mean values of five biological replicates for volatile compounds. SAS ver. 9.3 (SAS Institute, Gary, NC) was used to analyze the data, using analysis of variance (PROC ANOVA) with multi-comparison correction. Mean separation was determined by Duncan’s multiple range test at the 0.05 level. Principal component analysis (PCA) and average linkage Cluster analysis of HS-SPMEGC-MS and E-nose data was performed on covariance matrix with JMP® 9.0 (SAS Institute).

3. Results 3.1. Difference in volatile profile between FL 47 and Tasti-Lee tomato fruits Based on HS-SPME-GC-MS analysis, FL 47 and Tasti-Lee tomato fruits together possessed 42 volatile compounds, belonging to 8 chemical classes, including 15 aldehydes, 7 hydrocarbons, 6 alcohols, 5 ketones, 3 oxygen-containing heterocyclic compounds, 3 esters, 2 nitrogen compounds, and 1 sulfur- and nitrogen-containing heterocyclic compound (Table 1). Aldehydes comprised the largest percentage of the total volatile concentration in two cultivars, followed by alcohols and ketones, and the top three compound classes constituted more than 95% of total volatile concentration (Table 1). 42 compounds were listed in Table 1 along with their retention indexes, classifications, odor descriptions, and odor thresholds in water. Of these, cis-3-hexenal was the most abundant volatile compounds in both pericarp and locular gel of both cultivars, accounting for more than 12% of total volatile concentration. Furthermore, neral could not be detected in pericarp and locular gel of FL 47 tomato fruit, while locular gel did not demonstrate nonanal. For Tasti-Lee tomato fruit, 2-methylpropanal, 2-methylpropanol, 4-methylpentanol, and methyl salicylate were not identified in pericarp, while locular gel did not possess 5-ethyl-2(5H)-furanone, 4-methylpentanol, terpinolene, undecane, neral, and

2913

nonanal (Table 1). Besides the composition, differences in the concentrations of some volatile compounds were observed between two cultivars as well. For pericarp, the concentrations of ketones and oxygen-containing heterocyclic compounds in Tasti-Lee tomato fruit were 37 and 120%, respectively, higher than those in FL 47 tomato fruit, associated with higher abundances of 3-methylbutanal (166%), 2-phenylacetaldehyde (43%), neral (as the concentrations of neral in pericarp of FL 47 tomato fruit could not be detected base on HS-SPME-GC-MS analysis, there was no definite value), 3-methylbutanol (115%), 6-methyl-5-hepten-2-one (111%), geranyl acetone (215%), 2-methylfuran (305%), and 5-ethyl-2(5H)-furanone (109%) (Table 1). On the other hand, the levels of alcohols, esters and nitrogen compounds in Tasti-Lee fruit were 55, 20, and 52%, respectively, of those detected in FL 47 fruit, and the concentrations of 2-methylpropanal, 2-methylbutanal, 2-methyl-2-butenal, heptanal, 1-nitro-pentane, 1-nitro-2-phenylethane, methyl salicylate, 2-methylbutyl acetate, butyl acetate, 2-ethylfuran, 2-butanone, 2-methylpropanol, 2-methylbutanol, 4-methylpentanol, 3-methylpentanol, 2-phenylethanol, α-pinene, and dodecane in Tasti-Lee fruit were 0, 44, 34, 73, 51, 52, 0, 23, 33, 43, 44, 0, 27, 0, 40, 50, 58, and 64%, respectively, of those detected in FL 47 fruits. For locular gel, the concentration of ketones in Tasti-Lee tomato fruit was 42% higher than that in FL 47 tomato fruit, in accompany with 22, 92, 193, 44, 68, and 247%, respectively, higher concentrations of butanal, 3-methylbutanal, 3-methylbutanol, acetone, 6-methyl-5-hepten-2-one, and 2-methylfuran than those in FL 47 fruits (Table 1). On the other hand, the concentrations of alcohols and nitrogen compounds in FL 47 fruit were 22 and 188%, respectively, higher than those in Tasti-Lee fruits association with higher abundances of 2-methylpropanal (270%), 2-methylbutanal (226%), 2-methyl-2-butenal (107%), 2-phenylacetaldehyde (117%), 2-methylpropanol (279%), 2-methylbutanol (152%), 4-methylpentanol (as the concentrations of 4-methylpentanol in locular gel of Tasti-Lee tomato fruit could not be detected based on HS-SPME-GC-MS analysis, there was no definite value), 3-methylpentanol (50%), 2-butanone (50%), 1-nitro-pentane (436%), and 1-nitro-2-phenylethane (169%) than those in FL 47 fruit (Table 1).

