Review Cite This: J. Agric. Food Chem. 2018, 66, 3291−3301
pubs.acs.org/JAFC
What Do We Know about the Chemistry of Strawberry Aroma? Detlef Ulrich,*,† Steffen Kecke,‡ and Klaus Olbricht§,∥ †
Julius Kühn-Institute, Institute for Ecological Chemistry, Plant Analysis and Stored Product Protection, Quedlinburg 06484, Germany ‡ Julius Kühn-Institute, Data Processing Unit, Quedlinburg 06484, Germany § Hansabred GmbH Co. KG, Dresden 01108, Germany ∥ Humboldt-Universität zu Berlin, Albrecht Daniel Thaer-Institute, Berlin 10099, Germany S Supporting Information *
ABSTRACT: The strawberry, with its unique aroma, is one of the most popular fruits worldwide. The demand for specific knowledge of metabolism in strawberries is increasing. This knowledge is applicable for genetic studies, plant breeding, resistance research, nutritional science, and the processing industry. The molecular basis of strawberry aroma has been studied for more than 80 years. Thus far, hundreds of volatile organic compounds (VOC) have been identified. The qualitative composition of the strawberry volatilome remains controversial though considerable progress has been made during the past several decades. Between 1997 and 2016, 25 significant analytical studies were published. Qualitative VOC data were harmonized and digitized. In total, 979 VOC were identified, 590 of which were found since 1997. However, 659 VOC (67%) were only listed once (single entries). Interestingly, none of the identified compounds were consistently reported in all of the studies analyzed. The present need of data exchange between “omic” technologies requires high quality and robust metabolic data. Such data are unavailable for the strawberry volatilome thus far. This review discusses the divergence of published data regarding both the biological material and the analytical methods. The VOC extraction method is an essential step that restricts interlaboratory comparability. Finally, standardization of sample preparation and data documentation are suggested to improve consistency for VOC quantification and measurement. KEYWORDS: Fragaria × ananassa Duch., volatile organic compounds, gas chromatography, mass spectrometry, identification
■
INTRODUCTION The garden strawberry (Fragaria × ananassa) is one of the most popular fruits and represents a significant global economic market. In 2014, the worldwide production was 8.1 million tons (FAOSTAT).1 The unique flavor of strawberries is the primary reason for its high popularity.2 The garden strawberry emerged in the mid-1700s in Versailles, from an accidental hybridization of the American octoploids F. chiloensis and F. virginiana.3,4 The hybridization combined some of the most important characteristics of the garden strawberry: large fruits from the Chilean landrace and a unique, pleasant sweetish aroma deriving from the smallerfruited, red wild Virginia strawberry. This combination cannot be found in other species of the genus although Fragaria species have unique and diverse aroma patterns.5 Due to the combination of high sensory popularity and high nutrition, the health value of the garden strawberry can provide an important component for healthy human nutrition. The chemical basis of strawberry aroma has been a frequently researched topic. The volatile organic compounds (VOC) and their subset, the aroma compounds, were intensively identified and quantified. Early analytical investigations were performed by Coppens and Hoejenbos6 for F. moschata (syn. F. eliator) at the end of the 1920s and were published about 1939. In his compendium from 1991, Latrasse7 evaluated 54 studies on garden and wild strawberries and reported about 360 aroma compounds. For the garden strawberry, the review of Zabetakis and Holden8 lists over 80 studies with about 280 volatiles. Since © 2018 American Chemical Society
the end of the 1960s, the number of VOC identified has increased due to improved analytical technology, i.e., gas chromatography−mass spectrometry (GC-MS) coupling. The online database of the Nutrition and Food Research Institute of The Netherlands9 (TNO) currently lists 323 VOC for strawberries from 15 substance classes. These have been identified by quadrupole and ion trap mass spectrometers. From the intensity of activities in the field of metabolomics and the introduction of new powerful analytical techniques, the identification of more “new” metabolites is expected.10 A higher demand for metabolite data also results from modern breeding strategies. The goals of many breeding programs now include development of specific individual secondary metabolites or metabolite profiles responsible for the health value or sensory quality.11,12 Few of the strawberry VOC are consistently identified between studies. For example, in Ulrich and Olbricht11 and Schwieterman et al.,13 only 28 of 116 total VOC identified were reported in both studies. To date, a comprehensive inventory of the published strawberry volatilome is incomplete, despite more than 80 years of metabolite research. Today’s analytical techniques provide detailed chemical profiles, but compounds are inconsistent between reports. This lack of reliability is an Received: Revised: Accepted: Published: 3291
March March March March
2, 2018 8, 2018 13, 2018 13, 2018 DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry Table 1. Compilation of Material and Methodsa For F. × ananassa VOC Measurements no
ref
study
year
material
1
34
1997
1 unknown cultivar
fresh
yes, with CaCl2
1 berry to 500 g
LLE (diethyl ether)
2
35
Schieberle and Hofmann Ulrich et al.
storage
1997
fresh
yes, with NaCl
200 g
LLE (Freon)
3 4 5 6 7 8 9
33 32 31 30 29 28 27
Song et al. Gomes da Silva et al. Hakala et al. Fukuhara et al. de Boishebert Nuzzi et al. Jouquand et al.
1998 1999 2002 2005 2006 2008 2008
3 cultivars (and 1 F. vesca accession) unknown cultivar 3 cultivars 7 cultivars 1 cultivar 14 cultivars and 8 breeding clones 4 cultivars and 2 breeding clones 3 cultivars and 5 breeding clones
fresh frozen frozen frozen frozen fresh fresh
no, whole fruit yes yes yes yes quartering yes, with CaCl2
100 g 100 g 200 g 50 g 100 g 1500 g 80 g
10 11 12 13
26 25 24 23
Zhang, Y. T. et al. Li et al. Zhang, J. J. et al. Du et al.
2009 2009 2011 2011
4 1 1 2
frozen frozen frozen fresh
8.3 g ? 0.5 g 200 g
14 15
22 20
Ozcan and Barringer Vandendriessche et al.
2011 2012
unknown cultivar 1 cultivar
fresh fresh
yes yes quartering yes, with NaCl/ NaF yes, with SnCl2 no whole fruit
16 17
21 19
Ubeda et al. Vandendriessche et al.
2012 2013
4 cultivars 1 cultivar
fresh fresh
yes yes, with NaCl
80 g ?
