2017 Toledo Font Del Rio NIRS Glucos Eruca Molecules Publicado - 22-00851

molecules
Article
Rapid and Cost-Effective Quantification of
Glucosinolates and Total Phenolic Content in Rocket
Leaves by Visible/Near-Infrared Spectroscopy
Eva Mara Toledo-Martn 1 , Rafael Font 2 , Sara Obregn-Cano 3 , Antonio De Haro-Bailn 3 ,
Myriam Villatoro-Pulido 3 and Mercedes Del Ro-Celestino 1, *
1 Department of Genomics and Biotecnology, IFAPA Center La Mojonera, Camino San Nicols, La Mojonera 1,
04745 Almera, Spain; ortiztoledo@hotmail.com
2 Department of Food and Health, IFAPA Center La Mojonera, Camino San Nicols, La Mojonera 1,
04745 Almera, Spain; rafaelm.font@juntadeandalucia.es
3 Department of Agronomy and Plant Breeding, Institute of Sustainable Agriculture, (CSIC),
Alameda del Obispo s/n, 14080 Crdoba, Spain; saraobregon@ias.csic.es (S.O.-C );
adeharobailon@ias.csic.es (A.D.H.-B); mvillatoro@ias.csic.es (M.V.-P.)
* Correspondence: mercedes.rio.celestino@juntadeandalucia.es; Tel.: +34-950-156453
Academic Editor: Derek J. McPhee

Received: 27 April 2017; Accepted: 17 May 2017; Published: 20 May 2017
Abstract: The potential of visible-near infrared spectroscopy to predict glucosinolates and total
phenolic content in rocket (Eruca vesicaria) leaves has been evaluated. Accessions of the E. vesicaria
species were scanned by NIRS as ground leaf, and their reference values regressed against different
spectral transformations by modified partial least squares (MPLS) regression. The coefficients of
determination in the external validation (R2 VAL) for the different quality components analyzed
in rocket ranged from 0.59 to 0.84, which characterize those equations as having from good to
excellent quantitative information. These results show that the total glucosinolates, glucosativin and
glucoerucin equations obtained, can be used to identify those samples with low and high contents.
The glucoraphanin equation obtained can be used for rough predictions of samples and in case
of total phenolic content, the equation showed good correlation. The standard deviation (SD) to
standard error of prediction ratio (RPD) and SD to range (RER) were variable for the different quality
compounds and showed values that were characteristic of equations suitable for screening purposes
or to perform accurate analyses. From the study of the MPLS loadings of the first three terms of
the different equations, it can be concluded that some major cell components such as protein and
cellulose, highly participated in modelling the equations for glucosinolates.
Keywords: glucosinolates; rocket; Eruca vesicaria; visible spectroscopy; near infrared; spectroscopy;
chemometrics
1. Introduction
Rocket is a herb originated in the Mediterranean region [1], where the aerial part is mostly
consumed fresh as salad, but widely distributed all over the world [2]. The term rocket refers
mainly to Eruca and Diplotaxis genera within the Cruciferae family. Throughout more than 20 centuries,
traditional medicine has attributed to rocket plants a number of health-promoting or therapeutic
properties, such as antiphlogistic, depurative, diuretic, digestive, aphrodisiac and rubefacient [3].
Eruca contains a range of health-promoting phytochemicals including carotenoids, vitamin C,
fibres, polyphenols, and glucosinolates (GLs) [4]. GSLs are evolutionarily recent secondary metabolic
products having arisen million years ago [5], acting as stimulants or deterrents to insects and
Molecules 2017, 22, 851; doi:10.3390/molecules22050851 www.mdpi.com/journal/molecules

Molecules 2017, 22, 851 2 of 14
Molecules 2017, 22, 851 2 of 15
herbivores [6]. They are known in only a few angiosperm families of the order Brassicales, which
includes
Molecules
Molecules
Molecules
Molecules
Molecules
the
2017,
2017, 2017,
2017,
2017, 22,
22,22,Brassicaceae
22,
851
851
22,
851
851 851 family, and of which Eruca is a member [7]. 22of
2 2ofof1414
2ofof
1414
14
herbivores [6]. They are known in only a few angiosperm families of the order Brassicales, which
Glucosinolates have a well-defined structure with a side-chain (R-group) and D-glucopyranose
includes
herbivores the Brassicaceae
[6]. They are family,
known and
in of
only which
afew Eruca is a member [7].
as herbivores
herbivores [6].
a -thioglucoside
herbivores
herbivores [6].
[6]. [6].
They
They
They They
are
are are
known
attached
are known
known known to
in in
inincarbon
only
only
only only
aaafew
few
atomafewfew angiosperm
no. angiosperm
angiosperm families
families
families
0 in (Z)-N-hydroximine
angiosperm
angiosperm families
families of
ofof
ofofthe
the the
the the order
orderorder
order
order
Brassicales,
Brassicales,
Brassicales,
sulphate esters
Brassicales,
Brassicales,
which
which
[8]
which
which which
(Figure
Glucosinolates
includes the have family,
Brassicaceae a well-definedand of structure
which Eruca with is a amember
side-chain [7]. (R-group) and D-glucopyranose as
includes
includes
includes
1). The the
includes the
the the Brassicaceae
Brassicaceae
Brassicaceae
structural
Brassicaceaediversity family,
family,
family,
family, ofandand
and and
ofof
glucosinolatesof
which which
which
of which Eruca Eruca Eruca
Eruca is
is mainlyis a ais
is a member a
member member
member
due to[7]. [7].
[7].[7].
the different substituents possible at the
a -thioglucoside
Glucosinolates
Glucosinolates
attached
have towell-defined
carbon atom
awell-defined
awell-defined
no. 0 in
structure with (Z)-N-hydroximine
aside-chain
aside-chain
sulphate
(R-group) and Desters [8] (Figure 1).
-glucopyranose
Glucosinolates
Glucosinolates
Glucosinolates
side-chain position have
have R,have
have aawhich
well-defined
awell-defined
can bestructure structure
structure
structure
very with
with
with
variable with [9]. side-chain
aaaside-chain
side-chain (R-group)
(R-group)
(R-group)
More (R-group)
than 132 andand
and
types and -glucopyranose
-glucopyranose
D
-glucopyranose
DD-glucopyranose
D of GSLs have been
The structural diversity of glucosinolates is mainly due to the different substituents possible at the
asasaas
aas a-thioglucoside
a-thioglucoside
asuncovered -thioglucoside
-thioglucoside
a-thioglucoside attached
attached
attached
attached
attached
so far [10,11]. to
Table
to to
totocarboncarbon
carbon carbon
carbon
1 lists atomatom
atom
atom
all atom no.
no.
known
no.
no.0no.
00in0in
inGSL
in (Z)-N-hydroximine
0(Z)-N-hydroximine
in (Z)-N-hydroximine
(Z)-N-hydroximine
(Z)-N-hydroximine
compounds
sulphate
sulphate
sulphate
sulphate
sulphate
identified
esters
estersesters
esters
esters
to-date [8][8]
[8]
[8]
in [8] (Figure
(Figure
(Figure
(Figure
(Figure
Eruca species
side-chain
1). 1). The
1).
The The position
structural
structural
structural R, which
diversity
diversity
diversity ofofofcan be very variable
ofglucosinolates
glucosinolates
glucosinolates ismainly[9].due
ismainly
mainly More
due
due than
totothe
the 132 types
different
different of GSLs have
substituents
substituents beenatuncovered
possible
possible atthe
atthethe
1).
1). TheThe structural
structural
[7,12,13]. diversity
diversity of glucosinolates
glucosinolates isisismainly
mainly due due tototothe
thethe different
different
different substituents
substituents
substituents possible
possible
possible atat the
the
so far
side-chain
side-chain
side-chain
[10,11].
side-chain
side-chain positionTable
position
position
position
position R,
1 lists
R,R,
R,R,which
which
which all
which
which known
can
can
can
cancan be
be
be GSL
be
very
bevery very
very verycompounds
variable
variable
variable
variable
variable [9].
[9].
identified
[9].
[9].
More
[9].More More
MoreMorethan
thanthan to-date
than
than
132
132
132
132132 in
types
types
types
Eruca
types
types
of
of
of
ofofGSLs
GSLsspecies
GSLs
GSLs GSLs
have
have
have
[7,12,13].
have
havebeen been
been
been been
uncovered
uncovered
uncovered so
sosofarso
far far
far[10,11].
