- Email: [email protected]
Non-destructive flavour evaluation of red onion (Allium cepa L.) Ecotypes: An electronic-nose-based approach
Food Chemistry 141 (2013) 896–899
Contents lists available at SciVerse ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Non-destructive flavour evaluation of red onion (Allium cepa L.) Ecotypes: An electronic-nose-based approach Mariateresa Russo a,b,⇑, Rosa di Sanzo a, Vittoria Cefaly a, Sonia Carabetta a, Demetrio Serra b, Salvatore Fuda a a b
Department of Agriculture, Laboratory of Food Chemistry, University of Reggio Calabria, Feo di Vito, 89100 Reggio Calabria, Italy Research Laboratory ‘‘Quality and Food Safety and New Technologies’’ – Qua.Sic.A.Tec. MIUR Acc n. 28/2010, Area Industriale, 88046 Lamezia Terme (CZ), Italy
a r t i c l e Article history: Received 4 October Received in revised Accepted 14 March Available online 22
i n f o 2012 form 9 March 2013 2013 March 2013
Keywords: Alliums cepa L. ‘‘Tropea’’ red onion Artificial olfactory system Electronic nose Metal oxide semiconductor Pyruvic acid
a b s t r a c t This work reports preliminary results on the potential of a metal oxide sensor (MOS)-based electronic nose, as a non-destructive method to discriminate three ‘‘Tropea Red Onion’’ PGI ecotypes (TrT, TrMC and TrA) from each other and the common red onion (RO), which is usually used to counterfeit. The signals from the sensor array were processed using a canonical discriminant function analysis (DFA) pattern recognition technique. The DFA on onion samples showed a clear separation among the four onion groups with an overall correct classification rate (CR) of 97.5%. Onion flavour is closely linked to pungency and thus to the pyruvic acid content. The e-nose analysis results are in good agreement with pyruvic acid analysis. This work demonstrated that artificial olfactory systems have potential for use as an innovative, rapid and specific non-destructive technique, and may provide a method to protect food products against counterfeiting. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Onion (Allium cepa L.) is a species of great economic importance, which is widely cultivated all over the world. The health benefits of onion consumption include protection against cancer, coronary heart disease, diabetes and ageing. These effects have been mainly attributed to flavonoids, vitamins and organosulphur compounds (Goldman, Kopelberg, Debaene, & Schwartz, 1996; Kumari, Mathew, & Augusti, 1995). Although they possess significant nutritional value, onions are primarily consumed for their unique flavour and for their ability to enhance the taste of foods (Kopsell & Randle, 1997; Rodrigues et al., 2003). Onions are widely cultivated throughout the Mediterranean basin and in Italy. The ‘‘Tropea Red Onion’’ (TrO), a sweet red onion characterised by its elongated bulb and almost white flesh, is cultivated, in three different ecotypes, only in the Tyrrhenian coastal areas of Calabria region (Southern Italy). In March 2008, the European Union registered the Protected Geographical Indication (PGI) certification for the onions produced in this particular area (Commission Regulation (EC) No. 284/2008). The ‘Cipolla Rossa di Tropea – Calabria’ PGI denotes bulbs of the species A. cepa ⇑ Corresponding author at: Department of Agriculture, Laboratory of Food Chemistry, University of Reggio Calabria, Feo di Vito, 89100 Reggio Calabria, Italy. Tel.: +39 0968209606; fax: +39 0968209361. E-mail address: [email protected] (M. Russo). 0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2013.03.052
exclusively from the local ecotypes. These have a characteristic shape according to time of production: ‘Tonda Piatta’, the early crop; ‘Mezza Campana’, the mid to early crop; ‘Allungata’, the late crop. TrO, for its tenderness, crispness and sweetness is a highly regarded Italian horticultural product and one of the most counterfeited. Sensory evaluation of onion flavour is closely linked to its pungency, which shows a close correlation with its pyruvate content (Schwimmer & Weston, 1961). The pungent flavour of onions is produced by hydrolysis of precursor compounds, such as S-alk(en)yl-L-cysteine sulfoxides (Lancaster, Shaw, Joyce, McCallum, & McManus, 2000), when the cells are mechanically ruptured. The hydrolysis reaction is catalysed by allinase and is completed within 6 min (Schwimmer & Weston, 1961). This reaction produces thiopropanol S-oxide (a lacrymator), ammonia, sulfur volatiles and pyruvic acid (Fig. 1; Block, 1992). Pyruvic acid concentration (lmol/g FW) in macerated onion tissue is used as a quality assurance indicator of pungency (Abayomi, Terry, White, & Warner, 2006; Pineda, Lué- Merù, Rivas, et al., 2004). Pyruvic acid exists universally in the plant tissues as part of an intermediate metabolism (Goodwin & Mercer, 1983). For this reason, the background levels of pyruvic acid or control (PC) need to be subtracted from the total pyruvic acid (PT) concentrations to calculate the enzymatically produced pyruvate (PE). Background pyruvic acid can be measured after the alliinase is deactivated. The deactivation of alliinase has been achieved by heating onion tissues in a
897
M. Russo et al. / Food Chemistry 141 (2013) 896–899
using a canonical discriminant function analysis (DFA) pattern recognition tool. DFA has been used extensively to perform pattern recognition and it has been reported to produce good performance for the classification of foodstuffs (Russo et al., 2012). DFA was used to check the capability of the e-nose system in assigning onion samples to a specific group. Finally, the aim of this study was to discriminate flavour and aroma characteristics of the early, middle and late crops of Tropea Red Onion and also distinguish between them and the common red onion by the use of an electronic nose. Studies on different ecotypes, to the best of our knowledge, have not been carried out before. The analysis of pyruvic acid by HPLC with UV detection was also carried out.