18 19 20
17 16 15
N. N. Samykanno et al. Ulrich and Olbricht
2013 2013 2013
fresh frozen frozen
whole fruit ? yes, with NaCl
1 fruit 5 fruits 10 to 300 g
21 23 22 24 25
14 11 10 13 12
Mishra and Kar Schwieterman et al. Cannon et al. Oz et al. Ulrich and Olbricht
2014 2014 2015 2016 2016
1 cultivar 2 cultivars 5 cultivars (and 16 F. vesca accessions) 2 cultivars 35 cultivars and 3 breeding clones 1 cultivar 8 Cultivars 10 cultivars and 6 breeding clones For Strawberry VOC
HS-SPME (DVB) P&T (Tenax) P&T (OV-1 and OV-25) SPE (Porapak Q) HS-SPME (DVB) P&T (Anasorb CSC) HS-SPME (DVB/Car/ PDMS) HS(?)-SPME SPE (Porapak Q) LEe HS-SPME (DVB/Car/ PDMS) static-HS HS-SPME (DVB/Car/ PDMS) P&T (Lichrolut EN) HS-SPME (DVB/Car/ PDMS) P&T (Tenax) HS-SPME (PDMS/DVB) imm-SBSE
10 000 g 7 fruits or 100 g 100 000 g ? >1 000 g
(HS?)-SPME (PDMS) P&T (HaySep) SPE (silica) HS-SPME imm-SBSE
no
ref
GC separation
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
34 35 33 32 31 30 29 28 27 26 25 24 23 22 20 21 19 17
polar and mid polar polar mid polar polar and mid polar mid polar polar mid polar unpolar mid polar mid polar polar mid polar polar no unpolar polar unpolar midpolar
IT-MS Q-MS TOF-MS IT-MS Q-MS Q-MS IT-MS Q-MS Q-MS Q-MS Q-MS Q-MS Q-MS SIFT-MS Q-MS IT-MS Q-MS Q-MS and TOF-MS
1, 2, 1 1, 2 1 1, 2, 1, 2, 1, 2, 1, 2 1, 2 1, 2, 1, 2, 1, 2, 1, 2, 1g 1, 2 1, 2, 1, 2 1
19 20 21 23 22 24 25
16 15 14 11 10 13 12
unpolar and polar (2D) polar midpolar mid polar apolar and mid polar (2D) ? polar
TOF-MS Q-MS Q-MS Q-MS Q-MS Q-MS Q-MS
1, 2, 1, 2, 1, 2, 1, 2, 1h ? 1, 2,
cultivars cultivar cultivar cultivars
MS
identification 3
(3) (3) (3)
(3) (3) (3) (3)
(3)
(3) (3 partly) (3) (3)
(3 partly)
3292
homogenization
fresh yes fresh yes fresh yes, with CaCl2 ? ? fresh yes Measurements
sample size
55 g 1 fruit
quantitation μg/kg by SIDA SQ, IST SQ in % SQ in % SQ, relative peak areas no SQ, peak heights ESTc SQ, IST SQ in % μg/kg by ESTf μg/g by ISTf SQ, relative concentrations ppb on the basis of kinetic data SQ, absolute peak areas μg/kgc SQ, relative peak areas SQ, peak areas and “approximately conc.” in ng/g SQ, relative peak areas SQ, peak areas SQ, IST ng/gFW·h by ESTf SQ, relative percentage SQ in % SQ, peak areas
VOC isolation
sum of VOCb
new VOCb
15 23 31 95 39 48 23 32 69 50 20 50 54 41 97 28 62 74
0 1 5 31 3 9 1 6 7 12 1 43 5 4 22 7 3 14
124 65 24 75 553 63 64
40 21 2 6 322 23 6
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry Table 1. continued
Abbreviations: IT-MS: ion trap MS; Q-MS: quadrupol MS; SIFT-MS: flow tube MS; TOF-MS: time-of-flight MS; SQ: semiquantitation. Identification: 1: MS library search; 2: retention indices from literature; 3: coelution of authentic references with proof of origin of reference compounds. Number in parentheses means that no details and/or no origin are reported. bCounting of substance numbers is in accordance with the JKI database. These quantities sometimes differ from those which are given in the original publications. cNo details or recovery rates are given. d “Troyonoko” possibly is a scribal error of the Japanes cultivar ’Toyonaka’. eFreezing with liquid nitrogen and grinding following by liquid extraction of the frozen powder with petroleum ether and cyclohexane. fNo recovery rates are given. gIdentification on the basis of positive charged product ions. hOut of 563 VOC six new compounds were fully identified by synthesis and NMR characterization. a
added to the gamma-form. Substance names for which no CAS number was available were included in the overall list under the chemical names given in the original study. For statistical analysis, the association set (set union) was constructed from the 27 substance lists using an in-house web application (FindIntersection). From the viewpoint of accurate chemical analysis the data on substance identification were not strictly handled in the evaluated literature. For this review, we have adopted the guidelines from Molyneux and Schieberle (2007)37 for the nomenclature in the text (Table 1), even though these have been used in the original work in a different way. This guideline includes two steps for identification: (a) mass spectrometric fragmentation and retention indices must be determined on at least two separation columns of different polarity (Table 1, footnotes 1 and 2), and (b) comparison of the mass spectra and retention indices (RI) with those of authentic reference substances as a so-called coelution must be made (Table 1, footnote 3). Identification on the basis of a simple search in mass spectrometric libraries cannot be considered sufficient. Therefore, substances were “identified” when the requirements of points (a) and (b) were met on at least one separation column. Otherwise the substances were described as “tentatively identified”. For substance quantitation (which is not the main topic of this review), the published studies were also inconsistent. Using gas chromatography−flame ionization detector (GC-FID) or gas chromatography−mass spectrometry (GC-MS), the term “quantitation” could be used when accompanied by a stable isotope dilution analysis (SIDA) or by the standard addition method. In these cases, the terms “quantitation” or “quantified” were used. In other cases, the terms “semiquantitation” or “semiquantified” were applied. Thus, the specification of absolute, quantitative values, such as μg/kg or ng/g is inadmissible if based only on a single internal standard.
obstacle for plant genetics, research, and breeding programs. The identification of metabolic quantitative trait loci (QTL), the candidate genes for valuable metabolites and transcription studies, are based on metabolite analysis. Modern breeding strategies for creating new cultivars with a high resistance against diseases, high health value, and high sensory quality depend on the implementation of powerful and consistent chemical analyses. The aims of this review are to provide a detailed overview of strawberry volatilome identification and to highlight potential causes of the apparent irreproducibility. Twenty-five studies from the past 21 years that have not previously been summarized in review have been included here. Based on published methodological information, factors influencing the confirmation of the substance identification are discussed with regard to the sample collection, preparation, and analytical methods.
■
METHODS
From 1997 to 2017, more than 250 papers were published concerning strawberry VOC in the context of sensory quality (Web of Science search with descriptor ‘strawberr*’ AND (“aroma” OR “volatiles”); www.webofknowledge.com; web access 2017-10-20). Of the published manuscripts, 25 studies were chosen for evaluation.10,11,13−35 The studies were chosen based on (1) the VOC analyses performed by the authors and (2) reports of strawberry VOC identification. Studies that described the quantitation of few individual compounds or a compound class were excluded.36 This approach ensured that studies with comparable objectives were included and that the research focused on the elucidation of the strawberry volatilome. Because the evaluated literature covered the past 21 years, a review of Zabetakis and Holden8 and the TNO database9 were also included to contrast with previous findings. The TNO data contained references from 1956 to 1995, plus one publication from Du et al.23 Du et al. did not contain “new” compounds in comparison to Zabetakis and Holden.8 An approximation for the state of knowledge about VOC identification previous to 1996 was obtained by constructing an association set from both VOC lists. Most published substance sets listed the VOC by chemical name. Because the spelling of the chemical names was inconsistent, each entry was transferred manually into the internationally common chemical abstract service (CAS) registry numbers (http://www.cas. org/content/chemical-substances/). Online databases were searched for the CAS numbers (The Flavornet, http://www.flavornet.org/ flavornet.html, The Good Scents Company, www. thegoodscentscompany.com/data, The NIST WebBook, https:// www.nist.gov/, PubChem, https://pubchem.ncbi.nlm.nih.gov/, ChemSpider, http://www.chemspider.com/). Obvious literal errors in the names were corrected. For substances with stereoisomers, the unspecified substance name was listed if no reliable isomer determination was possible by mass spectrometric identification. In principle, both CAS numbers for the respective stereoisomers as well as a separate number for the unspecified form are available. If only the unspecified form was mentioned, e.g., hexenal, the compound was assigned to the transform, i.e., (E)-2-hexenal, to avoid the unspecified substance being counted as a separate entry. In analogy, unspecified lactones were
■
OBJECTIVES OF THE VOC ANALYSES Strawberry VOC analyses can be grouped into five categories (Table 2) 1. Sensory quality (aroma, flavor) 2. Interactions of genes with environment (GxE) 3. Bioactivity of VOC as signaling or defense substances 4. Metabolic analyses for plant genetics and plant cultivation and 5. Methodological work in the area of VOC analysis Schieberle and Hofmann34,39 determined the character impact compounds in strawberry juice by means of quantitative measurements (SIDA). Their sensory activity was assessed using the aroma value concept which was based on the comparison of metabolite concentrations with their odor thresholds.58 A total of 15 VOC were fully identified, with no new substances being published in comparison with that in the TNO database9 or the review by Zabetakis and Holden.8 For the comparison of the 25 publications from the methodological point of view,38 Schieberle and Hofmann34 was the only comprehensive assessment of the sensory quality of the VOC, including complete identification, exact quantitation, flavor concept, and recombination experiments. This study was based on a previous work by Schieberle39 using a gas chromatography-olfactometry (GC-O) study in which an identical VOC 3293
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry Table 2. Compilation of Scientific Aims for F. × ananassa VOC Measurements no
ref
1
34
2
35
3 4
study
year
aim
Schieberle and Hofmann Ulrich et al.
