Journal of the Science of Food and Agriculture
J Sci Food Agric 84:1333–1340 (online: 2004) DOI: 10.1002/jsfa.1828
Volatile metabolite profiling to discriminate diseases of McIntosh apple inoculated with fungal pathogens A Vikram, B Prithiviraj, H Hamzehzarghani and AC Kushalappa∗ Plant Science Department, McGill University, Ste-Anne-de-Bellevue, Quebec, Canada H9X 3V9
Abstract: Gas chromatography–mass spectrometry (GC/MS) technology was used to profile volatile metabolites from the headspace gas of McIntosh apple (Malus domestica Borkh.), which were noninoculated or inoculated with four different fungi, Botrytis cinerea Pers, Penicillium expansum Link, Mucor piriformis Fischer and Monilinia sp. The study yielded a total of 498 different volatile metabolites. Among them only 35 occurred relatively consistently in six replicates over three incubation periods. Of the consistent metabolites, 20 were specific to one or more inoculation agents/diseases, including seven that were unique to apples inoculated with different pathogens. Fluoroethene and 3,4-dimethyl-1-hexene were specific for Penicillium, while butanoic acid butyl ester, 4-methyl-1-hexene and 2-methyltetrazole were specific for Mucor. Similarly, acetic acid methyl ester and fluoroethane were specific to Botrytis and Monilinia, respectively. The method developed in this study can be used by storage managers to detect apple diseases at an early stage of disease progress and use this to manage apple diseases in storage, after further validation under commercial conditions. 2004 Society of Chemical Industry
Keywords: volatile compounds; metabolomics; disease diagnosis; post-harvest pathogens; electronic nose; disease detection
INTRODUCTION Apple is one of the major fruit crops cultivated throughout the world. China is the largest producer of apples, accounting for nearly half of the world production followed by the USA. Canada produces 514 333 metric tonnes of apple, the main appleproducing provinces being Quebec, Ontario, British Columbia, Nova Scotia and New Brunswick.1 Apples are often stored for up to 10 months after harvest and are subject to attack by numerous pathogens. Post-harvest diseases are a major problem in stored apples and the diseases are difficult to detect, as sporulating structures are often not produced at lowtemperature and controlled-atmosphere storage. Some of the important post-harvest pathogens of apples are Penicillium expansum Link, Botrytis cinerea Pers and Mucor piriformis Fischer.2 P. expansum causes blue mold rot, in which soft watery brown spots are seen which undergo rapid enlargement at temperatures between 20 and 25 ◦ C. B. cinerea is responsible for gray mold rot and the lesions are dry and firm at first, becoming soft as the rot advances. Under humid conditions production of gray brown conidia are observed. M. piriformis causes mucor rot, where dark brown lesions with a pale brown margin develop
on the fruit surface, especially at the stem and flower end. At an advanced stage the flesh becomes soft watery but odorless. The rate of disease development in stored apples depends mainly on the initial amount of invisible infection already acquired in the field and storage environmental conditions favoring lesion expansion. Controlled atmosphere is one of the major breakthroughs in mechanical cooling methods of fruit and vegetable storage. It involves monitoring and controlling oxygen and carbon dioxide together with temperature and relative humidity so as to provide optimal conditions to maintain product quality.3,4 It has proved to be very effective in extending the storage life of apples in particular.5 However, since the storage room is closed for a long period of time there is no way to detect diseases by visual observation. Thus there is a need to develop sensitive, rapid and cost effective methods to detect diseases and to identify pathogens, so that such a knowledge base can help reduce unexpected storage losses. The interest in apple volatiles began in the early part of last century, when a number of compounds, primarily acetaldehyde and esters of formic, acetic and hexanoic acid, were identified.6 Later Meigh7
∗
Correspondence to: AC Kushalappa, Plant Science Department, McGill University, Ste-Anne-de-Bellevue, Quebec, Canada H9X 3V9 E-mail:
[email protected] Contract/grant sponsor: Natural Sciences and Engineering Research Council of Canada (Received 18 July 2003; revised version received 31 December 2003; accepted 26 January 2004) Published online 21 June 2004
2004 Society of Chemical Industry. J Sci Food Agric 0022–5142/2004/$30.00
1333
A Vikram et al