[10,11].
[10,11]. Table
Table Table 1lists
1listsallall
lists allknown
known GSL
GSL compounds
compounds identified
identified to-date
to-date ininEruca
Eruca species
species
uncovered
uncovered so far [10,11].
[10,11]. Table
Table 111lists
lists all
all known
known
known GSL
GSL
GSL compounds
compounds
compounds identified
identified
identified to-date
to-date
to-date ininEruca
in Eruca
Eruca species
species
species
[7,12,13].
[7,12,13].
[7,12,13].
[7,12,13].
[7,12,13].
Figure 1. General
Generalstructure of of
structure glucosinolates. R denotes
glucosinolates. the variable
R denotes side chain
the variable derived
side chain from amino
derived from
acids. acids.
amino
Figure
Figure
Figure 1.1.General
Generalstructure
structure ofofglucosinolates.
glucosinolates.RRdenotes
denotes the variable
the variableside chain
side derived
chain derived from
fromamino
amino
Figure
Figure 1.1.1.
General
General
General structure
structure
structure ofof
of glucosinolates.
glucosinolates.
glucosinolates. RRR denotes
denotes
denotes the
the
the variable
variable
variable side
side
side chain
chain
chain derived
derived
derived from
from
from amino
amino
amino
The GLs
acids.
The are naturally hydrolyzed by the myrosinase enzyme or enteric
GLs are naturally hydrolyzed by the myrosinase enzyme or enteric microflora giving
acids.
acids.
microflora giving rise
rise toto
a
acids.
acids.
a numberofofproducts
number productsamong amongwhich whichisothiocyanates
isothiocyanatesare areknownknownfor fortheir
theirantibacterial
antibacterial and and antifungal
antifungal
properties
The The
The The
GLsGLs
GLs GLs
are are
applied
are are naturally
naturally
naturally hydrolyzed
to biofumigation
biofumigation
naturally hydrolyzed
hydrolyzed
hydrolyzed byby by
the
thethe
[12,14,15],
by the myrosinase
and also
myrosinase
myrosinase
myrosinase also
enzymeenzyme
enzymeenzyme oror or
to efficiently
efficiently
or enteric
entericenteric
enteric microflora
reduce risks
microflora
microflora
microflora giving
givingof
giving
giving rise
rise
rise totototo
degenerative
rise
The
properties GLs are
appliednaturally
to hydrolyzed by the myrosinase
[12,14,15], and enzyme to or enteric microflora
reduce risksgiving of rise to
degenerative
a number
diseases
a number of
such products
ofasas cancer
products among
[16].
among which
It which
has isothiocyanates
been speculated
isothiocyanates are
that
are known
thefor
known for their
isothiocyanates
for their antibacterial
(ITCs)
antibacterial and
like
and antifungal
sulforaphane
antifungal
aadiseases
anumber
number
number of
of ofproducts
products
products
such among
among
among
cancer [16]. which
which
which
It has isothiocyanates
isothiocyanates
isothiocyanates
been speculated are
areare known
known
known
that the forfor their
their
their
isothiocyanates antibacterial
antibacterial
antibacterial (ITCs) and
and andlikeantifungal
antifungal
antifungal
sulforaphane
properties
properties
properties
propertiesapplied applied
(4-methylsulfinylbutyl),applied
applied to toto biofumigation
biofumigation
obtained
biofumigation
appliedtotobiofumigation
biofumigation from [12,14,15],
[12,14,15],
[12,14,15], hydrolysis
[12,14,15], and andand
andalsoalsoalso
of
also also toto efficiently
efficiently
glucoraphanin
to efficiently
toefficiently
efficientlyreduce reduce
reduce reduce risks
risks of of degenerative
(4-(methylsulfinyl)butyl-GSL),
reduce risks of degenerative
degenerative
risksofofdegenerative
degenerative
properties
(4-methylsulfinylbutyl), obtained [12,14,15],
from and
hydrolysis ofto glucoraphanin risks
diseases
are diseases
in
diseases
diseases great
suchsuch
suchsuch
part
asasasas cancer
cancer cancer
responsible
cancer [16].[16].
[16].
[16]. ItItIthas
hasIthas
for hasthe
been been
been been speculated
speculated
protective
speculated
speculated that
that
effects
that
that the
thethe
ofthe isothiocyanates
isothiocyanates
cruciferous
isothiocyanates
isothiocyanates (ITCs)
(ITCs)
vegetables
(ITCs)
(ITCs) like like
like like
by sulforaphane
sulforaphane
the
sulforaphane induction
sulforaphane of
diseases
are such as cancer
in great part responsible [16]. It has been
for the from speculated
protective that the isothiocyanates
effectsofofglucoraphanin
cruciferous vegetables (ITCs) like sulforaphane
by the induction of Phase
(4-methylsulfinylbutyl),
(4-methylsulfinylbutyl), obtained
obtained
obtained
obtained from
from from hydrolysis
hydrolysis
hydrolysis
hydrolysis of glucoraphanin
ofofglucoraphanin
glucoraphanin
glucoraphanin (4-(methylsulfinyl)butyl-GSL),
Phase
IIare
enzymes. II enzymes.
(4-methylsulfinylbutyl),These These
obtained
enzymes have enzymes
from have
hydrolysis
a chemoprotective a chemoprotective
of effect, since they effect, since they
block chemical block chemical
carcinogenesis,
are are in in
iningreat
arein great great
great
great partpart
partpart responsible
responsible
responsible
responsible for
forfor
for
the
thethe
the protective
protective
protective
protective effects
effectseffects
effects of of
ofcancers cruciferous
cruciferous
cruciferous
ofcruciferous
cruciferous vegetables
vegetables
vegetables
vegetables byby byby
the the
the the induction
induction
induction
induction ofofofof
are
carcinogenesis, part responsible
thus helping for tothe protective
prevent the effects
onset of
of [1719]. vegetables by the induction of
thus
Phase helpingII to prevent
enzymes. Thesethe onset
enzymes of cancers
have a[1719].
chemoprotective effect, since they block chemical
Phase
Phase Phase
Phase The II enzymes.
IIIIIIenzymes.
enzymes.
enzymes.
glucoraphanin, These
These
These These enzymes enzymes
enzymes
enzymes
glucosativin have
havehave have a chemoprotective
aaachemoprotective
chemoprotective
chemoprotective
(4-mercaptobutyl-GSL) effect,effect,
effect,
effect,
and since
since
since since
glucoerucin they
they
they theyblock
block
block block chemical
chemical
chemical
chemical
(4-(methylthio)butyl-
The
carcinogenesis,glucoraphanin,
carcinogenesis, thus
thus helping glucosativin
helping toto prevent
prevent
(4-mercaptobutyl-GSL)
thethe onset
onset ofofcancers
cancers
and glucoerucin
[1719].
[1719].
carcinogenesis,
carcinogenesis,
carcinogenesis,
GSL) have been been thus
thus
thus helping
helping
helping
identified to
to to
as prevent
prevent
prevent
the most the
thethe onset
onset
onset
abundant ofof
of cancers
cancers
cancers
GLs [1719].
[1719].
in[1719].
leaf rocket [4]. [4]. Glucosativin
Glucosativin and and glucoerucin
glucoerucin
GSL) have
TheThe identified
glucoraphanin,
glucoraphanin, as the
glucosativin
glucosativin most abundant
(4-mercaptobutyl-GSL) GLs
(4-mercaptobutyl-GSL) in leaf rocket
and
and glucoerucin
glucoerucin (4-(methylthio)butyl-
The
TheThe glucoraphanin,
glucoraphanin,
glucoraphanin,
breakdown products glucosativin
glucosativin
glucosativin
are thought (4-mercaptobutyl-GSL)
totocontribute most to to and
andand glucoerucin
glucoerucin
glucoerucin
pungency andand (4-(methylthio)butyl-
flavour in rocket [13]. [13].
The
breakdown
GSL)
GSL) GSL)
have have
have
been products
been
been identified are
identified
identified asas thought
asas
the the
the
most most
most contribute
abundant
abundant
abundant GLs GLs
GLs most
ininin in
leafleaf
leaf
rocket pungency
rocket
rocket[4]. [4].
