Content uploaded by Luke Bell
Author content
All content in this area was uploaded by Luke Bell on May 02, 2018
Content may be subject to copyright.
REVIEW
Glucosinolates www.mnf-journal.com
Taste and Flavor Perceptions of Glucosinolates,
Isothiocyanates, and Related Compounds
Luke Bell, Omobolanle O. Oloyede, Stella Lignou, Carol Wagstaff, and Lisa Methven*
Brassicaceae plants are renowned for their taste, aroma and trigeminal
characteristics; predominantly bitter taste, sulfurous aroma, and pungency.
Compounds responsible for these sensations include the glucosinolates
(GSLs) and their hydrolysis products, particularly isothiocyanates (ITCs), but
also sulfur-containing volatile compounds. This article reviews the relative
importance of taste and flavor perceptions resulting from such compounds;
collating evidence from papers where findings are based on sensory analytical
correlations, and those that have extracted specific compounds prior to
sensory evaluation. Where specific GSLs impart bitterness and many ITCs
impart pungency, this is clearly not true for all GSLs and ITCs. Designing crop
improvement strategies for sensory traits based on total GSL content would
be flawed, as it does not consider the relative differences in sensory
characteristics of different GSLs and ITCs, nor the contribution from other
GSL hydrolysis products. In addition, some Brassicaceae plants are consumed
raw, whilst others are cooked; this affects not only the hydrolysis of GSLs, but
also the generation and release of sulfides. Therefore, in breeding new plant
varieties, it is prudent to consider the individual GSLs, the typical cooking
conditions the plant is subjected to, enzyme stability, and resultant
composition of both GSL hydrolysis products (including ITCs) and sulfides.
1. Introduction
Glucosinolates (GSLs) are secondary defense metabolites
present within the Brassicaceae family of plants. They are
β-thioglucoside-N-hydroxysulfates containing an amino acid–
derived side chain (R; see Figure 1). The structures and
hydrolysis of GSLs have previously been reviewed by Holst
and Williamson.[] GSLs are held within organelles in the
plant cytoplasm, whereas myrosinase enzymes that hydrolyze
these compounds are situated in the vacuole. When tissues
are damaged, by cutting or mastication, hydrolysis can lead
to the generation of numerous degradation products, which
include isothiocyanates (ITCs) and other compounds, such
as indoles and thiocyanates (Figure ).[–] The final product
composition depends greatly upon pH and temperature condi-
tions. Myrosinase enzyme is required for primary hydrolysis of
GSLs; however, products can be further modified by specifier
Dr. L. Bell, O. O. Oloyede, Dr. S. Lignou, Prof. C. Wagstaff, Dr. L. Methven
Department of Food and Nutritional Sciences
University of Reading
Whiteknights, Reading RG6 6AP, Berkshire, UK
E-mail: l.methven@reading.ac.uk
DOI: 10.1002/mnfr.201700990
proteins, such as epithiospecifier protein
(ESP), nitrile specifier protein (NSP), and
thiocyanate formation protein (TFP).[,]
ESP is a cofactor of myrosinase that pro-
motes the formation of epithionitriles
and nitriles from GSLs with a terminal
double bond. Where ESP is intact, there
is a tendency for greater nitrile produc-
tion and reduced ITC production; the re-
verse occurs where ESP is denatured, for
example, by mild temperature cooking.[]
Throughout the scientific literature, it
is generally accepted that GSLs and ITCs
contribute toward the distinctive tastes
and flavors associated with Brassicaceae
plants.[] However, there is relatively lit-
tle evidence relating taste and aroma per-
ception to individual compounds, and
few studies compare the relative ef-
fects of GSLs and ITCs alongside other
phytochemical components; or investi-
gate the extent of interactions between
such compounds on sensory perception.
In particular, the relative importance
of the ITC degradation products and
sulfur-containing volatile organic chemicals (SVOCs) is not well
understood.
It is accepted that bitterness is one of the key attributes
associated with Brassicaceae plants, and is proposed to result
from GSLs and ITCs.[] Numerous studies have indicated that
consumer preference and choice of vegetables in the diet is in
part determined by taste sensitivity to bitter compounds.[] It is
therefore important to understand how individual compounds
contribute toward sensations of taste, aroma, and trigeminal
sensations (such as pungency) in order to encourage greater
consumption of Brassicaceae vegetables. ITCs and indole degra-
dation products are well known for their health-beneficial prop-
erties (Table 1).[–] This is primarily due to anticarcinogenic
activity and chemoprotective mechanisms, and prevention of
cardiovascular and neurodegenerative disorders.[,] It is there-
fore beneficial to understand what tastes and flavors GSLs/ITCs
impart, and how they can be modulated without increasing
unpleasant sensations and reducing consumer acceptability.
Some authors have argued that good taste and health ben-
efits in a Brassicaceae crop product are incompatible goals.[]
However, as will be explored in this review, some potent health
beneficial GSL and ITC compounds have no distinct or ob-
jectionable taste. Furthermore, those compounds with bitter
attributes can be masked to a degree by other mechanisms. This
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (1 of 13)
www.advancedsciencenews.com www.mnf-journal.com
Figure 1. The glucosinolate (GSL)–myrosinase reaction produces numerous and diverse hydrolysis products, including epithionitriles, thiocyanates,
isothiocyanates (ITCs), nitriles, and oxazolidines. These products are influenced by many factors, including: ambient temperature, pH, and the presence
of enzyme co-factors (e.g., ESP, TFP, and NSP). Glucosinolate compounds can have a bitter taste, and oxazolidines (such as goitrin) impart extreme
bitterness. Isothiocyanate compounds are responsible for some bitterness (e.g., sinigrin), but primarily for the hot, pungent, and lachrymose sensations
that are typified by mustards, rocket, watercress, horseradish, and wasabi. Some ITCs are not known to impart strong flavor (e.g., erucin), whereas others
have not been previously described. This figure presents the chemical structures of ITCs present within commonly consumed Brassicales species. The
name of the precursor GSL is given in brackets, and bold numbers refer to those assigned in Fahey et al. and Agerbirk and Olsen. Sensory descriptors
for each compound are described in Table 1.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (2 of 13)
www.advancedsciencenews.com www.mnf-journal.com
Table 1 . Commonly reported glucosinolate and isothiocyanate compounds from plants within the Brassicales family, and their associated tastes, aromas, and trigeminal sensations. Taste and
trigeminal descriptors from isolation of either pure compounds and/or aroma descriptors according to GC-O analyses.
Glucosinolate Sensory
descriptor
Hydrolysis product Sensory descriptor Commonly found in Health-beneficial/toxic effects References
Sinigrin Bitter taste →Allyl ITC Bitter taste, pungent, sulfurous,
mustard-like, lachrymose,
horseradish-like
Brown mustard, horseradish,
Cabbage
Induces cell cycle arrest and
apoptosis in drug resistant
cancer cell lines
[11,22,40,50,60,62,65,99]
Allyl thiocyanate Musty, sulfurous, mustard-like Cabbage Unknown specific effects
Gluconapin Bitter taste →3-butenyl ITC Pungent, “green,” wasabi-like
flavor and heat, vegetable-like,
cabbage-like
Cook’s Scurvy grass, Brussels
sprouts, brown mustard,
broccoli, kale
Unknown specific effects [22,23,60,62,83,86]
Epi/Progoitrin Bitter taste →Goitrin Extreme bitter taste Brussels sprouts, salad rocket, sea
kale, turnips
Potentially toxic; may induce goiter
in thyroid-impaired or
iodine-deficient individuals
[22,50,51]
Glucobrassicin/
Neoglucobrassicin
Bitter taste →Indole-3-carbinol “Unpleasant” taste Chinese cabbage, broccoli, green
cabbage, red cabbage
Has been shown to inhibit prostate
cancer
[12,22]
Glucoraphanin No taste? →Sulforaphane No taste or flavor? Broccoli, salad rocket Prevents and suppresses cancer
formation; effective against
human prostate cancer
[13,30,46]
Glucotropaeolin ? →Benzyl ITC Pungent Garden cress, nasturtium, papaya Induces phase II detoxification
enzymes in human bladder and
bone cancer cell lines
[14,15,60]
Gluconasturtiin ? →Phenethyl ITC Pungent, radish-like,
watercress-like, produces a
“tingling” sensation
Watercress, horseradish, wasabi Causes cell cycle arrest and
apoptosis in human bone cancer
cell lines
[15,50,62]
Glucocapparin ? →Methyl ITC Horseradish-like, lachrymose,
pungent, vegetable-like
Capers Unknown specific effects [40,62,65,86]
Methyl thiocyanate Sulfurous
Glucoputranjivin ? →Isopropyl ITC Mustard-like, pungent Wasabi Unknown specific effects [60,62]
Glucobrassicanapin ? →4-pentenyl ITC Acrid, “green,” pungent, fragrant,
mustard-like, horseradish-like
Broccoli, rapeseed Unknown specific effects [60,62,99]
Glucosinalbin ? →4-hydroxybenzyl ITC Pungent Maca, white mustard Unknown specific effects [64]
Glucoraphasatin ? →Raphasatin Pungent Radish, Spanish black radish,
Japanese white radish
Unknown specific effects [65]
Glucosativin ? →Sativin Rocket-like Salad rocket, wild rocket Unknown specific effects [86]
Glucoiberverin ? →Iberverin Radish-like, pungent Cabbage, broccoli Unknown specific effects [60,65,86]
Glucoerucin ? →Erucin Radish-like, cabbage-like Salad rocket, radish, Chinese
cabbage, wild rocket
Inhibits cell proliferation in
prostate and adenocarcinoma
cell lines
[16,86,99]
?=unknown attribute; yet to be determined.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (3 of 13)
www.advancedsciencenews.com www.mnf-journal.com
can be achieved primarily through plant breeding and selection
to modify phytochemical constituents of crops—a practice that,
somewhat ironically, has traditionally been employed to remove
bitter- and unpleasant-tasting compounds from cultivars, and
increase sweetness. This has been largely due to safety concerns,
rather than being consumer driven.[] It is arguable that many of
these practices may have been misguided or unnecessary, partic-
ularly in the case of GSLs and ITCs within human food crops, as
healthy humans generally tolerate these with no adverse health
eects.[]
We aim to draw together information relating to the taste and
flavor perceptions of GSL, ITC, and related VOC compounds,
both in isolation, and within the food matrix. We will discuss
how these perceptions are modified according to human geno-
type, and describe the mechanisms responsible. We will high-
light compounds of interest for future study, as well as evaluate
the methods employed to link compounds with sensory percep-
tions.
