Introduction

Industrial processing of soft fried onions in Denmark includes cleaning, sorting, removal of root and top, and knife or steam peeling. Alternatively the root, top and dry outer leaves are removed by carborundum peeling only. The following unit operations are cutting of onions to smaller pieces, dipping of the pieces in batter, and frying in a bath of rape oil. During frying browning occurs by the Maillard reaction that is generally weak at the beginning of the season in September and most intense by the end of the season in May to June [1, 2]. In some years and for some deliveries at an industrial processing plant the browning of onions was recognised to be scarce on several occasions during the season. By experience the operators had recognised that a satisfactory brown/yellow colour could be obtained by storage of the peeled onions at 5 °C for a few hours to overnight. No explanation for this has been found.

The reactants in the browning reaction in onion are amino acids and the monosaccharides fructose and glucose that are released from proteins and carbohydrates, respectively. The latter is mainly made up by fructans [3] that consist of a series of homologous oligo- and polysaccharides of fructose (GFn, G=Glucosyl, F=Fructosyl; n>2), which can be considered as derivatives of sucrose. Fructans occur in higher plants such as members of the plant familics Asteraceae, Liliaceae and Gramineae. Fructose-oligomers are, for example, contained in onions, which is one of the most widely distributed liliaceous plants in the world. For example, the concentration of GF3, GF4, GF5 and GF6 in eight cultivars of common onion was 11, 10, 8 and 13 w/w% of water-soluble carbohydrates, respectively, at harvest [4].

During bulb growth fructans are synthesized from sucrose by stepwise addition of fructosyl residues leading to oligofructosides with increasing chain length from GF3 to GF20 or higher [5, 6, 7]. The basis for the synthesis of fructans in plants is the three trisaccharides 1-kestose (Inulin series), 6-kestose (6-kestose-, Phlein- and Levan series) and neokestose (Neokestose series) [8]. The synthesis of fructans is catalysed by three fructosyltransferases: 1. sucrose:sucrose 1F-β-d-fructosyltransferase that transfer a fructosyl residue from one molecule sucrose to another sucrose molecule by synthesis of the trisaccharide 1-kestose, 2. 1F-fructosyltransferase that promote the length of fructans by transferring a fructosyl residue from one fructan to another fructan and 3. 6G-fructosyltransferase that transfer a fructosyl residue from 1-kestose to C-6 of the glucose moiety of sucrose.

Mechanical impact of plants with or without tissue wounding may initiate synthesis of plant hormones such as ethylene, absisic acid, gibberellin, systemin or jasmonic acid [9, 10] that promote gene decoding and synthesis of several different enzymes such as phenylalanine lyase in potatoes [11], cellulase, leucine amino peptidase and superoxide dismutase in tomato fruits [10, 12], and peroxidase and cell wall degrading enzymes in cucumber fruits [13, 14]. Based on this, it was hypothesised that enzymes catalysing hydrolysis of fructans may be induced by mechanical impact at several steps included in processing of soft fried onion as found by processing of pre-peeled potatoes [15]. In fact previous research has shown that onion may be very susceptible to mechanical impact by falling [16, 17] and that suberin was deposited in intercellular air spaces [18]. Therefore initiation of enzyme synthesis by mechanical impact may lead to synthesis of exohydrolases and increased enzymatic hydrolysis of fructans that is the major source of the increased energy production necessary for a repair of damaged tissues [3]. As a result of this the fructans may be hydrolysed to a mixture of low molecular fructans and finally to fructose and glucose as found previously [19, 20]

Fructan hydrolysis during storage is catalysed by exohydrolases that remove fructosyl residues from the non-reducing end of the fructan molecules and occur with increasing activities from the internal towards the external outer leaf bases [21]. Exohydrolases are β-(2-6)-linkage specific [4, 20, 22, 23] and the distribution of fructans obtained by hydrolysis during storage of onions has previously been shown to be 65%, 25% and 10% for 1-kestose (GF2), nystose (GF3) and 1F-β-d-fructofuranosylnystose (GF4), respectively [20, 24, 25]. In addition, glucose, fructose, sucrose, and neokestose are released by enzymatic hydrolysis of onion tissue during storage [25].

The aim of the present study was to investigate the effect of field drying, cutting, peeling, and storage of raw and peeled onions on the concentration of non-structural carbohydrates such as glucose, fructose, sucrose and low molecular oligofructosides, and browning.