3.2. Difference in volatile profile between pericarp and locular gel of FL 47 and Tasti-Lee tomato fruits Although the volatile profiles were different between two cultivars as mentioned above, both showed the same differences in the concentrations of some compounds between pericarp and locular gel (Table 1). As shown in Table 1, the

2914

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

Table 1 Concentrations of volatile compounds detected in pericarp and locular gel of red FL 47 and Tasti-Lee tomato fruits Volatile compound Aldehydes 2-Methylpropanal Butanal 3-Methylbutanal3) 2-Methylbutanal3) 2-Methyl-2-butenal cis-3-Hexenal3) Hexanal3) trans-2-Hexenal3) Heptanal trans, trans-2,4Hexadienal Benzaldehyde Octanal 2-Phenylacetaldehyde3)

Retention index

Odor threshold in water (mg L–1)2)

Concentration (mg L–1) FL 47 Tasti-Lee Locular gel Pericarp Locular gel Pericarp

567 590 638 646 719 771 774 828 875 887

Pungent, malt, green 8.2–19 Pungent, green 0.009 Malt 0.00015–0.0002 Cocoa, almond, malt 0.003 Green, fruit 0.5 Leafy, green 0.00025 Grass, tallow, fat 0.0045–0.005 Green, leafy 0.017 Fat, citrus, rancid 0.003 Green 0.06

0.068 a 0.015 b 0.21 b 0.81 a 0.12 a 0.99 b 0.15 b 0.44 ab 0.0037 c 0.104 ab

0.016 b 0.0072 c 0.13 c 0.24 b 0.056 b 4.02 a 0.79 a 0.74 a 0.0075 a 1.306 a

0.018 b 0.018 a 0.41 a 0.25 b 0.056 b 0.51 b 0.11 b 0.28 b 0.0038 c 0.019 b

0c 0.0059 c 0.35 a 0.11 c 0.019 c 3.52 a 0.96 a 0.65 a 0.0055 b 0.393 ab

945 969

Almond, burnt sugar Fat, soap, lemon, green Hawthorne, honey, sweet Fat, citrus, green Lemon

0.35 0.0007

0.0043 ab 0.00046 b

0.0055 ab 0.00119 a

0.0038 b 0.00048 b

0.0063 a 0.00125 a

0.004

0.023 ab

0.018 b

0.010 c

0.025 a

0.001 0.03

0b 0b

0.0011 a 0b

0b 0b

0.0011 a 0.0069 a

Pine, turpentine Solvent, gasoline, citrus Lemon, orange Smokey, woody Alkane Alkane Alkane

0.006 0.15

0.00020 b 0.00020 b

0.00037 a 0.00041 a

0.00015 b 0.00017 b

0.00022 b 0.00031 a

0.01 0.2 10 10 –

0.061 b 0.000074 b 0.00076 b 0.0035 c 0.013 b

0.125 a 0.000320 a 0.00748 a 0.0145 a 0.031 a

0.049 b 0b 0b 0.0017 c 0.021 ab

0.100 ab 0.000231 a 0.00490 a 0.0093 b 0.026 a

Alcoholic, grassy, sweet Whiskey, malt, burnt Malt, wine, onion Pungent Pungent Honey, spice, rose, lilac