[4]. Glucosativin
Glucosativin
Glucosativin flavour
and andand inglucoerucin
rocket
glucoerucin
glucoerucin
GSL)
GSL) have
have
glucoerucin been
been identified
identified
(GER) could as thethe
undergo most
most abundant
abundant
a high GLsGLs
yielding in leafleaf rocket
rocket
transformation [4].
[4]. Glucosativin
Glucosativin
into GRA and andby glucoerucin
glucoerucin
chemoselective
The glucoerucin
breakdown
breakdown
breakdown
breakdown
breakdown products
products
products
(GER)
products
products are
arearearecould
are thought
thought
thought
thought
undergo
thought to to
totocontribute
to
a highmost
contribute
contribute
contribute
contribute yielding
most most
most most to totransformation
pungency
totopungency
to pungencypungency
pungency and
andandand
and into
flavour
flavour
flavour
flavour
GRA
flavour inin
ininrocket
in
by
rocket
rocket
rocket chemoselective
rocket [13].
[13]. [13].
[13].
[13]. The
The
The
The The
oxidation
glucoerucin
oxidation using
using (GER) hydrogen
could
hydrogen peroxide
undergo
peroxide a [20]
high
[20] and
and boosted
yielding
boosted our
transformation
our interest
interest for
into
for GER
GRA
GER and
andby its ITC-derivative,
chemoselective
its ITC-derivative,
glucoerucin
glucoerucin
glucoerucin
glucoerucin (GER) (GER)
(GER)
(GER) could
couldcould undergo
undergo
undergo
could undergo ITC). a a
a high a
high
high high yielding
yielding
yielding
yielding transformation
transformation
transformation
transformation into
intointo into
GRA
GRAGRA GRA by
by by by chemoselective
chemoselective
chemoselective
chemoselective
erucin
oxidation (ERN, 4-methylthiobutyl
using hydrogen peroxide Previous
[20] and investigation
boosted our has shown
interest shown
for GER that
and ERN induces apoptosis
apoptosis
erucin
oxidation
oxidation
oxidation
oxidation (ERN,
using
using
using 4-methylthiobutyl
using hydrogen
hydrogen
hydrogen
hydrogen peroxide
peroxide
peroxide
peroxide ITC).
[20]
[20][20] Previous
[20]
and
andand and
boosted
boosted
boosted investigation
boosted our
our our our
interest
interest
interest has
interest for
forfor for
GER
GERGER GER that
and
and
and and itsits
ERN
its
its its ITC-derivative,
induces
ITC-derivative,
ITC-derivative,
ITC-derivative,
ITC-derivative,
selectively
erucin
selectively (ERN, in
in human
human leukemia
4-methylthiobutyl
leukemia cells,
ITC).
cells, but
Previous
but not
not in
in nontransformed
investigation
nontransformed has shown T
T lymphocytes,
that ERN
lymphocytes, induces thereby being aa
apoptosis
thereby being
erucin
erucin
erucin
erucin (ERN,
(ERN,
(ERN, (ERN, 4-methylthiobutyl
4-methylthiobutyl
4-methylthiobutyl
4-methylthiobutyl ITC).
ITC).
ITC). ITC). Previous
Previous
Previous
Previous investigation
investigation
investigation
investigation has
has has has
shown
shown
shown shown that
that
that that
ERN
ERNERN ERNinducesinduces
induces
induces apoptosis
apoptosis
apoptosis
apoptosis
promising
selectively
promising new
in agent
human in cancer therapy
leukemia therapy
cells, but [21].
not in nontransformed Tlymphocytes,
Tlymphocytes, thereby being a
selectively
selectively
selectively
selectively innew
in in
human
inhuman agent
human
human in
leukemiacancer
leukemia
leukemia
leukemia cells,
cells,
cells, cells,
but
but but [21].
but
not
not not not in nontransformed
ininnontransformed
in nontransformed
nontransformed lymphocytes,
TTTlymphocytes,
lymphocytes, thereby thereby
thereby
thereby being being
being
being aaa a
promising
promising
promising
promising new
newnew
new agent
agent
agentagent
ininin incancer
cancercancer
cancer therapy
therapytherapy
therapy [21].[21].
[21]. [21].
promising new agent in cancer therapy [21].
Table
Table 1. 1. Identified
Identified glucosinolates
glucosinolates in in Eruca
Eruca species
species differing
differing only
only in in the
the side
side chain
chain R.R.
Table
Table
Table 1. Identified
1. Identifiedglucosinolates
glucosinolates in Eruca
in Erucaspecies
species differing
differingonly
only in the
inside
theside
sidechain
chainR.
R.R. R.
Table
RTable 1.1.1.
(Variable Identified
Identified
Identified
Side glucosinolates
glucosinolates
glucosinolates
Chain) inin
in Eruca
Eruca
Eruca species
species
species differing
differing
differing only
only
only inin
in the
thetheside
side chain
chain
chain R.
R Group
(Variable Side Chain) R-Group Trivial Name
Structure
RR(Variable
(Variable Side Chain)
Side Chain) R-Group Trivial Name
RRR(Variable
(Variable
(Variable
Group Side
Side
Side
Structure Chain)
Chain)
Chain) R-Group
R-Group Trivial
Trivial Name
Name
Group Structure R-Group
R-Group
R-Group
2-(Benzoyloxy) ethyl Trivial
Trivial
Trivial - Name
Name
Name
GroupGroup
Group
Group Structure
Structure
Structure
Structure
2-(Benzoyloxy)
2-(Benzoyloxy) ethyl
ethyl -
2-(Benzoyloxy)
2-(Benzoyloxy)
2-(Benzoyloxy) ethyl
ethyl
ethyl -- - -
3-Hydroxy-5-(methyl-
2-(Benzoyloxy) ethyl - -
sulfinyl)pentyl
3-Hydroxy-5-(methyl- - -
3-Hydroxy-5-(methyl- -- -
sulfinyl)pentyl
sulfinyl)pentyl
sulfinyl)pentyl
sulfinyl)pentyl
sulfinyl)pentyl
4-(-D-glucopyranosyldisulfanyl) Diglucothiobeinin
3-Hydroxy-5-(methyl-sulfinyl)pentyl -
4-(-
4-(-
butyl
4-(- DD-glucopyranosyldisulfanyl)
-glucopyranosyldisulfanyl)
-glucopyranosyldisulfanyl) Diglucothiobeinin
Diglucothiobeinin
Diglucothiobeinin
4-(-
4-(- D -glucopyranosyldisulfanyl)
D-glucopyranosyldisulfanyl)
D Diglucothiobeinin
Diglucothiobeinin
butyl
butylbutyl
butyl
butyl
4-(-D-glucopyranosyldisulfanyl)
5-(Methylsulfinyl)pentyl butyl Glucoalyssin
Diglucothiobeinin
5-(Methylsulfinyl)pentyl
5-(Methylsulfinyl)pentyl Glucoalyssin
Glucoalyssin
Glucoalyssin
Glucoalyssin
4-Phenylbutyl Glucoalyssin
Glucoamoracin
4-Phenylbutyl
4-Phenylbutyl
4-Phenylbutyl Glucoamoracin
Glucoamoracin
Glucoamoracin
4-Phenylbutyl
Glucoamoracin
4-Phenylbutyl
7-(Methylsulfinyl)heptyl Glucoamoracin
Glucoibarin
7-(Methylsulfinyl)heptyl
7-(Methylsulfinyl)heptyl Glucoibarin
Glucoibarin
Glucoibarin
Glucoibarin
Ethyl
Ethyl
Ethyl
Glucolepiidin
Glucolepiidin
Glucolepiidin
Ethyl
Ethyl
Ethyl Glucolepiidin
Glucolepiidin
Glucolepiidin
sulfinyl)pentyl
sulfinyl)pentyl
4-(-D4-(-
-glucopyranosyldisulfanyl) Diglucothiobeinin
D-glucopyranosyldisulfanyl) Diglucothiobeinin
butyl butyl
Molecules 2017, 22, 851 5-(Methylsulfinyl)pentyl Glucoalyssin 3 of 15

4-Phenylbutyl
Glucoamoracin
Table 1. Cont.