1997
33 32
Song et al. Gomes da Silva et al.
1998 1999
5
31
Hakala et al.
2002
6 7
30 29
Fukuhara et al. de Boishebert
2005 2006
8 9
28 27
Nuzzi et al. Jouquand et al.
2008 2008
10 11
26 25
Zhang,YT et al. Li et al.
2009 2009
12
24
Zhang, JJ et al.
2010
13 14
23 22
Du et al. Ozcan and Barringer
2011 2011
15
20
2012
16
21
Vandendriessche et al. Ubeda et al.
Quantitation of selected key flavor compounds VOC analysis for flavor breeding (key compounds) Test of SPME as rapid method VOC profile of “Oso Grande” in comparison to “Selva” and “Chandler” (aroma properties) VOC profile of “Senga sengana” in comparison to 5 others. Geographical origin, processing VOC profile of “Toyonoka” by SPE Characterization of varieties (SPME and data processing like Kohonen map) Comparison of GC-O with OAV Eating quality and harvest date GxE (genotype and harvest date) VOC profile comparison (aroma) Estimation of key compounds in “Toyonoka” Aroma development during maturation VOC-profiles of 2 varieties (aroma) VOC-profiles depending on varieties, storage, ripening stages. SIFT-MS SPME IMS and SPME-fastGC-MS
17
19
2013
18 19
17 16
Vandendriessche et al. N. N. Samykanno et al.
20 21 22 23
15 14 10 11
Ulrich and Olbricht Mishra and Kar Cannon et al. Schwieterman et al.
2013 2014 2014 2014
24 25
13 2
Oz et al. Ulrich and Olbricht
2016 2016
1997
2012
2013 2013
for VOC extraction. Schieberle and Hofmann,34 Ulrich et al.,35 and Zhang, J. J., et al.24 used this method. The bulk of research was carried out by adsorption methods such as solid phase microextraction (SPME), purge and trap (P&T), and stirbarsorptive extraction (SBSE). Li et al.25 and Cannon et al.10 used isolation of the VOC by solid phase extraction (SPE). Ozcan and Barringer22 applied static headspace extraction in combination with selected-ion flow-tube mass spectrometry (SIFT-MS) and without GC separation. However, the static headspace extraction is unsuitable for detecting aroma relevant VOC in the ppm range or below because this method has no concentration step. A further objective for VOC analysis of strawberries was the determination of specific methodological questions in sample preparation, separation, or detection. Nuzzi et al. (2008)28 evaluated six strawberry genotypes as a model system for comparative analysis of qualitative and semiquantitative results on the aroma activity of VOC obtained on the basis of GC-O, and the calculation of odor activity values. A total of 38 identified VOC were included in the comparison. Both methods for the determination of the aroma patterns yielded comparable results. Deviating from other studies, Ozcan and Barringer (2011)22 used SIFT-MS as a detection and identification method. A total of 41 VOC were examined depending on the variety, the frost storage, and the VOC release in mouthspace and nosespace. Measurement of the VOC in the respiratory air was possible because the SIFT-MS can be used without preconcentration and removal of water. Vandendriessche et al.20 used an unconventional separation and detection method for investigating the impact of an infection on VOC patterns by using the headspace multicapillary column-ion mobility spectrometry (HS MCC-IMS). With this approach, 97 VOC were semiquantified to identify biomarkers for infection by Botrytis cinera. Over the past 20 years, gas chromatography time-of-flight mass spectrometry (GC-TOFMS) systems have become increasingly useful as powerful analyzers for substance identification. For strawberry, three applications have been described using this technique. In the studies of Song,33 the application note N. N.,17 and Samykanno,18 31, 74, and 124 VOC were identified. A major component of strawberry metabolite patterns is genotypically determined. Therefore, comparative studies of strawberry varieties were performed (Table 3). In the 25 reviewed studies, 76 cultivars and 24 breeding clones were analyzed. Fragaria × ananassa ‘Camarosa’ (5 times), ‘Albion’, ‘Chandler’, ‘Festival’, and ‘Toyonoka’ (4 times each) were examined most frequently. The analyses of cultivars and breeding lines show that VOC analysis played a major role for breeding research and practical cultivation of strawberries. Further questions for VOC analysis that are related to both sensorial quality research and breeding include ontogenic effects22 and the interaction of genotype by environment (GxE). The latter was considered by Jouquand et al.27 and Samykanno et al.18 This work is a prerequisite for a metabolitedirected selection in the breeding process, because environmentally dependent metabolites are unsuitable as separate breeding objectives.