12.5

0.089 a

0.057 b

0.023 c

0d

0.25–0.3 0.25–0.3 0.82–4.1 0.83–4.1 1.0–1.1

0.22 c 0.58 a 0.020 b 0.0135 b 0.65 c

0.15 d 0.48 b 0.031 a 0.0196 a 1.41 a

0.65 a 0.23 c 0c 0.0090 c 0.37 c

0.33 b 0.13 d 0c 0.0079 c 0.71 b

0.79 b

0.67 b

1.14 a

0.71 b

7 0.0015 0.05 0.06

0.0166 a 0.012 a 0.027 c 0.038 b

0.0112 b 0.013 a 0.058 b 0.093 b

0.0110 b 0.020 a 0.045 b 0.043 b

0.0049 c 0.020 a 0.122 a 0.293 a

3.5–4.0 –

0.015 c 0.0044 b

0.016 c 0.0129 a

0.052 b 0.0028 b

0.065 a 0.0055 b

–

0.045 bc

0.058 b

0c

0.120 a

0.066 0.005–0.011 0.04

0.0011 b 0.00025 c 0.00184 ab

0.0035 a 0.00192 a 0.00260 a

0.0017 b 0.00015 c 0.00058 ab

0.0012 b 0.00044 b 0b

22

0.00313 b

0.00394 a

0.00058 d

0.00202 c

1 016

Nonanal Neral3) Hydrocarbons α-Pinene Cymene

1 059 1 180

Limonene Terpinolene Undecane Dodecane Tridecane Alcohols 2-Methylpropanol

998 1 048 1 051 1 137 1 222

3-Methylbutanol3) 2-Methylbutanol 4-Methylpentanol 3-Methylpentanol 2-Phenylethanol3)

707 711 809 817 1 084

Ketones Acetone

Odor description1)

910 994

612

533

Pungent, irritating, floral 2-Butanone 591 Sweet 665 Fruity, floral, green 1-Penten-3-one3) 950 Fruity, floral 6-Methyl-5-hepten-2-one3) 1 367 Sweet, floral, estery Geranyl acetone3) Oxygen-containing heterocyclic compounds 2-Methylfuran 602 Chocolate 2-Ethylfuran 676 Rum, coffee and chocolate 5-Ethyl-2(5H)-furanone 933 Caramellic Esters Butyl acetate 744 Pear 2-Methylbutyl acetate 847 Fruit 1 156 Peppermint Methyl salicylate3) Nitrogen compounds 1-Nitro-pentane 916 Pleasant, fruity

40

(Continued on next page)

2915

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

Table 1 (Continued from preceding page) Volatile compound

Retention index

Odor description1)

1 250 Flower, spice 1-Nitro-2-phenylethane3) Sulfur- and nitrogen-containing heterocyclic compounds 1 002 Tomato leafy, green 2-Isobutylthiazole3) Sum Aldehydes Hydrocarbons Alcohols Ketones Oxygen-containing heterocyclic compounds Esters Nitrogen compounds Sulfur- and nitrogen-containing heterocyclic compounds Total compounds

0.002

Concentration (mg L–1) FL 47 Tasti-Lee Locular gel Pericarp Locular gel Pericarp 0.0203 b 0.0377 a 0.0075 c 0.0198 b

0.0035

0.0049 a

0.0037 a

0.0035 a

0.0035 a

2.94 b 0.079 b 1.57 a 0.88 bc 0.064 b 0.0032 b 0.0234 b 0.0049 a 5.57 b

7.34 a 0.178 a 2.15 b 0.84 c 0.087 b 0.0080 a 0.0417 a 0.0037 a 10.65 a

1.70 b 0.071 b 1.29 c 1.26 a 0.055 b 0.0025 b 0.0081 c 0.0035 a 4.39 b

6.05 a 0.141 a 1.17 c 1.15 ab 0.191 a 0.0016 b 0.0218 b 0.0035 a 8.74 a

Odor threshold in water (mg L–1)2)

1)

Odor descriptions of 6-methyl-5-hepten-2-one and geranyl acetone adapted from Klee (2010), while others from Acree and Arn (2010). Odor threshold values in water adapted from van Gemert (2003). Key tomato aromatic volatile compounds recommended by Buttery (1993), Tandon et al. (2001), and Klee (2010). Data presented were the mean values of five biological replicates for volatile compounds. Mean values followed by the different letters within the same row showed significant difference using Duncan’s multiple range test (P<0.05). -, no data was reported.

2) 3)

concentrations of total volatile compounds in pericarp of FL 47 and Tasti-Lee tomato fruits were 91 and 99%, respectively, higher than those in locular gel. In association with this, pericarp possessed 249 and 355%, 225 and 199%, and 178 and 269%, respectively, higher levels of aldehydes, hydrocarbons, and nitrogen compounds in comparison with those in locular gel of FL 47 and Tasti-Lee tomato fruits. Furthermore, the abundances of cis-3-hexenal, hexanal, heptanal, octanal, nonanal, cymene, terpinolene, undecane, dodecane, 2-phenylethanol, 6-methyl-5-hepten-2-one, 2-methylbutyl acetate, 1-nitro-pentane, and 1-nitro-2-phenylethane in pericarps were higher than those in locular gel of FL 47 and Tasti-Lee tomato fruits, while locular gel possessed higher concentrations of butanal, 2-methylbutanal, 2-methyl-2-butenal, 2-methylpropanol, 3-methylbutanol, 2-methylbutanol, and 2-butanone (Table 1). Besides the common differences in the concentrations of some compounds between pericarp and locular gel in both varieties, the difference for some compounds seemed to be cultivar-dependent. For example, locular gel of FL 47 tomato fruit possessed higher level of alcohols in association with higher 3-methylbutanal; on the other hand, the concentrations of α-pinene, limonene, tridecane, 4-methylpentanol, 3-methylpentanol, 2-ethylfuran, and butyl acetate were higher in pericarp of FL 47 tomato fruit along with higher esters (Table 1). For Tasti-Lee tomato fruit, higher level of oxygen-containing heterocyclic compounds was observed in pericarp in accompany with higher contents of trans-2-hexenal, benzaldehyde, 2-phenylacetaldehyde, neral, geranyl acetone, 2-methylfuran and 5-ethyl-2(5H)-furanone, while acetone was higher in locular gel (Table 1).