Fig. 1. Schematic of the compounds formed in onion from the hydrolysis of flavour precursors (a: S-alk(en)yl-cysteinsulfoxide; b: alk(en)yl-sulfenic acid; c: pyruvic acid; d: ammonia).
microwave oven (Schwimmer & Weston, 1961; Yoo, Pike, & Hamilton, 1995) or by homogenising with trichloroacetic acid (Randle & Bussard, 1993). Since the PC level can substantially change the PE concentration, its accurate measurement is critical in measuring onion pungency (Boyhan, Schmidt, Woods, Himelrick, & Randle, 1999; Thomas, Parkin, & Simon, 1992; Yoo & Pike, 2001). Pyruvic acid content determination and organoleptic tests are used for assessing the eating characteristics of onions. The accumulation of organosulfur compounds in onions and pyruvic acid content depends upon many factors, especially sulfur-based fertilisation and environmental growing conditions, although genotype the most important factor (Chope, Terry, & White, 2007; Yoo, Pike, Crosby, Jones, & Leskovar, 2006). The genetic background seems to be directly related to the different abilities to control sulfur uptake and assimilation in the biosynthetic pathway that results in the onion’s flavour (Randle, 1992; Randle, Block, Littlejohn, Putman, & Bussard, 1994). Onion flavour is traditionally measured through the use of a combination of conventional analytical instrumentation and human organoleptic profiling panels. These methods are expensive not only in terms of time, but are also inaccurate because of a lack of either sensitivity or quantitative information. In this paper an investigation was carried out to determine the flavour of different onion samples using an electronic nose (e-nose) and thus to explore the possibility of replacing existing analytical methods. The e-nose is a relatively novel device used for volatile sensing. The first pioneering studies about the concept of an artificial nose system able to measure odours was reported in 1982 by Persaud and Dodd of the University of Warwick, Coventry, UK (Cho et al., 2010). The e-nose has been designed for automated detection and recognition of odours, vapours and gases (Cho et al., 2010). It does not separate the volatile fraction of the matrix into its constituents but supplies a global evaluation of aroma, mimicking the human olfactory system with instrumental objectivity. Electronic nose technology has been successfully used to discriminate quality and flavour of various products, including tomatoes, citrus, spices and onions (Abbey & Joyce, 2007; Russo, Serra, Suraci, & Postorino, 2012). The technology may also be used to assess quality of stored grain, fish, drugs, drinks and food spoilage (Abbey, Aked, & Joyce, 2001). The e-nose used in this work comprises an array of semiconductor gas sensors (MOS), each of which has an electrical resistance that has partial sensitivity to the headspace of onion. The signals from the sensor array are then conditioned by suitable interface circuitry, resulting in an onion data-set. The data were processed
2. Materials and methods 2.1. Plant materials In 2011 the three local ecotypes of Tropea red onion and one red onion cultivar usually sold as Tropea, were analysed and compared (Table 1). All onion types were grown in a field in IGP district and on the same farm, so that the evaluation was not influenced by environmental factors and/or cultural practices. Three different samples were collected for each type, consisting of five onions. Each analysis was performed in triplicate for each sampling. 2.2. Electronic nose An electronic nose (ISENose 2000, Labservice Analytica, Bologna, Italy) was used to discriminate aroma fingerprints of onion ecotypes according to the method of Abbey et al. (2001). The edible parts of the onions were homogenised in an Ultra-Turrax system (T50 basic, IKA Werke, Staufen, Germany) at room temperature. The slurry was filtered after 20 min and 20 ml of the filtrate were transferred into a 250-ml flask. Twenty millitres of 5% trichloroacetic acid (TCA; Sigma Chemical Co., Milano, Italy) were added to terminate alliinase activity. The mixture was agitated vigorously and allowed to stand for 30 min. Deionised water (10 ml) was added to 10 ml of filtrate/ TCA solution and mixed. One millilitre of this diluted solution was put into a 25-ml vial, and equilibrated for 10 min at 25 °C to allow the development of headspace before e-nose analysis. The instrument, equipped with 12 metal oxide semiconductor (MOS) sensors, was used to generate a pattern for the compounds present in the headspace of the extracted red onion samples. The operating conditions in the acquisition phase were: baseline correction 10 s; injection time 10 s; sampling 180 s; delay 300 s; purge 60 s; carrier chromatographic air (80% N2, 20% O2) flow 300 ml/min; vial volume 25 ml. 2.3. Pyruvic acid extraction procedure and HPLC analysis The quantification of the pyruvic acid content in the onion was conducted using a modified method of Yoo and Pike (2001). The neck, base and central leaf tissues (about 5 cm long) were removed and the remaining fresh scales were cut into 1–2 cm squares using
Table 1 Red Onion ecotypes. ID
Sample
TrT TrMC TrA RO
Tropea Ecotype ‘‘Tonda’’ Tropea Ecotype ‘‘Mezza Campana’’ Tropea Ecotype ‘‘Allungata’’ Red Onion
898
M. Russo et al. / Food Chemistry 141 (2013) 896–899
a razor blade. The cut pieces (10 kg) were mixed thoroughly, rinsed with water and drained to remove the soil residues. As a control to measure the background level of pyruvic acid (PC), 100 g tissue were placed in a plastic bag (17.5 cm 21 cm). The samples were heated in a microwave oven (AFM 442 Ultimys Duo Grill; Moulinex, Italy) for 3 min at 900 W to deactivate the alliinase. After microwaving, the sample was placed in a bag under ice to stop the cooking. The tissue was blended in an Ultra-Turrax system for 2 min and the juice was collected through centrifugation (Eppendorf 5804R; Milano, Italy) and filtration with a syringe filter. To measure total (background and alliinase-produced) pyruvic acid content (PT), 100 g tissue were blended in an Ultra–Turrax system for 2 min. The homogenised mixture was macerated for 30 min at room temperature and juice was collected through centrifugation and filtration with syringe filter. Chromatographic analysis on onion extracts was carried out using an HPLC (LC-10AD system; Shimadzu, Kyoto,Japan) with diode array detector. Each sample extract was injected in duplicate on a 300 7.8 mm i.d. Supelcogel C610H (Sigma–Aldrich, St Louis, MO) and eluted using 0.3 ml/l H2SO4 at a flow rate of 0.5 ml/min. The eluate was monitored at 210 nm. Standards were prepared with pyruvic acid, purchased from Sigma Chemical Co. (Milano, Italy). Stock and working standards were prepared in HPLC-grade methanol. Enzymatically produced (from alliinase) pyruvate in each sample was calculated from the difference of the two determinations. All analyses were run in triplicate. 