Study of glycosidic precursors and free aroma compounds VOC-analysis for flavor breeding Odor profiling with TOF-MS GxE interaction on VOC. Flavor breeding and production Metabolic diversity for breeding Quality changes during storage In-depht analysis of VOC, VSC VOC-analysis and acceptance for breeding VOC-profiles of 8 cultivars (aroma) VOC-profiles and acceptance for breeding
list was mentioned except for the mesifuran (2,5-dimethyl-4methoxy-3(2H)-furanone). Further studies by Ulrich et al.,35 Gomes da Silva et al.,32 Nuzzi et al.,28 Fukuhara et al.,30 Jouquand et al.,27 Zhang, Y.T. et al.,26 Li et al.,25 Du et al.,23 Vandendriessche et al.,19 Samykanno et al.,18 Cannon et al.,10 Schwieterman et al.,13 and Ulrich and Olbricht11 investigated the VOC patterns with regard to strawberry aroma. The two recent studies by Schwieterman et al.13 and Ulrich and Olbricht11 pursue an extended and comparable objective by using adequate instrumental analysis for VOC and non-VOC (or aggregate parameters). They also correlated data with a consumer test to obtain consumer preference (acceptance). GC is the method of choice for VOC analyses. For this purpose sample preparation is of crucial importance. In sample preparation, the analytes are isolated from a complex, aqueous matrix and transferred to a water-free GC-capable sample. Concentration of the aroma compounds is essential because the VOC occur in the parts per million (ppm) or subppm range. Initially, liquid−liquid extraction with organic solvents was used
■
GENERAL SURVEY OF THE IDENTIFIED VOC The TNO database9 and the review by Zabetakis and Holden8 were state of the art determinations for VOC identification until 1996, with 307 and 275 substances listed. By creating the 3294
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry Table 3. Cultivars of F. × ananassa Used for VOC Analysis
a
no
ref
study
year
1
34
1997
unknown cultivar (Spain) from local market
2 3 4
35 33 32
1997 1998 1999
Elsanta, Polka, Senga Gourmella (and 1 accession of F. vesca) unknown cultivar Chandler, Oso Grande, Selva
5 6 7
31 30 29
Schieberle and Hofmann Ulrich et al. Song et al. Gomes da Silva et al. Hakala et al. Fukuhara et al. de Boishebert
2002 2005 2006
8 9 10 11 12 13 14 15 16 17 18 19 20
28 27 26 25 24 23 22 20 21 19 7 6 15
Bounty, Honeoyea, Jonsoka, Korona, Polkaa, Senga Sengana Toyonaka Cal Giant3, Capitola, Ciflorette, Cifrance, Cigaline, Cigoulette, Cilady, Ciloe, Cireine, Darselect, Earliglow, Madeleine, Naiad, Pajaro and 8 breeding clones Alba, Darselect, Dora, Eva and 2 breeding clones Festival, Rubigem, Sugarbaby and 5 breeding clones Allstar, Toyonoka, Xingdu1, Xingdu2 Toyonaka Troyonokab Festival, Radiance Albion, Camarosa, Chandler, Sweet Charlie Elsanta Camarosa, Candonga, Fuentepina, Sabrina Charlotte unknown cultivar Albion, Juliette Alba, Elegance, Frau Mieze Schindler, Mara de Bois, Polka (and 16 F. vesca accessions)
21 22 23
14 10 11
24 25
13 12
Nuzzi et al. Jouquand et al. Zhang,YT et al. Li et al. Zhang, JJ et al. Du et al. Ozcan Vandendriessche Ubeda Vandendriessche N. N. Samykanno Ulrich and Olbricht Mishra Cannon Schwieterman
2008 2008 2009 2009 2010 2011 2011 2012 2012 2013 2013 2013 2013
Oz Ulrich and Olbricht
2016 2016
2014 2014 2014
material
Camarosa, Chandler Ciflorette Albion, Benicia, Camarosa, Camino Real, Chandler, Charlotte, Darselect, Elyana, Evie2, Festival, Galetta, Mara des Bois, Mojave, Monterrey, Portola, Proprietary1, Proprietary2, Proprietary3, Proprietary4, Proprietary5, Proprietary6, Radiance, Red Merlin, Rubygem, San Andreas, Sweet Anne, Sweet Charlie, Treasure, Ventana, Winter Dawn, Winterstar and 3 breeding clones Albion, Camarosa, Festival, Fortuna, Rubygem, Sabrosa, Sweet Ann Clery, Daroyal, Elegance, Elianny, Elsanta, Evie2, Frau Mieze Schindler, Honeoye, Rumba,Sonata and 6 breeding clones
In addition also from organic cultivation. bScribal error of “Toyonoka”
coelution and the remaining 547 VOC were tentatively identified by means of a library search and retention index (RI) comparison. Thus, Cannon et al.10 provided the most comprehensive identification of the strawberry volatilome, listing 322 compounds which were (tentatively) identified in strawberry for the first time. The complete strawberry volatilome to date, including those compounds found previous to 1996 (389 VOC), totals to 979 VOC. Thus, the strawberry is one of the most thoroughly studied fruits in the plant kingdom. The frequency of the 30 most often identified VOC is summarized (Table 4). The most frequently found substances in strawberries are the methyl and ethyl esters of hexanoic and butanoic acids with 24 to 22 entries. The 30 most frequently analyzed compounds included esters (17, 15 straight chain and 2 branched), acids (4), lactones (2), aldehydes (2), furans (2), alcohols (1), ketones (1), and terpenoids (1), while on the other extreme, 670 substances occurred only once. Surprisingly, none of the 979 VOC was comentioned in all of the 27 evaluated literature sources; 959 compounds were comentioned in fewer than half of the reports. Thus, only a partial consensus concerning the qualitative composition of the strawberry volatome was reached among researchers, despite their intensive analytical work. Possible causes for this discrepancy are subsequently discussed.
association set from the lists of both publications, 389 VOC were identified in strawberries previous to this review. The 25 studies evaluated for this review listed between 15 and 124 identified VOC. In contrast, Cannon et al.,10 observed 553 substances (Figure 1). For their analysis, 100 kg fruit of the
Figure 1. Cumulative sum of the number of VOC identified in strawberries.
cultivar C ̀ ifloretté were extracted by means of dichloromethane and then separated into 125 fractions using solid phase extraction (SPE). Subsequently, fractions were separated by means of two-dimensional GC (2D-GC) on two separation columns of different polarity and detected by quadrupole MS. Out of the 553 VOC, six substances could be fully identified by 3295
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry Table 4. Most Frequently Identified VOC in Strawberries #
substance
CAS
entriesa
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
methyl hexanoate ethyl hexanoate ethyl butanoate methyl butanoate linalool γ-decalactone hexyl acetate γ-dodecalactone DMMF (E)-2-hexenal butyl acetate DMHF ethyl 3-methylbutanoate ethyl 2-methylbutanoate hexanoaic acid methyl octanoate 2-methyl butanoic acid ethyl acetate hexanal butanoic acid (E)-2-hexen-1-ol ethyl octanoate butyl butanoate octyl butanoate 2-heptanone benzyl acetate (E)-2-hexenyl acetate methyl pentanoate pentyl acetate acetic acid
106-70-7 123-66-0 105-54-4 623-42-7 78-70-6 706-14-9 142-92-7 2305-05-7 4077-47-8 6728-26-3 123-86-4 3658-77-3 108-64-5 7452-79-1 142-62-1 111-11-5 116-53-0 141-78-6 66-25-1 107-92-6 928-95-0 106-32-1 109-21-7 110-39-4 110-43-0 140-11-4 2497-18-9 624-24-8 628-63-7 64-19-7
24 24 23 22 22 21 19 18 18 18 17 17 16 16 15 14 14 14 14 13 13 12 12 12 12 12 12 12 12 12
a
Sum of entries in 25 studies from 1997 to 2016, in the review of Zabetakis (1997)8 and in the TNO-database.9 DMMF, 2,5-dimethyl-4methoxy-3(2H)-furanone; DMHF, 2,5-dimethyl-4-hydroxy-3(2H)-furanone.
■
STUDIES USING GC-O AND THE FLAVOR VALUE CONCEPT In seven studies, the aim was to separate the aroma-active VOC, called character impact compounds or key compounds, from a larger number of identified substances.10,21,23,28,30,34,35 For this purpose, the aroma value concept58 including the determination of odor activity values as well as the GC-O approach were applied. Both methods lead to an improvement of the compound identification because reference substances must be used or because the odor quality is added as an additional filter for identification when using GC-O. Nuzzi et al.28 and Schieberle and Hofmann34 compared the odor activity values (OAV) using GC-O. These two studies showed that the different methods provided similar results for the character impact compounds. Nuzzi et al.28 confirmed this although only semiquantitative data were used to determine the OAV. Between 12 and 48 VOC were determined as character impact compounds in individual studies (Figure 2). Seven publications listed 105 character impact compounds. Only one substance, ethyl butanoate, was colisted in complete consensus in these seven studies. While 36 VOC were found more than in one study, 69 compounds were single entries. Thus, the research on character impact compounds also exhibited the same trend as for the general VOC identification without consideration of the olfactory properties. Only a few substances
Figure 2. Character impact compounds identified in strawberries using GC-O. Studies: A, Ulrich et al.;35 B, Schieberle and Hofmann;34 C, Fukuhara et al.;30 D, Nuzzi et al.;28 E, Du et al.;23 F, Cannon et al.;10 G, Ubeda et al.21 Red bars, tentatively or complete identified compounds.