Materials and methods

Field experiments and sampling

Four cultivars of common onion (Allium cepa L. var. cepa, cv. Hyduro, Hyton, Summit, and Wembley) and two cultivars of shallot (Allium cepa L. var. ascalonium, cv. Bonilla and Matador) were grown on sandy loam by normal practice on the experimental field at Research Centre Aarslev [26]. Harvesting was carried out at 80% fallen top and the onions were field cured for two weeks and dried at 24 °C before storage or processing. Onions of size 40–60 mm (o.d.) were used in the experiments described below.

Handling and processing

Onions (cv. Hyton) were harvested in September 1998. As shown in Table 1 the experimental design encompassed ten treatments, including field drying (A and B), storage of the raw unprocessed onions at 15 °C (E and F) for one or two weeks, removal of top and root (D and E), peeling by hand or industrial steam peeling (B and C), and storage of the peeled onions at 5 and 20 °C (B, C, E and F), respectively. For each treatment seven randomised samples of ten raw onions were stored for 0, 8, 24, 32, 48, 56 or 72 hours until analysis.

Table 1 The experimental design for the handling and processing experiment
Full size table

Damage and short-term storage

Onions were harvested in September 2001 (cv. Hyton) and the experiment included controlled damage of cured and dried non-peeled onions at three levels: 1) no damage, 2) six drops from a height of 35 cm to the flour made of concrete and 3) two drops from 105 cm to the same floor before cold storage at 5 °C, RH 75–80%, of three samples of 60 onions for 0, 3 and 6 months, respectively. By sampling from cold storage the onions were peeled using a handheld knife and split into 3 lots each with 20 onions that were stored at 20 °C for 0, 24 and 72 h until frying, sample preparation and analysis.

Frying

Onions from the experiment with mechanical damage and storage at 5 °C were cut into approximately 5-mm pieces using a kitchen knife and frozen and stored at −24 °C until frying. Five samples of 50 g raw onion were used for each experimental treatment by dipping in an experimental fryer (Varpelev, Aarslev, Denmark) with 140 L rape oil at 140 °C for 3 min as described previously [1]. The colour of the cooled soft fried onion was evaluated visually using a scale from zero to 10 for no browning and severe browning, respectively.

Cultivars and long-term storage

Four cultivars of common onion (Hyduro, Hyton, Summit, Wembley) and two shallots (Bonilla, Matador) were harvested in 1998, cured, dried and stored in a ventilated room at 1 °C, RH 75–80% for 6, 7, 8 and 10 months. Each experimental level included 25 onions.

Sample preparation for extraction and analysis

The peeled onions were cut into 5-mm pieces using a Robot Coupe SA kitchen cutter (CL 50 B, Montceau en Bourgogne, France) and then frozen in liquid nitrogen and freeze dried at 22 °C in a Martin Christ freeze drier (γ 1-20, Osterode, Germany). All samples were stored in a freezer at −24 °C until analysis. The dried material was treated in a micro hammer mill (Culatti, Bie and Berntsen, Rødovre, Denmark) equipped with a 1-mm sieve. Non-structural carbohydrates were extracted from 60 mg freeze dried material by adding 50 ml ultra pure boiling water generated by an Elgastat Maxima Analytica water purification system (Elga Ltd., England). The extracts were kept at room temperature for 3 h and then filtered through a 0.45-μm AcetatePlus Cameo filter. Samples for analysis of non-structural carbohydrates were diluted 12.5 times with ultra pure water; in this case 4 ml of onion extracts were diluted with 46 ml ultra pure water.

Analysis of non-structural carbohydrates

Fructose, glucose, sucrose and low molecular oligofructosides were measured by analytical high performance anion exchange chromatography (HPAEC) according to the method described by Campbell et al. [19] with a few modifications. Quantification of non-structural carbohydrates were performed on a Dionex Series 300DX ion chromatograph (Sonnyvale, US) coupled with pulsed amperometric detection (PAD) and equipped with a borate trap, a CarboPac PA1 guard (4×50 mm) and a CarboPac PA1 (4×250 mm) analytical column. Separations were performed by gradient elution with solvent A (200 mmol/l NaOH (JT Baker, 50%, Deventer, The Netherlands), solvent B (100 mmol/l NaOH and 600 mmol/l CH3CO2Na (Fluka, 71180, Steinheim, Germany) and solvent C (water). Elution profile: 0.0 min (10% A, 0% B), 10.0 min (20% A, 0% B), 38.0 min (38% A, 24% B), 42.0 min (38% A, 24% B), 47.0 min (25% A, 50% B), 60 min (0% A, 0% B), 70 min (10% A, 0% B). All changes in the gradient were linear programmed and all eluants were prepared with ultra pure carbonate-free water. Injection volume: 20 μl. Flow: 1.0 ml/min.