Molecules
Molecules2017,
Molecules 2017,22,
2017, 22,851
22, 851
851 33of of14
14
Molecules
Molecules
Molecules 2017,
2017,
Molecules 2017,
Molecules 22,
22,
2017, 22,
2017, 851
851
22, 851
22, 851
851 7-(Methylsulfinyl)heptyl
Glucoibarin 333333 of
of
of
of 14
14
14
of 14
of 14
14
Table
Table1.
Table 1.Cont.
1. Cont.
Cont.
EthylTable
Ethyl1.
Table
Table 1. Cont.
1. Cont.
Cont. Glucolepiidin
Glucolepiidin
Ethyl Glucolepiidin
2-Phenylethyl
2-Phenylethyl
2-Phenylethyl
2-Phenylethyl
2-Phenylethyl Gluconasturtiin
Gluconasturtiin
Gluconasturtiin
Gluconasturtiin
Gluconasturtiin
2-Phenylethyl
2-Phenylethyl Gluconasturtiin
Gluconasturtiin
4-(Methylsulfinyl)-3-butenyl
4-(Methylsulfinyl)-3-butenyl Glucoraphenin
Glucoraphenin
Glucoraphenin
Glucoraphenin
Glucoraphenin
4-(Methylsulfinyl)-3-butenyl Glucoraphenin
Glucoraphenin
Dimeric
Dimeric
Dimeric 4-mercaptobutyl
4-mercaptobutyl
4-mercaptobutyl
Dimeric 4-mercaptobutyl DMB
DMB
DMB DMB
Dimeric
Dimeric
Dimeric 4-mercaptobutyl
4-mercaptobutyl
4-mercaptobutyl DMB
DMB
DMB
4-Mercaptobutyl
4-Mercaptobutyl
4-Mercaptobutyl
4-Mercaptobutyl
4-Mercaptobutyl Glucosativin
Glucosativin
Glucosativin
Glucosativin
Glucosativin
4-Mercaptobutyl
4-Mercaptobutyl
4-Mercaptobutyl Glucosativin
Glucosativin
Glucosativin
4-Hydroxy-3-indolymethyl
4-Hydroxy-3-indolymethyl 4-Hydroxyglucobrassicin
4-Hydroxyglucobrassicin
4-(Methylthio)butyl
4-(Methylthio)butyl
4-(Methylthio)butyl Glucoerucin
Glucoerucin
Glucoerucin
4-(Methylthio)butyl
4-(Methylthio)butyl
4-(Methylthio)butyl
4-(Methylthio)butyl Glucoerucin
Glucoerucin
Glucoerucin
Glucoerucin
4-Hydroxybenzyl
4-Hydroxybenzyl
4-Hydroxybenzyl
4-Hydroxybenzyl Glucosinalbin
Glucosinalbin
Glucosinalbin
Glucosinalbin
4-Hydroxybenzyl
4-Hydroxybenzyl
4-Hydroxybenzyl Glucosinalbin Glucosinalbin
Glucosinalbin
(R,S)-2-Hydroxy-3-butenyl
(R,S)-2-Hydroxy-3-butenyl Progoitrin
Progoitrin
Progoitrin
Progoitrin
(R,S)-2-Hydroxy-3-butenyl Progoitrin Progoitrin
Progoitrin
3-Indolymethyl
3-Indolymethyl
3-Indolymethyl Glucobrassicin
Glucobrassicin
Glucobrassicin
3-Indolymethyl
3-Indolymethyl
3-Indolymethyl
3-Indolymethyl Glucobrassicin
Glucobrassicin
Glucobrassicin
Glucobrassicin
1-Methylpropyl
1-Methylpropyl
1-Methylpropyl
1-Methylpropyl Glucocochlearin
Glucocochlearin
Glucocochlearin
Glucocochlearin
1-Methylpropyl
1-Methylpropyl
1-Methylpropyl Glucocochlearin
Glucocochlearin
Glucocochlearin
2-Methylbutyl
2-Methylbutyl
2-Methylbutyl
2-Methylbutyl Glucojiabutin
Glucojiabutin
Glucojiabutin
Glucojiabutin
2-Methylbutyl
2-Methylbutyl
2-Methylbutyl Glucojiabutin
Glucojiabutin
Glucojiabutin
3-(Methylthio)propyl
3-(Methylthio)propyl Glucoiberverin
Glucoiberverin
Glucoiberverin
Glucoiberverin
3-(Methylthio)propyl Glucoiberverin
Glucoiberverin
Glucoiberverin
Glucoiberverin
3-Butenyl
3-Butenyl
3-Butenyl Gluconapin
Gluconapin
Gluconapin
3-Butenyl
3-Butenyl
3-Butenyl
3-Butenyl Gluconapin
Gluconapin
GluconapinGluconapin
Benzyl
Benzyl
Benzyl
Benzyl
Benzyl Glucotropaeolin
Glucotropaeolin
Glucotropaeolin
Glucotropaeolin
Glucotropaeolin
Benzyl
Benzyl Glucotropaeolin
Glucotropaeolin
1-Methoxyindol-3-ylmethyl
1-Methoxyindol-3-ylmethyl Neoglucobrassicin
Neoglucobrassicin
Neoglucobrassicin
Neoglucobrassicin
1-Methoxyindol-3-ylmethyl Neoglucobrassicin
Neoglucobrassicin
Neoglucobrassicin
4-(Methylsulfinyl)
4-(Methylsulfinyl)butyl
butylbutyl
4-(Methylsulfinyl)
4-(Methylsulfinyl)
4-(Methylsulfinyl) butyl Glucoraphanin
Glucoraphanin
Glucoraphanin
Glucoraphanin
4-(Methylsulfinyl) butyl
4-(Methylsulfinyl) butyl
butyl
Glucoraphanin
Glucoraphanin
Glucoraphanin
4-Methoxyindol-3-ylmethyl 4-Methoxyglucobrassicin
4-Methoxyglucobrassicin
ItIt is
It is possible,
is possible, although
possible, although speculative,
although speculative, that
speculative, that the
that the ability
the ability of
ability of ERN
of ERN to
ERN toto induce
induce apoptosis
induce apoptosis selectively
apoptosis selectively in
selectively in
in
It
It is
It It possible,
is
is possible,
is possible,
possible, although
although
although
although speculative,
speculative,
speculative,
speculative, that
that the
that the
that
the ability
ability of
of ERN
the ability
ability of ERNof
ERN to
to induce
toERN
induce
induce to apoptosis
apoptosis
apoptosis selectively in
selectivelyselectively
induce apoptosis
selectively in
in
cancer
cancer
cancer
cancer cells,
cells,
cells,
cells, unlike
unlike
unlike
unlike SFN,
SFN,
SFN,
SFN, may
may
may
may be
be
be
be at
at
at
at least
least
least
least partially
partially
partially
partially related
related
related
related to
to
to
to its
its
its
its unusual
unusual
unusual
unusual capability
capability
capability
capability toto
to
to act
act
act
act as
as
as
as aaa
cancer
cancer
in cancer
cancer cells,
cells, unlike
cells, cells,
unlikeunlike
unlike SFN,
SFN, may
SFN, may be
SFN,
may be
may
be at
at least
at least
be
least at partially
partially related
least partially
partially related to
to its
relatedrelated
to its unusual
its unusual
to its unusual
unusual capability
capability
capability to
to act
capability
to act to aaaaact as a
act as
as
as
hydroperoxide-scavenging
hydroperoxide-scavenging preventive
preventive
preventive antioxidant
antioxidant
antioxidant [22].
[22].
[22].
hydroperoxide-scavenging preventive
preventive
preventive antioxidant
antioxidant [22].
antioxidant [22].
[22].