identified six alcohols, five aldehydes, three ketones and six esterified acids in intact apples. Over 300 compounds contributing to flavor and aroma in different cultivars of apple have now been reported.8 However, no work has been published on volatiles produced by diseased apples. Volatile production by many other fruits and vegetables has been extensively studied with a view to detecting disease occurrence in order to reduce losses in storage. Volatile metabolites produced by diseased potato, onion, citrus, raspberry, peach and other crops have been studied using gas chromatography and gas chromatography–mass spectrometry.9 – 15 Many compounds were found to be disease-specific, eg potatoes infected with Phytophthora infestans produced butanal, 3-methyl butanal, undecane and verbenone, while those infected with Fusarium coeruleum produced 2-pentyl furan and capaene.16 However, compounds produced were often inconsistent. A modeling approach has been taken to discriminate diseases despite such variations, and neural network modeling of GC retention time data enabled discrimination of four potato diseases.10 However, the use of GC retention time does not give the identity of compounds, and such models must be backed up by compound identity for stability. Therefore a study was carried out to analyze volatiles of apple cv McIntosh inoculated with some major fungal pathogens using GC/MS technology with the aim of developing a disease warning system. EXPERIMENTAL Preparation of apple for inoculation Apples of cultivar McIntosh harvested from a single orchard were obtained from the Stevenson orchard (Quebec). Apparently disease-free apples were selected, surface-sterilized with 0.5% sodium hypochlorite solution for 3 min, rinsed with sterile water and dried under a laminar flow hood. Five wounds, 4 mm in diameter and 3 mm deep, were made around the equatorial region with a cork borer. Pathogen cultures Four fungal pathogens were used in this study. The cultures of Botrytis cinerea (isolate no. B27), Penicillium expansum (isolate No. 1790), Mucor piriformis (isolate No. 563) and Monilinia sp (isolate No. 1683) were obtained from the Agriculture and Agri-Food Canada, Summerland, British Columbia. They were subcultured on potato dextrose agar at 22 ◦ C for 7 days and were stored at 4 ◦ C. Subcultures from these, 7–10 days old, were used for inoculation. Spore suspensions were prepared by flooding the plates with 0.05% Tween 80 in distilled water and filtering through two layers of cheese cloth. The spore concentration in the suspension was adjusted to 105 ml−1 conidia using a hemocytometer. Inoculation and incubation The apples were inoculated with different pathogens by placing 20 µl of fungal spore suspension into each 1334
wound using a micro pipette and placed singly in wide-mouthed, 1 l, glass Mason jars. An aliquot of 10 ml of sterile distilled water was added to provide a saturated atmosphere to facilitate initial establishment of the infection. A stainless steel support was used to suspend the apple above the water level. The apples in the bottles were incubated at 20 ◦ C for disease development. After 24 h, the bottle air was continuously flushed with laboratory air, which passed through a flask containing water to provide air with high relative humidity, at the rate of 0.04 m3 h−1 .
Volatile accumulation and GC/MS analysis The apples were removed from Mason jars at 2, 4 and 6 days after inoculation (DAI) and placed in 1 l widemouthed glass bottles, cap-lined with Teflon-coated septa, flushed with pure dry air and incubated at 20 ◦ C for 30 min for volatile accumulation. The headspace gas was sampled and the volatiles were analyzed using HAPSITE (INFICON, Syracuse, NY, USA). The HAPSITE was programmed to sample headspace gas for 30 s at the rate of 100 ml min−1 . Following sampling the apples were returned to Mason jars and air with high humidity was flushed continuously. The entire experiment was conducted six times. HAPSITE is a GC/MS system equipped with a hand-held sniffing probe attached to a terminal stainless steel needle, 21 gauge and 10 cm long, which could be inserted into a septum bottle. The headspace air was preconcentrated in a tube trap containing 15 mg carboxen and the sample was then thermally desorbed at 225 ◦ C. The gas passed through a GC with a capillary column SPB-5 (30 m, 0.32 mm internal diameter with 1.0 µm film coating, and configured in a heating coil, INFICON-930-489G8). The carrier gas used was volatile organic-free nitrogen at a flow rate of 3 ml min−1 . The temperature of the column was maintained at 50 ◦ C for 4 min followed by a ramping of 3 ◦ C min−1 for 26 min and 14.4 ◦ C min−1 for 5 min. The GC was interphased to an MS equipped with a quadrupole analyzer. The mass spectrum was scanned at the rate of 1.04 s per mass decade over a mass range 46–300 m/z. The NIST mass spectral search program (version 2.0) was used for tentative identification of the compounds detected. The amounts of volatiles were expressed as mass–ion abundance (relative response of quadrupole detector). A mixture of bromopentafluorobenzene and 1,3,5 Tris (trifluoromethyl) benzene, programmed to inject a known amount automatically from a canister (INFICON, Syracuse, NY, USA), was used as an internal standard to calibrate the GC/MS.