For quantification and identification of non-structural carbohydrates standard solutions containing 0, 2, 4, 8 and 12 μg/100 ml of glucose, fructose, sucrose (Sigma-Aldrich, Steinheim, Germany), 6-kestose, neokestose (produced according to the method of Chatterton et al. [27]), 1-kestose, nystose and 1F-β-d-fructofuranosylnystose (Wako Pure Chemical Industries, Ltd. Osaka, Japan) were used. Unidentified carbohydrates were quantified from the calibration curve of the known carbohydrate nearest to the unknown compound. All determinations were carried out in two replicates.

For the determination of the total oligofructosides the difference between hydrolysed and non-hydrolysed carbohydrates was used [2] and total oligofructosides was calculated according Hoebregs [28]. A filtered sample of 500 μl extract was hydrolysed with 50 μl 2 N trifluoroacetic acid (TFA) at 40 °C for 60 min [29]. Quantification of carbohydrates before and after hydrolysis was performed by HPAEC-PAD using a CarboPac PA-10 guard (4×50 mm) and a CarboPac PA-10 (4×250 mm) analytical column. Separations were performed by gradient elution with solvent A (200 mmol/l NaOH) and solvent B (water). Elution profile: 0.0 min (12% A), 10.0 min (30% A), 15.0 min (40% A, 24% B), 17.0 min (100% A), 29.4 min (100% A), 29.9 min (12% A), 35 min (12% A). All changes in the gradient were linear programmed and all eluants were prepared with ultra pure carbonate-free water. Injection volume: 20 μl. Flow: 1.0 ml/min.

Statistics

All treatments and analyses were carried out in two replicates. The statistical methods used were general linear methods (GLM) by analysis of variance (AOV) and linear regression with F-test for the linear model (P<0.05). Simple linear regression (y=a+bx) was applied for sugars and low molecular fructans and a square root model (y=a+bx0.5) was used for fructan fructose (fructose released by hydrolysis of fructans), fructan glucose (glucose released by hydrolysis of fructans) and total fructans (fructan fructose+fructan glucose). The constants and regression coefficients were tested using t-tests. Averages were separated using multiple range tests and significant differences are indicated by letters. All statistical analyses were carried out using a Statgraphic Statistical Package (Statistical Graphics Corporation, Version 4, Rockville, USA).

Results and discussion

Analysis of non-structural carbohydrates in onion bulbs

The water extracts of onion bulbs of various cultivars subjected to different storage times and processing treatments were analysed by HPAEC-PAD. Figure 1 represents a typical HPAEC chromatogram of an onion bulb extract from shallot and common onion, respectively. Peak separation was excellent with reasonable retention times (Table 2), allowing all carbohydrates to be resolved to baseline, identified and quantified. Peaks corresponding to glucose, fructose, sucrose and the oligofructosides 1-kestose (1F-β-d-fructofuranosylsucrose, GF2), neokestose (6G-β-d-fructofuranosylsucrose, GF2), nystose (1F-(1-β-d-fructofuranosyl)2 sucrose, GF3) and 1F-β-d-fructofuranosylnystose (GF4) were detected in extracts of onion bulbs together with four fructan oligomers of unknown structures (unknown F1–F4). Fructose, glucose, sucrose, 1-kestose, neokestose, nystose and 1F-β-d-fructofuranosylnystose, that are well-known constituents of onion bulbs [25, 30], were identified by spiking with authentic standards. Detectability was also very good for these non-structural carbohydrates as well as linear detectability for the PAD detector. The latter depends upon many factors such as column type, sensitivity settings, and the component being detected. The linearity for the settings and conditions described under Materials and Methods was determined to be in the range 0–12 μg/100 ml. The relative standard deviation (RSD) from ten injections to determine precision were 0.80, 0.71, 1.77, 1.28, 1.39, 1.70, 1.34, 1.90, 1.56, 1.80 and 1.60% for glucose, fructose, sucrose, 1-kestose, nystose, unknown F1, neokestose, unknown F2, unknown F3, unknown F4 and 1F-β-d-fructofuranosylnystose, respectively. The practical lower limit of the peak area was evaluated to be 3 SD, which is far below the area of any carbohydrate measured.

