Development of aroma-synthesising capacity throughout fruit ...

Journal of Horticultural Science & Biotechnology (2008) 83 (2) 253–259

Development of aroma-synthesising capacity throughout fruit maturation of ‘Mondial Gala’ apples By I. LARA*, A. ORTIZ, G. ECHEVERRÍA, M.L. LÓPEZ and J. GRAELL Àrea de Post-Collita, UdL-IRTA, Alcalde Rovira Roure 191, 25198 Lleida, Spain (e-mail: [email protected]) (Accepted 25 October 2007) SUMMARY Emission of aroma volatile compounds (AVCs) and the activity of some related enzymes were monitored during ontree maturation of ‘Mondial Gala’ apples. Volatile esters were quantitatively prominent among the AVCs identified throughout the experiment and, in most cases, their production increased noticeably during the later stages of fruit development. However, the activity of alcohol o-acyltransferase (AAT), the enzyme directly responsible for ester production, was detectable at almost constant levels throughout the experimental period, and therefore it was not in accordance with changes in the emission of volatile esters. Multivariate analysis indicated that the increase in the production of volatile esters during maturation arose mainly from the greater availability of substrates for esterification reactions, rather than from increased AAT activity. The role of some aroma-related enzyme activities in controlling the supply of precursors for ester production is discussed.

pple (Malus
domestica Borkh.) cultivars in the ‘Gala’ group represent the second most important dessert apple variety in Spain. The group is also the second most important in Lleida (NE Spain) in terms of production, and this is expected to increase in the near future. Commercial interest in the cultivars in this group arises from their early harvest time and marketing possibilities, in combination with their excellent organoleptic properties – including a crisp, juicy flesh and a strong, pleasant aroma and flavour (White, 1991). Although standard quality specifications for ‘Gala’ apples marketed in the EU refer mainly to size and surface colour, eating quality and consumer acceptance of apples are determined by other attributes such as firmness, soluble solids content, titratable acidity, and the production of aroma volatiles. Post-harvest synthesis of aroma volatile compounds (AVCs) depends on the developmental stage at harvest, as the process is associated with fruit ripening (Dirinck and Schamp, 1989) and the aroma profile changes continuously throughout fruit development. Too early a harvest, aimed at improving fruit storage potential or resistance to the usual post-harvest handling procedures, will often result in a lack of flavour development; whereas delaying the picking date to improve the aroma and red colour of the fruit skin will lead to poorer storage potential. A deeper understanding of the changes in the biosynthesis of AVCs taking place during fruit maturation may thus be of help when determining the optimal harvest date. Fruit aroma is a complex trait. The identity, concentration, and total number of AVCs are specific for each cultivar, and depend on the developmental stage as well as on the activity and substrate specificity of the key enzymes involved in the biosynthetic pathways (Dixon and Hewett, 2000). In addition, the contribution of each particular AVC to the aroma profile will also depend on its odour threshold and the presence of other

A

*Author for correspondence.

compounds (Buttery, 1993). Volatile esters, associated with the “fruity” attributes of fruit flavour, are the most conspicuous contributors to the aroma profile of intact apples, where they can account for up to 98% of the total volatiles emitted by ripe fruit (López et al., 1998). Volatile esters are generated by the esterification of alcohols and acyl-CoAs derived primarily from fatty acids and amino acids (Sanz et al., 1997). The esterification reaction is catalysed by alcohol oacyltransferase (AAT; EC 2.3.1.84), which is thus directly responsible for the production of these compounds by fruit tissues. However, in previous work, no large variations were found in AAT activity during on-tree maturation of ‘Fuji’ (Echeverría et al., 2004) or ‘Pink Lady’® (Villatoro et al., in press) apples, in spite of sharply increased ester emissions at advanced stages of maturity. Although regulatory differences may exist for different cultivars, resulting in dissimilar patterns of aroma-related enzyme activity throughout fruit development, those observations suggest that the availability of substrates for AAT may play a major role in determining the concentration and identity of the specific esters emitted by fruit. Therefore, some critical steps in ester emission may be located upstream of AAT in the biosynthetic pathway. It has been suggested that the limited availability of fatty acid-derived precursors may be a major factor restricting the production of volatile esters in immature apple fruit (Song and Bangerth, 1994; 2003). Accordingly, partial inhibition of lipoxygenase (LOX; EC 1.13.11.12) activity in ‘Fuji’ (Lara et al., 2006) and in ‘Mondial Gala’ (Lara et al., 2007) apples stored under controlled atmospheres led to the development of abnormal aroma profiles after transfer of the fruit to air. The purpose of this work was to study the progress of the ability to produce aroma volatiles throughout ontree maturation of ‘Mondial Gala’ apples, and to identify the enzyme activities that may control this ripeningrelated process in this apple cultivar.