2. Methods of Associating Taste and Flavor with
Specific Compounds
2.1. Chemical Extraction and Analysis
GSLs, as nonvolatile compounds, are usually extracted from
ground plant material in aqueous methanol, then identified
and quantified by liquid chromatography mass spectrometry
(LC-MS).[] Relating GSL type and concentration to bitter taste
is then often done through correlating changes to sensory
data.[–] However, such correlations can be misleading, and in
order to confirm the contribution of GSLs and/or ITCs to taste
perception, food grade extraction and fractionation is required
to provide pure isolates for human testing. This can be done
through ethanol extraction and semi-preparative high pressure
liquid chromatography (HPLC), using water and ethanol as sol-
vents, as described by Zabaras et al.[] in their characterization
of taste-active extracts from raw Brassica vegetables.
Analytically, volatile flavor compounds can be extracted
using solvents (solvent-assisted flavor evaporation [SAFE]),
headspace techniques such as solid phase micro extraction
(SPME), dynamic headspace extraction (DHE), or solid phase
extraction (SPE) for the analysis of more polar compounds in
low-fat foods.[,] Extracted compounds can then be identified
and quantified using gas chromatography mass spectrometry
(GS-MS), and assigned an odor characteristic by employing GC-
olfactometry (GC-O),[,,] or by the presentation of isolated
compounds to sensory assessors. This can be a challenging pro-
cess, requiring a high degree of optimization for the foodstu in
question.[] It also relies upon adequate compound library iden-
tifications or reference standards, which for some compounds
are not readily available or aordable. Identification alone does
not indicate whether a volatile compound has any impact on
the flavor of a sample; odor impact compounds are usually
denoted by odor activity values (OAVs), which are calculated as
the ratio between the concentration of the aroma compound in
each respective food sample matrix and its sensory threshold
concentration, which has usually been measured in water.[,]
2.2. Sensory Analysis and Relating Analytical and Sensory Data
How individual GSL, ITC, or other compounds are perceived (if
at all) within a whole food matrix is, however, a very dierent
question to asking how pure compounds are perceived when pro-
vided individually to human subjects. There are many dierent
factors that will contribute toward perception, such as the relative
concentrations of volatile compounds (i.e., the overall flavor com-
position), if they degrade quickly, if processing modifies them,
or if they act in competition for specific olfactory or taste recep-
tors. Typically, sensory panels are used to evaluate and score foods
according to the intensity of sensory perceptions (bitter, sweet,
sour, salty, pungent, etc.). Such sensory data can be correlated
to chemical data using principal component analysis (PCA) or
partial least squares regression (PLS).[,] This produces a sep-
aration within a multidimensional space where associations be-
tween sensory and chemical data can be inferred and assigned
levels of significance or relatedness.
Humans are typically poor at distinguishing between multi-
ple compounds within a mixture, as these are “combined” by the
brain to form a dierent sensation that would otherwise be per-
ceived if each of the compounds was sampled individually. Stud-
ies that have reconstituted taste and flavors have illustrated that
characteristic “profiles” associated with foods and beverages are
a result of complex interactions between volatile components, as
well as with nonvolatiles.[] It is for this reason, therefore, that
correlation analyses be interpreted cautiously, as these do not
necessarily reveal a complete picture of taste and flavor genera-
tion on a molecular level. All food samples are composed of many
dierent tastes and volatile aromas (“natural mixtures”), and hu-
mans vary in their ability to detect, identify, describe, and cog-
nitively “separate” these from each other.[,] Many of the com-
pounds found in Brassicales crops are likely to act synergistically
to create distinctive flavors; however, due to the lack of available
food grade ITC, thiocyanate, nitrile, and indole standards, this
hypothesis has never been tested.
In some Brassicales crops, GSL compounds are significantly
associated with bitterness,[] but in others no significant relation-
ships have been determined.[] Such lack of consensus between
studies may be indicative of the eects that dierent preparation
and cooking procedures have on bitter-producing compounds.
However, as ITCs and/or nitriles are rarely quantified in such
studies, it is not explicitly clear whether it is the precursor
or hydrolysis products that are the predominant cause of the
presence/absence of bitter sensations, or if they both infer inde-
pendent sensory properties. Often, GSLs are used as a proxy mea-
surement for hydrolysis products, and correlations (or their ab-
sence) with sensory traits (such as bitterness) are often conflated
with each other. Caution must be taken in the interpretation of
such analyses, as correlation is not equal to causation. For exam-
ple, some GSLs may hypothetically correlate with pungency, but
it is not the GSLs themselves that cause this sensation (Table ).
Baik et al.[] analyzed the GSL content of broccoli cultivars
to determine any associations with sensory traits. Broccoli (Bras-
sica oleracea var. italica) does not contain high concentrations
of GSLs that produce pungent or acrid ITCs (e.g. sinigrin,
gluconapin, glucobrassicanapin, gluconasturtiin[]), and so it
was therefore unsurprising that GSLs alone did not overtly
contribute significantly to the overall taste profile. The major
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (4 of 13)
www.advancedsciencenews.com www.mnf-journal.com
GSL constituent of broccoli is glucoraphanin, which is thought
to have little perceptible taste (however no known studies have
isolated and explicitly tested this compound, to our knowledge).
The authors (Baik et al.[]) concluded that sulfides and other
VOCs contributed more to the taste and flavor than GSLs.
Cartea et al.[] also drew similar conclusions: that GSLs and
their breakdown products do not contribute significantly to
taste and aroma of broccoli cultivars. It is however possible that
GSLs and ITCs may be perceptible at very low concentrations
and do contribute more than has been previously realized.
Until adequate studies are conducted to ascertain the detection
thresholds of these compounds, much of the discussion found
in such papers will remain conjecture.
In neither of the aforementioned studies were ITCs, nitriles,
or other hydrolysis products quantified, and so the lack of corre-
lations between GSLs and sensory attributes is not the same as
there being no correlations between GSL hydrolysis products and
sensory attributes. GSLs are not indicative of hydrolysis product
type or abundance, and are merely a convenient proxy for the
types that may be produced under certain conditions.[] In order
to determine this properly, dedicated analyses of ITCs and nitriles
should be performed alongside GSL analysis. More thorough in-
vestigation into the tastes, flavors, and abundances of hydrolysis
products is needed before conclusions can be drawn of whether
GSL content is unimportant for sensory attributes. In one of the
few studies to have isolated and tasted nonvolatile fractions from
Brassica vegetables, several compounds within each extract were
found to contribute to the bitterness (such as phenolics), and nei-
ther total nor individual GSL content alone could fully explain
the perceived bitterness.[] This may only be applicable to B. oler-
acea however, as many other species have very dierent GSL and
ITC profiles. Exploration of other diverse species and genotypes
will determine the impact of these compounds on sensory traits
within their own respective food matrices.
3. Bitter Taste of Glucosinolates and
Isothiocyanates
3.1. General
The presence of some intact GSLs is commonly associated with
bitter sensations within the literature. From an evolutionary
perspective, humans have evolved to be sensitive to bitter
compounds and to broadly reject them when perceived to be
excessive. In terms of survival, this is an essential protective
mechanism to prevent poisoning. Alkaloids, phenols, flavonoids,
terpenes, isoflavones, and oxazolidines are all known to have
bitterness associated with them, and all are present in the
human diet; but they are also known to be toxic (in some form)
in high doses.[] It is interesting therefore that Brassicaceae
crops have become so popular across the globe, despite them
containing these bitter phytochemicals.