3.3. Principal component analysis (PCA) and cluster analysis of E-nose data of pericarp and locular gel of FL 47 and Tasti-Lee tomato fruits E-nose, which acts as an alternative objective method to determine differences in volatile profiles, has been considered as a replacement for panelists in quality control for ease of analysis, reproducibility and convenience, although various factors such as room temperature, light, humidity and static electricity can affect the quality of data obtained (Baldwin et al. 2011a; Raithore et al. 2014). Because the basic underlying principle behind E-nose and human smell perception is similar, electronic nose with its eighteen sensors and data processing component, can discriminate one set of samples from another with different volatile profiles (Tan et al. 2001). As the eighteen sensors raw data obtained by E-nose analysis are multidimensional. In order to simplify the data and extract relevant information, a PCA was conducted based on covariance (Fig. 2). As a result, the data obtained from the eighteen components were simplified and the eighteen dimensional space made with the raw data was converted into a two dimensional space; and the first two principal components (PC) explained 96.7% of the data variability. As shown in Fig. 2, E-nose could discriminate volatile profile in pericarp of FL 47/Tasti-Lee tomato fruit from that in locular gel along with principal components 1 (PC1), which explained 75.4% of total variability. Furthermore, E-nose also could discriminate the volatile profile in pericarp/locular gel of FL 47 tomato fruit from that in pericarp/locular gel of Tasti-Lee tomato fruit along with PC2, which explained

2916

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

21.3% of total variability (Fig. 2). In order to clarify the relationship between GC-MS and E-nose results, PCA & cluster analysis were performed using 43 volatile data detected by HS-SPME-GC-MS. Fig. 3 shows the projection on the main first two principle components revealing in total 68.4% (40.9% on PC1 and 27.5% on PC2) of the total aroma variation. The score plot shows the positions of the replicate samples inside the reduced aroma space as defined by PC1 and PC2, with their mutual distances reflecting the differences in their volatility. As shown in Fig. 3, PC1 separated the pericarp of FL 47/ Tasti-Lee tomato fruit from locular gel, while PCA2 separated the pericarp/locular gel of FL 47 tomato fruit from pericarp/ locular gel of Tasti-Lee tomato fruit, which were in agreement with the E-nose results (Figs. 2 and 3).

2-phenylacetaldehyde, neral, 3-methylbutanol, 2-phenylethanol, 1-penten-3-one, 6-methyl-5-hepten-2-one, geranyl acetone, methyl salicylate, 1-nitro-2-phenylethane, and 2-isobutylthiazole (Table 1). They could be divided into four groups based on different derivations (Lewinsohn et al. 2005; Klee 2010; Klee and Giovannoni 2011; Ilg et al. 2014; Rambla et al. 2014): 1-penten-3-one, cis-3-hexenal, hexanal, and trans-2-hexenal are fatty acid derivatives; 6-methyl-5-hepten-2-one, geranyl acetone, and neral come from carotenoids; the branched chain amino acids are the precursors for 2-methylbutanal, 3-methylbutanal, 3-methylbutanol, and 2-isobutylthiazole; on the other hand, 2-phenylacetaldehyde, 2-phenylethanol, methyl salicylate, and 1-nitro-2-phenylethane are synthesized from phenylalanine. Similar to the study of Baldwin et al. (2015), the 3 compounds of 3-methylbutanal, 6-methyl-5-hepten-2-one and 3-methylbutanol were higher in pericarp and locular gel of Tasti-Lee tomato fruit when compared with those in FL 47. On the other hand, the levels of cis-3-hexenal and trans-2-hexenal were higher in pericarp and locular gel of FL 47 fruits, although they were not significant at P<0.05 level (Table 1). The genetic make-up plays a pivotal role in determining the volatile profile in tomato fruit (Wang et al. 2016). 152 S. lycopersicum heirloom varieties of different genetic back-