2.4. Data analysis The pyruvic acid content in onion extracts was identified by its retention time characteristics. The linearity of the detector response for the prepared standards was assessed by means of linear regression, based on the amounts of each standard and the area of the corresponding peak on the chromatogram. The limit of detection (LOD) and quantification (LOQ) were determined by calculation of the signal-to-noise ratio. Signal-to-noise ratios of approximately 3:1 and 10:1 were used for estimating the detection limit and quantification limit, respectively. LOD and LOQ values of pyruvic acid were 0.029 and 0.088 mmol/kg respectively; r2 Values of the compounds were higher than 0.999, thus confirming the linearity of the method. The precision of injection was demonstrated by replicate injections of the standard solution and the relative standard deviation (RSD) of peak area was 0.38% (n = 8). The sensors’ responses were evaluated using discriminant function analysis (DFA) by IseNose OCS software (version DT 01837 Rev. 02A, ISENose 2000; Labservice Analytica, Bologna, Italy). 3. Results and discussion The sensor responses were evaluated using Discriminant Function Analysis (DFA). The four samples (three local ecotypes and one common red onion) were chosen as grouping variables and sensors outputs were used as independent variables. A comprehensive database (500 data points) was developed to increase the reliability of the analysis. Four different sets of e-nose data clusters for onion headspace volatiles were identified in the DFA analysis (Fig. 2). Three clusters matched the three ecotypes of Tropea red onion (TrT red, TrA azure and TrMC blue) and one cluster matched the red onion (RO green). E-nose data set clusters for headspace volatiles of three types of Tropea red onion and common red onion are shown in Fig. 2. The e-nose discriminates between the three Tropea Red Onion ecotypes and between them and common RO (Fig. 2). Discrimination
Fig. 2. The three-dimensional Discriminant Function Analysis (DFA) plot for e-nose data set clusters of headspace volatiles of the different onion ecotypes (TrT: red, TrA: azure, TrMC: blue and RO: green).
is considered acceptable when it reaches values up to 80%; values around 85% are considered a sign of good discrimination, while values up to 90% indicated excellent discrimination. The DFA on onion samples showed a clear separation among the four onion groups, with an overall correct classification rate (CR) of 97.5%. These results prove that the e-nose is capable of discriminating between the flavours of onions. The e-nose results were compared with pyruvic acid content estimates. It is generally accepted that there is a high correlation between levels of enzymatically-produced pyruvate and perception of pungency (Schwimmer & Weston, 1961). The amou nt of pyruvic acid produced due to alliinase activity following tissue damage was evaluated in all samples. The theoretical 1:1 M ratio between pyruvate and its flavour precursors (alk(en)ylcysteinesulphoxides) (Yoo & Pike, 2001) suggested that the HPLC analysis of pyruvic acid content may represent a useful estimation of onion pungency (Bacon et al., 1999). Previous study reports considerable variation in background pyruvic acid concentrations. Schwimmer and Weston (1961) reported background levels of 2.1–4.1 lmol/ml from White Grano, Southport Red Globe, and five other cultivars. Randle and Bussard (1993) reported 0.1–1.0 lmol/ml pyruvate from 16 shortday onion cultivars, using the original Schwimmer and Weston method. Yoo et al. (1995) reported less than 0.5 lmol/ml using undiluted onion juice and suggested that measurement of the background level could be disregarded. Luo and Ewart (1995) reported background levels of 0.79 lmol/ml. The background pyruvic acid concentrations in our analysed samples showed levels of 0.22 ± 0.12 to 0.30 ± 0.11 lmol/g, lower than previously reported. The sweet onion industry in USA classifies onions on the basis of pungency as low pungency/sweet (0–3 lmol pyruvic acid/g FW), medium pungency (3–7 lmol pyruvic acid/g FW), and high pungency (above 7 lmol pyruvic acid/g FW) (Dhumal, Datir, & Pandey, 2007). Based on this classification, the sample RO can be considered as a medium pungent cultivar (3.79 ± 0.86 lmol/g) and all the three Tropea ecotypes as low pungency cultivars (Table 2). Of the three Tropea ecotypes, TrA shows the highest concentration of pyruvic acid (2.57 ± 0.81 lmol/g), TrMC shows a middle value (2.45 ± 0.30 lmol/g) while TrT shows the lowest concentration
M. Russo et al. / Food Chemistry 141 (2013) 896–899 Table 2 Content of pyruvic acid in onion juice of different onion ecotypes by HPLC-UV analysis (PC: background levels of pyruvic acid, PT: total pyruvic acid and PE: enzymatically produced pyruvic acid). ID
PT (lmol/g fw)
PC (lmol/g fw)
PE (lmol/g fw)
TrT TrA TrMC RO
2.26 ± 0.61 2.57 ± 0.81 2.45 ± 0.30 3.79 ± 0.86
0.30 ± 0.11 0.26 ± 0.16 0.22 ± 0.12 0.27 ± 0.12
1.96 ± 0.06 2.31 ± 0.81 2.23 ± 0.17 3.52 ± 0.37
(2.26 ± 0.61 lmol/g). The latter results are in agreement with data obtained by the e-nose analysis and with those reported by consumers identifying the ecotype ‘‘Tonda’’ (TrT) as the sweetest. This work showed that the e-nose can discriminate the Tropea red onion from other red onion and the different ecotypes of Tropea Red Onion with respect to their flavour and pungency. 4. Conclusion It has been shown that over 80% of onion pungency depends on genetic factors, while the remaining part is related to external factors, such as the habitat and cultivation practices (Yoo et al., 2006). Therefore the choice of cultivars appears to be a very important factor in the production of onions with a low level of pungency, while the habitat also influences the organoleptic characteristics. The present work showed the potential use of an innovative, rapid, specific and non-destructive technique able to discriminate effectively not only among ecotypes but, especially, between Tropea Red Onion PGI and other red onions. The importance of genetics compared to the habitat and cultivation techniques in the determination of the specific quality characteristics of a food product was demonstrated. The onion samples under investigation were grown under the same conditions in the same field with the same cultivation techniques. Nevertheless, their quality attributes were significantly different. Further studies are in progress to determine which onion headspace volatile chemicals interact with the e-nose sensors, in order to allow a clear correlation between onion quality and e-nose data sets. Acknowledgement This work is part of the ongoing project Dolce Rossa supported by the APQ ‘‘Scientific Research and Technological Innovation’’ (Az. 2 and Az. 3) financed by Calabria Region-Department of Research, European Community, Italian Ministry of Economic Development (MISE) and Italian Ministry of Research Science and Technology (MIUR). References Abayomi, L. A., Terry, L. A., White, S. F., & Warner, P. J. (2006). Development of a disposable pyruvate biosensor to determine pungency in onions (Allium cepa L.). Biosensors and Bioelectronics, 21, 2176–2179.