3.2. Human Bitter Taste Perception and Consumer Acceptance
Humans detect bitterness as (arguably) a uniform sensation;
this is despite the fact that many dierent compound classes
interact with many dierent taste receptors to construct bitter
sensation.[] To complicate matters further, intensity of percep-
tion also varies with age; being greater in infants and children,
and much reduced in the elderly.[]
The predominant mode of action is via compound interac-
tion with G-proteins coupled with TASR taste receptors on the
tongue (Figure 2). Due to genetic linkage and recombination,
humans have a diverse array of genotypes for any given taste
receptor, being either homozygous (containing either two func-
tioning or two nonfunctioning copies of a gene) or heterozygous
(containing one functioning and one nonfunctioning copy). Such
haplotypes can give rise to dierent phenotypes; sometimes re-
ferred to as “supertaster,” “medium taster,” or “non-taster,” for
any given receptor.[] Dierences in the intensity of taste per-
ception may influence liking, and consequently, consumption of
bitter-tasting foods such as Brassicas.[]
Bitter “blindness” or “non-tasters” to the synthetic compounds
-n-propylthiouracil (PROP) and phenylthiocarbamide (PTC) is
well documented. These compounds contain a thiourea group,
which is also present in GSLs and ITCs, and preferentially bind
to the TR bitter taste receptor.[] Bitterness sensitivity to such
compounds has been linked to the Mendelian inheritance of
TASR haplotypes. Three functional single nucleotide poly-
morphisms (SNPs) within the TASR gene encode amino acid
substitutions which result in two common haplotypes: proline
alanine valine (Pro-Ala-Val; PAV), the dominant (sensitive) vari-
ant, and alanine valine isoleucine (Ala-Val-Ile;AVI), the recessive
(insensitive) one.[] Previous papers have found PAV/PAV indi-
viduals to rate bitter intensity of Brassicaceae vegetables signifi-
cantly higher than AVI/AVI individuals.[,] Bitter sensitivity is
also influenced by receptor cell abundance, a surrogate marker
for which is fungiform papillae density (FPD) on the tongue.
Therefore, the extent of bitter response of PAV/PAV individuals
to Brassicaceae vegetables will be modulated by FPD.
However, some research has shown that this does not neces-
sarily translate into whether an individual will base their liking
upon this sensitivity.[] A study by Bell et al.[] which determined
TASR diplotypes of individuals reported that “supertaster”
status did not have a significant impact upon liking of dierent
rocket salad accessions. There was a general trend observed that
“non-tasters” liked the taste of leaves more overall, but this was
not significantly dierent from either “supertasters” or “medium
tasters.” Similar observations were made in broccoli and white
cabbage by Shen et al.,[] where it was suggested that liking was
also determined by numerous other factors (such as food famil-
iarity, cultural acceptance, fungiform papillae density, and the
genotype of other TASR receptors).
3.3. Glucosinolate Bitterness
Compounds such as sinigrin, gluconapin, progoitrin, and indole
GSLs (glucobrassicin, neoglucobrassicin) are cited as imparting
bitterness in crops such as broccoli, and other varieties of B.
oleracea;[,–] however, this eect is not universal, as taste pan-
elists do not uniformly detect or describe these as bitter in iso-
lation (Table ).[] This suggests that some individuals may be
“non-tasters” or “bitter blind” for these GSL compounds, as with
PROP and PTC.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (5 of 13)
www.advancedsciencenews.com www.mnf-journal.com
Figure 2. Sensorial detection of glucosinolates (GSLs), isothiocyanates (ITCs), and sulfur-containing volatile organic compounds (SVOCs) by the human
nervous system. Glucosinolate compounds are exclusively detected upon the tongue, predominantly by TAS2R bitter taste receptors such as TAS2R38,
upon ingestion. The action of mastication breaks plant tissues and brings GSLs and myrosinase enzymes into contact, releasing hydrolysis products
that include ITCs. Many of these compounds are volatile and interact with ANKTM1 TRP ion channels to infer pungency within mucosal membranes
in the mouth, nose, and throat. Olfactory receptors also perceive ITCs and SVOCs, producing aromas characteristic of mustard, rocket, radish, and
cabbage. Other classes of compound are also likely to produce odors and flavors associated with Brassicaceae, Caricaceae, and Moringaceae crops,
such as sulfides, SVOCs, as well as nitriles.
In Brussels sprouts, high sinigrin and progoitrin concen-
trations have been linked with consumer rejection and poor
taste.[] The threshold of detection for sinigrin has been re-
ported as low as mg L−, and goitrin (the main progoitrin
hydrolysis product; -vinyloxazolidine--thione) mg L−,[]
but values for other GSLs and hydrolysis products are scarce in
the literature. This is perhaps in part due to the lack of aordable
standards in quantities larger than a few milligrams, and the
diculty in isolating compounds to a food grade standard. The
extreme bitterness of goitrin (Table ) is perhaps linked with
its reported toxic eects in individuals who are iodine deficient
or thyroid impaired. The compound can cause the condition
known as goiter, where it acts as an iodine competitor, and
thus can inhibit uptake. However, there is little evidence for
this compound being linked with illness in otherwise healthy
individuals.[]
In both Diplotaxis tenuifolia and Eruca sativa, which are com-
monly referred to as “wild rocket” and “salad rocket,” respec-
tively, the compounds progoitrin and dimeric--mercaptobutyl
GSL (DMB) have also been associated with bitter taste.[] It has
been reported that the presence of some GSLs contributes to a
pleasant or distinct taste to certain cultivars of rocket. Both Bell
et al.[] and D’Antuono et al.[] reported that the presence of -
hydroxyglucobrassicin (an indolic GSL) within rocket leaf tissues
was associated with such pleasant traits, and Bell et al.[] also re-
ported that relatively “minor” GSLs contributed toward increased
liking for a subset of consumers.
3.4. Isothiocyanate Bitterness
Due to the various factors involved in GSL hydrolysis product for-
mation, and the eciency of myrosinases under diering pH,
ionic, and temperature conditions, it is not certain that ITCs
will be produced.[] In some Brassicaceae crops, nitriles are the
predominant breakdown product, not ITCs; such as in white
cabbage (B. oleracea var. capitata).[] It is unknown how ITC–
nitrile ratios aect perceptions of bitterness, or indeed if nitriles
and other degradation products are bitter tasting too. Studies re-
port nitriles being present in SPME experiments, but olfactom-
etry has not been performed adequately to determine any flavor
attributes.[]
In studies that have quantified GSLs and ITCs separately, cor-
relation analyses indicate that some specific compounds individ-
ually contribute to bitterness, and others do not.[] Such evidence
supports particular compounds leading to bitter traits, but stops
short of establishing a causal link. It is highly probable that some
ITC compounds may be perceived as more bitter (and/or pun-
gent) than others, but the scarcity of food grade standards for
such compounds makes testing such hypotheses dicult.