254

Aroma biosynthesis during apple maturation

MATERIALS AND METHODS Plant material Apple fruit (Malus
domestica Borkh. ‘Mondial Gala’) were collected from an experimental orchard (IRTA-Estació Experimental de Lleida) at Gimenells (NE Spain), at approx. 1-week intervals during 1 month prior to commercial harvest, then selected for uniformity and the absence of defects. The sampling period was 11 July - 12 August 2002, corresponding to 105 and 137 days after full bloom (DAFB), respectively. Commercial harvest in the area took place 130 DAFB. Maturity and quality parameters at each sampling date were measured, as described elsewhere (Echeverría et al., 2004). Fruit were also used for analyses of the emission of AVCs, measurements of acetaldehyde concentrations, and determinations of some aroma-related enzyme activities, as described below. Analysis of aroma volatile compounds (AVCs) At each sampling date, 8 kg of apples (2 kg per replicate
4 replicates) were taken for analysis of AVCs. The extraction was performed from each 2-kg sample of intact fruit according to the dynamic headspace method. Each fruit sample was placed in an 8 l Pyrex glass container, and an air-stream (900 ml min–1) was passed through for 4 h. The effluent was then passed through an ORBO-32 adsorption tube (Supelco, Bellefonte, PA, USA) filled with 100 mg activated charcoal (20/40 mesh), from which AVCs were desorbed by agitation for 40 min in 0.5 ml diethyl ether. The identification and quantitation of AVCs were achieved on a Hewlett Packard 5890 gas chromatograph (Hewlett-Packard Co., Barcelona, Spain) equipped with a flame ionisation detector (FID) and a cross-linked free fatty acid phase (FFAP; 50 m
0.2 mm i.d.
0.33 µm) capillary column (Agilent Technologies, S.L., Madrid, Spain). One µl of the extract was injected in all the analyses. Helium was used as the carrier gas (at 0.013 ml s–1), with a split ratio of 40:1. The injector and detector were held at 220ºC and 240ºC, respectively. The analysis was conducted according to the following programme: 70ºC (1 min); 70ºC to 142ºC (at 3ºC min–1); 142ºC to 225ºC (at 5ºC min–1); and 225ºC (for 10 min), as described elsewhere (Echeverría et al., 2002). AVCs were identified by comparing their retention indices with those of standards, and by enriching the apple extract with authentic samples. The quantification was made using butylbenzene [> 99.5% (v/v); Fluka Chemie, Buchs, Switzerland] as the internal standard. A GC-MS system (Hewlett-Packard 5890) was used for compound confirmation, in which the same capillary column was used as in the GC analyses. Mass spectra were obtained by electron impact ionisation at 70 eV. Helium was used as the carrier gas (at 0.013 ml s–1), according to the same temperature gradient programme as described above. Spectrometric data were recorded (Hewlett Packard 3398GC Chemstation) and compared with those from the original NIST HP59943C library of massspectra (Hewlett-Packard). Results were expressed as µg AVC kg–1. Analysis of acetaldehyde concentration The juice from 20 fruit from each harvest date was obtained, individually, and frozen at –20ºC until analyses

of acetaldehyde contents, as described by Ke et al. (1994). Frozen juice from each fruit was thawed, and a 5 ml sample was placed in a 10 ml test-tube, which was closed with an elastic cap and incubated at 65ºC for 1 h. A 1 ml headspace gas sample was removed with a syringe and injected into a Hewlett Packard 5890 gas chromatograph, equipped with a column containing 5% Carbowax on Carbopack (60/80; 2 m
2 mm i.d.) as the stationary phase (Supelco), and a FID. Nitrogen was used as the carrier gas (at 0.75 ml s–1), and the operating conditions were as follows: oven temperature 110ºC, injector temperature 180ºC, detector temperature 220ºC. Acetaldehyde was identified and quantified by comparison with an external standard, and the results were expressed as µl l–1. Extraction and assay of aroma-related enzyme activities Lipoxygenase (LOX), hydroperoxide lyase (HPL), pyruvate decarboxylase (PDC), alcohol dehydrogenase (ADH), and alcohol o-acyltransferase (AAT) activities were determined at each sampling date. Samples of both skin and flesh tissue were taken separately from four apples, frozen in liquid nitrogen, lyophilised, powdered, and kept at –80ºC until processing. One hundred mg of lyophilised powdered tissue was used for each determination. The extraction and assay of LOX, PDC, ADH, and AAT activities in crude enzyme extracts were performed as described elsewhere (Lara et al., 2003). HPL activity was extracted and assayed according to Vick (1991). The total protein content of each enzyme extract was determined by the Bradford method (1976), with modifications (BioRad Protein Assay Kit; Bio-Rad Laboratories Inc., Hercules, CA, USA) according to the manufacturer’s instructions, using BSA as a standard. In all cases, 1 activity unit (U) was defined as a variation of 1 unit of absorbance min–1. Each determination was done in triplicate, and the results were expressed as specific activity (U mg–1 protein). Statistical and multivariate analyses A factorial design, with sampling date and replicate as factors, was used to statistically analyse the results. All data were tested by analysis of variance (GLMANOVA), using the SAS programme package (SAS Institute, Inc., 1987). Means were separated by LSD test at P ≤ 0.05. Partial least-square regression (PLSR) was used as a predictive method to relate a matrix of several dependent variables (Y) to a set of explanatory variables (X) in a single estimation procedure. Unscrambler Version 6.11a software (CAMO ASA, 1997) was used for developing these models. Samples were coded H1 to H6, corresponding to fruit picked between 105 and 137 DAFB, respectively. As a pre-treatment, data were centred and weighted by the inverse of the standard deviation of each variable in order to avoid dependence on measured units (Martens and Naes, 1989). Leverage correction was run as a validation procedure.