Disease severity assessment The diameter of diseased apple tissue was measured 2, 4 and 6 DAI. In addition, at 6 DAI the apple was cut and the depth of diseased tissue was measured, from which the volume of diseased tissue was calculated. J Sci Food Agric 84:1333–1340 (online: 2004)
Apple volatile metabolites for disease discrimination
Experimental design and data analysis The experiment was designed as factorial, with six main factors of inoculation agents/pathogens: nonwounded–inoculated with sterile water = N-control; wounded–inoculated with sterile water = W-control; inoculated with pathogens Botrytis, Penicillium, Mucor and Monilinia and three sub-factors of incubation times, namely 2, 4 and 6 DAI. The entire experiment was conducted six times. The data output consisted of compounds and their abundance of mass ions. The metabolites which occurred in two or more replicates were used to calculate the frequency of metabolite occurrence (out of six replicates) and average abundance of metabolites for different treatments. The compounds identified were also grouped according to their chemical functional group, for which average abundance was calculated, which was divided by the average total abundance and multiplied by 100 to derive normalized or percent age abundance. The compounds occurring in one or more inoculation-agents, but not in all six, were sorted and designated as inoculation-agent or disease discriminatory metabolites. Statistical analysis The proportions of abundance of different disease discriminatory compounds for different inoculation treatments were subjected to factor analysis using principal component method by means of procedure FACTOR in SAS (version 8.02, SAS Institute Inc., Cary, NC, USA). A set of variables that are linear combinations of original proportions of abundance of metabolites was produced. The new variables were independent of each other and ranked according to the amount of variance. After initial factor extraction an orthogonal varimax rotation was used to estimate the factor loadings. Factor scores were calculated and projected for all treatments on the plane of factor 1 and 2. The first five factors associated with the inoculationagent discriminatory metabolites were used to describe the different inoculation-agents/pathogens. A scatter plot was developed based only on factors 1–3.
RESULTS Disease progress Visible rotting symptoms were observed in apples inoculated with Penicillium and Botrytis at 2 DAI which increased over the days of incubation. The lesions from Botrytis and Penicillium were 1.6 and 1.1 cm in diameter, respectively, at 2 DAI, which increased to 2.8 and 1.9 cm at 4 DAI, and the lesions coalesced at 6 DAI. The inoculation of Mucor showed brownish lesions 0.4–0.8 cm in diameter, whereas Monilinia produced only a slight discoloration around the inoculation site at 6 DAI. The non-wounded control remained without any disease throughout the 6 day period, while the wounded control showed slight browning at the inoculation site, but no disease. The range of volumes of disease observed was: J Sci Food Agric 84:1333–1340 (online: 2004)
2.4–28.5 cm3 for Botrytis; 1.1–17 cm3 for Penicillium; and 0.4–4 cm3 for Mucor. Volatile metabolic profile The headspace gas analyses of apples of cv McIntosh, sterile water-inoculated controls and inoculated with pathogens, yielded a total of 498 different volatile metabolites. The abundance of metabolites varied from 75 × 103 to 8.8 × 109 . Most of the volatiles were eluted within 20 min and 99% within 25 min. The number and abundance of compounds varied with treatments, incubation times and replicates. The incubation time (2, 4 and 6 DAI) did not show a particular general trend in volatile production although there were some increasing or decreasing trends in abundance of certain compounds within a disease, the occurrence of a compound was inconsistent among replicates (data not shown). Thus the volatile compounds produced at 2, 4 and 6 DAI were combined and were classified into different chemical groups. The average abundances and frequencies of occurrence of metabolites which belonged to different chemical groups are presented in Table 1. The highest number of volatile compounds, 147, was detected from wounded control treatment. The other treatments, Botrytis, non-wounded control, Penicillium, Mucor and Monilinia yielded 134, 127, 111, 108 and 100 volatile compounds, respectively. In general, all the treatments recorded the highest number of esters. The maximum numbers of alcohols, aromatics, aliphatics and esters were detected in wounded control, while the largest numbers of sulfurs and alkenes were found in Botrytis-inoculated ones. The largest number of heterocyclic compounds was found in the non-wounded control. The normalized abundance of chemical groups of compounds varied among the inoculationagents/diseases (Table 1). In all the treatments the major fraction of total volatiles came from esters. Higher proportions of aliphatics (28.3%) and alkenes (4.9%), were detected in Botrytis, while higher proportions of esters (65.7%) and alcohols (7.8%) were found in Monilinia-inoculated apples. Non-wounded control treatment recorded the normalized abundance of aromatics (10.7%) and heterocyclic (9.9%) compounds in proportions higher than the other treatments. Inoculation-agent/disease-specific volatile metabolites The volatile metabolites which occurred at least in three out of 18 experiments, six replications over three incubation periods, are presented in Table 2. A total of 35 volatile compounds were relatively consistent. The compound with the highest abundance (45 × 109 ) was butanoic acid butyl ester and was found in the non-wounded control. The frequency of occurrence of different compounds varied among inoculation agents. The disease-specific compounds that occurred most frequently were 2-methyltetrazole, butanoic acid butyl ester, butyl 2-methylbutanoate, 1335