were listed in consensus. Between 66% and 67% of the identified compounds are single entries; that is, they were only identified in one study (69 out of 105 in the seven GC-O studies and 659 out of 979 of all 27 reviewed literature sources). 3296
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry
■
patterns, proper freezing at −25 °C is a viable option because, contrary to the sensory quality, the patterns of the volatiles are less influenced by this process.45 Homogenization. For the cleanup step, the fruits were mostly homogenized and mixed with inorganic salts (calcium chloride, sodium chloride, tin chloride, and sodium flouride) to suppress enzyme activity. Some studies used quartered and some undamaged fruits. With whole fruits used for VOC analysis, the analogy to sensory quality in consumption was lost, because the process of homogenization affects the VOC patterns, e. g. by influencing lipoxygenase activity (LOX). In this case,17,33 the VOC pattern corresponds more to the orthonasal perception (smell) rather than the flavor when consumed (retronasal). VOC Isolation. The method of choice for VOC analysis is GC. The preparation of a GC-capable sample is associated with isolation and concentration as well as transfer into a water-free matrix. The cleanup process is the decisive and most complex step regarding the result of the analysis. For this reason, a number of attempts were made to develop effective methods for VOC analysis without time-consuming cleanup. These techniques have not been established for a well-founded metabolome analysis, but they can be used for certain purposes, such as the determination of maturity.46,47 The classical method for VOC isolation is liquid−liquid extraction (LLE). Here the extraction power depends on the polarity of the extraction solvent. The disadvantages are the high workload, the lack of automation, and the extraction of nonvolatile compounds. The LLE is unsuitable for a high throughput method and is therefore used for basic investigations with a small sample number. Nevertheless, the best results with regard to quality (number of extracted VOC) and quantity (high recovery rates) can be achieved by means of LLE. In addition to LLE, adsorption methods are used for VOC isolation. These include purge-and-trap methods (dynamic headspace), solid-phase extraction (SPE), and stirbar-sorptive extraction (SBSE). Since the market launch of solid phase microextraction (SPME) for water analysis in 1993, this technique has been used in many applications and is widely used in the VOC analysis of plant material. Ten out of the 25 evaluated studies used this technique for isolation. The wide distribution of the SPME technique is due to its easy handling without solvent use, including the possibility of automation. A disadvantage of the SPME, as with all adsorption methods, is the strong discrimination effect for individual substances or substance classes in complex matrices. This effect is particularly pronounced for SPME due to its design and leads to insufficient extraction of polar compounds such as acids and furanones because of very low recovery rates. Exact quantitation of VOC in complex matrices is impossible due to the bias in combination with the limited adsorption capacity of the SPME fibers. A comparison of SPME with other isolation techniques was published for strawberry and other crops.48−50 Although the SPME technique is subject to severe restrictions for complex systems, such as the strawberry matrix, this technique is often used in a completely uncritical manner, without accounting for the limitations on the recovery rates or substance identification. Separation. Apolar, medium-polar, and polar separation columns were used for chromatographic separation. For a complete substance identification by means of MS in the sense of a good analytic practice, independent separation and
INFLUENCING FACTORS ON SUBSTANCE IDENTIFICATION The quantitative and qualitative composition of the plant metabolome is subject to complex influencing factors. These include, on the one hand, factors which determine the quality of the test material, such as the genotype and the environmental influences. The environmental influences (“outside” influences) include cultivation, harvesting, storage conditions, maturity, and phytopathological status. On the other hand, the results of metabolic analyses are also known to be influenced by the analytical method. Important parameters are VOC extraction (cleanup), gas chromatographic separation, detection, and data processing. Some of the essential parameters are summarized in Table 1. Genotype. VOC data from 71 cultivars and 30 breeding clones have been published since 1997 (Table 3). In all studies that investigate several genotypes, the genotype has been described as a key influencing factor on the quality and quantity of the aroma patterns. The most commonly analyzed cultivars were F. × ananassa “Albion”,11,13,16,22 “Chandler”,11,14,22,32 “Festival”,11,13,14,23,27 and “Toyonoka”24−26,30 which were evaluated in four independent studies and “Camarosa”11,13,14,21,22 in five. However, the lists of the identified VOC for the respective cultivars also show only a low qualitative correspondence. This is an indication that other variables beside the genotype influenced the results. Sampling Size. Obtaining a consistent representative sample is an important prerequisite for chemical analysis. The minimum sample size in chemical analysis depends on the accuracy of the method, particle size, and homogeneity.40,41 If guidelines from the area of solids analytics were to be adopted for a typical strawberry fruit size of about 30 mm, a sample in the range of ten kilograms to several hundred kilograms would be needed.41 The actual sample sizes reported were between 0.5 g and 100 kg (Table 1). Four of the 25 studies used a sample size in the kilogram range, but other sample sizes were in the range of a single fruit or the size of few grams. Fruit-tofruit variations have been published for VOC content and dry mass.42,43 These effects are more quantitative than qualitative. However, qualitative effects may occur if a sample size was too small and results dropped below the detection limit. Maturity. Fruit composition varies during maturation.22,57 Differences in the results may be due to differential fruit maturity. Strawberries are not commonly evaluated using Brix value. Though Nuzzi28 evaluated the total soluble solids (TSS) values as “ripening index”, he did not use this as a consistent harvest criteria. In several studies18,20,22−25,28 fruit color was judged prior to harvest. Harvest Influence within the Season. Environmental influences cause qualitative and quantitative effects on the metabolite patterns and also on the results of substance identification.44 A partial solution for this special problem can be the analysis of a batch sample containing all mature and healthy fruits of the season.16,44 Thus, the influence of the harvesting time during the season could be eliminated. Investigations on material of unknown provenance, species designation, and cultivation site (Table 3) are basically problematic with regard to reproducibility. Storage. Studies on the aroma of the strawberries were mostly performed on fresh fruit. In 9 of 24 studies, frozen material was used. Freezing was used to bridge the period between harvest and instrumental analysis. Regarding the VOC 3297
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry
None of the studies provided full results of identifications on two separation columns of different polarities. Furthermore, some of the studies are based solely on the comparison of mass spectrometric fragmentation with library data or do not provide any information for identification (Table 1).10,13,17,22,32,35 The inclusion of retention data for the qualification of the MS identification was carried out in the remaining papers. The coelution of authentic reference substances is mentioned in 15 studies, but only a part of the published VOC could be covered in some papers. Information on the reference substances used and their origin or synthesis is only available in two studies.10,34 More detailed identification studies for chiral compounds are not reported. The addition of odor qualities for flavor-active VOC, as discussed for GC-O, is another way to improve the identification of flavor substances.
identification was required on two separation columns of different polarities.37 This approach reduces the likelihood of false identifications due to peak overlapping, particularly in the case of automatic library searches. These requirements were not fully met by any of the 25 studies. Mass Spectrometric Detection. For identification, quadrupole-, ion trap-, TOF-, and a SIFT-MS were used. Due to its high mass resolution and high scan rate, the TOFMS technique is preferred for the identification of a large number of mass fractions. At the same time, the limitation of the chromatographic separation capacity of the GC column can be counteracted by deconvolution. GC-TOF-MS systems were used in three studies in which 31, 74, and 124 VOC were identified.17,18,33 However, the largest number of identified VOC (553) was obtained by Cannon et al.10 using a quadrupole-MS detector, by means of a complex cleanup method in combination with a 2D-GC. Details on the quality of the reported substance identification are given in the following section. Quantitation. An exact quantitation in gas chromatography is only possible by coelution of isotope-labeled references (stable isotope dilution analysis) or by the standard addition method. In the studies considered here, Schieberle and Hofmann (1997)34 and Li et al. (2009)25 achieved these conditions. The remaining reports share semiquantitative data, even if the original work did not provide any information or (incorrectly) specify a concentration unit. Artifacts. One problem is the distinction between genuine VOC and artifacts. For this purpose, the execution of suitable blank analyses, possibly using statistical evaluation methods, must be carried out. Some lists contain ethanol and acetone, which are possible artifacts. Thermal reactions can also be a cause of artifact formation in the GC injector. Thus, when the SPME technique is not applied appropriately (fruit particles on the fiber surface), by thermal degradation, furanones may be formed from sugars51 adhering to the fiber or injection needle of the SPME device. At >160 °C, furanones may decompose into a variety of small molecules, including acetone and other ketones, as well as alkylfuranones.52 The formation of furans was reported as artifacts when using SPME as sample preparation method.59 The reviewed lists of identified VOC also contain phthalates and biphenyl which may originate from plasticizers, agrochemicals, or food additives. Because no information is given on the blank-analyses in most studies, it cannot be ruled out that the list of the 979 identified VOC in strawberry contains a series of artifacts.