899
Abbey, L., Aked, J., & Joyce, D. C. (2001). Discrimination amongst Allium using an electronic nose. Annals of Applied Biology, 139, 337–342. Abbey, L., & Joyce, D. C. (2007). Quality assessment of diced onion (Allium cepa L.) using an electronic nose. Ghana Journal of Science, 47, 83–90. Bacon, J. R., Moates, G. K., Neg, A., Rhodes, M. J. C., Smith, A. C., & Waldron, K. W. (1999). Quantitative analysis of flavor precursors and pyruvate levels in different tissues and cultivars of onion (Allium cepa). Food Chemistry, 64, 257–261. Block, E. (1992). The organosulfur chemistry of the genus Allium – Implications for the organic chemistry of sulfur. Angewandte Chemie, International Edition in English, 1135–1178. Boyhan, G. E., Schmidt, N. E., Woods, F. M., Himelrick, D. G., & Randle, W. M. (1999). Adaption of a spectrophotometric assay for pungency in onion to a microplate reader. Journal of Food Quality, 22, 225–233. Cho, H. J., Lim, S. S., Lee, Y. S., Kim, J., Lee, C. H., Kwon, D. Y., et al. (2010). Hexane/ ethanol extract of Glycyrrhiza uralensis licorice exerts potent anti-inflammatory effects in murine macrophages and in mouse skin. Food Chemistry, 121, 956–966. Chope, G. A., Terry, L. A., & White, P. J. (2007). Preharvest application of exogenous abscisic acid (ABA) or an ABA analogue does not affect endogenous ABA concentration of onion bulbs. Plant Growth Regulation, 52, 117–129. Dhumal, K., Datir, S., & Pandey, R. (2007). Assessment of bulb pungency level in different Indian cultivars of onion (Allium cepa L.). Food Chemistry, 100, 1328–1330. Goldman, I. L., Kopelberg, M., Debaene, J. E., & Schwartz, B. S. (1996). Antiplatelet activity in onion (Allium cepa) is sulfur dependent. Thrombosis and Haemostasis, 76, 450–452. Goodwin, T. W., & Mercer, E. I. (1983). In Introduction to plant biochemistry (2nd ed., pp. 204–207). New York: Pergamon Press. Kopsell, D. E., & Randle, W. M. (1997). Onion cultivars differ in pungency and bulb quality changes during storage. HortScience, 32, 1260–1263. Kumari, K., Mathew, B. C., & Augusti, K. T. (1995). Antidiabetic and hypolipidemic effects of S-methyl cysteine sulfoxide isolated from Allium cepa Linn. Indian Journal of Biochemistry & Biophysics, 32, 49–54. Lancaster, J. E., Shaw, M. L., Joyce, M. D. P., McCallum, J. A., & McManus, M. T. (2000). A novel alliinase from onion roots. Biochemical characterization and cDNA cloning. Plant Physiology, 122, 1269–1280. Luo, C., Ewart, L. (1995). A new pyruvic acid (PA) procedure for rapid evaluation of onion pungency. Abstract and program of the 1995 onion research conference. Wisconsin Center, Madison, WI, December 6–9, p. 26. Pineda, M., Lué- Merù, M. P., Rivas, R., Gallignani, M., Valero, M., et al. (2004). Pungency evaluation of onion cultivars from the Venezuelan West- Center region by flow injection analysis-UV–visible spectroscopy pyruvate determination. Talanta, 64, 1299–1303. Randle, W. M. (1992). Onion germplasm interacts with sulfur fertility for plant sulfur utilization and bulb pungency. Euphytica, 59, 151–156. Randle, W. M., Block, E., Littlejohn, M. H., Putman, D., & Bussard, M. L. (1994). Onion (Allium cepa L.) thiosulfinates respond to increasing sulfur fertility. Journal of Agriculture and Food Chemistry, 42, 2085–2088. Randle, W. M., & Bussard, M. L. (1993). Streamlining onion pungency analyses. HortScience, 28, 60. Rodrigues, A. S., Fogliano, V., Graziani, G., Mendes, S., Vale, A. P., & Gonçalves, C. (2003). Nutritional value of onion regional varieties in northwest Portugal. The Electronic Journal of Environmental, Agricultural and Food Chemistry, 2, 519–524. Russo, Mt., Serra, D., Suraci, F., & Postorino, S. (2012). Effectiveness of electronic nose systems to detect bergamot (Citrus Bergamia). The Journal of Essential Oil Research, 24, 137–151. Schwimmer, S. S., & Weston, W. J. (1961). Enzymatic development of pyruvic acid in onion as a measure of pungency. Journal of Agriculture and Food Chemistry, 9, 301–304. Thomas, D. J., Parkin, K. L., & Simon, P. W. (1992). Development of a simple pungency indicator test for onions. Journal of the Science of Food and Agriculture, 60, 499–504. Yoo, K. S., Pike, L., Crosby, K., Jones, R., & Leskovar, D. (2006). Differences in onion pungency due to cultivars, growth environment, and bulb sizes. Scientia Horticulturae, 110, 144–149. Yoo, K. S., & Pike, M. L. (2001). Determinazion of background pyruvic acid concentrations in onions, Allium species, and other vegetables. Scientia Horticulture, 89, 249–256. Yoo, K. S., Pike, L. M., & Hamilton, B. K. (1995). A simplified pyruvic acid analysis suitable for onion breeding programs. HortScience, 30, 1306.