A Vikram et al Table 1. Number and normalized abundance (%) of different chemical groups of compoundsa detected in McIntosh apples inoculatedb with different agents/pathogens
Chemical group
N-control
W-control
Botrytis
Mucor
Monilinia
Penicillium
Alcohol Aliphatic Alkene Aromatic Ester Heterocycle Sulfur Other Total
9 (2.5) 23 (11.8) 5 (1.8) 22 (10.7) 33 (63.2) 22 (9.9) 3 (T) 10 (0.1) 127 (315)
14 (5.9) 30 (27.4) 7 (4.4) 28 (0.2) 40 (55.4) 18 (6.5) 3 (T) 7 (T) 147 (105)
11 (5.5) 19 (28.3) 8 (4.9) 19 (7.0) 37 (51.9) 13 (1.5) 11 (T) 16 (0.7) 134 (210)
7 (3.6) 23 (24.1) 6 (0.5) 11 (2.5) 35 (60.7) 12 (7.8) 5 (T) 9 (0.8) 108 (278)
10 (7.8) 24 (21.4) 6 (1.1) 9 (T) 28 (65.7) 10 (3.9) 3 (T) 10 (0.1) 100 (213)
9 (4.4) 17 (21.9) 8 (1.2) 14 (0.2) 31 (64.8) 13 (1.5) 6 (T) 13 (5.9) 111 (206)
a Normalized abundance = (average abundance/total abundance for all compounds)100 − the values in parenthesis; T = trace < 105 ; total = total number of relatively consistent compounds detected and in parenthesis are the total mass ion abundance (quadrupole detector output) = ×108 , based on average per replicate × incubation time. b N-control = non-wounded water-inoculated control; W-control = wounded water-inoculated control.
2-butenoic acid ethyl ester, propanoic acid hexyl ester, heptanoic acid ethyl ester, 2-methylbutanoic acid methyl ester and thiirane. Although compounds like 1-butanol, 1-hexanol, butanoic acid propyl ester, ethyl acetate, hexanoic acid ethyl ester, hexanoic acid methyl ester and isobutyric acid ethyl ester were not specific to a given inoculation agent, the pathogeninoculated treatments produced two to eight times higher abundances than the control (Table 2). Out of the 35 most frequent compounds, only 20 were specific to one or more inoculation agents, including seven that were unique to single disease, which were not found in the non-wounded or wounded controls. Mucor produced 4-methyl-1-hexene, 2methyltetrazole and butanoic acid butyl ester, while Penicillium produced fluoroethene and 3,4-dimethyl1-hexene. Acetic acid methyl ester and fluoroethane were specific for Botrytis and Monilinia, respectively. 1-Hexene was detected only in wounded control and Mucor, whereas acetic acid hexyl ester, butyl 2methylbutanoate and hexanoic acid butyl ester were identified in the non-wounded and wounded controls. Propanoic acid hexyl ester and 1-octene-3-methoxy methoxy were found only in the non-wounded and wounded controls and Monilinia, while 2-butenoic acid ethyl ester was detected in the wounded control, Mucor and Penicillium and propanoic acid methyl ester in Botrytis, Mucor and Penicillium. Disease discrimination based on proportion of abundance Factor analysis, using the principle component method, based on proportion of total abundance of 20 inoculation-agent/disease discriminatory metabolites (observations 1–20 from Table 2), discriminated all six different inoculation-agents/diseases (Fig 1). Five factors explained 100% of the total variance, with a maximum of 38% by factor 1. The factor scores varied for different inoculation-agents/diseases (scores for factors 1–3 shown in Fig 1): factor 1 for Wcontrol; factor 2 for Botrytis; factor 3 for Penicillium; factor 4 for Mucor, and factor 5 for Monilinia. Factors 1336
1–3 loaded low for N-control, Monilinia and Mucor, while factors 4 and 5 loaded low for N-control, Wcontrol, Botrytis and Penicillium. The inspection of eigenvector loadings for 20 compounds (Table 3) showed that the production of compounds like 2-methyltetrazole, 4-methyl-1-hexene, fluoroethene, hexanoic acid butyl ester, 2-butenoic acid ethyl ester, thiirane and chloroform are governed by factor 1 (Wcontrol), while factor 2 (Botrytis) explains another aspect of the pathosystem, the production of acetic acid hexyl ester, propanoic acid hexyl ester, propanoic acid methyl ester and hexanoic acid propyl ester. Similary, the production of fluoroethane and 1-hexene are dependent on factor 3 (Penicillium), while acetic acid methyl ester, butyl 2-methylbutanoate and 2methylbutanoic acid methyl ester are dependent on factor 4 (Mucor) and 3,4-dimethyl-1-hexene on factor 5 (mainly Monilinia). Interestingly the differences between non-wounded control, wounded control and pathogen inoculated as a whole are quite large. The apples inoculated with Botrytis can be differentiated from those inoculated with other pathogens by increased production of acetic acid hexyl ester, propanoic acid methyl ester and hexanoic acid propyl ester (Table 3).