4. Discussion Of more than 400 volatile compounds identified in the ripening tomato fruit, only a small number are suggested to have contribution to tomato aroma (Klee 2010). According to the reports/studies of Buttery (1993), Tandon et al. (2001), and Klee (2010), 15 important compounds, proposed to have contribution to tomato aroma, were detected in this study, including 3-methylbutanal, 2-methylbutanal, cis-3-hexenal, hexanal, trans-2-hexenal,

B

A 8

FL 47 locular gel FL 47 pericarp Tasti-Lee locular gel Tasti-Lee pericarp

6

Component 2 (21.3%)

4 2 0 –2 –4 –6 –8 –8

–6

–4

–2

0

2

4

6

FL 47 locular gel FL 47 locular gel FL 47 locular gel FL 47 locular gel FL 47 locular gel FL 47 pericarp FL 47 pericarp FL 47 pericarp FL 47 pericarp FL 47 pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee locular gel Tasti-Lee locular gel Tasti-Lee locular gel Tasti-Lee locular gel Tasti-Lee locular gel

8

Component 1 (75.4%)

Fig. 2 Principal component analysis (PCA, A) and cluster analysis (B) of electronicnose nose (E-nose) data in pericarp and locular gel of FL 47 and Tasti-Lee tomato fruits.

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

B

A FL 47 locular gel FL 47 pericarp Tasti-Lee locular gel Tasti-Lee pericarp

8 6 4 Component 2 (27.5%)

2917

2 0 –2 –4 –6 –8 –8

–6

–4

–2

0

2

4

6

FL 47 locular gel FL 47 locular gel FL 47 locular gel FL 47 locular gel FL 47 locular gel Tasti-Lee locular gel Tasti-Lee locular gel Tasti-Lee locular gel Tasti-Lee locular gel Tasti-Lee locular gel FL 47 pericarp FL 47 pericarp FL 47 pericarp FL 47 pericarp FL 47 pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee pericarp Tasti-Lee pericarp

8

Component 1 (40.9%)

Fig. 3 Principal component analysis (PCA, A) and cluster analysis (B) of 42 aromatic volatile compounds detected by HS-SPMEGC-MS data in pericarp and locular gel of FL 47 and Tasti-Lee tomato fruits.

grounds demonstrated different volatile profile in association with high variability for some branched-chain and phenolic volatiles (Rambla et al. 2014). The difference in volatile profile between Tasti-Lee and FL 47 fruits might be due to different genotypes (Baldwin et al. 2015). Besides genotypes, the volatile content is affected by many other factors such as harvest maturity, environmental temperature, and cultivation conditions (Wang et al. 2016). In both studies mentioned above, some factor such as cultivation conditions was ignored. Thus, further investigation was needed to eliminate these issues. In spite of the difference in composition/concentration of the volatile compounds between two cultivars, both FL 47 and Tasti-Lee tomato fruits demonstrated the same trend on the difference in the volatile profile between pericarp and locular gel. In consistent with the result of Maul and Sargent (1998), total volatile concentrations in pericarp of two cultivars are higher than those in locular gel (Table 1). Besides, for the 15 important compounds higher abundances of cis-3-hexenal, hexanal, 6-methyl-5-hepten-2-one, 2-phenylethanol, and 1-nitro-2-phenylethane were observed in pericarp, while locular gel had higher concentrations of 2-methylbutanal and 3-methylbutanol (Table 1). These results suggested that if we sampled the whole fruit the abundances of 2-methylbutanal and 3-methylbutanol might be higher in comparison to these in pericarp, which was

confirmed by a previous study that the whole Beefsteak tomato fruit produced more 2-methylbutanal and 3-methylbutanol than that of pericarp (data not shown). Next we will mainly discuss the mechanisms underlying these differences according to precursor groups.