RESULTS AND DISCUSSION Modifications in the emission of AVCs during tree maturation of ‘Mondial Gala’ apples Maturity and quality parameters of fruit at each sampling date (Table I), were used to estimate the

255

I. LARA, A. ORTIZ, G. ECHEVERRÍA, M. L. LÓPEZ and J. GRAELL TABLE I Maturity and quality parameters of ‘Mondial Gala’ apples at each sampling date (H1 – H6) Parameter

H1 105 DAFBa

H2 113 DAFB

H3 119 DAFB

H4 125 DAFB

H5 130 DAFB

H6 137 DAFB

0.1 c 29.2 c

0.1 c 30.3 c

0.1 c 28.7 c

0.3 c 30.4 c

1.1 b 53.7 b

2.3 a 69.7 a

63.9 f 112.2 f 124.6 a 9.6 e 5.3 a 1.0 d 86.3 a 113.4 a

67.0 e 130.1 e 113.8 b 10.2 d 4.4 bc 1.5 c 75.6 a 112.2 a

72.8 d 163.2 d 101.0 c 10.7 cd 4.3 c 1.8 c 77.1 a 111.4 a

76.8 c 197.9 c 93.2 d 11.0 c 4.2 c 1.7 c 62.2 a 109.8 b

79.8 b 212.0 b 89.8 d 11.8 b 4.5 bc 2.3 b 36.7 b 96.0 c

81.4 a 229.6 a 78.5 e 12.9 a 4.7 b 2.8 a 27.2 c 82.7 c

Ethylene production (µ kg–1 h–1) Acetaldehyde content (µl l–1) Fruit Diameter (mm) Weight (g) Firmness (N) SSC (ºBrix)a TA (g malic acid l–1)b SI (1 – 6)c Hue (ES)d Hue (SS)d

Values represent means of four (ethylene production) or 20 (acetaldehyde content and fruit quality parameters) replicates. Means followed by different lower-case letters within the same row are significantly different at P ≤ 0.05 (LSD Tukey’s test). a DAFB, days after full bloom. b SSC, soluble solids content. c TA, titratable acidity. d SI, starch index (1, maximum starch; 6, no starch). e ES, exposed side; SS, shaded side.

maturity stage at each time point considered. Up to 29 AVCs (20 esters, eight alcohols, and one terpene), depending on harvest date, were identified and quantified in the volatile fraction emitted by ‘Mondial Gala’ fruit throughout the experimental period (Table II). Esters (eight acetates, four propanoates, three butanoates, three 2-methylbutanoates, and two hexanoates) were quantitatively prominent among the AVCs produced, accounting for 69 – 89% of total production. Three acetate esters (butyl acetate, hexyl acetate, and 2-methylbutyl acetate) stood out quantitatively, followed by two hexyl esters (hexyl 2-

methylbutanoate and hexyl propanoate), pentyl acetate, and three butyl esters (butyl 2-methylbutanoate, butyl propanoate and butyl butanoate). However, the amount produced was not representative of the contribution of each of these compounds to the aroma profile of the fruit. On the basis of the log10 odour units present (Buttery, 1993), ethyl 2-methylbutanoate was found to be the volatile compound with the greatest impact on overall fruit flavour at commercial maturity, in spite of the low production observed (Table II). Positive log10 odour units were also found for ethyl butanoate, although emission of this compound was also very low at

TABLE II Aroma volatile production (µg kg–1) by ‘Mondial Gala’ apples at each sampling date (DAFB) Compound Methyl acetate Ethyl acetate Ethanol t-Butyl propanoate Propyl acetate 2-Methylpropyl acetate 1-Propanol Ethyl butanoate Ethyl 2-methylbutanoate Butyl acetate 2-Methyl-1-propanol 2-Methylbutyl acetate 1-Butanol Butyl propanoate 4-Methyl-2-pentanol Pentyl acetate 2-Methylbutyl propanoate 2-Methyl-1-butanol D-Limonene Butyl butanoate Butyl 2-methylbutanoate 1-Pentanol Hexyl acetate Hexyl propanoate 1-Hexanol Butyl hexanoate Hexyl butanoate Hexyl 2-methylbutanoate Hexyl hexanoate Total aroma emission

Code ma ea etOH tbpr pra 2mpra prOH eb e2mb ba 2mprOH 2mba bOH bpr 4m2pOH pa 2mbpr 2mbOH limon bb b2mb pOH ha hpr hOH bh hb h2mb hh

KIa 834 898 932 964 984 1,020 1,036 1,043 1,059 1,082 1,091 1,131 1,144 1,148 1,163 1,183 1,199 1,210 1,219 1,228 1,240 1,253 1,283 1,349 1,358 1,423 1,426 1,436 1,581

OTHb 8,300 (h) 13,500 (c) 100,000 (a) 19 (h) 2,000 (c) 65 (b) 9,000 (a) 1 (d) 0.006 (b) 66 (c) 1,000 (d) 11 (b) 500 (a) 25 (a) – 43 (c) 19 (h) 250 (d) 34 (d) 100 (e) 17 (e) 4,000 (f) 2 (f) 8 (g) 500 (f) 700 (e) 250 (b) 6 (e) –