3.5. Non-Bitter Glucosinolates and Isothiocyanates
Some GSL compounds, such as glucoraphanin, have been
consistently noted for the lack of significant association with
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (6 of 13)
www.advancedsciencenews.com www.mnf-journal.com
bitterness in vegetable tissues. The ITCs and degradation prod-
ucts of glucoraphanin and glucoiberin are only semi-volatile, and
so are unlikely to contribute overtly to the flavor of crops such as
broccoli.[] Breeders have taken advantage of this aspect of the
chemistry to increase glucoraphanin concentrations within new
cultivars and varieties such as Benefort´
e.[] Bell et al.[] noted
that neither glucoraphanin or glucoerucin shared any significant
correlations with bitterness or consumer acceptance/rejection in
rocket salad. The health benefits associated with glucoraphanin
and its ITC sulforaphane (SF) are well documented in the
literature and could therefore be selectively enhanced without
significant increases in bitterness.[,]
4. Pungency of Isothiocyanates
4.1. Human Pungency Perception
A key attribute of ITC compounds is the pungent or burning sen-
sation that is caused upon consumption (Table ). This is due
to the interaction with pain receptors feeding into the trigem-
inal nerve, and is distinct from flavor and taste (Figure ).[]
These trigeminal nerve endings within the mouth, nose, and eyes
also convey inflammatory and thermal stimuli to the brain. ITCs
achieve such trigeminal stimulation by activating transient recep-
tor potential (TRP) ion channels.[] AllylITC(AITC),benzylITC
(BITC), and phenethyl ITC (PEITC; Figure ) have all been shown
to activate the ANKTM TRP ion channel within mammalian
cells, which has also been linked with the burning perception of
noxious cold. Other pungent chemicals, such as capsaicin (from
chili peppers), also interact with this ion channel family and in-
duce inflammatory responses in addition to the perception of in-
tense heat.[]
4.2. Isothiocyanate Compounds
In terms of plant defense, ITCs dissuade herbivores and re-
duce insect proliferation as a result if no other feed source is
available.[] AITC is perhaps the best-studied ITC compound,
and is well known for its hot, lachrymose, and pungent eects
in crop species such as mustards (Brassica juncea and Brassica
nigra), horseradish (Armoracia rusticana), and wasabi (Eutrema
japonicum; Table ). These properties have led to the prevalence
of these crops as the main ingredients in condiment sauces in
various cultures across the world; mustards and horseradish are
common in European cuisine, and wasabi paste is widely used in
Japanese dishes such as sushi.[,,]
Wasabi is perhaps the most infamous pungent Brassicaceae
crop, known for its intense lachrymose eects when eaten. The
ITCs of glucocapparin (methyl ITC), glucoputranjivin (isopropyl
ITC), sinigrin (AITC), gluconapin (-butenyl ITC), and gluco-
brassicanapin (-pentenyl ITC) have all been identified within
wasabi roots, and all are described as having pungent or acrid
flavors and aromas (Table ).[] In horseradish by contrast, the
pungency of ITCs has also been interpreted as spicy, or even as
imparting a “cooling” sensation on the palate. This is a descriptor
that at first seems unlikely, yet the function of ANKTM TRP ion
channels also serves as a means of detecting noxious cold, and so
ITCs could be interpreted this way by the brain.[]
Horseradish root is commonly added to fermented foods in
Eastern Europe, as well as to bread as flavoring. It may even be
added to drinks, such as the “Bloody Mary,” as a substitute for
Tabasco sauce. The spiciness of horseradish is a fundamental
component of its uses and, generally, the more pungent it is, the
better from a culinary perspective.[] A study by D’Auria et al.[]
showed that the pungency of horseradish declines after grating
and during refrigeration, probably due to the cell damage in-
curred and the hydrolysis of GSLs when exposed to myrosinase.
Pungency can be preserved with the addition of fat or oil which
limits ITC degradation.[]
In Sinapis alba (white mustard), pungency is primarily caused
by -hydroxybenzyl ITC, the hydrolysis product of the aromatic
GSL glucosinalbin. This compound can cause intense burning
sensations that are overwhelming for many people; however, the
trait is considered enjoyable to mustard consumers after a pe-
riod of familiarization.[] Another pungent compound that is less
well known and studied is raphasatin (-methylthio--butenyl
ITC [MTBITC]), the ITC of glucoraphasatin (also known as
dehydroerucin).[] This compound is abundant in radish species
(Raphanus spp.) and is especially strong in varieties such as
daikon white radish.[] The degree of pungency of roots is linked
to growth season, and is highly dependent upon environmental
factors.[]
4.3. Consumer Acceptance of Pungency
Pungency has been linked with improved consumer acceptance
of radish in some cohorts.[,] Wills and Coogan[] demon-
strated that dierent cultural populations find the pungency of
radish, and the ITC raphasatin, to be more acceptable than oth-
ers. A Japanese cohort (who would be more familiar with radish
in traditional dishes) preferred the pungent sensations, whereas
an Australian cohort broadly rejected the sensation. There is ev-
idence to suggest that taster status for genes such as TASR
dier between ethnic groups. For example, populations of Asian
descent typically have fewer bitter nontasters (to PROP and PTC)
than Caucasians and South Asians.[] This is somewhat contrary
to what one would expect (as bitterness in Brassicaceae is typically
regarded as a negative trait), but Asian populations generally have
a much larger amount of these crops in their diets than other pop-
ulations. This is evidenced by reduced breast cancer risk in Asian
populations compared with Western, and due to the higher abun-
dance of ITCs present.[] It is however unknown how olfactory
receptor (OR) gene haplotypes dier between ethnic populations,
and therefore it is unknown if this is also a factor in determin-
ing the liking of pungency in cultures where such sensations are
common.
This perhaps demonstrates the diculty faced by plant
breeders in producing new cultivars with improved sensory
traits: what one population prefers may be completely rejected by
another, and for completely dierent reasons. It may therefore
be prudent for breeders to produce multiple varieties suited to
the tastes of dierent populations or markets, rather than having
a “one-size-fits-all” approach. As repeated exposure to sensory
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (7 of 13)
www.advancedsciencenews.com www.mnf-journal.com
signals generally increases liking, this also paves the way for
varieties increasing in strength, allowing for the developing
tastes of consumers—a commercial strategy that has been
adopted by chilli marketers.
Bell et al.[] noted that “hotness” (pungency) of salad rocket
leaves negatively impacted the liking of a large proportion of con-
sumers. They identified three consumer clusters with varying
reasons for their liking/disliking of the accessions tested: many
preferred milder and sweeter leaves (but were generally accepting
of all types; .%), while only a comparative minority preferred
the hot and peppery leaves (.%). The last group broadly re-
jected all of the samples presented, but slightly favored those that
were milder (.%). In this instance, it was pungency that was
the basis for consumer liking, and not bitterness.
In other species, pungency is not a desirable trait. Papaya fruit
contains glucotropaeolin, and its ITC (BITC) is noted for creating
unpalatable flavor which can be disliked by consumers.[] The
same is true of watercress, where excess concentrations of PEITC
can lead to increased pungency and consumer rejection.[] Fur-
ther, detailed consumer studies of other Brassicaceae are needed
in order to properly determine preferences and the concentra-
tions at which pungency becomes unpalatable.
5. Flavor of Isothiocyanates
5.1. General
Whereas pungency is a trigeminal sensation, isothiocyanates
in mustard, radish, and other pungent Brassicaceae can be
described as flavors or aromas that are perceived by olfactory
receptors (OR; Figure ). Humans are known to have around
genes encoding OR,[] meaning that there are far more
receptors dedicated to the sense of smell than there are to taste.
There are a few known individual dierences in the perception
of some odors caused by genotypic variation; for example,
β-ionone can be perceived as either floral or sour according to a
genetic variation in ORA, and the ability to detect green aroma
from -hexen--ol is aected by the genotype of ORJ.[,]
However, other aroma characteristics are (thought to be) per-
ceived consistently across a population. Some authors have
reported that the amount of genetic variance reported for some
ORs in the ability to perceive certain odors is low, suggesting that
many aromas and flavors are not perceived dierently between
individuals.[] However, Keller et al.[] demonstrated that SNPs
in genes encoding human odorant receptors partly accounted
for the variation in odor perception between individuals, and
other authors have reported that there are substantial copy
number variants in ORs, suggesting that the variation in odor
perception (interpretation or intensity) may be greater than we
are currently aware of ref. []. It is thus far unreported whether
the ability to perceive the smell of certain ITCs is heritable or
not. It is proposed that most odors are perceived through a
pattern recognition system, where a single compound can bind
with dierent anities to more than one OR;[] however, the
specific OR(s) which respond to ITCs are as yet unreported.
The intensity of ITC flavors and aromas in radish, mustard,
horseradish, and wasabi are often conflated with the intense
trigeminal pungency they also produce. These attributes are
more subtle and easier to “separate” cognitively in species of
rocket (E. sativa and D. tenuifolia[]) and watercress (Nasturtium
officinale),[] where pungency is not as intense. The intensity of
ITC flavor strength has been linked with agronomic practices,
such as the application of nutrient sulfur in radish,[] and may
be an agronomic means of increasing pungency without inten-
sive breeding in the short term.
Typical flavors of Brassicaceae have traditionally been linked
with GSLs and their hydrolysis products, such as sinigrin, glu-
coiberin, and glucoraphanin.[] Some studies have reported that
a higher GSL content is associated with a stronger flavor in
crops such as broccoli;[] however, above a certain threshold,
GSL and ITC content can become detrimental to consumer
acceptance.[,] As will be discussed, not all ITCs contribute
equally to flavor profiles; nor indeed are peppery, sulfurous, mus-
tard, and burnt flavors exclusively produced by these compounds.