4.1. Fatty acid derivatives C6-aldehydes, including hexanal, trans-2-hexenal, and cis-3-hexenal are among the most abundant volatiles in tomato fruit. They provide the fruit with “green”, “grassy”, “tallow”, “fat” or “leafy” notes (Table 1). Maul et al. (2000) reported that hexanal was positively correlated with sweetness and ripe tomato ratings. In tomato fruit, they are synthesized from C-18 fatty acids, linoleic acid and linolenic acid, which are initially acted upon by TomloxC to produce 13-hydroperoxides (13-HPOs). Then 13-HPOs are subsequently cleaved by 13-hydroperoxide lyase (13-HPL), releasing C6-aldehydes, hexanal and cis-3-hexenal. The latter could further be converted into trans-2-hexenal, either non-enzymatically or by means of a cis-3, trans-2-enal isomerase. HPL is a key enzyme for C6-aldehydes synthesis in tomato fruit (Canoles et al. 2005; Wang et al. 2015c). In our study the abundances of cis-3-hexenal and hexanal were higher in tomato pericarp than those in locular gel, which was in consistent with results of Maul and Sargent (1998).

2918

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

4.2. Apocarotenoid volatiles Apocarotenoid volatiles are among the most important contributors to tomato aroma (Klee 2010). Three apocarotenoid volatiles were identified in this study, including neral, 6-methyl-5-hepten-2-one, and geranyl acetone, which were characterized as ‘lemon’, ‘sweet’, ‘fruity’, ‘estery’ or ‘floral’ (Table 1). In tomato fruit, they are directly synthesized from carotenoid oxidative cleavage by the action of carotenoid cleavage dioxygenases (CCD), encoded by LeCCD1A and LeCCD1B (Klee 2010; Ilg et al. 2014); and their productions correlate strongly with the levels of their direct precursor carotenoid compositions (Lewinsohn et al. 2005; Klee and Giovannoni 2011): 6-methyl-5-hepten-2-one directly comes from lycopene, ζ-carotenoid is the direct precursor for geranylactone, while neral is mainly derived from lycopene and its tetraterpene precursors (Lewinsohn et al. 2005; Klee 2010). Previously, Moretti et al. (1998) found that the abundances of carotenoids were lower in pericarp in comparison with that of the locular gel. Thus, higher abundance of 6-methyl-5-hepten-2-one in pericarp might be due to higher lycopene content. A similar result was found by Maul and Sargent (1998) that the concentrations of 6-methyl-5hepten-2-one and geranyl acetone were higher in Solimar tomato pericarp than those in locular gel.

and 1-nitro-2-phenylethane in pericarp might be due to higher transcriptional levels of LeAADC1A, LeAADC1B, and LeAADC2.

4.4. Branched-chain derivatives 2-Methylbutanal and 3-methylbutanol, described as ‘cocoa’, ‘almond’, ‘whiskey’, ‘malt’ or ‘burnt’ (Table 1), are synthesized from isoleucine and leucine, respectively. Although their bio-pathway is not well established, the first step for their biosynthesis is hypothesized to be catalyzed by branched-chain aminotransferases (BCATs), which remove the amino groups from the respective amino acids. Subsequently, there is a decarboxylation to produce the aldehydes and a reduction to form the alcohols (Klee 2010). Six BCAT genes have been identified in tomato fruit, but the specific isomer for their biosynthesis is still unknown. For branched-chain volatiles, much of the regulation of their production is suggested to occur downstream of precursor supply (Kochevenko et al. 2012). This suggests that higher levels of 2-methylbutanal and 3-methylbutanol in locular gel might be due to higher BCAT activity. In agreement with this, Yang et al. (2011) previously found that high expression level of BanBCAT was correlated well to higher productions of branched-chain volatiles in banana fruit.

4.3. Phenolic volatiles

5. Conclusion

Phenylpropanoids, including 2-phenylacetaldehyde, 2-phenylethanol, and 1-nitro-2-phenylethane, are important contributors to tomato aroma as well (Klee 2010), imparting fruit its “flower”, “honey”, “hawthorne”, “lilac”, “rose”, “spice” or “sweet” notes (Table 1). Furthermore, 2-phenylacetaldehyde and 2-phenylethanol could enhance tropical and fruity aromas when spiked in combination with sugar or sugar plus acid in a tomato puree, as well as overall aftertaste in combination with acid (Baldwin et al. 2008). They biosynthesis starts with the conversion of phenylalanine to phenylethylamine by the action of aromatic amino acid decarboxylases (AADCs), encoded by LeAADC1A, LeAADC1B, and LeAADC2 (Tieman et al. 2006). Phenylethylamine is then metabolized to 2-phenylacetonitrile or 1-nitro-2-phenylethane by a series of not fully characterized reactions or to 2-phenylacetaldehyde by an, as yet unidentified, amine oxidase. Finally, a small family of 2-phenylacetaldehyde reductases (PAR), encoded by LePAR1 and LePAR2, catalyze the reduction of 2-phenylacetaldehyde to 2-phenylethanol. The first and rate-limiting step for their biosynthesis is performed by AADCs (Klee 2010), which is regulated at the transcriptional level (Facchini et al. 2000). This suggests that the higher levels of 2-phenylethanol