105 DAFBc 66.1 ab 47.8 a 72.0 a 19.4 a 14.6 b 14.1 b 13.0 b 17.6 a 32.7 a 9.9 c 8.7 b 11.3 c 1.9 c 1.5 b 1.5 b 15.6 bc 1.0 b n.d. n.d. 2.2 b n.d. 1.4 b 16.1 c 15.1 bc 1.3 c n.d. n.d. 7.3 b 5.4 b 401.5 c

†

113 DAFB

119 DAFB

125 DAFB

130 DAFB

137 DAFB

81.6 a 34.1 b 56.5 ab 22.6 a 16.8 b 12.1 bc 16.4 b 7.5 b 22.9 b 10.9 c 7.7 b 9.4 c 7.2 c n.d. n.d. 5.8 c n.d. n.d. n.d. n.d. n.d. n.d. 4.3 c 4.1 c 1.5 c 1.7 c 2.1 c 4.9 b 2.1 b 331.5 c

25.3 c 31.3 b 32.4 bc 3.7 b 1.4 c 1.0 c 2.3 b 1.0 c 1.0 c 20.8 c 0.7 c 8.0 c 9.3 c 1.8 b n.d. 2.1 c 0.8 b 2.5 c n.d. 2.0 b n.d. Traces 15.4 c 2.8 c 1.9 c 2.7 c 1.9 c 3.0 b 2.1 b 178.0 c

57.3 abc 36.7 bc 52.3 abc 42.3 ab 14.4 c 32.5 b 57.5 ab 17.2 c 54.7 a 16.8 a 4.9 b 8.3 b 12.1 bc 7.9 bc 39.0 a 9.0 bc 6.5 bc 25.4 a 11.2 b 11.7 b 44.7 a 12.7 a 1.1 c 1.9 c 23.7 b 3.0 c 7.2 c 17.5 c 325.6 b 1,091.0 a 6.2 bc 5.7 bc 16.9 a 20.5 c 241.1 b 587.2 a 13.9 c 130.5 b 346.6 a 4.0 b 16.6 b 84.3 a n.d. 0.5 b 22.5 a 13.0 bc 34.8 b 92.1 a 4.2 ab 1.9 b 15.0 a 5.6 c 37.0 b 81.9 a n.d. 1.5 a 1.6 a 6.5 b 14.2 b 69.4 a n.d. 17.6 b 90.4 a n.d. 3.6 b 8.7 a 22.9 c 296.4 b 974.0 a 10.0 bc 34.5 b 142.4 a 3.0 c 21.4 b 62.6 a 3.1 c 35.0 b 169.6 a 4.5 c 40.0 b 130.8 a 6.2 b 87.2 b 329.1 a 5.1 b 45.1 b 104.2 a 423.3 c 1,493.8 b 4,611.8 a

Log (OU)d

0.28 3.08 1.22 1.73 0.53 0.53

0.73 2.69 1.25

1.74

Values represent means of four replicates (n.d.: non-detectable; Traces: < 0.5 µg kg–1). † Means followed by different lower-case letters within the same row for a given compound are significantly different at P ≤ 0.05 (LSD Tukey’s test). a Kovats retention index in a cross-linked FFAP column (Poole and Poole, 1993). b Odour thresholds reported by: (a): Flath et al. (1967); (b): Takeoka et al. (1992); (c): Takeoka et al. (1996); (d): Rychlik et al. (1998); (e): Takeoka et al. (1990); (f): Buttery, (1993); (g): Van Gemert and Nettenbreijer, (1977); and (h): Schnabel et al. (1988). (–): not found. c DAFB, days after full bloom. d Odour unit (OU) value = amount / OTH. Only positive values at 137 DAFB are indicated.

Aroma biosynthesis during apple maturation

Modifications in aroma-related enzyme activities during on-tree maturation of ‘Mondial Gala’ apples All the esters identified were already measurable in early-harvested (105 DAFB) fruit, indicating that the capacity for ester biosynthesis was present in immature tissues. Indeed, AAT activity was detectable throughout the experimental period (Figure 1A). Some increase in AAT activity was apparent between 105 and 113 DAFB, both in skin and flesh tissues, possibly signalling the onset of the metabolic modifications leading to ripening. However, a decrease in extractable AAT activity was found for more mature fruit, in accordance with previous reports on apple (Fellman et al., 2000). Because no sharp or continued enhancement of AAT activity was observed during the experimental period, it is suggested that the increase in the emission of volatile esters throughout fruit maturation arose mainly from the greater availability of substrates for esterification, rather than from increased enzyme activity, in agreement with previous work on ‘Fuji’ (Echeverría et al., 2004) and ‘Pink Lady’® (Villatoro et al., in press) apples. Good correlations were found between the emission of specific ester families and their corresponding alcohol precursors. The plot of ethyl esters vs. ethanol emitted is given, as an example (Figure 1B), showing an r2 value of 0.83. For butyl and hexyl esters, the correlations with 1-butanol and 1-hexanol, respectively, were even higher (r2 = 0.99; data not shown). When a PLSR model was developed for volatile esters emitted (Y variables), with precursors as the X variables, this dependence was further illustrated. This model explained 89% of the total variability in ester emission by the availability of precursors, as indicated by the agreement between the on-plot distribution of samples