5.2. Flavors Associated with Pungency
The plant genus Lepidium contains species commonly known
as pepperwort, peppercress, and peppergrass (and colloquially
in New Zealand as Cook’s Scurvy grass). Unsurprisingly, this
is because of the plants’ flavor and the ITC compounds they
contain (Table ). Sansom et al.[] performed informal sensory
evaluations of dried Lepidium species powders, determining
that the predominant ITC responsible for the “wasabi-like”
flavor of leaves is -butenyl ITC (BITC; hydrolysis product
of gluconapin). The Lepidium species tested contained diverse
GSL profiles, including sinigrin, glucotropaeolin, glucosinalbin,
glucolimnathin, glucocochlearin, and glucoputranjivin. The
ITC products of these latter three compounds in particular
are not well studied from a sensory perspective, and further
study is needed to elucidate their contribution to the overall
flavor profiles of these species and other crops. Ezo-wasabi
(Cardamine fauriei) is another such example, which contains
similar “wasabi-like” flavor attributes.[]
The presence of ITCs is regularly linked with peppery and
mustard characteristics of the leaf flavor profiles of rocket
(Table ). Until relatively recently, this was merely
speculation,[,] with virtually no experimental evidence to
link ITC compounds to this attribute. Bell et al.[] identified
a compound within the headspace of rocket, which was later
associated with pungent odor, sulfur and mustard flavors.[]
This compound was incorrectly identified, but follow-up work
by Rao et al.[] confirmed it to be the ITC of glucosativin
(sativin; -mercaptobutyl ITC).[] Odor extracts confirmed this
compound to be responsible for the typical rocket aroma that is
experienced when leaves are crushed or chewed.
5.3. Radish-Like Flavor
Erucin was characterized in rocket by Rao et al.[] as having a
“radish” aroma; it is unknown what olfactory receptor(s) is/are
responsible for this sensation. The lower relative intensity of
this compound and SF in isolation, compared to sativin, may ex-
plain why they are generally imperceptible to sensory assessors
within the whole tissue matrix during mastication. Phenethyl
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (8 of 13)
www.advancedsciencenews.com www.mnf-journal.com
ITC (PEITC) has also been described as having a radish-like flavor
in low concentrations. It is likely that ITCs in lower abundance
in pungent crops such as horseradish and wasabi subtly deter-
mine flavor,[,] and this may also vary significantly according to
cultivar.
5.4. Determination of Relative Flavor Contribution
As pointed out by Rao et al.,[] the method of extraction and its
duration seems to aect both the abundance and the suggested
contribution of specific ITC compounds to the overall aroma. The
variety of methods used to extract and characterize VOCs (GC-O,
aroma extract dilution analysis [AEDA], headspace-SPME, stir-
bar sorptive extraction [SBSE]) by Rao et al.[] demonstrated this
to great eect. Future studies on VOCs and ITCs should there-
fore be mindful of these experimental factors (such as extraction
eciency; as determined by the polarity of the compounds of in-
terest) as these could inadvertently bias the relative compound
intensities reported to be responsible for the perception of a par-
ticular attribute.
The strategy for performing multiple extracts and aroma anal-
yses as set out by Rao et al.[] may give the best overall represen-
tation of ITC and VOC contributions to flavor and aroma of Bras-
sicaceae species. As they highlighted, multiple cultivars should
be analyzed to determine the genotypic variability and abundance
of odor-active compounds. This could also feasibly be linked to
genetic components, and therefore allow selection for improved
flavor and aroma traits with a high degree of specificity.
6. Glucosinolate and Isothiocyanate Sensory
Relationships with Sweetness
6.1. Free Sugars
Several studies have observed that an increase in the abundance
of free sugars (e.g., glucose, fructose, galactose, sucrose) within
tissues confers a reduction in the intensity of bitterness[]
and also influences aftertaste.[] The bitterness of Brussels
sprouts is oset by perceptions of sweetness, for example,[] and
sensory analyses of cauliflower have shown that sweet-tasting
compounds modulate bitter perception.[] This has led to the
hypothesis that cultivar taste can be modified through the
manipulation of sugar-GSL/ITC ratios.[] As pointed out by
Padilla et al.,[] cultivars can have varying concentrations of
GSLs (high or low) but are perceived as being just as bitter as
one another. The important question then is to determine if
other compounds or hydrolysis products cause this, and/or if
the ratio with free sugars influences this perception.
It is interesting to note that cultivars of green and purple
cauliflowers are known not to have the characteristic bitter-
ness of the regular white variety. Instead, green and purple
cauliflowers have a sweet, mild, and “nutty” profile.[] Astudy
by Schonhof et al.[] compared the sensory profiles of white, Ro-
manesco (“pyramidal”), green, and purple cauliflower cultivars.
Analysis of GSL and sugar profiles showed that there were large
dierences in the sensory properties of each according to the
relative abundances of alkenyl GSLs (e.g., sinigrin, gluconapin,
progoitrin), indole GSLs (glucobrassicin, neoglucobrassicin,
-hydroxyglucobrassicin, and -methoxyglucobrassicin), and
sucrose concentrations. Consumers generally preferred the
cultivars with brighter colors (green and purple) and those
that had fewer of the aforementioned bitter-tasting GSLs. The
colored cauliflower types were higher in glucoraphanin, but
lower in glucoerucin compared to the conventional white variety.
However, all of the cauliflower cultivars were lower in abundance
for glucoraphanin compared to broccoli. Interestingly, neither
of these GSLs was strongly associated with bitterness.
The authors suggested that the modification of sugar profiles
to counteract bitterness created by GSLs is a strategy for in-
creasing consumer acceptance and enhancing the amounts of
GSLs/ITCs/indoles within the diet. The trend for species belong-
ing to a single genus, and cultivars within a single species, to have
dierent taste and flavor profiles to mainstream crops is also ap-
parent in rape kale (Brassica napus var. pabularia; “nabicol” and
“couve nabica” cultivars), which are also said to be less bitter than
conventional kale (B. oleracea var. acephala).[,]
6.2. Free Amino Acids
It has been hypothesized that free amino acid concentrations
within plant material may also modulate perception of bitterness
and pungency. Bell et al.[] observed that accessions of rocket
salad were perceived to be less bitter and pungent where con-
centrations of free amino acids were higher.
Compounds such as alanine, threonine, serine, and proline
are known to infer sweetness in foodstus.[,] Park et al.[] ob-
served that glycine and alanine contributed to sweetness in in-
bred cabbage lines, and that valine and leucine infer bitter tastes.
Some authors have even suggested that the primary cause of bit-
terness in mizuna (B. juncea var. japonica) leaves comes from low
concentrations of -glutamine and -asparagine, and high con-
centrations of malic acid, and is not GSL- or ITC-derived, despite
the presence of AITC.[] Further research is needed to adequately
determine the eects of free amino acids on Brassicaceae taste
and flavor profiles, and to separate their relative eects in rela-
tion to GSL/ITC-derived bitterness.
7. Contribution of Sulfur-Containing Volatile
Organic Compounds to the Taste of Brassicaceae
7.1. Nitriles and Sulfur-Containing VOCs
The aforementioned sensory attributes of bitterness, pungency,
and pepperiness are all typically assumed to be characteristics
solely of GSLs or their ITCs. However other compounds are also
known to contribute toward these perceptions.[] For example,
the presence of nitrile compounds in nakajimana (Brassica rapa
cv. nakajimana), a Japanese leafy vegetable, have been linked with
pungent and bitter sensory attributes.[]
Three independent studies of rocket salad VOCs have reported
tetrahydrothiophene as a pungent odorant of rocket.[,,] Rao
et al.[] described the compound as having a “gas-like” odor, and
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (9 of 13)
www.advancedsciencenews.com www.mnf-journal.com
Bell et al.[] observed strong correlations of this compound with
sensory descriptive attributes such as “burnt rubber” odor, as well
as with perceptions of heat, tingling, and bitter taste. It is un-
known how this compound is synthesized in planta,butasyn-
thetic version of the compound is used routinely as an odorant
in the gas industry because of its strong aroma.[] It may be an
ITC degradation product, but this has not been demonstrated ex-
perimentally.
7.2. Sulfides
O-odors, such as “sulfurous,” “earthy,” or “musty” are of-
ten attributed to Brassicaceae crops and GSL/ITC content.[]
These traits predominantly stem from sulfur-compound degra-
dation products, such as from S-methyl-L-cysteine sulfoxide
(SMCSO),[] and formation can be facilitated by factors such as
bacterial metabolism, plant senescence, cooking, and enzymatic
breakdown as a result of tissue damage (e.g., cutting).[–]
Sulfides are generally undesirable flavor attributes,[] and
compounds like methanethiol, dimethyl sulfide (DMS), dimethyl
trisulfide (DMTS), and dimethyl disulfide (DMDS) are regularly
linked with sulfurous aromas and overcooked o-flavors.[,,]
DMTS is produced during the cooking of cauliflower and other
Brassica varieties, and is likely to be greater in cultivars with high
sulfur content.[] These compounds are particularly problematic
in Brassicaceae foods because they are detectable by humans at
very low concentrations (. ppb).[] Methanethiol has been
noted for its presence in anaerobic atmospheres, and its forma-
tion is potentially promoted when modified atmospheric pack-
aging (MAP) is used for preparing bagged salads and leafy Bras-
sicas. For cooked Brassicas, the amount of sulfur volatiles formed
is highly dependent upon duration, temperature, and total tissue
water content.[] Storage and processing conditions post harvest
are also likely to impact the abundance and types of VOCs, GSLs,
and ITCs produced,[] and therefore the relative abundance of
sulfides.