DISCUSSION Headspace volatile metabolites of McIntosh apple, non-inoculated or inoculated with pathogens, were used to discriminate six inoculation agents (diseases and control) using GC/MS technology. This is the first systematic study where volatiles produced by diseased apples have been identified. All the inoculation agents studied here can be discriminated qualitatively based on the unique volatile compounds detected and also quantitatively based on factor scores for proportion of abundance values of certain combinations of metabolites. In our studies all the pathogen inoculated apples produced at least one unique metabolite, whereas Mucor and Penicillium produced three and two compounds, respectively. In addition, there were several metabolites that were specific to two to J Sci Food Agric 84:1333–1340 (online: 2004)
J Sci Food Agric 84:1333–1340 (online: 2004)
07:25 03:18 02:13 01:19 13:17 01:05 01:07 07:37 14:09 15:30 22:43 21:54 06:22 18:36 01:54 18:23 04:15 18:09 01:25 01:45 02:10 03:20 07:35 07:37 06:42 04:58 03:06 08:38 01:05 01:45 13:24 09:56 08:51 03:45 02:50
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Alkene Alkene Heterocycle Ester Ester Aliphatic Aliphatic Alkene Ester Ester Ester Alkene Ester Ester Ester Ester Ester Ester Sulfur Aliphatic Alcohol Alcohol Alcohol Heterocycle Ester Ester Ester Ester Alcohol Ester Ester Ester Ester Ester Ester
Chem-Group
3,4-Dimethyl-1-hexene 4-Methyl-1-hexene 2-Methyltetrazole Acetic acid methyl ester Butanoic acid butyl ester Fluoroethane Fluoroethene 1-Hexene Acetic acid hexyl ester Butyl 2-methylbutanoate Hexanoic acid butyl ester 1-Octene, 3-(methoxymethoxy)2-Butenoic acid ethyl ester Propanoic acid hexyl ester Propanoic acid methyl ester Heptanoic acid ethyl ester 2-Methylbutanoic acid methyl ester Hexanoic acid propyl ester Thiirane Chloroform 1-Butanol 2-Methyl-1-butanol 1-Hexanol 2H-Pyran-2-one, tetrahydro-3,6-dimethyl 2-Methylbutanoic acid ethyl ester Butanoic acid ethyl ester Butanoic acid methyl ester Butanoic acid propyl ester Ethanol Ethyl acetate Hexanoic acid ethyl ester Hexanoic acid methyl ester Pentanoic acid ethyl ester Isobutyric acid, ethyl ester Propanoic acid ethyl ester
Compounds
1 (3) 8 (15) 61 (15) 55 (9) 16 (5) 122 (18) 174 (18) 42 (18) 242 (7) 1 (5) 14 (16) 50 (18) 31 (6) 126 (12) 28 (9) 88 (18)
10 (3) 230 (7)
26 (8)
1 (13) 36 (10) 4 (3) 18 (3)
N-control
8 (3) 1 (4) 1 (4) 3 (15) 2 (13) 98 (10) 18 (4) 93 (18) 72 (18) 36 (17) 2 (9) 2 (4) 39 (10) 110 (18) 34 (5) 50 (7) 30 (6) 82 (18)
27 (3)
16 (3) 1 (9) 1 (5) 1 (3) 7 (3) 1 (5) 4 (4)
W-control
1 (9) 1 (3) 64 (17) 2 (13) 78 (8) 2 (6) 162 (18) 128 (18) 69 (18) 267 (5) 2 (6) 35 (17) 115 (18) 46 (13) 1 (5) 29 (11) 8 (18)
15 (11)
8 (8)
1 (7)
Botrytis
1 (4) 44 (5) 14 (4) 58 (3) 1 (5) 1 (3) 55 (14) 1 (13) 103 (10) 5 (4) 175 (18) 109 (18) 39 (16) 318 (4) 1 (8) 26 (16) 153 (18) 36 (9) 1 (13) 55 (13) 3 (17)
1 (5)
5 (3)
124 (4)
1 (3) 57 (4)
Mucor
75 (3) 1 (6) 66 (3) 1 (7) 1 (3) 5 (14) 81 (15) 305 (10) 26 (7) 71 (18) 121 (18) 28 (18) 124 (4) 1 (4) 22 (16) 240 (18) 37 (7) 116 (12) 1 (12) 7 (18)
20 (3)
40 (4)
1 (3)
Monilinia
8 (16) 1 (13) 78 (10) 14 (5) 161 (18) 75 (18) 67 (18) 2 (3) 2 (4) 42 (12) 112 (18) 69 (8) 132 (5) 77 (14) 9 (18)
1 (5) 53 (4) 11 (9) 18 (3) 1 (7)
1 (7)
1 (3)
3 (3)
Penicillium
P U U B U M P WU NW NW NW NWM WUP NWM BUP WUMP NBUMP NWUMP WBUMP NWBUM NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP NWBUMP
Agentb
b
a
Average abundance = average of 18 replicates × incubation periods; abundance ×106 . Agent: N-control = non-wounded control; W-control = wounded control; B = Botrytis; U = Mucor; M = Monilinia; P = Penicillium. The numbers outside the brackets are abundance and those inside the brackets are frequency of compounds. Some discriminatory compounds found were detected once or twice in other groups and their frequencies are not included in the above list.