(Figure 2A) and of compounds characterising fruit at each maturity stage (Figure 2B). Principal component 1 (PC1) alone accounted for 66% of the total variance, and allowed the differentiation of samples from immature to more mature stages. Early (H1 – H4) samples were characterised by higher emissions of ethyl esters, as well as ethanol, their alcohol precursor (Figure 2B), consistent with the evolution of these compounds throughout fruit maturation (Table II). Samples H5 and H6, corresponding to fruit picked at 130 and 137 DAFB, respectively, were characterised by a higher availability of most substrates, resulting in higher emissions of butyl, 2-methylbutyl, propyl, pentyl, and hexyl esters. Therefore, AAT activity alone was apparently not a good predictor for ester production by ‘Mondial Gala’ apples during on-tree maturation. AAT catalyses the esterification of acyl-CoAs and alcohols to generate volatile esters. One important acyl-CoA compound is acetyl-CoA, derived from acetaldehyde by the combined action of aldehyde dehydrogenase and acetate-CoA ligase (MacDonald and Kimmerer, 1993), which leads to the production of acetate-type esters. Acetaldehyde can also be reduced to ethanol, which is a precursor for ethyl esters by an ADH-mediated reaction. Thus, the availability of 1.0

A

LSD0.05 (skin) LSD0.05 (flesh)

AAT activity (U mg-1 protein)

commercial harvest. Generally, with the exception of a transient increase at 125 DAFB, production of all three ethyl esters identified in the volatile fraction of ‘Mondial Gala’ apples (ethyl 2-methylbutanoate, ethyl butanoate and ethyl acetate) decreased during the experimental period, concomitantly with that of ethanol, their alcohol precursor (Table II). The decline in the production of ethyl 2-methylbutanoate during maturation is in agreement with previous reports for ‘Fuji’ (Echeverría et al., 2004), but contrasts with observations for ‘Pink Lady’® apples (Villatoro et al., in press). In addition to ethyl 2-methylbutanoate and ethyl butanoate, eight more AVCs were found to have positive log10 odour units at commercial maturity, and thus to be likely to contribute to the overall fruit aroma: namely, hexyl acetate, hexyl 2-methylbutanoate, 2-methylbutyl acetate, hexyl propanoate, butyl acetate, butyl 2methylbutanoate, butyl propanoate, and pentyl acetate (Table II). Some of these volatile esters (ethyl 2methylbutanoate, ethyl butanoate, 2-methylbutyl acetate, hexyl acetate, butyl acetate and butyl propanoate) have been reported to confer “fruity” notes to the overall flavour (Mehinagic et al., 2006). In contrast to observations for ethyl esters, the production of contributing hexyl, 2-methylbutyl, butyl, and pentyl esters increased with maturity stage. In all instances, changes observed in the emission of these compounds reflected those in their alcohol precursors, 1-hexanol, 2-methyl-1butanol, 1-butanol, and 1-pentanol, respectively.

0.5

0.0 100

110

120

130

140

Days after full bloom

July

August

100

B

y = 1.37x - 10.61 r2 = 0.83

Ethyl esters -1 (µg kg )

256

10

1

0

20

40

60

80

100

Ethanol content (µl l-1) FIG. 1 Panel A, alcohol o-acyltransferase (ATT) specific activity in the skin (
) and flesh () of ‘Mondial Gala’ apples on different sampling dates. Values represent the means of three replicates. Vertical bars indicate LSD0.05. Panel B, correlation between the emissions of ethanol and ethyl esters by ‘Mondial Gala’ apples during on-tree maturation. Points represent the means of four replicates.

I. LARA, A. ORTIZ, G. ECHEVERRÍA, M. L. LÓPEZ and J. GRAELL 10000 10,000

257

2 Ry2 ==0.015x 0.92 + 29.01

Acetate esters (µg kg-1)

r = 0.92

1000 1,000

100

AA

10 0

20

40

60

80

100

Acetaldehyde content (µl l-1) FIG. 3 Correlation between the acetaldehyde content and the emission of acetate esters in ‘Mondial Gala’ apples during on-tree maturation.

A

PDC activity (U mg-1 protein)

acetaldehyde is also a factor to take into account when analysing the biosynthesis of aroma volatile esters, and will be partially dependent upon PDC activity, which synthesises acetaldehyde from pyruvic acid. The acetaldehyde content of ‘Mondial Gala’ apple juice remained steady during the early stages of fruit maturation, followed by a significant increase after 125 DAFB (Table I). These results contrast with previous observations for other apple cultivars such as ‘Fuji’, where only small differences in acetaldehyde content were found throughout the experimental period (Echeverría et al., 2004). These observations were suggested to arise from low PDC activity in combination with ADH-mediated reduction of acetaldehyde to ethanol. Such differences may be related to the physiological characteristics of both cultivars. Whereas ‘Fuji’ apples show a slow, attenuated climacteric, ‘Gala’ fruit undergo rapid climacteric and ripening changes. For ‘Pink Lady’® apples (Villatoro et al., in press), the acetaldehyde content rose from approx. 1 month before commercial harvest, followed by a decline just before the commercial picking date. The increase in acetaldehyde concentration at advanced stages of maturity in ‘Mondial Gala’ fruit was concomitant with a significantly enhanced emission of acetate esters (Table II), with a high linear correlation (r2 = 0.92; Figure 3). However, changes in PDC activity (Figure 4A) did not reflect those in acetaldehyde content, which might indicate that PDC is not a ratelimiting factor for the accumulation of this precursor