Further
Further investigation
Further
Further investigation is
investigation
investigation is preventive
is
is in
in
in progressantioxidant
in progress
progress
progress to
to
to disclose[22].
to disclose
disclose
disclose possible
possible relationships
possible
possible relationships between
relationships
relationships between the
between
between the fine
the
the fine
fine
Further
Further
Further investigation
investigation
investigation is in
is inis progress
progress
in progress to disclose
to disclose
to possible
possible
disclose relationships
relationships
possible relationshipsbetween
between the fine
the
between fine
fine the fine
antioxidant/prooxidant
antioxidant/prooxidantbalance
antioxidant/prooxidant balance of
balance
balance ofERN
of
of ERNand
ERN and its
and itsselective
its selective pro-apoptotic
selective pro-apoptotic behaviour.
pro-apoptotic behaviour. Other
behaviour. Other GSLs
Other GSLshave
GSLs have
have
antioxidant/prooxidant balance of ERN
balance ERNof and
and its
ERNits selective
selective
and
pro-apoptotic
pro-apoptotic
its selective
behaviour.
behaviour. Other
pro-apoptotic Other GSLs
GSLs have
behaviour. have Other
also
also been
also been identified
been identified within
identified within rocket
within rocket tissue,
rocket tissue, for
tissue, for example
for example diglucothiobeinin
example diglucothiobeinin [23],
diglucothiobeinin [23], 4-
[23], 4-
4-
also
also
also been
been
been identified
identified within
within rocket
rocket tissue,
tissue, for
for example
example diglucothiobeinin
diglucothiobeinin [23],
[23], 4-
4-
GSLs haveidentified
hydroxyglucobrassicin)
hydroxyglucobrassicin) also been [24]
within
[24]
[24] identified
and
and
and
rocket tissue,
within
4-methoxyglucobrassicin
4-methoxyglucobrassicin
4-methoxyglucobrassicin rocketfor example
tissue,
[25].
[25].
[25]. for diglucothiobeinin
example diglucothiobeinin [23], 4- [23],
hydroxyglucobrassicin) [24]
hydroxyglucobrassicin) [24]
[24] and
[24] and 4-methoxyglucobrassicin
and 4-methoxyglucobrassicin
and 4-methoxyglucobrassicin [25].
4-methoxyglucobrassicin [25].
[25].
[25].
4-hydroxyglucobrassicin)
InIn relation
In relation to
relation to the
to the total
the total [24] andacid
total phenolic
phenolic
phenolic 4-methoxyglucobrassicin
acid
acid content,
content, rocket
content, rocket tissues[25].
rocket tissues
tissues also
also contain
also contain significant
contain significant levels
significant levels of
levels of
of
In
In relation
In relation to
relation to the
to the total
the total phenolic
total phenolic acid
phenolic acid content,
acid content, rocket
content, rocket tissues
rocket tissues also
tissues also contain
also contain significant
contain significant levels
significant levels of
levels of
of
In
polyglycosylated
polyglycosylated
polyglycosylated
polyglycosylated relation to the
flavonoids,
flavonoids,
flavonoids, total
which
which
which phenolic
are
are
are acid
antioxidants
antioxidants
antioxidants content,
poorly
poorly
poorly rocket
absorbedtissues
absorbed
absorbed in
in
in thealso
the
the contain
proximal
proximal
proximal significant
gastrointestinal
gastrointestinal
gastrointestinal levels of
polyglycosylated flavonoids,
polyglycosylated flavonoids, which
flavonoids, whichare
which areantioxidants
are antioxidantspoorly
antioxidants poorly absorbed
poorly absorbed in
absorbed in the
in the proximal
the proximal gastrointestinal
proximal gastrointestinal
gastrointestinal
polyglycosylated
tract
tract and
tract
tract and are
and are therefore
are therefore flavonoids,
therefore likely likely to
likely to which
to reach
reach the
reach are
the antioxidants
the colon
colon in
colon poorly
in substantial
in absorbed
substantial quantities,
substantial in the
quantities, protecting
quantities, proximal
protecting the
protecting gastrointestinal
the colonic
the colonic
colonic
tract andand are are therefore
therefore likely likely to to reach
reach the the colon
colon in in substantial
substantial quantities,
quantities, protecting
protecting the the colonic
colonic
epithelium
epithelium
tract
epithelium and from
from
are
from free
free radical
therefore
free radical
radical attack
attack
likely
attack to [26].
[26].
reach
[26]. Previous
Previous
the
Previous works
colonworks
works in showed
showed variability
substantial
showed variability for
quantities,
variability for the
for the
the total
total phenolic
protecting
total phenolic
the colonic
phenolic
epithelium
epithelium from
epithelium from free
from free radical
free radical attack
radical attack [26].
attack [26]. Previous
[26]. Previous works
Previous works showed
works showed variability
showed variability for
variability for the
for the total
the total phenolic
total phenolic
phenolic
compounds
compounds
compounds
compounds (4474.5
(4474.5
(4474.5
(4474.5 toto
to
to 32700
32700
32700
32700 g
g
g
g g g
g1
11dw)
1
gg1
1
1 dw)
dw)
dw) and
and
and
and demonstrated
demonstrated
demonstrated
demonstrated that
that
that
that rocket
rocket
rocket
rocket leaves
leaves
leaves
leaves are
are
are
are an
an
an
an excellent
excellent
excellent
excellent source
source
source
source
compounds
compounds (4474.5
compounds (4474.5 to
(4474.5 to 32700
to 32700 g
32700 g gg
g 1 dw)
1
1 dw) and
dw) and demonstrated
and demonstrated that
demonstrated that rocket
that rocket leaves
rocket leaves are
leaves are an
are an excellent
an excellent source
excellent source
source
of
of these
of these compounds
these compounds [4],
compounds [4], being
[4], being even
being even richer
even richer than
richer than the
than the commercial
the commercial broccoli
commercial broccoli florets
broccoli florets [27].
florets [27]. Taking
[27]. Taking into
Taking into
into
of
of these compounds [4], being even richer than the commercial broccoli florets [27]. Taking into
of these
these compounds
compounds [4],
[4], being
being even
even richer
richer than
than the
the commercial
commercial broccoli
broccoli florets
florets [27].
[27]. Taking
Taking into
into
Molecules 2017, 22, 851 4 of 15
epithelium from free radical attack [26]. Previous works showed variability for the total phenolic
compounds (4474.5 to 32,700 gg1 dw) and demonstrated that rocket leaves are an excellent source
of these compounds [4], being even richer than the commercial broccoli florets [27]. Taking into account
the aforementioned it is therefore important to characterise the content of bioactive compounds in
new or reintroduced cultivars. However, the quantification of the glucosinolate and/or phenolic
acid content by the standard methods is expensive and time-consuming, and in addition, specialised
personnel are needed.
Numerous techniques have been employed for quantification of these compounds, such as
thin-layer chromatography, palladium test, UV spectroscopy, reverse phase by high performance
liquid chromatography (LC), thermospray LC with tandem MS in the two most common interfaces
(ESI or APCI), capillary gas chromatography (GC-MS and GC-MS-MS), high speed counter current
chromatography or supercritical fluid chromatography [2830]. The high cost and labour input
required for obtaining the glucosinolate content by LC or total phenolic acid content by UVvisible
spectrometry are serious handicaps to analyse large sets of samples, which is usually necessary to
identify the target genotypes in screening programmes. In contrast, the use of fast analytical techniques
such as near-infrared spectroscopy (NIRS) results in many advantages, since analysis can be carried
out with a considerable saving of time, at a low cost and without using hazardous chemicals. NIRS
has been widely used for decades for qualitative and quantitative analysis in agriculture and food
research, and many authors have used this technique for determining the glucosinolate content in
seeds, [29,3133], leaves [12,34,35] and flowers [36] from a wide range of cruciferous species. Therefore,
it may be expected that this technique coupled with chemometric tools could provide an alternative
method to undertake the analysis of glucosinolates in rocket leaves.
This study is part of an ongoing breeding programme focused on obtaining varieties of rocket
with enhanced health properties for human nutrition [30]. The main objective was to test the potential
of NIRS for predicting the total and individual major glucosinolate content, as well as total phenolic
acid content found in the rocket leaves. In addition, we provide some knowledge about the mechanism
used by NIRS for determining these compounds successfully in the leaves of this species.
2. Results and Discussion
2.1. Reference Values

Thirteen glucosinolates belonging to three chemical classes were detected in rocket leaves:
seven aliphatic compounds (glucoerucin, gluraphanin, gluconapin, glucoiberverin, progoitrin,
gluconapoleiferin, glucobrassicanapin and mercaptobutyl), one aromatic (gluconasturtiin) and
four indole compounds (4-hydroxyglucobrassicin, 4-methoxyglucobrassicin, glucobrassicin and
neoglucobrassicin). The GL profile found in these accessions was similar to the profiles reported
by other authors in rocket leaves [4,25,3739].