RT(m:s)
OBS
Table 2. Volatile metabolites, abundancea and frequency, detected in McIntosh apples not inoculated and inoculated with four different fungal pathogens
Apple volatile metabolites for disease discrimination
1337
A Vikram et al Factor3
P
1.45 M
0.57
W
B
-0.31 1.85 0.83 -1.19 1.96
U
-0.19 Factor1
N
1.03 Factor2
-1.20
0.10 -0.83
Figure 1. Projections of factor scores, in the space of factors 1–3, from factor analysis (using principal component method) of the proportion of abundance values (based on mass–ion abundances of compounds) for McIntosh apples inoculated with different agents/pathogens. N = non-wounded water-inoculated control; W = wounded water-inoculated control; B = inoculated with Botrytis; M = Monilinia; U = Mucor; P = Penicillium. Table 3. Eigenvector loadings and eigenvalues of principal components derived from proportion of abundance of inoculation/disease discriminatory metabolites
Metabolites
Factor
Factor 2
Factor 3
Factor 4
Factor 5
1. 3,4-Dimethyl-1-hexene 2. 4-Methyl-1-hexene 3. 2-Methyltetrazole 4. Acetic acid methyl ester 5. Butanoic acid butyl ester 6. Fluoroethane 7. Fluoroethene 8. 1-Hexene 9. Acetic acid hexyl ester 10. Butyl 2-methylbutanoate 11. Hexanoic acid butyl ester 12. 3-(Methoxymethoxy)-1-Octene 13. 2-Butenoic acid ethyl ester 14. Propanoic acid hexyl ester 15. Propanoic acid methyl ester 16. Heptanoic acid ethyl ester 17. 2-Methylbutanoic acid methyl ester 18. Hexanoic acid propyl ester 19. Thiirane 20. Chloroform Eigenvalues Variance Cumulative variance explained (%)b
0.11 0.89a 0.98a −0.19 −0.68 0.20 0.87a 0.01 −0.01 −0.09 0.85a 0.68 0.98a −0.20 −0.23 −0.42 −0.15 −0.11 0.97a 0.97a 7.60 0.38 0.38
0.32 −0.25 −0.04 0.08 −0.19 0.08 0.11 −0.29 0.97a 0.24 −0.20 0.33 −0.10 0.82a 0.95a −0.68 0.27 0.91a −0.09 −0.10 5.47 0.27 0.65
0.41 −0.02 0.06 0.22 −0.58 0.96a −0.48 0.92a −0.22 −0.61 −0.44 −0.62 0.17 0.01 0.16 −0.22 0.60 −0.23 0.21 0.21 4.48 0.23 0.88
−0.01 −0.25 −0.20 0.95a −0.41 0.06 −0.04 0.24 −0.02 0.74 −0.18 0.21 −0.03 0.54 −0.07 −0.56 0.73 0.33 0.01 −0.01 1.94 0.10 0.98
0.85a 0.30 0.02 −0.02 −0.10 0.18 −0.07 0.12 0.07 −0.14 −0.09 −0.09 0.02 0.05 0.14 −0.07 0.09 0.03 0.03 0.03 0.49 0.02 1.00
a b
The high levels of eigenvector loadings of the compounds on the corresponding factor. Cumulative percentage of total variance explained by factors.