This study provides strong evidence that pericarp possessed a higher abundance of total volatile compounds than that in locular gel in tomato fruit, associated with higher levels of aldehydes, hydrocarbons, nitrogen compounds. Both PCA and cluster analysis by using GC-MS and E-nose data could discriminate pericarp from locular gel samples.

Acknowledgements We wish to thank the financial support to this experiment from the Public Welfare Research Projects of the Ministry of Agriculture of China (2014030232).

References Acree T, Arn H. 2010. Flavornet and human odor space. Gas chromatography-olfactometry (GCO) of natural products. Cornell University, USA. [2015-12-03]. http://www.flavornet. org/flavornet.html Bai J H, Baldwin E A, Imahori Y, Kostenyuk I, Burns J, Brecht J K. 2011. Chilling and heating may regulate C6 volatile aroma production by different mechanisms in tomato (Solanum lycopersicum) fruit. Postharvest Biology and Technology, 60, 111–120.

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

Baldwin E A, Bai J H, Plotto A, Cameron R, Luzio G, Narciso J, Manthey J, Widmer W, Ford B L. 2012. Effect of extraction method on quality of orange juice: hand-squeezed, commercial-fresh squeezed and processed. Journal of the Science of Food and Agriculture, 92, 2029–2042. Baldwin E A, Bai J H, Plotto A, Dea S. 2011a. Electronic noses and tongues: Applications for the food and pharmaceutical Industries. Sensors, 11, 4744–4766. Baldwin E A, Goodner K, Plotto A. 2008. Interaction of volatiles, sugars, and acids on perception of tomato aroma and flavor descriptors. Journal of Food Science, 73, S294–S307. Baldwin E A, Plotto A, Manthey J, McCollum G, Bai J, Irey M. 2009. Effect of Liberibacter infection (Huanglongbing disease) of citrus on orange fruit physiology and fruit/fruit juice quality: Chemical and physical analyses. Journal of Agricultural and Food Chemistry, 58, 1247–1262. Baldwin E A, Plotto A, Narciso J, Bai J H. 2011b. Effect of 1-methylcyclopropene on tomato flavour components, shelf life and decay as influenced by harvest maturity and storage temperature. Journal of the Science of Food and Agriculture, 91, 969–980. Baldwin E A, Scott J W, Bai J H. 2015. Sensory and chemical flavor analyses of tomato genotypes grown in Florida during three different growing seasons in multiple years. Journal of the American Society for Horticultural Science, 140, 490–503. Buttery R. 1993. Quantitative and sensory aspects of flavor of tomato and other vegetables and fruits. In: Acree T, Teranishi R, eds., Flavor Science: Sensible Principles and Techniques. ACS, Washington, D.C. pp. 259–286. Canoles M, Soto M, Beaudry R. 2005. Hydroperoxide lyase activity necessary for normal aroma volatile biosynthesis of tomato fruit, impacting sensory perception and preference. HortScience, 40, 1130–1131. Facchini P J, Huber-Allanach K L, Tari L W. 2000. Plant aromatic L-amino acid decarboxylases: Evolution, biochemistry, regulation, and metabolic engineering applications. Phytochemistry, 54, 121–138. van Gemert L. 2003. Odour Thresholds-compilations of Odour Thresholds in Air, Water and Other Media. Oliemans Punter & Partners BV, Utrecht. Ilg A, Bruno M, Beyer P, Al-Babili S. 2014. Tomato carotenoid cleavage dioxygenases 1A and 1B: Relaxed double bond specificity leads to a plenitude of dialdehydes, monoapocarotenoids and isoprenoid volatiles. FEBS Open Bio, 4, 584–593. Klee H J. 2010. Improving the flavor of fresh fruits: Genomics, biochemistry, and biotechnology. New Phytologist, 187, 44–56. Klee H J, Giovannoni J J. 2011. Genetics and control of tomato fruit ripening and quality attributes. Annual Review of Genetics, 45, 41–59. Kochevenko A, Araújo W L, Maloney G S, Tieman D M, Do P T, Taylor M G, Klee H J, Fernie A R. 2012. Catabolism of branched chain amino acids supports respiration but