■
CONCLUSION In conclusion the most important reasons for the low conformity in the substance identification which was documented here are the following: • use of different genotypes • influence of different environmental settings (GxE) • sample preparation inconsistencies • use of different MS types • uncritical use of identification methods • artifacts The progress in analytical techniques and bioinformatics has led to the development of metabolomics and thus to the increased application of this approach in many areas. The exchange of data with other “omics approaches” is currently boosting scientific progress. A key issue for chemical analysis is substance identification, which represents a challenge for the analysis of VOC in complex biological systems. In the last 20 years, many more VOC substances in strawberries have been identified (Figure 1). The volatilome of the strawberry is one of the most frequently investigated plants. Prior to 2017, more than 979 VOCs were identified. As the literature has shown, however, publications have little consensus on defining the volatile compounds. This situation is scientifically unsatisfactory and leads to inconsistent results in analyses using metabolite data. This inconsistent information is disadvantageous, especially for plant genetics and breeding. Metabolic data are increasingly being sought for functional genomics and breeding for flavor. Because approximately 67% of the published VOCs were reported in only one study and no single substance was found in all of the 27 evaluated sources, the reproducibility becomes a serious question. The qualitative and quantitative factors influencing the results of the VOC analyzes, as discussed above, can be grouped into two classes: (a) The biological system through internal (genotype) and external (environment) factors causes a considerable diversity of the metabolite patterns. b) The results are significantly influenced by the analytical method used. (a) The influences on biological variability are complex because of the gene-by-environment interactions (GxE) inherent to the system. Both variables (GxE) can produce both quantitative and qualitative variability, individually and in their combination. This influence cannot be eliminated and must be statistically verified by means of an adequate test design (multiyear, multiple-order, number of biological repetitions). In the 25 studies, 18 used only a single sample
■
QUALITY OF GC-MS IDENTIFICATION IN THE EVALUATED STUDIES Coupling GC and MS is the method of choice for VOC identification. Identification techniques provide different confidence levels. The highest level is reached when the identity of the molecule is validated by coelution of an authentic standard substance and subsequent MS analysis (confirmed identification or fully identified). A lower level is achieved when a structure is proposed only by spectral similarities present in a database (tentative, provisional, or putative identification). Table 1 shows the level of mass spectrometric identification in three stages. Using coelution of authentic references with proof of origin of reference compounds (Table 1, level 3) represents the highest level and corresponds to the requirements of Molyneux and Schieberle37 for an exact identification on one column type. 3298
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry
The level of substance identification by MS (confirmed or tentative identification) should be indexed. Also results which are cited in secondary literature should indicate the quality of substance identification to prevent an inflationary increase of tentatively identified substances in substance lists and databases. The consequence is repeated misidentification and misapplication in other areas of science. These errors would misdirect the mapping of metabolic QTLs, the study of candidate genes, the transcriptomics, and the marker-assisted selection in plant breeding.
from a single harvest date for analysis which can be the reason for qualitative differences.44 A partial solution for this problem was suggested by Ulrich and Olbricht16,44 using the analysis of a batch sample containing all mature and healthy fruits of the season (pooling). Thus, the influence of the harvesting time during the season could be mitigated. Investigations on material of unknown provenance, species designation, and cultivation site (Table 3) are basically problematic with regard to reproducibility. (b) The influences introduced by the chosen analytical method are significant. All of the influencing factors discussed above can cause quantitative and qualitative effects on the result and can reinforce each other. To minimize the methodological problems, an adequate test design has to be chosen. The basis for reproducible analysis is taken during sampling and cleanup. Errors that are introduced in this step cannot be eliminated in the further process of analysis by applying sophisticated detectors and algorithms for data processing. If sufficient material is available, a maximum sample size should be selected with a subsequent sample division. When investigating the strawberry volatiles, sample sizes of 0.5 g to 100 kg were used. However, sometimes analytical experiments with very small sample sizes, especially in genetics and breeding, in single plant experiments of a population or of wild material are necessary. Often, biological replicates may not be possible. If metabolic analysis is carried out, the analytical reliability of the results must be discussed. The extraction method appeared to exert the greatest influence. A reference to this was the large number of identified substances in the work of Cannon et al.,10 in which more than 500 VOC could be detected by applying a large sample quantity in combination with an elaborated extraction technique. In comparison, the use of fast and simple isolation methods such as the headspace SPME must be viewed critically. Even the most widely used detection by MS can be seen as a cause of divergent results. The technically complex MS detectors often have lower stability than GC standard detectors, e.g., flame ionization detector. Even mass spectrometers of the same manufacturer and type can produce different results depending on the degree of pollution and tune-set. In the totality of the methodological variables discussed above, these factors are a cause of the high interlaboratory variability. Completely standardized methods such as proposed in the Metabolomics Standards Initiative (MSI) have so far not been used in the field of aroma analysis.53,54 However, the scientific methodology requires the development and exchange of reproducible and falsifiable data, which is obviously not the case for the 670 single entries in the strawberry volatilome. The compound patterns recorded by instrumental analysis are highly dependent on the procedures used. Therefore, the use of a standardized method for the production of test material (genotype, cultivation method) can be useful for studies on genetics and breeding (marker, transcriptomics). Distinct analytical techniques cannot cover the full metabolome and any other characteristic of the plant material. Until now standard operating procedures (SOP) for analysis have not been widely accepted.55,56 The standardization of protocols to guide data production, quality, and robustness is central to coordinating efforts between scientists working in different laboratories. In chemical analysis, SOP can help narrow down the divergence. The focus here must be on sample preparation and VOC extraction. The occurrence of artifacts must be examined by careful methodological experiments. Finally, the quality of generated data must be evaluated.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.8b01115. Tables of the VOCs found and their frequencies (XLS)
■
AUTHOR INFORMATION
Corresponding Author
*(D.U.) Phone: (49) 3946-47231. Fax: (49) 3946-47300. Email:
[email protected]. ORCID
Detlef Ulrich: 0000-0003-3726-3691 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors wish to thank Kirsten Weiß for the careful analysis of the original data. The authors would like to thank Kim Hummer for the excellent support.