five inoculation agents. They can also be used for discriminating diseases. The differences (factor scores) among non-wounded control, wounded control and pathogen inoculated as a whole are quite large, meaning the occurrence of diseases can be easily detected. Many volatile compounds have been reported from apples. Fifty-six volatile compounds have been detected in Delicious type apples.17 More than 300 1338
compounds contributing to apple flavor and aroma from different cultivars have been reported.8 In our study the sterile water-inoculated apples produced 147 volatiles. In pathogen-inoculated apples, however, the number of compounds was relatively low, but some were distinctly different. In addition to qualitative discrimination based on volatile types, the normalized abundance of functional groups of compounds could also be used to J Sci Food Agric 84:1333–1340 (online: 2004)
Apple volatile metabolites for disease discrimination
discriminate different diseases. The normalized abundance of esters was the highest among groups of compounds with more than 50% of the total abundance, and their abundance varied among diseases, as observed in apples by many researchers.17 – 20 In Renetta Canada apples, 12 of the 14 volatile compounds identified were esters.20 This included six different butanoic acid esters while we were able to identify four different butanoic acid esters. Detection of aroma volatiles of apple has been the main interest of many workers. Hexanal, 2-hexanal, ethyl propionate, ethyl 2-methyl propionate, methyl butyrate, methyl 2-methyl butyrate, ethyl butyrate, ethyl 2-methyl butyrate and ethyl pentanoate have been attributed to the aroma of McIntosh apples.18 Several esters, alcohols and aldehydes were identified in our experiment with McIntosh apples. The three volatile esters from apple—butyl acetate, 2-methyl butyl acetate and hexyl acetate—are considered major contributors to the characteristic apple-like aroma and flavor in most cultivars.8 There exists some similarity in the major esters, although volatile profiles vary among cultivars.21 The major ester components in the volatile profile of Rome apples are ethyl acetate, hexyl acetate, hexyl butyrate and butyl 2-methyl butyrate. The major ester constituents in Bisbee delicious apples are hexyl acetate, butyl acetate, 2-methyl butyl acetate, ethyl acetate, ethyl 2-methyl butyrate and pentyl acetate,19,22 while the first three compounds are also prominent in Golden Delicious apples.23 In the present investigation many volatile metabolites having antimicrobial properties were found, but they were inconsistent over replications. The presence of the sulfur compound thiirane was observed in all the pathogen inoculated apples but not in the control. The production of sulfur as a defense response during plant pathogen interactions has been reported previously.24 – 26 It is known that sulfur compounds are potent antimicrobial agents and the antimicrobial activity of sulfur compounds like thiadiazole, thiols and thiocyanates has been reported against various Gram-negative bacteria.27,28 It should be noted that, even though we have reported many inoculation agent/disease discriminatory volatile metabolites, there is about a 10% possibility that those compounds could occur in other inoculations. Thus, two or more samples must be collected to increase the probability of identity. The occurrence of compounds was inconsistent among replicates. Such variations are not new in metabolic studies as they have been reported previously.29 Such changes in metabolite production are a result of variations in host, pathogen and environmental factors.24,26 Another reason which can be attributed to the problem is the protocols used in the identification of compounds, based on the NIST library search, in which it was difficult to find a perfect match. The method developed here for the detection and discrimination of apple diseases, based on inoculation-agent/disease discriminatory volatile J Sci Food Agric 84:1333–1340 (online: 2004)
metabolites, abundance and frequency indices of certain sets of disease-discriminatory compounds and normalized abundance of chemical groups of volatile compounds, could be used by storage managers to detect apple diseases at an early stage of disease progress. Such a knowledge base can be used to manage apple diseases in storage. Important management decisions like length of product storage and selection of lots for longer storage could be taken using this knowledge of disease potential. It is often noted that, although many diseases are important in storage, the outbreaks are often from a single disease. The disease-specific volatile metabolites, unique and common to only a few, reported here, could be used as markers to discriminate diseases even when more than one disease is present, but this has to be tested before commercial application. For practical purposes it is advisable that most of the diseases known to occur commonly in a particular orchard or storage facility should be studied. Scaleup studies and models to automate the detection procedures, including calculation of normalised values for 20 compounds and factor scores, are required to apply such a tool to the detection of diseases in large storage facilities.
ACKNOWLEDGEMENT The study was funded by a strategic grant initiative of the Natural Sciences and Engineering Research Council of Canada. The authors are thankful to Dr PL Sholberg, Agriculture and Agri-Food Canada, Summerland, BC for providing the cultures of pathogens used here and to Mr W Stevenson, Stevenson Orchard, QC, for providing apples.