GSLs and SMCSO in Brassica species both rely on available
sulfur for their formation. As mentioned previously, the GSL glu-
coraphanin is present in greater concentrations in new cultivars
of broccoli, such as Benefort´
e. This upregulation of GSL forma-
tion competes with SMCSO for sulfur; hence an increase in the
proportion of sulfur allocated to methionine-derived glucosino-
lates results in a decrease in that allocated to SMCSO.[] It is
perhaps unsurprising therefore that such high GSL broccoli has
proved acceptable to consumers, as neither glucoraphanin nor
sulforaphane are bitter-tasting compounds, and such varieties
could theoretically result in less sulfide generation. This has not
been explicitly tested in this variety, however, and the concentra-
tions of sulfides produced are still likely to exceed the minimum
detection threshold.
7.3. Thiocyanates
GSLs produce diverse hydrolysis products, but very few other
than ITCs have been sensorially characterized or studied in great
detail. Some Brassicaceae produce thiocyanates as a GSL hydrol-
ysis product, as well as ITCs. Plants such as the aptly named
stinkweed (Thlaspi arvense), and land cress (Coronopus didymus),
produce these o-smelling compounds as a defense strategy
against herbivory.[]
Allyl thiocyanate (ATC) is another of the hydrolysis products
of sinigrin. Chin et al.[] reported that it was problematic to de-
termine the odor characteristics of ATC because it is relatively
unstable, and dicult to separate from AITC; during authentic
compound synthesis distillation conditions, ATC readily converts
to AITC. They were however able to determine the odor of ATC as
musty, sulfurous, and mustard-like. Relatively little else is known
about these compounds and their eects upon odors within com-
mercial crops.
8. Modification of Taste and Flavor Profiles
through Cooking
As emphasized in the introduction, processing substantially
modifies taste and flavor profiles of Brassica vegetables. The re-
spective thermal stabilities of myrosinase and the specifier pro-
teins are of key importance. As discussed earlier, there are a
variety of end products that can result from GSL hydrolysis.
Where myrosinase enzyme and ESP are both intact, there is a
tendency for greater nitrile production and reduced ITC produc-
tion; the reverse occurs when ESP is denatured but myrosinase
remains intact. For example, in broccoli, ESP stimulates gluco-
raphanin conversion toward sulforaphane nitrile at the expense
of sulforaphane.[] It is unknown how nitrile formation impacts
taste and flavor, if at all, in major crops.
Thermal stability of myrosinase and ESP is reported to vary
in dierent Brassica varieties and between cultivars;[,] how-
ever, ESP is generally denatured at lower temperatures compared
to myrosinase.[,] Therefore, in cases with mild cooking condi-
tions, there can be a greater concentration of resulting ITC than
in the raw, macerated plant; but under more typical cooking con-
ditions (e.g., boiling or steaming) both ESP and myrosinase may
be denatured, resulting in intact GSL being present in the cooked
product but few hydrolysis products. Cooking will also aect the
generation of sulfur volatiles, and the characteristic sulfurous
odor of cooked Brassica (e.g. boiled cabbage) that results from the
release of sulfides, such as DMS. No published research has been
conducted to determine how ITC–nitrile ratios influence taste,
flavor, and consumer acceptance of Brassica cultivars. Further in-
vestigation of cooking practices to optimize both nutritional and
sensory quality of Brassica types is needed.
9. Modification of Taste and Flavor Profiles
through Breeding
Taste and flavor profiles can be modified through selective breed-
ing, but this often comes at the expense of nutritional health
benefits.[] There is renewed focus for breeding Brassicaceae
crops with greater GSL content, whilst maintaining consumer ac-
ceptability. This must be approached cautiously, as focusing on
only one particular set of compounds (such as only GSLs) may
bias interpretations of sensory-chemical data.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (10 of 13)
www.advancedsciencenews.com www.mnf-journal.com
It would be prudent for breeding eorts to determine the ITC
profile and myrosinase activity of cultivars in combination with
the abundance of free sugars, VOCs, and (possibly) amino acids
to modify taste and flavor. This would be a wiser track than se-
lecting plants only for milder taste, and naively assuming that
ITC content would be unaected in direct response to unchanged
GSL content. In this way, both health benefits and sensory ac-
ceptability can be monitored and preserved, and eventually im-
proved over several generations. The sensations of pungency and
hotness are not a reliable indicator of the quantity of health-
beneficial ITCs present, as evidenced by the fact that some (such
as SF) have no discernable flavor. Others may be masked by high
relative sugar content, and there is a danger that sweetness is
selected for at the expense of bitter and pungent ITCs, which
are potentially health beneficial. It is also not sucient to as-
sume that GSL/ITC content will be preserved over generations if
they are not constantly and quantitatively measured. This also ap-
plies across growing environments and climates, where changes
to GSL biosynthesis (as well as other phytochemicals) can have
a large impact upon taste, flavor, and pungency of Brassicales
crops.[] Breeding of Brassica plants to optimize GSL or ITC con-
tent must also consider how the plant will be processed before
consumption. Accounting for all of these influencing factors is
a considerable challenge, and requires large amounts of invest-
ment in both breeding programs and conducting the required
phytochemical analyses.
Many papers highlighting the potential for breeding new cul-
tivars with improved health benefits neglect the importance of
co-selecting for taste and flavor attributes in tandem with phyto-
chemical traits, which are arguably just as important. Many con-
sumers will reject a product on the first instance of consumption
if it does not taste good to them, no matter how “healthy” it is
purported to be. This needs to be a fundamental consideration
in breeding research eorts.[]
10. Summary and Conclusions
It is generally accepted that many GSL compounds impart bitter-
ness and many ITCs impart pungency. As has been discussed,
however, this is not universally applicable, and more detailed
studies are needed to establish perception thresholds of isolated
compounds. This will have to involve either the food grade syn-
thesis of compounds, or isolation from plant material. The lat-
ter option is likely to be the preferred method for nonexpert
chemists, and it would be desirable for food grade standards to
be commercially available from specialist suppliers. There is po-
tentially a large market for such a service for these compounds.
If the isolated tastes and/or flavors can be experimentally deter-
mined (e.g., by taste extraction dilution analysis for nonvolatiles,
and GC-O for volatiles), this would allow a better understand-
ing of their contribution to the sensory profiles of Brassicaceae
vegetables, and why perceptions vary between cultivars. Individ-
ual ITCs have distinct aromas/flavors as highlighted in Table ;
but determining their relative contribution to a sensation in the
whole food matrix has not been quantitatively performed. This
would perhaps require quantitative reconstruction of an extract
in a similar way to what has been achieved in wine;[] but until
stable food grade standards become more widely available, this
will remain an unexplored area. Much more research is needed
to characterize the sensory attributes of nitriles, thiocyanates, and
indoles in dierent crop species, and how they aect taste and/or
flavor.
The human olfactory system is far more sensitive than any GC-
MS,[] hence employing the use of sensory and consumer panels
in the development and research of crop varieties and cultivars
for enhanced traits should be integral until more sophisticated
approaches can be developed. Additionally, more fundamental
research is required to understand the trigeminal and olfactory
mechanisms responsible for ITC perception in humans.
It is also important not to conflate the sensory eects and con-
tributions of GSLs and ITCs when conducting correlation anal-
yses. GSLs are phytochemical precursors to many compounds,
not just ITCs, and basing crop improvement choices solely upon
GSL content is therefore not guaranteed to be indicative of down-
stream sensory perceptions generated by hydrolysis products. Se-
lective breeding programs should take an informed and analytical
approach when basing selections upon taste and flavor. Without
chemical analysis, it is impossible to determine the ITC status of
Brassicales crops, and breeders should endeavor to incorporate
these into breeding programs wherever possible.
To complicate the picture further, the abundance of free sug-
ars, free amino acids, sulfides, and phenolics also need to be ac-
counted for in determining the taste and flavor of Brassicaceae
and other GSL/ITC-producing crops such as Moringa and Pa-
paya. The eects imparted by sweet or bitter compounds within
the food matrix are likely to depend greatly on the individual
GSL/ITC profile of each respective species, and the relative bit-
terness/pungency they in turn produce. This interaction could
be fundamental for producing new cultivars with both enhanced
sensory and health-promoting properties in future.
Conflict of Interest
The authors declare no conflict of interest.
Keywords
aroma, glucosinolates, isothiocyanates, pungency, sulfide, taste
Received: November 29, 2017
Revised: February 5, 2018
Published online:
[1] B. Holst, G. Williamson, Nat. Prod. Rep. 2004,21, 425.
[2] N. Agerbirk, C. E. Olsen, Phytochem.2012,77, 16.
[3] J. W. Fahey, A. T. Zalcmann, P. Talalay, Phytochem.2001,56,5.
[4] J. C. Kuchernig, M. Burow, U. Wittstock, BMC Evol. Biol. 2012,12,
127.
[5] N. V. Matusheski, E. H. Jeffery, J. Agric. Food Chem. 2001,49, 5743.