2919

not volatile synthesis in tomato fruits. Molecular Plant, 5, 366–375. Lewinsohn E, Sitrit Y, Bar E, Azulay Y, Meir A, Zamir D, Tadmor Y. 2005. Carotenoid pigmentation affects the volatile composition of tomato and watermelon fruits, as revealed by comparative genetic analyses. Journal of Agricultural and Food Chemistry, 53, 3142–3148. Maul F, Sargent S A. 1998. Aroma volatile profiles from ripe tomatoes are influenced by physiological maturity at harvest: An application for electronic nose technology. Journal of the American Society for Horticultural Science, 123, 1094–1101. Maul F, Sargent S A, Sims C, Baldwin E, Balaban M, Huber D. 2000. Tomato flavor and aroma quality as affected by storage temperature. Journal of Food Science, 65, 1228–1237. Moretti C L, Sargent S A, Huber D J, Calbo A G, Puschmann R. 1998. Chemical composition and physical properties of pericarp, locule, and placental tissues of tomatoes with internal bruising. Journal of the American Society for Horticultural Science, 123, 656–660. van de Poel B, Vandenzavel N, Smet C, Nicolay T, Bulens I, Mellidou I, Vandoninck S, Hertog M L, Derua R, Spaepen S. 2014. Tissue specific analysis reveals a differential organization and regulation of both ethylene biosynthesis and E8 during climacteric ripening of tomato. BMC Plant Biology, 14, 1–15. Raithore S, Dea S, Plotto A, Bai J H, Manthey J, Narciso J, Irey M, Baldwin E A. 2014. Effect of blending Huanglongbing (HLB) disease affected orange juice with juice from healthy orange on flavor quality. LWT-Food Science and Technology, 62, 868–874. Rambla J L, Tikunov Y M, Monforte A J, Bovy A G, Granell A. 2014. The expanded tomato fruit volatile landscape. Journal of Experimental Botany, 65, 4613–4623. Tan T T, Schmitt V, Lucas O, Isz S. 2001. Electronic noses and electronic tongues. LabPlus International, 13, 16–19. Tandon K S, Jordan M, Goodner K L, Baldwin E A. 2001. Characterization of fresh tomato aroma volatiles using GColfactometry. Proceedings of the Florida State Horticultural Society, 114, 142–144. Tieman D, Taylor M, Schauer N, Fernie A R, Hanson A D, Klee H J. 2006. Tomato aromatic amino acid decarboxylases participate in synthesis of the flavor volatiles 2-phenylethanol and 2-phenylacetaldehyde. Proceedings of the National Academy of Sciences of the United States of America, 103, 8287–8292. USDA (United States Department of Agriculture). 1997. United States standards for grades of fresh tomatoes. [201512-03]. http://www.hort.purdue.edu/prod_quality/quality/ tomatfrh.pdf Wang L B, Baldwin E A, Bai J H. 2016. Recent advance in aromatic volatile research in tomato fruit: The metabolisms and regulations. Food and Bioprocess Technology, 9, 203–216.

2920

WANG Li-bin et al. Journal of Integrative Agriculture 2016, 15(12): 2911–2920

Wang L B, Baldwin E A, Plotto A, Luo W Q, Raithore S, Yu Z F, Bai J H. 2015a. Effect of methyl salicylate and methyl jasmonate pre-treatment on the volatile profile in tomato fruit subjected to chilling temperature. Postharvest Biology and Technology, 108, 28–38. Wang L B, Baldwin E A, Yu Z F, Bai J H. 2015b. The impact of kitchen and food service preparation practices on the volatile aroma profile in ripe tomatoes: Effects of refrigeration and blanching. Hortscience, 50, 1358–1364. Wang L B, Baldwin E A, Zhao W, Plotto A, Sun X X, Wang Z,

Brecht J K, Bai J H, Yu Z F. 2015c. Suppression of volatile production in tomato fruit exposed to chilling temperature and alleviation of chilling injury by a pre-chilling heat treatment. LWT-Food Science and Technology, 62, 115–121. Yang X T, Song J, Fillmore S, Pang X Q, Zhang Z Q. 2011. Effect of high temperature on color, chlorophyll fluorescence and volatile biosynthesis in green-ripe banana fruit. Postharvest Biology and Technology, 62, 246–257. (Managing editor WENG Ling-yun)