Contents lists available at SciVerse ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Non-destructive flavour evaluation of red onion (Allium cepa L.) Ecotypes: An electronic-nose-based approach Mariateresa Russo a,b,⇑, Rosa di Sanzo a, Vittoria Cefaly a, Sonia Carabetta a, Demetrio Serra b, Salvatore Fuda a a b
Department of Agriculture, Laboratory of Food Chemistry, University of Reggio Calabria, Feo di Vito, 89100 Reggio Calabria, Italy Research Laboratory ‘‘Quality and Food Safety and New Technologies’’ – Qua.Sic.A.Tec. MIUR Acc n. 28/2010, Area Industriale, 88046 Lamezia Terme (CZ), Italy
a r t i c l e Article history: Received 4 October Received in revised Accepted 14 March Available online 22
i n f o 2012 form 9 March 2013 2013 March 2013
Keywords: Alliums cepa L. ‘‘Tropea’’ red onion Artificial olfactory system Electronic nose Metal oxide semiconductor Pyruvic acid
a b s t r a c t This work reports preliminary results on the potential of a metal oxide sensor (MOS)-based electronic nose, as a non-destructive method to discriminate three ‘‘Tropea Red Onion’’ PGI ecotypes (TrT, TrMC and TrA) from each other and the common red onion (RO), which is usually used to counterfeit. The signals from the sensor array were processed using a canonical discriminant function analysis (DFA) pattern recognition technique. The DFA on onion samples showed a clear separation among the four onion groups with an overall correct classification rate (CR) of 97.5%. Onion flavour is closely linked to pungency and thus to the pyruvic acid content. The e-nose analysis results are in good agreement with pyruvic acid analysis. This work demonstrated that artificial olfactory systems have potential for use as an innovative, rapid and specific non-destructive technique, and may provide a method to protect food products against counterfeiting. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Onion (Allium cepa L.) is a species of great economic importance, which is widely cultivated all over the world. The health benefits of onion consumption include protection against cancer, coronary heart disease, diabetes and ageing. These effects have been mainly attributed to flavonoids, vitamins and organosulphur compounds (Goldman, Kopelberg, Debaene, & Schwartz, 1996; Kumari, Mathew, & Augusti, 1995). Although they possess significant nutritional value, onions are primarily consumed for their unique flavour and for their ability to enhance the taste of foods (Kopsell & Randle, 1997; Rodrigues et al., 2003). Onions are widely cultivated throughout the Mediterranean basin and in Italy. The ‘‘Tropea Red Onion’’ (TrO), a sweet red onion characterised by its elongated bulb and almost white flesh, is cultivated, in three different ecotypes, only in the Tyrrhenian coastal areas of Calabria region (Southern Italy). In March 2008, the European Union registered the Protected Geographical Indication (PGI) certification for the onions produced in this particular area (Commission Regulation (EC) No. 284/2008). The ‘Cipolla Rossa di Tropea – Calabria’ PGI denotes bulbs of the species A. cepa ⇑ Corresponding author at: Department of Agriculture, Laboratory of Food Chemistry, University of Reggio Calabria, Feo di Vito, 89100 Reggio Calabria, Italy. Tel.: +39 0968209606; fax: +39 0968209361. E-mail address: [email protected] (M. Russo). 0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2013.03.052
exclusively from the local ecotypes. These have a characteristic shape according to time of production: ‘Tonda Piatta’, the early crop; ‘Mezza Campana’, the mid to early crop; ‘Allungata’, the late crop. TrO, for its tenderness, crispness and sweetness is a highly regarded Italian horticultural product and one of the most counterfeited. Sensory evaluation of onion flavour is closely linked to its pungency, which shows a close correlation with its pyruvate content (Schwimmer & Weston, 1961). The pungent flavour of onions is produced by hydrolysis of precursor compounds, such as S-alk(en)yl-L-cysteine sulfoxides (Lancaster, Shaw, Joyce, McCallum, & McManus, 2000), when the cells are mechanically ruptured. The hydrolysis reaction is catalysed by allinase and is completed within 6 min (Schwimmer & Weston, 1961). This reaction produces thiopropanol S-oxide (a lacrymator), ammonia, sulfur volatiles and pyruvic acid (Fig. 1; Block, 1992). Pyruvic acid concentration (lmol/g FW) in macerated onion tissue is used as a quality assurance indicator of pungency (Abayomi, Terry, White, & Warner, 2006; Pineda, Lué- Merù, Rivas, et al., 2004). Pyruvic acid exists universally in the plant tissues as part of an intermediate metabolism (Goodwin & Mercer, 1983). For this reason, the background levels of pyruvic acid or control (PC) need to be subtracted from the total pyruvic acid (PT) concentrations to calculate the enzymatically produced pyruvate (PE). Background pyruvic acid can be measured after the alliinase is deactivated. The deactivation of alliinase has been achieved by heating onion tissues in a
897
M. Russo et al. / Food Chemistry 141 (2013) 896–899
using a canonical discriminant function analysis (DFA) pattern recognition tool. DFA has been used extensively to perform pattern recognition and it has been reported to produce good performance for the classification of foodstuffs (Russo et al., 2012). DFA was used to check the capability of the e-nose system in assigning onion samples to a specific group. Finally, the aim of this study was to discriminate flavour and aroma characteristics of the early, middle and late crops of Tropea Red Onion and also distinguish between them and the common red onion by the use of an electronic nose. Studies on different ecotypes, to the best of our knowledge, have not been carried out before. The analysis of pyruvic acid by HPLC with UV detection was also carried out.