[6] N. V. Matusheski, J. A. Juvik, E. H. Jeffery, Phytochem.2004,65, 1273.
[7] B. A. Halkier, J. Gershenzon, Ann. Rev. Pl. Biol. 2006,57, 303.
[8] G. Bonnema, R. Verkerk, presented at 3rd Intl. Glucosinolate Conf.,
Wageningen Universtiy, Wageningen, Netherlands 2014.
[9] A. Drewnowski, C. Gomez-Carneros, Am. J. Clin. Nutr. 2000,72,
1424.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (11 of 13)
www.advancedsciencenews.com www.mnf-journal.com
[10] A. T. Dinkova-Kostova, R. V. Kostov, Trends Mol. Med. 2012,18, 337.
[11] J. Jakubikova, Y. P. Bao, J. Sedlak, Anticancer Res.2005,25, 3375.
[12] F. H. Sarkar, Y. W. Li, J. Nutr. 2004,134, 3493S.
[13] L. M. Beaver, R. Kuintzle, A. Buchanan, M. W. Wiley, S. T. Glasser,
C. P. Wong, G. S. Johnson, J. H. Chang, C. V. L¨
ohr, D. E. Williams, R.
H. Dashwood, D. A. Hendrix, E. Ho, J. Nutr. Biochem. 2017,42, 72.
[14] J. D. Hayes, M. O. Kelleher, I. M. Eggleston, Eur. J. Nutr. 2008,47,
73.
[15] C.-L. Wu, A.-C. Huang, J.-S. Yang, C.-L. Liao, H.-F. Lu, S.-T. Chou,
C.-Y. Ma, T.-C. Hsia, Y.-C. Ko, J.-G. Chung, J. Orthop. Res. 2011,29,
1199.
[16] A. Melchini, M. H. Traka, S. Catania, N. Miceli, F. Taviano, P. Mai-
mone, M. Francisco, R. F. Mithen, C. Costa, Nutr. Cancer 2013,65,
132.
[17] S. Giacoppo, M. Galuppo, S. Montaut, R. Iori, P. Rollin, P. Bramanti,
E. Mazzon, Fitoterapia 2015,106, 12.
[18] T. A. Shapiro, J. W. Fahey, A. T. Dinkova-Kostova, W. D. Holtzclaw, K.
K. Stephenson, K. L. Wade, L. Ye, P. Talalay, Nutr. Cancer 2006,55,
53.
[19] L. Bell, M. J. Oruna-Concha, C. Wagstaff,Food Chem.2015,172, 852.
[20] I. Schonhof, A. Krumbein, B. Br¨
uckner, Food/Nahrung 2004,48, 25.
[21] H. Y. Baik, J. Juvik, E. H. Jeffery, M. A. Wallig, M. Kushad, B. P. Klein,
J. Food Sci. 2003,68, 1043.
[22] G. R. Fenwick, N. M. Griffiths, R. K. Heaney, J. Sci. Food Agric. 1983,
34, 73.
[23] D. Zabaras, M. Roohani, R. Krishnamurthy, M. Cochet, C. M. De-
lahunty, Food Funct.2013,4, 592.
[24] J. S. Elmore, in Flavour Development, Analysis and Perception in Food
and Beverages, 1st ed. (Eds: J. K. Parker, J. S. Elmore, L. Methven),
Woodhead Publishing, Cambridge, UK 2014, p. 31.
[25] J. S. Elmore, in Flavour Development, Analysis and Perception in Food
and Beverages (Eds: J. K. Parker, J. S. Elmore, L. Methven), Woodhead
Publishing, Cambridge, UK 2014, p. 47.
[26] J. K. Parker, in Flavour Development, Analysis and Perception in Food
and Beverages (Eds: J. K. Parker, J. S. Elmore, L. Methven), Woodhead
Publishing, Cambridge, UK 2014,p.3.
[27] K. Ridgway, in Flavour Development, Analysis and Perception in Food
and Beverages (Eds: J. K. Parker, J. S. Elmore, L. Methven), Woodhead
Publishing, Cambridge, UK 2014, p. 63.
[28] H. T. Lawless, H. Heymann (Eds.), Sensory Evaluation of Food,
Springer, New York 2010.
[29] B. Siegmund, in Flavour Development, Analysis and Perception in Food
and Beverages (Eds: J. K. Parker, J. S. Elmore, L. Methven), Woodhead
Publishing, Cambridge, UK 2014, p. 127.
[30] L. Bell, L. Methven, A. Signore, M. Jose Oruna-Concha, C. Wagstaff,
Food Chem.2017,218, 181.
[31] H. Wold, in Encyclopedia of Statistical Sciences (Eds: S. Kotz, N. John-
son),John Wiley & Sons, Inc., Hoboken, NJ 2006, p. 581.
[32] S. Frank, N. Wollmann, P. Schieberle, T. Hofmann, J. Agric. Food
Chem. 2011,59, 8866.
[33] J. Prescott, E. Monteleone, in Flavour Development, Analysis and Per-
ception in Food and Beverages (Eds: J. K. Parker, J. S. Elmore, L.
Methven), Woodhead Publishing, Cambridge, UK 2014, p. 369.
[34] L. Methven, in Flavour Development, Analysis and Perception in Food
and Beverages (Eds: J. K. Parker, J. S. Elmore, L. Methven), Woodhead
Publishing, Cambridge, UK 2014, p. 353.
[35] L. Bell, C. Wagstaff, J. Agric. Food Chem. 2017,65, 9379.
[36] M. E. Cartea, V. M. Rodr
´
ıguez, A. de Haro, P. Velasco, A. Ord´
as,
Euphytica 2008,159, 111.
[37] F. S. Hanschen, M. Schreiner, Front. Plant Sci. 2017,8, 1095.
[38] M. Behrens, S. Foerster, F. Staehler, J.-D. Raguse, W. Meyerhof, J.
Neurosci. 2007,27, 12630.
[39] J. A. Mennella, M. Y. Pepino, F. F. Duke, D. R. Reed, BMC Genet.
2010,11, 60.
[40] E. Engel, C. ¨
A. Line Baty, D. Le Corre, I. Souchon, N. Martin, J. Agric.
Food Chem. 2002,50, 6459.
[41] W. Meyerhof, C. Batram, C. Kuhn, A. Brockhoff, E. Chudoba, B. Bufe,
G. Appendino, M. Behrens, Chem. Senses 2010,35, 157.
[42] U.-K. Kim, E. Jorgenson, H. Coon, M. Leppert, N. Risch, D. Drayna,
Science 2003,299, 1221.
[43] M. A. Sandell, P. A. S. Breslin, Curr. Biol. 2006,16, R792.
[44] Y. Shen, O. B. Kennedy, L. Methven, Food Qual. Prefer. 2016,50, 71.
[45] J. E. Hayes, R. S. J. Keast, Physiol. Behav. 2011,104, 1072.
[46] L. Bell, L. Methven, C. Wagstaff, Food Chem.2017,222,6.
[47] R. B. Jones, J. D. Faragher, S. Winkler, Postharvest Biol. Technol. 2006,
41,1.
[48] H. B. Frandsen, K. E. Markedal, O. Mart
´
ın-Belloso, R. S´
anchez-Vega,
R. Soliva-Fortuny, H. Sørensen, S. Sørensen, J. C. Sørensen, Polish
J. Food Nutr. Sci. 2014,64, 17.
[49] B. C. Pence, F. Buddingh, S. P. Yang, J. Natl. Cancer Inst. 1986,77,
269.
[50] H. E. Van Doorn, Ph.D. Thesis,University of Wageningen,1999.
[51] K. A. Steinmetz, J. D. Potter, Cancer Causes Control 1991,2, 427.
[52] F. Pasini, V. Verardo, L. Cerretani, M. F. Caboni, L. F. D’Antuono, J.
Sci. Food Agric. 2011,91, 2858.
[53] L. F. D’Antuono, S. Elementi, R. Neri, Phytochem 2008,69, 187.
[54] L. Ludikhuyze, L. Rodrigo, M. Hendrickx, J. Food Prot. 2000,63, 400.
[55] D. Y. Zhao, J. Tang, X. L. Ding, LWT Food Sci. Technol. 2007,40,
439.
[56] M. Traka, R. Mithen, Phytochem. Rev. 2009,8, 269.
[57] M. H. Traka, S. Saha, S. Huseby, S. Kopriva, P. G. Walley, G. C. Barker,
J. Moore, G. Mero, F. van den Bosch, H. Constant, L. Kelly, H. Schep-
ers, S. Boddupalli, R. F. Mithen, New Phytol.2013,198, 1085.
[58] S.-E.Jordt,D.M.Bautista,H.Chuang,D.D.McKemy,P.M.Zyg-
munt, E. D. H¨
ogest¨
att, I. D. Meng, D. Julius, Nature 2004,427, 260.