The ranges, means and standard deviations of the total and individual glucosinolates used in this
study are summarised in Table 2. Individual plants exhibited TGSL concentrations that ranged from
4.86 to 44.65 molg1 dw, and a mean value of 15.6 molg1 dw. These concentrations are similar
to those contents previously found in rocket with values varying from 11 to 28.24 molg1 dw but
lower than those found in rocket sprouts which reached up to 55.4 molg1 dw [39].
The GRA aliphatic glucosinolate varied from 0.13 to 27.53 molg1 dw, with a mean content
of 9.11 molg1 dw and representing 38.66% of the TGSL. These results agree with those reported
by Kim and Ishii [25] and Villatoro-Pulido [4] which found values from 1.25 to 6.1 molg1 dw, but
higher values have also been reported by Bennet et al [37,38] with 32 molg1 dw.
The GRA aliphatic glucosinolate varied from 0.13 to 27.53 molg1 dw, with a mean content
of 9.11 molg1 dw and representing 38.66% of the TGSL. These results agree with those reported
by Kim and Ishii [25] and Villatoro-Pulido [4] which found values from 1.25 to 6.1 molg1 dw, but
higher values have also been reported by Bennet et al [37,38] with 32 molg1 dw.
Molecules 2017, 22, 851 5 of 15
Table 2. Calibration and cross-validation statistics of glucosinolates (molg1 dw) and total phenolic
content (mgGAEg1 dw) for rocket leaves measured by FNS-6500. GRA: Glucoraphanin; GLSAT:
Glucosativin; GER: Glucoerucin; TPC: Total phenolic content
Parameter Range Mean SD 1 R 2C 2 SEC 3 R 2 CV 4 SECV 5 RPDcv 6

Total GLs 4.8644.65 23.56 7.32 0.79 3.34 0.70 4.02 1.83
GRA 0.1327.85 9.11 6.41 0.94 1.52 0.82 2.72 2.37
GLSAT 0.2518.61 9.66 3.92 0.86 1.44 0.64 2.42 1.62
GER 0.148.01 1.45 1.6 0.97 0.26 0.93 0.41 3.99
TPC 2.3012.30 8.73 1.99 0.88 0.69 0.85 0.79 2.55
1 SD: standard deviation; 2 R2 C: coefficient of determination in calibration. 3 SEC: standard error in calibration.
4R2 CV: coefficient of determination in cross-validation. 5 SECV: standard error of cross-validation. 6 RPDcv: ratio
of the standard deviation to standard error of cross-validation.
Glucosativin was the indole glucosinolate that showed the highest mean content of all,
representing 41% of the TGSL, ranging from 0.25 to 18.61 molg1 dw followed by glucoerucin
and glucoraphanin.
The GER aliphatic glucosinolate varied from 0.14 to 8.01 molg1 dw. These concentrations are
similar to values previously reported in rocket which ranged from 0.14 to 4.03 molg1 dw [4,25].
Total phenol content ranged from 2.30 to 12.3 mgGAEg1 dw. The values found in this
study were lower than those previously reported in rocket leaves where the concentrations varied
from 4.47 to 32.7 mgGAEg1 dw, and in seeds and sprouts from commercial broccoli cultivars with
values reaching 27.97 mgGAE g1 dw [39].
Table 3 shows the correlations among TGLs, individual GLs and TPC and their associated p-values.
The TGLs were correlated significantly with GRA and GLSAT (R2 = 0.83 and R2 = 0.70 at p < 0.01,
respectively), which were the most abundant glucosinolates found in Eruca accessions (Tables 2 and 3).
Table 3. Pearson correlation coefficient (R2 ) among TGLs, individual GL and TPC in Eruca accessions.
Parameter TGLs GRA GLSAT GER TPC

TGLs 1 0.83 ** 0.70 ** 0.37 ** 0.04 n.s.
GRA 1 0.48 ** 0.53 ** 0.09 n.s.
GLSAT 1 0.14 n.s. 0.09 n.s.
GER 1 0.33 *
TPC 1
TGLs: Total glucosinolates; GRA: glucoraphanin; GLSAT: glucosativin; GER: glucoerucin; TPC: Total phenolic
content. n.s.: not significant; *: p < 0.05; **: p < 0.01.
A moderate negative correlation was found between GER and GRA (R2 = 0.53, p < 0.01),
suggesting that the biosynthesis of both molecules is interrelated. In E. sativa, GRA is derived by a
side-chain modification of GER [38]. The structure of both compounds is identical, except for the
presence of an oxygen atom on the side-chain sulfur (Table 1). GRA also had a moderate positive
correlation with GLSAT (R2 = 0.476, p < 0.01). A weak positive correlation was found between GER
and TPC (R2 = 0.33, p < 0.05).
The presence of a negative correlation between GRA and GER, and a positive correlation between
GRA and GLSAT is also of potential importance. These results suggest that indirect selection for high
GRA and GLSAT is possible, but in detriment of GER.
In disagreement with our work, a recent study performed on accessions of Eruca sativa [40]
indicated that no significant correlation was found between GRA and GER. Furthermore, no significant
correlations were observed for these GLs with any sensory attribute.
These results underline the importance of knowing the relationships between phytochemical
compounds in order to be taken into account in further breeding programs to find the adequate profile
according to the different market preferences.
Molecules 2017, 22, 851 6 of 14
These results underline the importance of knowing the relationships between phytochemical
compounds in order to be taken into account in further breeding programs to find the adequate
Molecules 2017, 22, 851 6 of 15
profile according to the different market preferences.
2.2.
2.2. Spectral
Spectral Data
Data Pre-Treatments and Equation
Pre-Treatments and Equation Performances
Performances
2.2.1.
2.2.1. Second
Second Derivative
Derivative Spectra
Spectra of
of Rocket
Rocket Leaf
The applicationofof
The application thethe second
second derivative
derivative and standard
and standard normaland
normal variate variate and de-trending
de-trending algorithms
algorithms to the raw
to the raw spectra (Logspectra (Log 1/R,
1/R, Figure Figure in
2), resulted 2),substantial
resulted incorrection
substantial correction
(Figure (Figure
2) of the 2) ofshift
baseline the
baseline
caused byshift caused by
differences differences
in particle in particle
size and size and
path length. path
Peaks andlength. Peaks
troughs and troughs
in Figure in Figure
2 correspond 2
to the
correspond to the points
points of maximum of in
curvature maximum curvatureand
the raw spectrum, in it
the
hasraw spectrum,
a trough and it has
corresponding a trough
to each peak
corresponding
in the original. to each
The peak in the complexity
increase original. Theofincrease in the spectra
the derivative complexity of the
resulted inderivative spectra
a clear separation
resulted
between in a clear
peaks separation
which overlapbetween
in the rawpeaks which overlap in the raw spectra.
spectra.
Figure 2. Second derivative spectra (2, 5, 5, 2; SNV + DT) of the raw optical data for rocket samples in
Figure 2. Second derivative spectra (2, 5, 5, 2; SNV + DT) of the raw optical data for rocket samples in
the range from 400 to 2500 nm.
the range from 400 to 2500 nm.
The bands
The bands inin the
the visible
visible region
region at at 548
548 andand 670
670 nmnm are
are due
due toto electronic
electronic transitions
transitions in in the
the green
green
and red, respectively. Thus, the band at 670 nm has been assigned to chlorophyll [41], which nearnear
and red, respectively. Thus, the band at 670 nm has been assigned to chlorophyll [41], which 680
680 has
nm nm ahas a strong
strong inverse
inverse correlation
correlation withwith
sugar sugar content
content [42].[42]. InNIR
In the the NIR segment
segment of theofspectrum,
the spectrum,
the
the main absorption bands were displayed at 1920 nm, which has been attributed
main absorption bands were displayed at 1920 nm, which has been attributed to O-H stretch plus O- to O-H stretch plus
O-H
H deformation;
deformation; 2054
2054 nmnm related
related toto N-H
N-H stretchofofamides;
stretch amides;2270
2270nmnmwhich
whichhas hasbeen
beenassigned
assigned toto O-H
O-H
plus C-C
plus C-C stretch
stretch groups
groups [43]
[43] of
of cellulose,
cellulose, and and atat 2310
2310 nm
nm related
related to
to C-H
C-H stretching
stretching and
and combination
combination
bands of the methylene groups [44]. Other minor absorptions were due to
bands of the methylene groups [44]. Other minor absorptions were due to the first overtone of the first overtone of O-H
O-H
stretching (1432
stretching (1432 nm),
nm), S-H
S-H stretch
stretch first
first overtone
overtone or or C-H
C-H stretch
stretch first
first overtone
overtone of of CH
CH33 groups
groups(1696(1696nm),
nm),
and C-H stretching by methylene groups
and C-H stretching by methylene groups (1730 nm). (1730 nm).