Fig. 1. Schematic of the compounds formed in onion from the hydrolysis of flavour precursors (a: S-alk(en)yl-cysteinsulfoxide; b: alk(en)yl-sulfenic acid; c: pyruvic acid; d: ammonia).
microwave oven (Schwimmer & Weston, 1961; Yoo, Pike, & Hamilton, 1995) or by homogenising with trichloroacetic acid (Randle & Bussard, 1993). Since the PC level can substantially change the PE concentration, its accurate measurement is critical in measuring onion pungency (Boyhan, Schmidt, Woods, Himelrick, & Randle, 1999; Thomas, Parkin, & Simon, 1992; Yoo & Pike, 2001). Pyruvic acid content determination and organoleptic tests are used for assessing the eating characteristics of onions. The accumulation of organosulfur compounds in onions and pyruvic acid content depends upon many factors, especially sulfur-based fertilisation and environmental growing conditions, although genotype the most important factor (Chope, Terry, & White, 2007; Yoo, Pike, Crosby, Jones, & Leskovar, 2006). The genetic background seems to be directly related to the different abilities to control sulfur uptake and assimilation in the biosynthetic pathway that results in the onion’s flavour (Randle, 1992; Randle, Block, Littlejohn, Putman, & Bussard, 1994). Onion flavour is traditionally measured through the use of a combination of conventional analytical instrumentation and human organoleptic profiling panels. These methods are expensive not only in terms of time, but are also inaccurate because of a lack of either sensitivity or quantitative information. In this paper an investigation was carried out to determine the flavour of different onion samples using an electronic nose (e-nose) and thus to explore the possibility of replacing existing analytical methods. The e-nose is a relatively novel device used for volatile sensing. The first pioneering studies about the concept of an artificial nose system able to measure odours was reported in 1982 by Persaud and Dodd of the University of Warwick, Coventry, UK (Cho et al., 2010). The e-nose has been designed for automated detection and recognition of odours, vapours and gases (Cho et al., 2010). It does not separate the volatile fraction of the matrix into its constituents but supplies a global evaluation of aroma, mimicking the human olfactory system with instrumental objectivity. Electronic nose technology has been successfully used to discriminate quality and flavour of various products, including tomatoes, citrus, spices and onions (Abbey & Joyce, 2007; Russo, Serra, Suraci, & Postorino, 2012). The technology may also be used to assess quality of stored grain, fish, drugs, drinks and food spoilage (Abbey, Aked, & Joyce, 2001). The e-nose used in this work comprises an array of semiconductor gas sensors (MOS), each of which has an electrical resistance that has partial sensitivity to the headspace of onion. The signals from the sensor array are then conditioned by suitable interface circuitry, resulting in an onion data-set. The data were processed
2. Materials and methods 2.1. Plant materials In 2011 the three local ecotypes of Tropea red onion and one red onion cultivar usually sold as Tropea, were analysed and compared (Table 1). All onion types were grown in a field in IGP district and on the same farm, so that the evaluation was not influenced by environmental factors and/or cultural practices. Three different samples were collected for each type, consisting of five onions. Each analysis was performed in triplicate for each sampling. 2.2. Electronic nose An electronic nose (ISENose 2000, Labservice Analytica, Bologna, Italy) was used to discriminate aroma fingerprints of onion ecotypes according to the method of Abbey et al. (2001). The edible parts of the onions were homogenised in an Ultra-Turrax system (T50 basic, IKA Werke, Staufen, Germany) at room temperature. The slurry was filtered after 20 min and 20 ml of the filtrate were transferred into a 250-ml flask. Twenty millitres of 5% trichloroacetic acid (TCA; Sigma Chemical Co., Milano, Italy) were added to terminate alliinase activity. The mixture was agitated vigorously and allowed to stand for 30 min. Deionised water (10 ml) was added to 10 ml of filtrate/ TCA solution and mixed. One millilitre of this diluted solution was put into a 25-ml vial, and equilibrated for 10 min at 25 °C to allow the development of headspace before e-nose analysis. The instrument, equipped with 12 metal oxide semiconductor (MOS) sensors, was used to generate a pattern for the compounds present in the headspace of the extracted red onion samples. The operating conditions in the acquisition phase were: baseline correction 10 s; injection time 10 s; sampling 180 s; delay 300 s; purge 60 s; carrier chromatographic air (80% N2, 20% O2) flow 300 ml/min; vial volume 25 ml. 2.3. Pyruvic acid extraction procedure and HPLC analysis The quantification of the pyruvic acid content in the onion was conducted using a modified method of Yoo and Pike (2001). The neck, base and central leaf tissues (about 5 cm long) were removed and the remaining fresh scales were cut into 1–2 cm squares using
Table 1 Red Onion ecotypes. ID
Sample
TrT TrMC TrA RO
Tropea Ecotype ‘‘Tonda’’ Tropea Ecotype ‘‘Mezza Campana’’ Tropea Ecotype ‘‘Allungata’’ Red Onion
898
M. Russo et al. / Food Chemistry 141 (2013) 896–899
a razor blade. The cut pieces (10 kg) were mixed thoroughly, rinsed with water and drained to remove the soil residues. As a control to measure the background level of pyruvic acid (PC), 100 g tissue were placed in a plastic bag (17.5 cm 21 cm). The samples were heated in a microwave oven (AFM 442 Ultimys Duo Grill; Moulinex, Italy) for 3 min at 900 W to deactivate the alliinase. After microwaving, the sample was placed in a bag under ice to stop the cooking. The tissue was blended in an Ultra-Turrax system for 2 min and the juice was collected through centrifugation (Eppendorf 5804R; Milano, Italy) and filtration with a syringe filter. To measure total (background and alliinase-produced) pyruvic acid content (PT), 100 g tissue were blended in an Ultra–Turrax system for 2 min. The homogenised mixture was macerated for 30 min at room temperature and juice was collected through centrifugation and filtration with syringe filter. Chromatographic analysis on onion extracts was carried out using an HPLC (LC-10AD system; Shimadzu, Kyoto,Japan) with diode array detector. Each sample extract was injected in duplicate on a 300 7.8 mm i.d. Supelcogel C610H (Sigma–Aldrich, St Louis, MO) and eluted using 0.3 ml/l H2SO4 at a flow rate of 0.5 ml/min. The eluate was monitored at 210 nm. Standards were prepared with pyruvic acid, purchased from Sigma Chemical Co. (Milano, Italy). Stock and working standards were prepared in HPLC-grade methanol. Enzymatically produced (from alliinase) pyruvate in each sample was calculated from the difference of the two determinations. All analyses were run in triplicate. 