[59] A. A. Agrawal, N. S. Kurashige, J. Chem. Ecol. 2003,29, 1403.
[60] T. Sultana, G. P. Savage, D. L. McNeil, G. P. Porter, B. Clark, J. Food
Agric. Environ. 2003,1, 117.
[61] R. Agneta, C. M¨
ollers, A. R. Rivelli, Genet. Resour. Crop Evol. 2013,
60, 1923.
[62] J. A. Depree, T. M. Howard, G. P. Savage, Food Res. Int. 1999,31,
329.
[63] M. D’Auria, G. R. Mauriello, Ital. J. Food Sci. 2004,39, 1079.
[64] S. K. Ghawi, Y. Shen, K. Niranjan, L. Methven, J. Food Sci. 2014,79,
S1756.
[65] W. Chen, E. Karangwa, J. Yu, S. Xia, B. Feng, X. Zhang, C. Jia, Food
Bioprocess Technol.2017,10,1.
[66] C. Suzuki, M. Ohnishi-Kameyama, K. Sasaki, T. Murata, M. Yoshida,
J. Agric. Food Chem. 2006,54, 9430.
[67] R. C. Coogan, R. B. H. Wills, V. Q. Nguyen, Food Chem.2001,72,1.
[68] R. B. H. Wills, R. C. Coogan, J. Sens. Stud. 2003,18, 83.
[69] N. H. Khataan, L. Stewart, D. M. Brenner, M. C. Cornelis, A. El-
Sohemy, J. Nutrigenet. Nutrigenomics 2009,2, 251.
[70] J. H. Fowke, F. L. Chung, F. Jin, D. Qi, Q. Cai, C. Conaway,J.-R. Cheng,
X.-O. Shu, Y.-T. Gao, W. Zheng, Cancer Res.2003,63, 3980.
[71] M. Possenti, S. Baima, A. Raffo, A. Durazzo, A. M. Giusti, F. Natella,
Glucosinolates, Springer International Publishing, Cham, Switzer-
land 2016,p.1.
[72] U. R. Palaniswamy, Horttechnology 2001,11, 622.
[73] D. R. Reed, A. Knaapila, Prog. Mol. Biol. Transl. Sci. 2010,94, 213.
[74] S. R. Jaeger, J. F. McRae, C. M. Bava, M. K. Beresford, D. Hunter,
Y. Jia, S. L. Chheang, D. Jin, M. Peng, J. C. Gamble, K. R. Atkinson,
L. G. Axten, A. G. Paisley, L. Tooman, B. Pineau, S. A. Rouse, R. D.
Newcomb, Curr. Biol. 2013,23, 1601.
[75] J. F. McRae, J. D. Mainland, S. R. Jaeger, K. A. Adipietro, H. Mat-
sunami, R. D. Newcomb, Chem. Senses 2012,37, 585.
[76] A. Keller, H. Zhuang, Q. Chi, L. B. Vosshall, H. Matsunami, Nature
2007,449, 468.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (12 of 13)
www.advancedsciencenews.com www.mnf-journal.com
[77] Y. Hasin, T. Olender, M. Khen, C. Gonzaga-Jauregui, P. M. Kim, A.
E. Urban, M. Snyder, M. B. Gerstein, D. Lancet, J. O. Korbel, PLoS
Genet.2008,4, e1000249.
[78] L. Buck, R. Axel, Cell 1991,65, 175.
[79] J. Jin, O. A. Koroleva, T. Gibson, J. Swanston, J. Magan, Y. Zhang, I.
R. Rowland, C. Wagstaff, J. Agric. Food Chem. 2009,57, 5227.
[80] N. Voutsina, A. C. Payne, R. D. Hancock, G. J. J. Clarkson, S. D. Roth-
well, M. A. Chapman, G. Taylor, BMC Genom.2016,17, 378.
[81] G. G. Freeman, N. Mossadeghi, J. Sci. Food Agric. 1972,23, 387.
[82] M. Hansen, A. M. Laustsen, C. E. Olsen, L. Poll, H. Sørensen, J. Food
Qual. 1997,20, 441.
[83] C. E. Sansom, V. S. Jones, N. I. Joyce, B. M. Smallfield, N. B. Perry,
J. W. van Klink, J. Agric. Food Chem. 2015,63, 1833.
[84] K. Abe, S. Kido, T. Maeda, D. Kami, H. Matsuura, H. Shimura, T.
Suzuki, Sci. Hortic. (Amsterdam) 2015,189, 12.
[85] R. N. Bennett, R. Carvalho, F. A. Mellon, J. Eagles, E. A. S. Rosa, J.
Agric. Food Chem. 2007,55, 67.
[86] R. N. Bennett, F. A. Mellon, N. P. Botting, J. Eagles, E. A. S. Rosa, G.
Williamson, Phytochem.2002,61, 25.
[87] L. Bell, N. D. Spadafora, C. T. M¨
uller, C. Wagstaff, H. J. Rogers, Food
Chem.2016,194, 626.
[88] A. Raffo, M. Masci, E. Moneta, S. Nicoli, J. S. del Pulgar, F. Paoletti,
Food Chem.2018,240, 1161.
[89] M. S. Cerny, E. Taube, R. Battaglia, J. Agric. Food Chem. 1996,44,
3835.
[90] T. Sultana, G. P. Savage, D. L. McNeil, N. G. Porter, R. J. Martin, B.
Deo, J. Sci. Food Agric. 2002,82, 1477.
[91] L.H.Levine,P.A.Bisbee,J.T.Richards,M.N.Birmele,R.L.Prior,
M.Perchonok,M.Dixon,N.C.Yorio,G.W.Stutte,R.M.Wheeler,
Adv. Sp. Res. 2008,41, 754.
[92] G. Padilla, M. E. Cartea, P. Velasco, A. de Haro, A. Ord´
as, Phytochem.
2006,68, 536.
[93] J. Kirimura, A. Shimizu, A. Kimizuka, T. Ninomiya, N. Katsuya, J.
Agric. Food Chem. 1969,17, 689.
[94] T. Nishimura, H. Kato, Food Rev. Int. 1988,4, 175.
[95] S. Park, M. V. Arasu, M.-K. Lee, J.-H. Chun, J. M. Seo, S.-W. Lee, N.
A. Al-Dhabi, S.-J. Kim, Food Chem.2014,145, 77.
[96] T. Fukuda, K. Okazaki, A. Watanabe, T. Shinano, N. Oka,
Metabolomics 2015,12, 20.
[97] F. S. Hanschen, E. Lamy, M. Schreiner, S. Rohn, Angew. Chemie Int.
Ed. 2014,53, 11430.
[98] M. Kato, Y. Imayoshi, H. Iwabuchi, K. Shimomura, J. Agric. Food
Chem. 2011,59, 11034.
[99] L. Jirovetz, D. Smith, G. Buchbauer, J. Agric. Food Chem. 2002,50,
4643.
[100] J. Swanston, Ullmann’s Encyclopedia of Industrial Chemistry,Wiley-
VCH Verlag GmbH & Co. KGaA, Weinheim, Germany 2000.
[101] G. S. Stoewsand, Food Chem. Toxicol. 1995,33, 537.
[102] M. Palermo, N. Pellegrini, V. Fogliano, J. Sci. Food Agric. 2014,94,
1057.
[103] R. Kubec, V. Drhov´
a, J. Vel
´
ıˇ
sek, J. Agric. Food Chem. 1998,46, 4334.
[104] M. Hansen, R. G. Buttery, D. J. Stern, M. I. Cantwell, L. C. Ling, J.
Agric. Food Chem. 1992,40, 850.
[105] H.-W. Chin, R. C. Lindsay, J. Food Sci. 1993,58, 835.
[106] R. C. Lindsay, J. K. Rippe, Biogeneration of Aromas, American Chem-
ical Society, Washington D.C., USA 1986, p. 286.
[107] N. Rangkadilok, B. Tomkins, M. E. Nicolas, R. R. Premier, R. N. Ben-
nett, D. R. Eagling, P. W. J. Taylor, J. Agric. Food Chem. 2002,50,
7386.
[108] G. R. Fenwick, R. K. Heaney, Food Chem.1983,11, 249.
[109] H.-W. Chin, Q. Zeng, R. C. Lindsay, J. Food Sci. 1996,61, 101.
[110] S. K. Ghawi, L. Methven, K. Niranjan, Food Chem.2013,138,
1734.
[111] O. A. Okunade, S. K. Ghawi, L. Methven, K. Niranjan, Food Chem.
2015,187, 485.
[112] M. Ishida, M. Hara, N. Fukino, T. Kakizaki, Y. Morimitsu, Breed. Sci.
2014,64, 48.
[113] J. C. Hufnagel, T. Hofmann, J. Agric. Food Chem.2008,56, 9190.
Mol. Nutr. Food Res. 2018, 1700990 C2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700990 (13 of 13)