Fig. 1a,b
figure 1

Non-structural carbohydrate HPAEC profile after extraction of: a shallot onion bulbs (cv. Bonilla); b common onion bulbs (cv. Heyton): 1. glucose, 2. fructose, 3. sucrose, 4. 1-kestose, 5. unknown F1, 6. neokestose, 7. nystose, 8. unknown F2, 9. unknown F3, 10. unknown F4, 11. 1F-β-d-fructofuranosylnystose. nC=nanocoulomb

Full size image
Table 2 The effect of field drying of raw oniona for one and two weeks on dry matter, non-structural carbohydrates and total fructans
Full size table

Besides the carbohydrates identified in this study a further six carbohydrates have previously been reported to be present in onion bulbs including the trisaccharide 6-kestose [31], and the tetra- and pentasaccharides 6G-(1-β-d-fructofuranosyl)nsucrose (n=2 and 3) and 1F-(1-β-d-fructofuranosyl)m-6G-(1-β-d-fructofuranosyl)n sucrose (m=1, n=1; m=2, n=1; and m=1, n=2) [25, 30]. Some of these non-structural carbohydrates may correspond to the unidentified fructan oligomers described as unknown F1-F4 (Fig. 1, Table 2). However, spiking experiments with an authentic standard of 6-kestose excluded the presence of this oligofructoside in the onion bulb extracts.

The composition of non-structural carbohydrates in the investigated common onions with relative high levels of glucose, fructose and sucrose compared to the fructans 1-kestose, nystose and neokestose, and very small quantities of penta- and hexasaccharides were as found in previous investigations [3, 6, 19, 21, 32, 33, 34, 35].

Field drying

Fructose was significantly lower after field drying for two weeks compared to one week (Table 2). The other sugars, oligofructosides, and total fructans were unaffected by field drying time. Storage of onion at 15 °C for one or two weeks before drying did not influence non-structural carbohydrates (data not shown).

Removal of top and root and peeling

When the top and root was removed before storage of the onions for one week at 15 °C, significant decreases was found for glucose, fructose, and unknown F1, whereas significant increases in 1-kestose, neokestose, nystose and unknown F2 were detected (Table 3). The other unknown compounds, sucrose, 1F-β-d-fructofuranosylnystose and total fructans were unaffected. The decreases probably occur because of respiration and wound healing at the cut surface including deposition of suberin [18]. In spite of the non-significant difference for total fructans it was supposed that these biochemical processes require increased synthesis of hydrolytic enzymes and hydrolysis of high molecular fructans and thereby an increase in 1-kestose, neokestose, nystose and unknown F2 [36]. The changes caused by removal of top and root may be related to increased gene expression, synthesis of RNA, synthesis of enzymes and enzymatic hydrolysis of storage fructans in onion as reported previously for starch in potatoes [11].

Table 3 The effect on dry matter, non-structural carbohydrates and total fructans in common oniona with top and root being present or removed when stored for one week at 15 °C
Full size table

Degradation of fructans by exohydrolytic enzymes have been found in grass (Lolium perenne L.) and Jerusalem artichoke that degrade fructan polymers by removing the terminal fructose residue from fructans resulting in release of fructose [22, 23] and it has been shown that the major change during storage of onions is hydrolysis of higher fructans to lower oligofructosides [36, 37].

Peeling

Dry matter increased while fructose and unknown F1 decreased significantly by steam peeling; however, glucose, sucrose and fructans were not affected by peeling (Table 4).

Table 4 The effect on dry matter, non-structural carbohydrates and total fructans in hand peeled or industrially peeled onionsa using steam at 120 °C for 30 s
Full size table

Storage of peeled onions

Increases in dry matter by storage of peeled onion (Table 5) may be due to production of wound tissues and deposition of suberin or similar compounds in the air tissue as found for onions and potatoes [15, 18]. No significant differences were found for glucose, sucrose, nystose, unknown F1, F3 and F4. The high positive and significant regression coefficient for fructose and significant negative regression constants for 1-kestose, neokestose, unknown F2 and 1F-β-d-fructofuranosylnystose shows that the increase in fructose may be due to hydrolysis of fructans according to the model y=a+bx were y is fructose and x is fructans.

Table 5 The effect of storage on non-structural carbohydrates in peeled onionsa
Full size table

Fructan fructose (fructose released by hydrolysis of fructans) and fructan glucose (glucose released by hydrolysis of fructans) contribute with 67 and 33% of total fructans, respectively (Table 5), which indicates that trisaccharides constitute a major part of fructans in onion as pointed out by Jaime et al. [36]. The hydrolysis rate for high molecular fructans followed a square root model: y=a+bx0.5 with relatively large negative regression coefficients showing hydrolysis of fructans to fructose as found previously [3, 6, 19].