2.4. Data analysis The pyruvic acid content in onion extracts was identified by its retention time characteristics. The linearity of the detector response for the prepared standards was assessed by means of linear regression, based on the amounts of each standard and the area of the corresponding peak on the chromatogram. The limit of detection (LOD) and quantification (LOQ) were determined by calculation of the signal-to-noise ratio. Signal-to-noise ratios of approximately 3:1 and 10:1 were used for estimating the detection limit and quantification limit, respectively. LOD and LOQ values of pyruvic acid were 0.029 and 0.088 mmol/kg respectively; r2 Values of the compounds were higher than 0.999, thus confirming the linearity of the method. The precision of injection was demonstrated by replicate injections of the standard solution and the relative standard deviation (RSD) of peak area was 0.38% (n = 8). The sensors’ responses were evaluated using discriminant function analysis (DFA) by IseNose OCS software (version DT 01837 Rev. 02A, ISENose 2000; Labservice Analytica, Bologna, Italy). 3. Results and discussion The sensor responses were evaluated using Discriminant Function Analysis (DFA). The four samples (three local ecotypes and one common red onion) were chosen as grouping variables and sensors outputs were used as independent variables. A comprehensive database (500 data points) was developed to increase the reliability of the analysis. Four different sets of e-nose data clusters for onion headspace volatiles were identified in the DFA analysis (Fig. 2). Three clusters matched the three ecotypes of Tropea red onion (TrT red, TrA azure and TrMC blue) and one cluster matched the red onion (RO green). E-nose data set clusters for headspace volatiles of three types of Tropea red onion and common red onion are shown in Fig. 2. The e-nose discriminates between the three Tropea Red Onion ecotypes and between them and common RO (Fig. 2). Discrimination
Fig. 2. The three-dimensional Discriminant Function Analysis (DFA) plot for e-nose data set clusters of headspace volatiles of the different onion ecotypes (TrT: red, TrA: azure, TrMC: blue and RO: green).
is considered acceptable when it reaches values up to 80%; values around 85% are considered a sign of good discrimination, while values up to 90% indicated excellent discrimination. The DFA on onion samples showed a clear separation among the four onion groups, with an overall correct classification rate (CR) of 97.5%. These results prove that the e-nose is capable of discriminating between the flavours of onions. The e-nose results were compared with pyruvic acid content estimates. It is generally accepted that there is a high correlation between levels of enzymatically-produced pyruvate and perception of pungency (Schwimmer & Weston, 1961). The amou nt of pyruvic acid produced due to alliinase activity following tissue damage was evaluated in all samples. The theoretical 1:1 M ratio between pyruvate and its flavour precursors (alk(en)ylcysteinesulphoxides) (Yoo & Pike, 2001) suggested that the HPLC analysis of pyruvic acid content may represent a useful estimation of onion pungency (Bacon et al., 1999). Previous study reports considerable variation in background pyruvic acid concentrations. Schwimmer and Weston (1961) reported background levels of 2.1–4.1 lmol/ml from White Grano, Southport Red Globe, and five other cultivars. Randle and Bussard (1993) reported 0.1–1.0 lmol/ml pyruvate from 16 shortday onion cultivars, using the original Schwimmer and Weston method. Yoo et al. (1995) reported less than 0.5 lmol/ml using undiluted onion juice and suggested that measurement of the background level could be disregarded. Luo and Ewart (1995) reported background levels of 0.79 lmol/ml. The background pyruvic acid concentrations in our analysed samples showed levels of 0.22 ± 0.12 to 0.30 ± 0.11 lmol/g, lower than previously reported. The sweet onion industry in USA classifies onions on the basis of pungency as low pungency/sweet (0–3 lmol pyruvic acid/g FW), medium pungency (3–7 lmol pyruvic acid/g FW), and high pungency (above 7 lmol pyruvic acid/g FW) (Dhumal, Datir, & Pandey, 2007). Based on this classification, the sample RO can be considered as a medium pungent cultivar (3.79 ± 0.86 lmol/g) and all the three Tropea ecotypes as low pungency cultivars (Table 2). Of the three Tropea ecotypes, TrA shows the highest concentration of pyruvic acid (2.57 ± 0.81 lmol/g), TrMC shows a middle value (2.45 ± 0.30 lmol/g) while TrT shows the lowest concentration
M. Russo et al. / Food Chemistry 141 (2013) 896–899 Table 2 Content of pyruvic acid in onion juice of different onion ecotypes by HPLC-UV analysis (PC: background levels of pyruvic acid, PT: total pyruvic acid and PE: enzymatically produced pyruvic acid). ID
PT (lmol/g fw)
PC (lmol/g fw)
PE (lmol/g fw)
TrT TrA TrMC RO
2.26 ± 0.61 2.57 ± 0.81 2.45 ± 0.30 3.79 ± 0.86
0.30 ± 0.11 0.26 ± 0.16 0.22 ± 0.12 0.27 ± 0.12
1.96 ± 0.06 2.31 ± 0.81 2.23 ± 0.17 3.52 ± 0.37
(2.26 ± 0.61 lmol/g). The latter results are in agreement with data obtained by the e-nose analysis and with those reported by consumers identifying the ecotype ‘‘Tonda’’ (TrT) as the sweetest. This work showed that the e-nose can discriminate the Tropea red onion from other red onion and the different ecotypes of Tropea Red Onion with respect to their flavour and pungency. 4. Conclusion It has been shown that over 80% of onion pungency depends on genetic factors, while the remaining part is related to external factors, such as the habitat and cultivation practices (Yoo et al., 2006). Therefore the choice of cultivars appears to be a very important factor in the production of onions with a low level of pungency, while the habitat also influences the organoleptic characteristics. The present work showed the potential use of an innovative, rapid, specific and non-destructive technique able to discriminate effectively not only among ecotypes but, especially, between Tropea Red Onion PGI and other red onions. The importance of genetics compared to the habitat and cultivation techniques in the determination of the specific quality characteristics of a food product was demonstrated. The onion samples under investigation were grown under the same conditions in the same field with the same cultivation techniques. Nevertheless, their quality attributes were significantly different. Further studies are in progress to determine which onion headspace volatile chemicals interact with the e-nose sensors, in order to allow a clear correlation between onion quality and e-nose data sets. Acknowledgement This work is part of the ongoing project Dolce Rossa supported by the APQ ‘‘Scientific Research and Technological Innovation’’ (Az. 2 and Az. 3) financed by Calabria Region-Department of Research, European Community, Italian Ministry of Economic Development (MISE) and Italian Ministry of Research Science and Technology (MIUR). References Abayomi, L. A., Terry, L. A., White, S. F., & Warner, P. J. (2006). Development of a disposable pyruvate biosensor to determine pungency in onions (Allium cepa L.). Biosensors and Bioelectronics, 21, 2176–2179.