■
ABBREVIATIONS USED 2D-GC, two-dimensional gas chromatography; CAS, Chemical Abstracts Service; FID, flame ionization detector; GxE, gene by environment interaction; GC, gas chromatography; GC-O, gas chromatography olfactometry; LLE, liquid−liquid-extraction; LOX, lipoxygenase; MS, mass spectrometry; OAV, odor activity values; P&T, purge and trap; Q-MS, quadrupol mass spectrometry; QTL, quantitative trait loci; RI, retention index; SBSE, stir bar sorptive extraction; SIFT-MS, selected ion flow tube-mass spectrometry; SPE, solid phase extraction; SPME, solid phase microextraction; SOP, standard operation procedure; SQ, semiquantitation; TOF, time-of-flight; VOC, volatile organic compounds
■
REFERENCES
(1) N.N. Internet Communication. http://www.fao.org/faostat/en/ #home, access date 2017-10-16. (2) Bhat, R.; Geppert, J.; Funken, E.; Stamminger, R. Consumers Perceptions and Preference for StrawberriesA Case Study from Germany. Int. J. Fruit Sci. 2015, 15, 405−424. (3) Darrow, G. M. M. The Strawberry: history, breeding and physiology, 1st ed.; Holt, Rinehart and Winston: New York, Chicago, San Francisco, 1966. (4) Staudt, G. The origin and history of the large-fruited garden strawberry, Fragaria x ananassa. Theor. Appl. Genet. 1961, 31, 212−218. (5) Ulrich, D.; Komes, D.; Olbricht, K.; Hoberg, E. Diversity of aroma patterns in wild and cultivated Fragaria accessions. Genet. Resour. Crop Evol. 2007, 54, 1185−1196. (6) Coppens, A.; Hoejenbos, L. Investigation of the volatile constituents of strawberry juice (Fragaria elatior Ehrh.). J. Rec Trav Chim Pays Bas 1939, 58, 680−690. 3299
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry (7) Latrasse, A. Fruits III. In Volatile compounds in Foods and beverages; Maarse, H., Ed.; Marcel Dekker Inc.: 1991; pp 329−387. (8) Zabetakis, I.; Holden, M. A. Strawberry flavour: Analysis and biosynthesis. J. Sci. Food Agric. 1997, 74, 421−434. (9) N.N. Internet Communication TNO Nutrition and Food Research Institute; Centraal Instituut voor Voedingsonderzoek Zeist, http:// www.vcf-online.nl, access date 2016-09-15. (10) Cannon, R. J.; Agyemang, D.; Curto, N. L.; Yusuf, A.; Chen, M. Z.; Janczuk, A. J. In-depth analysis of Ciflorette strawberries (Fragaria x ananassa ‘Ciflorette’) by multidimensional gas chromatography and gas chromatography-olfactometry. Flavour Fragrance J. 2015, 30, 302− 319. (11) Ulrich, D.; Olbricht, K. A search for the ideal flavor of strawberry - Comparison of consumer acceptance and metabolite patterns in Fragaria x ananassa Duch. J. Appl. Bot. Food Qual. 2016, 89, 223−234. (12) Folta, K. M.; Klee, H. Sensory sacrifices when we mass-produce mass produce. Hortic. Res. 2016, 3, 16032. (13) Schwieterman, M. L.; Colquhoun, T. A.; Jaworski, E. A.; Bartoshuk, L. M.; Gilbert, J. L.; Tieman, D. M.; Odabasi, A. Z.; Moskowitz, H. R.; Folta, K. M.; Klee, H. J.; Sims, C. A.; Whitaker, V. M.; Clark, D. G. Strawberry Flavor: Diverse Chemical Compositions, a Seasonal Influence, and Effects on Sensory Perception. PLoS One 2014, 9 (2), No. e88446. (14) Oz, A.; Baktemur, G.; Kargi, S. P.; Kafkas, E. Volatile Compounds of Strawberry Varieties. Chem. Nat. Compd. 2016, 52, 507−509. (15) Mishra, R.; Kar, A. Effect of Storage on the Physicochemical and Flavour Attributes of Two Cultivars of Strawberry Cultivated in Northern India. Sci. World J. 2014, 2014, 1−7. (16) Ulrich, D.; Olbricht, K. Diversity of volatile patterns in sixteen Fragaria vesca L. accessions in comparison to cultivars of Fragaria xananassa. J. Appl. Bot. Food Qual. 2013, 86, 37−46. (17) N.N. Enhanced odour profiling of strawberries using TD and TOF MS. Markes International Appl. Note 515; 2013; pp 1−7. https:// www.markes.com/Resources/Mass-Spec-application-notes/default. aspx. (18) Samykanno, K.; Pang, E.; Marriott, P. J. Genotypic and environmental effects on flavor attributes of ’Albion’ and ’Juliette’ strawberry fruits. Sci. Hortic. 2013, 164, 633−642. (19) Vandendriessche, T.; Geerts, P.; Membrebe, B. N.; Keulemans, J.; Nicola, B. M.; Hertog, M. L. Journeys through aroma space: a novel approach towards the selection of aroma-enriched strawberry cultivars in breeding programmes. Plant Breed. 2013, 132, 217−223. (20) Vandendriessche, T.; Keulemans, J.; Geeraerd, A.; Nicolai, B. M.; Hertog, M. L. Evaluation of fast volatile analysis for detection of Botrytis cinerea infections in strawberry. Food Microbiol. 2012, 32, 406−414. (21) Ubeda, C.; San-Juan, F.; Concejero, B.; Callejon, R. M.; Troncoso, A. M.; Lourdes Morales, M.; Ferreira, V.; Hernandez-Orte, P. Glycosidically Bound Aroma Compounds and Impact Odorants of Four Strawberry Varieties. J. Agric. Food Chem. 2012, 60, 6095−6102. (22) Ozcan, G.; Barringer, S. Effect of Enzymes on Strawberry Volatiles during Storage, at Different Ripeness Level, in Different Cultivars, and during Eating. J. Food Sci. 2011, 76, C324−C333. (23) Du, X.; Plotto, A.; Baldwin, E.; Rouseff, R. Evaluation of Volatiles from Two Subtropical Strawberry Cultivars Using GCOlfactometry, GC-MS Odor Activity Values, and Sensory Analysis. J. Agric. Food Chem. 2011, 59, 12569−12577. (24) Zhang, J. J.; Wang, X.; Yu, O.; Tang, J. J.; Gu, X. G.; Wan, X. C.; Fang, C. B. Metabolic profiling of strawberry (Fragaria x ananassa Duch.) during fruit development and maturation. J. Exp. Bot. 2011, 62 (3), 1103−1118. (25) Li, X. X.; Fukuhara, K.; Hayata, Y. Concentrations of Character Impact Odorants in ’Toyonoka’ Strawberries Quantified by Standard Addition Method and PQ Column Extraction with GC-MS Analysis. J. Jpn. Soc. Hortic. Sci. 2009, 78, 200−205. (26) Zhang, Y. T.; Wang, G. X.; Dong, J.; Zhong, C. F.; Kong, J.; Li, T. Z.; Han, Z. H. Analysis of Volatile Components in Strawberry