2.2.2. Calibration
2.2.2. Calibration Equation
Equation
Table 22 shows
Table showsthe thecross-validation
cross-validation statistics
statistics for for
the the best-fitting
best-fitting equations
equations obtained
obtained using using
FNS-
FNS-6500.The coefficient of determination for cross-validation (R 2 CV) oscillated from 0.64 (GLSAT) to
6500.The coefficient of determination for cross-validation (R2CV) oscillated from 0.64 (GLSAT) to 0.94
0.94 (GER).
(GER). For RPD
For RPD values,
values, the values
the values obtained
obtained varied
varied between
between 1.62 (GLSAT)
1.62 (GLSAT) to 3.99
to 3.99 (GER).
(GER).
The statistics
The statistics of
of the
the external
external validation
validation for for the
the different
different quality
quality compounds,
compounds, including
including standard
standard
errors of performance (SEP) and R 2 VAL values for the equations of best fit obtained for each of the
errors of performance (SEP) and R2VAL values for the equations of best fit obtained for each of the
traits are
traits are shown
shown in in Table
Table 4.4. The
The SEP
SEP values
values obtained
obtained in in the
thevalidation
validation were
were lower
lower than
than their
their respective
respective
SD, indicating that NIRS is able to determine these traits in rocket leaves.
SD, indicating that NIRS is able to determine these traits in rocket leaves. The standard errors The standard errors of
of the
the glucosinolate predictions reported in previous studies ranged from 0.05 to 18.74 mol g 1 dw
glucosinolate predictions reported in previous studies ranged from 0.05 to 18.74 molg1 dw for
for Brassica
Brassica cultivars
cultivars andand Indian
Indian mustard
mustard seeds.
seeds. TheThe percentage
percentage ofoferror
errorfor
fortotal
total and
and individual
individual
glucosinolates in this work is similar to the errors previously reported for
glucosinolates in this work is similar to the errors previously reported for the aforementioned the aforementioned matrices
such as leaves, flowers and seeds [3336,45].
matrices such as leaves, flowers and seeds [3336,45].
According to
According to Williams
Williams and and Norris
Norris [46]
[46] depending
depending on on the
the RR22 value
valuefrom
fromthetheexternal
externalvalidations,
validations,
models can be classified as models that can be used to discriminate between
models can be classified as models that can be used to discriminate between low and high sample low and high sample
values, in our study those obtained for TGLs, GSAT and GER; models that can be used for rough
Molecules 2017, 22, 851 7 of 14
values,
Molecules in our
2017, 22,study
851 those obtained for TGLs, GSAT and GER; models that can be used for rough 7 of 15
predictions of samples, calibrations obtained for GRA and models with good correlations, as was the
case of the calibration obtained for TPC (Table 4).
predictions of samples, calibrations obtained for GRA and models with good correlations, as was the
case Table
of the4.calibration obtained
Reference values andfor TPC (Table
external 4). statistics of the NIRS calibrations for glucosinolate
validation
(molg1dw) and total phenolic content (mgGAE g1 dw) in rocket leaves.
Table 4. Reference values and external validation statistics of the NIRS calibrations for glucosinolate
(molg1 dw) and
Reference (n = 30)(mgGAEg1 dw) in rocket
Valuescontent
total phenolic External
leaves. Validation
Parameter Range Mean SD 1 R2VAL 2 SEP(C) 3 RPDp 4 RER 5
Reference Values (n = 30) External Validation
Total GLs 13.52-39.88 23.56 7.63 0.61 4.80 1.59 6.48
Parameter
Range Mean SD 1 R2 VAL 2 SEP(C) 3 RPDp 4 RER 5
GRA 0.15-27.53 9.11 6.82 0.79 2.75 2.48 9.95
GLSATTotal GLs 13.5239.88
0.35-17.52 23.56
9.66 7.63
4.53 0.610.60 4.80 2.36 1.59 1.92 6.48 7.78
GRA 0.1527.53 9.11 6.82 0.79 2.75 2.48 9.95
GERGLSAT 0.20-7.34
0.3517.52 3.20
9.66 1.90
4.53 0.600.59 2.36 1.22 1.92 1.56 7.78 14.89
TPC GER 0.207.34
5.80-11.80 3.20
8.73 1.90
1.57 0.590.84 1.22 0.48 1.56 3.27 14.89 12.5
1 SD: TPC 5.8011.80 8.73 1.57 0.84 0.48
standard deviation; R VAL: coefficient of determination in external validation;
2 2 3.27 312.5
SEP(C):
1 SD: standard deviation; 2 R2 VAL: coefficient of determination in external validation; 3 SEP(C): standard error of
standard error of prediction corrected for bias; 4 RPDp: ratio of the standard deviation to standard
prediction corrected for bias; 4 RPDp: ratio5of the standard deviation to standard error of prediction (performance);
error
5 RER: of
ratioprediction to standard error of RER:
of the range(performance); ratio(performance).
prediction of the range to standard error of prediction
(performance).
On the
On the basis
basis of guidelines
guidelines for
for interpretation
interpretation of of RPD
RPD from
from external
external validation
validation [46],
[46], if
if this
this ratio
ratio
exceeds aa value
exceeds value of
of 3 the calibration
calibration equation
equation is is very
very significant,
significant, this
this being
being obtained
obtained in
in this
this study
study for
for
total phenolic compounds; the ratios between 1.5 < RPDpp <<2.5
total 2.5 characterize
characterize the
the equations
equations asas suitable
suitable
for screening purposes, which was obtained for GER, TGLs GLSAT
for GLSAT andand GRA
GRA (Table
(Table4).
4).
Figure 33 shows
Figure shows the relationship
relationship between
between the the predicted
predicted reflectance
reflectance spectroscopy
spectroscopy in in the
the near
near
infrared (NIRS)
infrared (NIRS)and
andreference values
reference for the
values for glucosinolates and total
the glucosinolates andphenolic content incontent
total phenolic the validation
in the
set samples.
validation set samples.
40 30
TGLs GRA
35 25
Predicted by NIRS
Predicted by NIRS
20
30
15
25
10
20
5
15
0
10
-5
10 20 30 40 50 -5 0 5 10 15 20 25 30 35
Determined by reference Detemined by reference
20 10
GLSAT GER
15
Predicted by NIRS
Predicted by NIRS
5
10
0
0
-5 0 5 10 15 20 25 0 5 10
Determined by reference Determined by reference
TPC
12
Predicted by NIRS
10
4
4 6 8 10 12
Determined by reference
Figure
Figure 3.
3. External validation scatter
External validation scatterplot
plotfor
fornear
nearinfrared
infraredpredicted
predicted values
values versus
versus reference
reference values
values for
for glucosinolates and total phenolic acid content in rocket
glucosinolates and total phenolic acid content in rocket leaves. leaves.
Molecules 2017, 22, 851 8 of 15
In terms of RER coefficients, the predictive ability of the equations in this work extended from
6.48 to 14.89 (Table 4). For GRA, GER and TPC, the validation yielded RER (9.9514.95) values, which
indicated models with good precision. For TGLs and GLSAT the external validation yielded RER
(6.487.78) values, indicative of models that could be used for screening purposes, which could be very
useful as a selection tool in rocket breeding programmes and quality control [47].