899
Abbey, L., Aked, J., & Joyce, D. C. (2001). Discrimination amongst Allium using an electronic nose. Annals of Applied Biology, 139, 337–342. Abbey, L., & Joyce, D. C. (2007). Quality assessment of diced onion (Allium cepa L.) using an electronic nose. Ghana Journal of Science, 47, 83–90. Bacon, J. R., Moates, G. K., Neg, A., Rhodes, M. J. C., Smith, A. C., & Waldron, K. W. (1999). Quantitative analysis of flavor precursors and pyruvate levels in different tissues and cultivars of onion (Allium cepa). Food Chemistry, 64, 257–261. Block, E. (1992). The organosulfur chemistry of the genus Allium – Implications for the organic chemistry of sulfur. Angewandte Chemie, International Edition in English, 1135–1178. Boyhan, G. E., Schmidt, N. E., Woods, F. M., Himelrick, D. G., & Randle, W. M. (1999). Adaption of a spectrophotometric assay for pungency in onion to a microplate reader. Journal of Food Quality, 22, 225–233. Cho, H. J., Lim, S. S., Lee, Y. S., Kim, J., Lee, C. H., Kwon, D. Y., et al. (2010). Hexane/ ethanol extract of Glycyrrhiza uralensis licorice exerts potent anti-inflammatory effects in murine macrophages and in mouse skin. Food Chemistry, 121, 956–966. Chope, G. A., Terry, L. A., & White, P. J. (2007). Preharvest application of exogenous abscisic acid (ABA) or an ABA analogue does not affect endogenous ABA concentration of onion bulbs. Plant Growth Regulation, 52, 117–129. Dhumal, K., Datir, S., & Pandey, R. (2007). Assessment of bulb pungency level in different Indian cultivars of onion (Allium cepa L.). Food Chemistry, 100, 1328–1330. Goldman, I. L., Kopelberg, M., Debaene, J. E., & Schwartz, B. S. (1996). Antiplatelet activity in onion (Allium cepa) is sulfur dependent. Thrombosis and Haemostasis, 76, 450–452. Goodwin, T. W., & Mercer, E. I. (1983). In Introduction to plant biochemistry (2nd ed., pp. 204–207). New York: Pergamon Press. Kopsell, D. E., & Randle, W. M. (1997). Onion cultivars differ in pungency and bulb quality changes during storage. HortScience, 32, 1260–1263. Kumari, K., Mathew, B. C., & Augusti, K. T. (1995). Antidiabetic and hypolipidemic effects of S-methyl cysteine sulfoxide isolated from Allium cepa Linn. Indian Journal of Biochemistry & Biophysics, 32, 49–54. Lancaster, J. E., Shaw, M. L., Joyce, M. D. P., McCallum, J. A., & McManus, M. T. (2000). A novel alliinase from onion roots. Biochemical characterization and cDNA cloning. Plant Physiology, 122, 1269–1280. Luo, C., Ewart, L. (1995). A new pyruvic acid (PA) procedure for rapid evaluation of onion pungency. Abstract and program of the 1995 onion research conference. Wisconsin Center, Madison, WI, December 6–9, p. 26. Pineda, M., Lué- Merù, M. P., Rivas, R., Gallignani, M., Valero, M., et al. (2004). Pungency evaluation of onion cultivars from the Venezuelan West- Center region by flow injection analysis-UV–visible spectroscopy pyruvate determination. Talanta, 64, 1299–1303. Randle, W. M. (1992). Onion germplasm interacts with sulfur fertility for plant sulfur utilization and bulb pungency. Euphytica, 59, 151–156. Randle, W. M., Block, E., Littlejohn, M. H., Putman, D., & Bussard, M. L. (1994). Onion (Allium cepa L.) thiosulfinates respond to increasing sulfur fertility. Journal of Agriculture and Food Chemistry, 42, 2085–2088. Randle, W. M., & Bussard, M. L. (1993). Streamlining onion pungency analyses. HortScience, 28, 60. Rodrigues, A. S., Fogliano, V., Graziani, G., Mendes, S., Vale, A. P., & Gonçalves, C. (2003). Nutritional value of onion regional varieties in northwest Portugal. The Electronic Journal of Environmental, Agricultural and Food Chemistry, 2, 519–524. Russo, Mt., Serra, D., Suraci, F., & Postorino, S. (2012). Effectiveness of electronic nose systems to detect bergamot (Citrus Bergamia). The Journal of Essential Oil Research, 24, 137–151. Schwimmer, S. S., & Weston, W. J. (1961). Enzymatic development of pyruvic acid in onion as a measure of pungency. Journal of Agriculture and Food Chemistry, 9, 301–304. Thomas, D. J., Parkin, K. L., & Simon, P. W. (1992). Development of a simple pungency indicator test for onions. Journal of the Science of Food and Agriculture, 60, 499–504. Yoo, K. S., Pike, L., Crosby, K., Jones, R., & Leskovar, D. (2006). Differences in onion pungency due to cultivars, growth environment, and bulb sizes. Scientia Horticulturae, 110, 144–149. Yoo, K. S., & Pike, M. L. (2001). Determinazion of background pyruvic acid concentrations in onions, Allium species, and other vegetables. Scientia Horticulture, 89, 249–256. Yoo, K. S., Pike, L. M., & Hamilton, B. K. (1995). A simplified pyruvic acid analysis suitable for onion breeding programs. HortScience, 30, 1306.