REFERENCES 1 Statistics Canada, Canadian fruit and vegetable production. Catalogue number 22-003-XIB (2003). 2 Snowdon AL, A Colour Atlas of Post-harvest Diseases and Disorders of Fruits and Vegetables, Vol 1. Wolfe Scientific, Barcelona, pp 170–216 (1990). 3 Blankenship SM, Controlled atmospheres for storage and transport of perishable agricultural commodities, in Proc 4th National Controlled Atmosphere Research Conference. North Carolina State University Horticultural Report no. 126, p 511 (1985). 4 Blanpied GD, Factors that influence commodity response to controlled atmosphere storage, Hort Sci 22:761–794 (1987). 5 Hardenburg RE, Watada AE and Wang CY, The commercial storage of fruits, vegetables, and florist and nursery stocks, in Agriculture Handbook of the United States Department of Agriculture, no 66. USDA, Washington, DC, p 130 (1986). 6 Power FB and Chestnut VK, The odorous constituents of apple. Emanation of acetaldehyde from the ripe fruit, J Am Chem Soc 42:1509 (1920). 7 Meigh DF, Volatile compounds produced by apples. I. Aldehydes and ketones, J Sci Food Agric 6:396 (1956). 8 Dimick PS and Hoskin JC, Review of apple flavor—state of the art, CRC Crit Rev Food Sci Nutr 18:387–409 (1983). 9 De Lacy Costello BPJ, Evans P, Ewen RJ, Gunson HE, Ratcliffe NM and Spencer-Phillips PTN, Identification of volatiles generated by potato tubers (Solanum tuberosum cv: 1339
A Vikram et al
10
11
12
13
14
15
16
17
18
19
Maris Piper) infected by Erwinia carotovora, Bacillus polymyxa and Arthrobacter sp, Plant Pathol 48:345–351 (1999). Kushalappa AC, Lui LH, Chen CR and Lee B, Volatile fingerprinting (SPME-GC-FID) to detect and discriminate diseases of potato tubers, Plant Dis 86:131–137 (2002). Kallio H and Salorinne L, Comparison of onion varieties by headspace gas chromatography-mass spectrometry, J Agric Food Chem 38:1560–1564 (1990). Davis PL and Smoot JJ, Germination of Penicillium digitatum spores as affected by solution of volatile components of citrus fruits, Phytopathology 62:488–489 (1972). Pauli A and Knoblauch K, Inhibitory effects of essential oil components on growth of food contaminating fungi, Z Lebensm Unters Forsch 185:10–13 (1987). Wilson CL and Wisniewski ME, Biological control of postharvest diseases of fruits and vegetables: an emerging technology, A Rev Phytopathol 27:425–441 (1989). Ouellette E, Raghavan GSV and Reeleder RD, Volatile profiles for disease detection in stored carrots, Can Agric Eng 32:255–261 (1990). De Lacy Costello BPJ, Evans P, Ewen RJ, Gunson HE, Jones PRH, Ratcliffe NM and Spencer-Phillips PTN, Gas chromatography-mass spectrometry analyses of volatile organic compounds from potato tubers inoculated with Phytophthora infestans or Fusarium coeruleum, Plant Pathol 50:489–496 (2001). Flath RA, Black DR, Guadagni DG, McFadden WH and Schultz TH, Identification and organoleptic evaluation of compounds in Delicious apple essence, J Agric Food Chem 15:29 (1967). Panasiuk O, Talley FB and Sapers GM, Correlation between aroma and volatile composition of McIntosh apples, J Food Sci 45:989 (1980). Mattheis JP, Buchanan DA and Fellman JK, Volatile compound production by Bisbee Delicious apples after sequential
1340
20
21
22
23
24
25 26
27
28
29
atmosphere storage, J Agric Food Chem 39:1602–1605 (1991). Thedy L, Ferrandino A, Mombelloni A, Pramotton R and Duverney C, Volatile composition of Renetta Canada fruit at harvest and after CA storage by SPME and GC/MS, Acta Hort 599:647–651 (2003). Kakiguchi N, Moriguchi S, Fukuda H, Ichimura N, Kato Y and Banba Y, Composition of volatile compounds of apple fruits in relation to cultivar, J Jpn Soc Hort Sci 55:280–289 (1986). Mattheis JP, Fellman JK, Chen PM and Patterson ME, Changes in headspace volatiles during physiological development of Bisbee Delicious apple fruit, J Agric Food Chem 39:1902–1906 (1991). Brackmann A, Streif J and Bangerth F, Relationship between a reduced aroma production and lipid metabolism of apples after long term controlled atmosphere storage, J Am Soc Hort Sci 118:243–247 (1993). Dixon RA, Achnine L, Kota P, Liu C, Reddy MSS and Wang L, The phenylpropanoid pathway and plant defense—a genomics perspective, Mol Plant Pathol 3:371–390 (2002). Fiehn O, Metabolomics—the link between genotypes and phenotypes, Plant Mol Biol 48:155–171 (2002). Roessner U, Luedemann A, Brust D, Fiehn O, Linke T, Willmitzer L and Fernie AR, Metabolic profiling allows comprehensive phenotyping of genetically or environmentally modified plant systems, Plant Cell 13:11–29 (2001). Fitton AO and Ridgway F, Synthesis and antimicrobial activity of 5,7- dichloroquinoline-8-thiol and its derivatives, J Med Chem 13:1008–1009 (1970). Kyung KH and Fleming HP, Antimicrobial activity of sulfur compounds derived from cabbage, J Food Prot 60:67–71 (1997). Fiehn O, Kopka J, Dormann P, Altman T, Trethewey RN and Willmitzer L, Metabolic profiling for plant functional genomics, Nat Biotech 18:1157–1161 (2000).
J Sci Food Agric 84:1333–1340 (online: 2004)