40 LSD0.05 (skin) LSD0.05 (flesh)

20

0

B

ADH activity (U mg-1 protein)

FIG. 2 Scores (Panel A) and loadings (Panel B) plots corresponding to a PLSR model of volatile esters emitted (Y variables) vs. precursor availability (X variables) in ‘Mondial Gala’ fruit on different sampling dates. Samples are coded as defined in Table I. Variables are labelled as indicated in Table II. AA, acetaldehyde.

during maturation of ‘Mondial Gala’ apple. Although the freezing of fruit juice for subsequent analysis might have modified enzyme activity in the juice samples, thus giving rise to this discrepancy, similar results have been reported for other fruits such as orange (Citrus sinensis L.; Bruemmer and Roe, 1985), tomato (Lycopersicon esculentum Mill.; Chen and Chase, 1993), pear (Pyrus communis L.; Chervin and Truett, 1999), and grape (Vitis vinifera L.; Or et al., 2000).

LSD0.05 (skin) LSD0.05 (flesh)

20

0 100

110

120

130

140

Days after full bloom

July

August

FIG. 4 Pyruvate decarboxylase (PDC; Panel A) and alcohol dehydrogenase (ADH; Panel B) specific activities in the skin (
) and flesh () of ‘Mondial Gala’ apples on different sampling dates. Values represent the means of three replicates. Vertical bars indicate LSD0.05.

258

Aroma biosynthesis during apple maturation

It has been suggested that, in fruits in which PDC activity is more-or-less constitutively available, the induction of ADH expression may be the trigger for ethanol production during fruit ripening (Or et al., 2000). Nevertheless, neither ethanol production (Table II), nor ADH activity (Figure 4B), underwent noticeable increases during on-tree maturation of ‘Mondial Gala’ apples. This might indicate that the pool of available acetaldehyde was being preferentially metabolised to acetate, and thus diverted to the production of acetate esters, rather than ethanol. Thus, the fact that earlyharvested fruit were characterised by higher emissions of ethanol and ethyl esters (Figure 2) might have arisen from the higher ADH activity in these samples (Figure 4B). Alternatively, the seeming lack of correlation between PDC activity and acetaldehyde content suggests that this precursor was provided via other biochemical pathways. Acetaldehyde can be produced either from pyruvic acid by PDC and the enzymatic oxidation of ethanol, the reverse reaction of alcoholic fermentation catalysed by ADH; or from fatty acids via the LOX/HPL pathway. Fatty acids are major precursors for the biosynthesis of volatiles through a number of different metabolic pathways (Sanz et al., 1997; Dixon and Hewett, 2000). Low oxidation rates for fatty acids have been suggested to account for a shortage of precursors for ester production (Brackmann et al., 1993; Fellman et al., 1993). At immature stages, lipid-derived AVC biosynthesis by fruits is thought to take place mainly through -oxidation (Sanz et al., 1997), as the enzymes and substrates for the LOX pathway have different sub-cellular locations with no chance of reaction. However, membrane fluidity, as well as lipid synthesis, increases as ripening proceeds (Bartley, 1985), leading to the onset of volatile production through this pathway. The relevance of this metabolic route for aroma production by ripe apples is illustrated by reports on ‘Fuji’ (Lara et al., 2006) and ‘Mondial Gala’ (Lara et al., 2007) fruit, showing that the inhibition of LOX activity, induced by controlled atmosphere storage, led to a significantly altered aroma profile after transfer from hypoxia to air. Therefore, the cleavage of fatty acid hydroperoxides to aldehydes by HPL may be a key step controlling the availability of acetaldehyde for ester biosynthesis in mature fruit. In order to assess the relationships, if any, between the availability of precursors and the enzyme activities considered here, a PLSR model was developed for the alcohols and acetaldehyde emitted (Y variables), with enzyme activities both in the skin and flesh as the X variables. This model (Figure 5) explained up to 78% of the total variation in precursor availability. Mature (H5H6) samples were situated on the right side of the PC1 axis (Figure 5A), and were characterised by higher emissions of most precursors. The loadings plot (Figure 5B) shows that the availability of these precursors was associated with higher levels of HPL activity in the flesh (HPLf) and LOX activity in the skin (LOXs). The model also shows that the acetaldehyde contents of fruit harvested at advanced stages of maturity were related to HPL rather than to PDC activity, which characterised early-harvested samples. LOX and HPL activities were inversely correlated in both skin and flesh tissues, which could indicate that these activities are coordinated.