Cultivars Xingdu 1 and Xingdu 2 and Their Parents. Agric. Sci. China 2009, 8, 441−446. (27) Jouquand, C.; Chandler, C.; Plotto, A.; Goodner, K. A Sensory and Chemical Analysis of Fresh Strawberries Over Harvest Dates and Seasons Reveals Factors That Affect Eating Quality. J. Am. Soc. Hortic. Sci. 2008, 133, 859−867. (28) Nuzzi, M.; Lo Scalzo, R.; Testoni, A.; Rizzolo, A. Evaluation of Fruit Aroma Quality: Comparison Between Gas ChromatographyOlfactometry (GC-O) and Odour Activity Value (OAV) Aroma Patterns of Strawberries. Food Analytical Methods 2008, 1, 270−282. (29) De Boishebert, V.; Giraudel, J. L.; Montury, M. Characterization of strawberry varieties by SPME-GC-MS and Kohonen self-organizing map. Chemom. Intell. Lab. Syst. 2006, 80, 13−23. (30) Fukuhara, K.; Li, X. X.; Okamura, M.; Nakahara, K.; Hayata, Y. Evaluation of odorants contributing to ’Toyonoka’ strawberry aroma in extracts using an adsorptive column and aroma dilution analysis. J. Jpn. Soc. Hortic. Sci. 2005, 74, 300−305. (31) Hakala, M. A.; Lapvetelainen, A. T.; Kallio, H. P. Volatile compounds of selected strawberry varieties analyzed by purge-and-trap headspace GC-MS. J. Agric. Food Chem. 2002, 50, 1133−1142. (32) Gomes da Silva, M. D.; Chaves das Neves, H. J. Complementary use of hyphenated purge-and-trap gas chromatography techniques and sensory analysis in the aroma profiling of strawberries (Fragaria x ananassa). J. Agric. Food Chem. 1999, 47, 4568−4573. (33) Song, J.; Fan, L. H.; Beaudry, R. M. Application of solid phase microextraction and gas chromatography time-of-flight mass spectrometry for rapid analysis of flavor volatiles in tomato and strawberry fruits. J. Agric. Food Chem. 1998, 46, 3721−3726. (34) Schieberle, P.; Hofmann, T. Evaluation of the character impact odorants in fresh strawberry juice by quantitative measurements and sensory studies on model mixtures. J. Agric. Food Chem. 1997, 45, 227−232. (35) Ulrich, D.; Hoberg, E.; Rapp, A.; Kecke, S. Analysis of strawberry flavour - discrimination of aroma types by quantification of volatile compounds. Eur. Food Res. Technol. 1997, 205, 218−223. (36) Olbricht, K.; Ulrich, D.; Weiss, K.; Grafe, C. Variation in the Amounts of Selected Volatiles in a Model Population of Fragaria x ananassa Duch. as Influenced by Harvest Year. J. Agric. Food Chem. 2011, 59, 944−952. (37) Molyneux, R. J.; Schieberle, P. Compound Identification: A Journal of Agricultural and Food Chemistry Perspective. J. Agric. Food Chem. 2007, 55, 4625. (38) Schieberle, P.; Hofmann, T. It comes down to taste - The molecular world in enjoying food. Chem. Unserer Zeit 2003, 37, 388− 401. (39) Schieberle, P. Heat-induced changes in the most odour-active volatiles of strawberries. In Trends in Flavour Research; Maarse, H., van der Heij, Eds.; Elsevier: Amsterdam, London, New York, Tokyo, 1994; pp 345−351. (40) Proctor, A.; Meullenet, J. F. Sampling and Sample Preparation. In Food Analysis, 2 ed.; Nielsen, S. S., Ed.; Aspen Publishers, 1998; pp 78−79. (41) Schwedt, G. Taschenatlas der Analytik., 3 ed.; Wiley-VCH Verlag GmbH & Co: KGaA, 2007; pp 20−21. (42) Gunness, P.; Kravchuk, O.; Nottingham, S. M.; D’Arcy, B. R.; Gidley, M. J. Sensory analysis of individual strawberry fruit and comparison with instrumental analysis. Postharvest Biol. Technol. 2009, 52, 164−172. (43) Vitten, M. D. Breeding Aspects of Freeze-dry Processing in Fragaria L. Dissertation Technical University Munich (TUM), 2008; Identifier: URN: urn:nbn:de:bvb:91-diss-20070914−627666−1-7; http://nbnresolving.de/urn:nbn:de:bvb:91-diss-20070914-627666-1-7. (44) Ulrich, D.; Olbricht, K. Diversity of metabolite patterns and sensory characters in wild and cultivated strawberries. J. Berry Res. 2014, 4, 11−17. (45) Hoberg, E.; Ulrich, D. Influence of freezing on flavour and healthy value of strawberry; Proceedings of the Deutsche Gesellschaft fur Qualitatsforschung (Pflanzliche Nahrungsmittel); DGQ e. V. XXXII. Vortragstagung: Wädenswil, 1997; pp 127−138. 3300
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
Review
Journal of Agricultural and Food Chemistry (46) Krumbein, A.; Ulrich, D. Assessment of flavour volatiles in vegetables - from comprehensive to rapid methods. Acta Hortic. 2003, 604, 425−429. (47) Baietto, M.; Wilson, A. D. Electronic-Nose Applications for Fruit Identification, Ripeness and Quality Grading. Sensors 2015, 15, 899−931. (48) Krumbein, A.; Ulrich, D. Comparison of three sample preparation techniques for the determination of fresh tomato aroma volatiles. In Flavour Science. Recent Developments; Taylor, A. J., Mottram, A. S., Eds.; The Royal Society of Chemistry: Cambridge, 1996; pp 289−292. (49) Ulrich, D.; Krumbein, A.; Schonhof, I.; Hoberg, E. Comparison of two sample preparation techniques for sniffing experiments with broccoli (Brassica oleracea var. italica Plenck). Nahrung 1998, 42, 392− 394. (50) Komes, D.; Ulrich, D.; Lovric, T.; Schippel, K. Isolation of white wine volatiles using different sample preparation methods. Vitis 2005, 44, 187−193. (51) Blank, I.; Fay, L. B. Formation of 4-hydroxy-2,5-dimethyl3(2H)-furanone and 4-hydroxy-2(or 5)-ethyl-5(or 2)-methyl-3(2H)furanone through Maillard reaction based on pentose sugars. J. Agric. Food Chem. 1996, 44, 531−536. (52) Shu, C. K.; Mookherjee, B. D.; Ho, C. T. Volatile Components of the Thermal-Degradation of 2,5-Dimethyl-4-Hydroxy-3 (H-2)Furanone. J. Agric. Food Chem. 1985, 33, 446−448. (53) Fiehn, O.; Robertson, D.; Griffin, J.; van der Werf, M.; Nikolau, B.; Morrison, N.; Sumner, L. W.; Goodacre, R.; Hardy, N. W.; Taylor, C.; Fostel, J.; Kristal, B.; Kaddurah-Daouk, R.; Mendes, P.; van Ommen, B.; Lindon, J. C.; Sansone, S. A. The metabolomics standards initiative (MSI). Metabolomics 2007, 3, 175−178. (54) Sansone, S. A.; Fan, T.; Goodacre, R.; Griffin, J. L.; Hardy, N. W.; Kaddurah-Daouk, R.; Kristal, B. S.; Lindon, J.; Mendes, P.; Morrison, N.; Nikolau, B.; Robertson, D.; Sumner, L. W.; Taylor, C.; van der Werf, M.; van Ommen, B.; Fiehn, O. The metabolomics standards initiative. Nat. Biotechnol. 2007, 25, 844−848. (55) Manach, C.; Brennan, L.; Dragsted, L. O. Using metabolomics to evaluate food intake: applications in nutritional epidemiology. In Metabolomics as a Tool in Nutrition Research.; Sebedio, J. L., Brennan, L., Eds.; Elsevier: 2014; pp 167−196. (56) Ferrara, M.; Sebedio, J. L. Challenges in nutritional metabolomics: from experimental design to interpretation of data sets. In Metabolomics as a Tool in Nutrition Research; Sebedio, J. L., Brennan, L., Eds.; Elsevier: 2014; pp 3−16. (57) Aharoni, A.; Giri, A. P.; Verstappen, A. P. G.; Bertea, C. M.; Sevenier, R.; Sun, Z.; Jongsma, M. A.; Schwab, W.; Bouwmeester, H. J. Gain and Loss of Fruit Flavor Compounds Produced by Wild and Cultivated Strawberry Species. Plant Cell 2004, 16, 3110−3131. (58) Rothe, M. Aroma values − a useful concept? Nahrung 1976, 20, 259−266. (59) Adams, A.; Van Lancker, F.; De Meulenaer, B.; OwzarekFendor, A.; De-Kimpe, N. On-fiber furan formation from volatile precursors: a critical example of artefact formation during Solid-Phase Microextraction. J Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2012, 897, 37−41.
3301
DOI: 10.1021/acs.jafc.8b01115 J. Agric. Food Chem. 2018, 66, 3291−3301