Short-term storage of non-peeled onions at 5 °C

Table 6 shows the effect of storage of raw onions at 5 °C on dry matter and non-structural carbohydrates during a six-month storage period. Dry matter and sucrose were unaffected, while glucose decreased as well as the concentration of all fructans because of hydrolysis to fructose that increased from the start to three months of storage (Table 6). This is in contrast to other results, which have shown that a maximum content of fructose in raw onions occurred after approximately seven months of storage [33]. Glucose may also be released by hydrolysis of fructans; however, this compound decreased probably due respiration and use in other biochemical processes during storage. The increases in fructose resulted in increased browning (Table 6 and 7) as expected [38].

Table 6 The effect of storage (at 5 °C) of raw onionsa on dry matter and non-structural carbohydrates
Full size table
Table 7 The effect of mechanical impact and storage on the colour of soft fried onionsa
Full size table

Effect of mechanical impact

Common onion was found to be very susceptible to mechanical impact by dropping 1 cm that gave rise to bruising [17]; however, there is some discrepancy with this because is has been found that drops between 40 and 100 cm were necessary in order to produce severe damage [16].

Table 8 shows the effects of three levels of mechanical impact by dropping of the raw onion onto a floor made of concrete. Glucose and fructose decreased while 1-kestose and sucrose increased by mechanical impact before storage. The remaining carbohydrates were unaffected. In spite of the clear changes by short-term storage, the applied mechanical impact only affected non-structural carbohydrates slightly (Table 8). The reason for this could be: 1) that the impact by falling was too low in comparison to peeling, 2) that the damage was repaired by wound reactions or 3) that onion may have the largest resistance against mechanical impact in the beginning of the storage period [39]. The changes in carbohydrate composition, primarily fructose, did not increase enough by mechanical impact to affect significantly browning by frying (Table 7).

Table 8 The effect of mechanical impact before storage (at 5 °C) of raw onionsa on dry matter and non-structural carbohydrates
Full size table

Storage of peeled and damaged onions at 20 °C

Significant changes in the concentration of non-structural carbohydrates occurred by storage of the damaged and peeled onions at 20 °C (Table 9). Dry matter, fructose, 1F-β-d-fructofuranosylnystose and unknown F4 increased significantly in the peeled onions, while glucose, sucrose, nystose, unknown F2 and F3 decreased. Only unknown F1 remained unaffected. These changes resulted in significantly increased browning by storage for up to 72 h (Table 7).

Table 9 The effect on dry matter and non-structural carbohydrates of peeled onionsa stored at 20 °C
Full size table

Cultivars and long-term storage

The concentrations of the non-structural carbohydrates glucose, fructose, sucrose, 1-kestose, neokestose, nystose and fructofuranosylnystose found in the bulbs of common onion and shallot cultivars (Table 10) were within the concentration range found previously [3, 19, 20, 33, 34]. The common onion cultivars were significantly different with respect to dry matter, glucose, fructose, sucrose, fructan fructose and total fructans; however, the pattern of non-structural carbohydrates was similar to that found previously in common onions (Hyduro, Hyton, Summit, Wembley) [19, 35, 36]. The shallot cultivars (Bonilla, Matador) were significantly different with respect to dry matter and nearly all non-structural carbohydrates, and also had a significantly higher content of dry matter and nearly all non-structural carbohydrates compared to common onions (Table 10).

Table 10 Dry matter and non-structural carbohydrates in stored onionsa
Full size table

The changes during long-term storage of raw onions at 1 °C (Table 11) showed that dry matter, glucose and fructose were almost constant until the last sampling after storage of the raw onions for ten months. Fructan glucose, fructan fructose and total fructans decreased during the whole sampling period following a square root model. This is in accordance with other findings [34] that showed a decrease in tri-, tetra-, penta- and higher carbohydrates from beginning of the storage period to a few weeks before sprouting.

Table 11 Carbohydrates in raw onionsa during storage at 1 °C
Full size table

Storability of onions has previously been shown to be inversely related to the content of fructose and exohydrolase activity [34] and increases with increasing dry matter [4, 34]. Mechanical damage by transport may decrease storability; however, this may be diminished by proper curing and ventilation [40, 41, 42].