FIG. 5 Scores (Panel A) and loadings (Panel B) plots corresponding to a PLSR model of precursor availability (Y variables) vs. enzyme activities (X variables) in ‘Mondial Gala’ fruit on different sampling dates. Samples are coded as defined in Table I. Precursors are labelled as indicated in Table II (AA, acetaldehyde). For enzyme labels, the suffix ‘s’ or ‘f’ refers to the activity in the skin or the flesh, respectively.

These data also suggest differences in this coordination between skin and flesh tissues. In flesh, for example, lower HPL activity in early samples might have led to a lower availability of acetaldehyde, despite higher levels of LOX activity providing the necessary hydroperoxide substrates (Figure 5B). As regards skin, higher HPL activity at immature stages might have resulted in the activation of LOX in mature samples as a mechanism to restore the hydroperoxide pool consumed by HPL.

CONCLUSIONS These results confirm that AAT activity alone cannot explain the increase in the emission of volatile esters during the later stages of development of ‘Mondial Gala’ apple fruit, as reported previously for ‘Fuji’ and ‘Pink Lady’® cultivars. Multivariate analysis showed that an adequate supply of precursors for the esterification reactions accounted for a large amount of the total variation in ester production. Accordingly, the availability of most precursors was higher in fruit picked close to the commercial harvest date. The results suggest that acetaldehyde, an important precursor for ester biosynthesis, was related to HPL rather than to PDC activity. This work was funded through Project RTA02-072, financed by the Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), Spain. A.

I. LARA, A. ORTIZ, G. ECHEVERRÍA, M. L. LÓPEZ and J. GRAELL Ortiz is the recipient of a Ph.D. grant from the University of Lleida, Spain. The authors are also indebted to T.

259

Fuentes for technical assistance, and to J. Voltas for useful comments on multivariate analysis.

REFERENCES BARTLEY, I. M. (1985). Lipid metabolism of ripening apples. Phytochemistry, 12, 2857–2859. BRACKMANN, A., STREIF, J. and BANGERTH, F. (1993). Relationship between a reduced aroma production and lipid metabolism of apple after long-term controlled atmosphere storage. Journal of the American Society for Horticultural Science, 118, 243–247. BRADFORD, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Analytical Biochemistry, 72, 248–254. BRUEMMER, J. H. and ROE, B. (1985). Pyruvate dehydrogenase activity during ripening of ‘Hamlin’ oranges. Phytochemistry, 24, 2105–2106. BUTTERY, R. G. (1993). Quantitative and sensory aspects of flavor of tomato and other vegetables and fruits. In: Flavor Science: Sensible Principles and Techniques. (Acree, T. E. and Teranishi, R., Eds.). ACS Professional Reference Book, Washington DC, USA. 259–286. CAMO ASA (1997). Unscrambler Users Guide. Ver. 6.11a. Programme Package for Multivariate Calibration. Trondheim, Norway. 312 pp. CHEN, A. S. and CHASE, T. (1993). Alcohol dehydrogenase-2 and pyruvate decarboxylase induction in ripening and hypoxic tomato fruit. Plant Physiology, 31, 875–885. CHERVIN, C. and TRUETT, J. K. (1999). Alcohol dehydrogenase expression and alcohol production during pear ripening. Journal of the American Society for Horticultural Science, 124, 71–75. DIRINCK, P. J. and SCHAMP, N. (1989). Instrumental aroma analysis for objective evaluation of parameters influencing aroma formation in apples and for prediction of the optimum picking date. Acta Horticulturae, 258, 421–428. DIXON, J. and HEWETT, E. W. (2000). Factors affecting apple aroma/flavour volatile concentration: a review. New Zealand Journal of Crop and Horticultural Science, 28, 155–173. ECHEVERRÍA, G., GRAELL, J. and LÓPEZ, M. L. (2002). Effect of harvest date and storage conditions on quality and aroma production of ‘Fuji’ apples. Food Science and Technology International, 8, 351–360. ECHEVERRÍA, G., GRAELL, J., LÓPEZ, M. L. and LARA, I. (2004). Volatile production, quality and aroma-related enzyme activities during maturation of ‘Fuji’ apples. Postharvest Biology and Technology, 31, 217–227. FELLMAN, J. K., MATTINSON, D. S., BOSTICK, B., MATTHEIS, J. P. and PATTERSON, M. (1993). Ester biosynthesis in ‘Rome’ apples subjected to low-oxygen atmospheres. Postharvest Biology and Technology, 3, 201–214. FELLMAN, J. K., MILLER, T. W., MATTINSON, D. S. and MATTHEIS, J. P. (2000). Factors that influence biosynthesis of volatile flavor compounds in apple fruits. HortScience, 35, 1026–1033. FLATH, R. A., BLACK, D. R., GUADAGNI, D. G., MCFADDEN, W. H. and SCHULTZ, T. H. (1967). Identification and organoleptic evaluation of compounds in ‘Delicious’ apple essence. Journal of Agricultural and Food Chemistry, 15, 457–461. KE, D., YAHIA, E. M., MATEOS, M. and KADER, A. A. (1994). Ethanolic fermentation of ‘Bartlett’ pears as influenced by ripening stage and atmospheric composition. Journal of the American Society for Horticultural Science, 119, 976–982. LARA, I., MIRÓ, R. M., FUENTES, T., SAYEZ, G., GRAELL, J. and LÓPEZ, M. L. (2003). Biosynthesis of volatile aroma compounds in pear fruit stored under long-term controlled-atmosphere conditions. Postharvest Biology and Technology, 29, 29–39. LARA, I., GRAELL, J., LÓPEZ, M. L. and ECHEVERRÍA, G. (2006). Multivariate analysis of modifications in biosynthesis of volatile compounds after CA storage of ‘Fuji’ apples. Postharvest Biology and Technology, 39, 19–28. LARA, I., ECHEVERRÍA, G., GRAELL, J. and LÓPEZ, M. L. (2007). Volatile emission after controlled atmosphere storage of ‘Mondial Gala’ apples (Malus
domestica): Relationship to some related enzyme activities. Journal of Agricultural and Food Chemistry, 55, 6087–6095.

LÓPEZ, M. L., LAVILLA, T., RECASENS, I., RIBA, M. and VENDRELL, M. (1998). Influence of different oxygen and carbon dioxide concentrations during storage on production of volatile compounds by ‘Starking Delicious’ apples. Journal of Agricultural and Food Chemistry, 46, 634–643. MACDONALD, R. C. and KIMMERER, T. W. (1993). Metabolism of transpired ethanol by eastern cottonwood (Populus deltoides Bartr.). Plant Physiology, 102, 173–179. MARTENS, H. and NAES, T. (1989). Partial least squares regression. In: Multivariate Calibration. Wiley, Chichester, UK. 116–165. MEHINAGIC, E., ROYER, G., SYMONEAUX, R., JOURJON, F. and PROST, C. (2006). Characterization of odor-active volatiles in apples: Influence of cultivars and maturity stage. Journal of Agricultural and Food Chemistry, 54, 2678–2687. OR, E., BAYBIK, J., SADKA, A. and OGRODOVITCH, A. (2000). Fermentative metabolism in grape berries: isolation and characterization of pyruvate decarboxylase cDNA and analysis of its expression throughout berry development. Plant Science, 156, 151–158. POOLE, C. L. and POOLE, S. K. (1993). The column in gas chromatography. In: Chromatography Today. Elsevier, Amsterdam, The Netherlands. 176–181. RYCHLIK, M., SCHIEBERLE, P. and GROSCH, W. (1998). Compilation of Odor Thresholds, Odor Qualities and Retention Indices of Key Food Odorants. Deutsche Forschungsanstalt für Lebensmittelchemie and Institut für Lebensmittelchemie der Technischen Universität München, Garching, Germany. 63 pp. SANZ, C., OLÍAS, J. M. and PÉREZ, A. G. (1997). Aroma biochemistry of fruits and vegetables. In: Phytochemistry of Fruits and Vegetables. (Tomás-Barberán, F. A. and Robins, R. J., Eds.). Clarendon Press, Oxford, UK. 125–155. SAS INSTITUTE, INC. (1987). SAS/STAT Guide for Personal Computers. 6th Edition. SAS Institute, Inc., Cary, NC, USA. 1,848 pp. SCHNABEL, K. O., BELITZ, H. D. and VON RANSON, C. (1988). Investigations on the structure-activity relationships of odorous substances. Part 1. Detection thresholds and odour qualities of aliphatic and alicyclic compounds containing oxygen functions. Zeitschrift für Lebensmitteluntersuchung und Forschung A, 187, 215–223. SONG, J. and BANGERTH, F. (1994). Production and development of volatile aroma compounds of apple fruits at different times of maturity. Acta Horticulturae, 368, 150–159. SONG, J. and BANGERTH, F. (2003). Fatty acids as precursors for aroma volatile biosynthesis in pre-climacteric and climacteric apple fruit. Postharvest Biology and Technology, 30, 113–121. TAKEOKA, G. R., FLATH, R. A., GUNTERT, T. R., TERANISHI, R. and GUENTERT, M. (1990). Volatile constituents of apricot (Prunus armeniaca). Journal of Agricultural and Food Chemistry, 38, 471–477. TAKEOKA, G. R., BUTTERY, R. G. and FLATH, R. A. (1992). Volatile constituents of Asian pear (Pyrus serotina). Journal of Agricultural and Food Chemistry, 40, 1925–1929. TAKEOKA, G. R., BUTTERY, R. G. and LING, L. (1996). Odour thresholds of various branched and straight chain acetates. Lebensmittel Wissenschaft und Technologie, 29, 677–680. VAN GEMERT, L. J. and NETTENBREIJER, A. H. (1977). Compilation of Odour Threshold Values in Air and Water. National Institute for Water Supply, Voorburg, The Netherlands. 79 pp. VICK, B. A. (1991). A spectrophotometric assay for hydroperoxide lyase. Lipids, 26, 315–320. VILLATORO, C., ALTISENT, R., ECHEVERRÍA, G., GRAELL, J., LÓPEZ, M. L. and LARA, I. (2007). Changes in biosynthesis of aroma volatile compounds during on-tree maturation of ‘Pink Lady’® apples. Postharvest Biology and Technology, in press. WHITE, A. G. (1991). The ‘Gala’ apple. Fruit Varieties Journal, 45, 2–3.