Previous studies reporting NIRS calibrations of glucosinolates in Brassica species have been
performed mainly on rapeseed seed, because of the commercial interest of this species. Thus, Biston [31]
reported R2 VAL values of 0.99 for total glucosinolates, independently of the different reference
methods used (palladium, glucose, gas-liquid chromatography or LC). Daun et al. [32] also reported
predictions for total glucosinolates with coefficients of determination that varied from 0.74 to 0.82,
and ratios of the standard error of prediction (SEP) performed on external validation to the SD of the
reference values (RPD), that ranged from 1.36 to 2.29. Other authors [29] developed multi-product
calibrations for individual and total glucosinolates considering simultaneously different Brassica
species. These authors demonstrated the validity of the technique in approaches like that, reporting
high R2 CV values for individual (gluconapin = 0.89; sinigrin = 0.90; progoitrin = 0.86) and total
glucosinolates (0.99) predictions. Mika [48] in a work performed on Brassica napus, reported R2 VAL
value for total glucosinolates on external validation (0.84) and RER value (7.5) slightly higher than
obtained in this work. In a previous work performed on Brassica juncea (L. Czern. & Coss.) seed [33],
they obtained R2 VAL values that ranged from 0.82 to 0.95 for total and individual glucosinolates.
It has to be noted that all the above referenced works, showed prediction accuracies that were similar
or slightly higher than those reported for leaves in this work, in spite of the lower concentrations of
these compounds in rocket leaves.
The determination coefficients of cross-validation obtained in this study for GER (0.93) and
GRA (0.82) were higher than those reported on Brassica napus [34] and Brassica oleracea leaves [35]
(gluconapin: 0.73; glucobrassicin: 0.81; progoitrin: 0.78; glucoalyssin: 0.37; glucobrassicin: 0.41;
gluconapin: 0.70; gluconasturtiin: 0.62; neoglucobrassicin: 0.60), similar to reported for GLSAT (0.65),
and lower than those reported for total glucosinolates (TGLs > 0.83) in our work (TGLs: 0.70).
The potential of vis-NIRS for determining these compounds in freeze-dried broccoli has also been
examined [36]. The R2 CV reported in the aforementioned study for different GLs (glucobrassicin:
0.89; methoxyglucobrassicin: 0.69; neoglucobrassicin: 0.68; total glucosinolates: 0.73) were similar to
those obtained in our work, but lower than those reported for glucoraphanin (GRA: 0.40). This fact is
remarkable for the importance to select accessions with high GRA contents for the anticarcinogenic
properties of sulforaphane (an isothiocyanate obtained from hydrolysis of glucoraphanin).
Previous calibration models have also been developed for non-destructive estimation of total
phenol content in intact rapeseed mustard seed by Fourier transform near infrared spectroscopy [49].
The optimal models for total phenol were achieved with coefficient of determination R2 CV of 0.96 and
RPDCV values of 4.98, these coefficients were higher than those obtained in the present work. The
lower total phenolic concentration in rocket leaf (2.3012.30 mgGAEg1 dry weight) in comparison
with those detected in rapeseed-mustard seeds (7.823.9 mgGAEg1 fresh weight) could be on the
basis of the lower accuracy of the calibration model for these compounds.
2.2.3. Modified Partial Least Square Loadings

Figure 4 shows the three MPLS loading plots of the TGL and TPC equations. The first PLS term of
the equations was strongly influenced by absorption bands characteristic of pigments localized between
400 and 700 nm. Thus, carotenoids found in the genus Eruca might be influencing the absorption band
at max = 470 nm including -carotene, and xanthophylls like lutein [4]. High absorbance observed
at 670 nm is indicative of red absorbing pigments, particularly chlorophyll that gives the fruit its
characteristic green color [50]. The NIR region of the spectrum showed characteristic absorption
bands at 1212 nm which is in the spectral segment assigned to C-H stretch second overtone and water
and also by absorption bands at 1724 nm and 1764 nm corresponding to C-H stretch first overtone
Molecules 2017, 22, 851 9 of 15
Molecules 2017, 22, 851 9 of 14
and absorptions
C-H combinations
by plantatpigments
2308 nm (624
and 2348
nm), nm.
C-H Those wavelengths
stretching (1740 nm), corresponding
N-H stretchingto by
absorptions
amides (2052
by plant pigments (624 nm), C-H stretching (1740 nm), N-H stretching by amides
nm) and O-H stretch/OH deformation hydroxyl (1916 nm) [35] highly influenced the second (2052 nm) and
factor
O-Hofstretch/OH deformation
the equation hydroxyl
(Figure 4). (1916factor
The third nm) [35]
washighly
mainlyinfluenced
modelled thewith
second factor
those of the
wavelengths
equation (Figure 4).toThe
corresponding third factor
pigments was mainly
(720 nm), modelledofwith
N-H stretching those
amides wavelengths
(2084 nm), C-H corresponding
combinations oftoCH2
pigments (720 nm), N-H stretching of amides (2084 nm), C-H combinations
groups (2324 nm), O-H stretch/OH deformation hydroxyl (1924 nm) and C-H of CH groups (2324 nm),
2 stretching (1756 nm
O-Hand
stretch/OH
1212 nm).deformation hydroxyl (1924 nm) and C-H stretching (1756 nm and 1212 nm).
Figure 4. MPLS loading plots for factor 1, 2 and 3 of the 2, 5, 5, 2 (SNV + DT) transformation for total
Figure 4. MPLS loading plots for factor 1, 2 and 3 of the 2, 5, 5, 2 (SNV + DT) transformation for total
glucosinolate and phenolic acid content using near infrared reflectance spectroscopy.
glucosinolate and phenolic acid content using near infrared reflectance spectroscopy.
Because the R
Because group
the of glucosinolates
R group (Figure
of glucosinolates 1) is1)
(Figure derived fromfrom
is derived amino acids,
amino it is it
acids, possible that that
is possible
some correlation
some between
correlation the protein
between and the
the protein total
and theglucosinolate contentcontent
total glucosinolate of the leaf
of exists
the leaf(peak at 2084
exists (peak at
nm, 2084
Figure 4). In the case of phenolic content, these compounds possess one or more aromatic
nm, Figure 4). In the case of phenolic content, these compounds possess one or more aromatic rings
withrings
one or more
with onehydroxyl
or moregroups, mainly
hydroxyl related
groups, with related
mainly combination
with bands of the OH
combination bandsfunctional
of the OH
group, CH aromatic
functional group, second overtones
CH aromatic and CH
second third overtones
overtones and CH (regions from 1415
third overtones to 1512from
(regions nm and
1415 to
from1512
1512nm
to and
2035from
nm) [43,51].
1512 to 2035 nm)[43,51].
3. Materials andand
3. Materials Methods
Methods
3.1. Plant Material and Greenhouse Experiments
3.1. Plant material and Greenhouse Experiments
Fifty-two accessions of Eruca were acquired from different European genebank collections and
Fifty-two accessions of Eruca were acquired from different European genebank collections and
collected from different countries of the world. The vegetal material consisted of: one accession of
collected from different countries of the world. The vegetal material consisted of: one accession of
Eruca stenocarpa, one accession of Eruca vesicaria subsp. longirostris, 10 accessions of Eruca vesicaria
Eruca stenocarpa, one accession of Eruca vesicaria subsp. longirostris, 10 accessions of Eruca vesicaria
subsp. vesicaria and 40 accessions of Eruca vesicaria subsp. sativa.
subsp. vesicaria and 40 accessions of Eruca vesicaria subsp. sativa.
These accessions are part of a germplasm collection located at the IFAPA-La Mojonera, Almera
These accessions are part of a germplasm collection located at the IFAPA-La Mojonera, Almera
(Southern Spain). Seeds were germinated in Petri dishes for 48 h at 25 C. Pots were placed in a
(Southern Spain). Seeds were germinated in Petri dishes for 48 h at 25 C. Pots were placed in a
greenhouse under natural light at 27/18 C (day/night) and a relative humidity of 50/70% (day/night).
greenhouse under natural light at 27/18 C (day/night) and a relative humidity of 50/70% (day/night).
When the plants reached 812 cm, they were transferred to a field in Crdoba, Spain (37 5100 420 N,
When the plants reached 812 cm, they were transferred to a field in Crdoba, Spain (37"51'42'N,
04 4800 000 W; 220 m a.s.l.). The experiment was designed as a randomized complete block consisting
04'48'00'W; 220 m a.s.l.). The experiment was designed as a randomized complete block consisting of
of rows of 5 meters in length with three replicates each.
rows of 5 meters in length with three replicates each.
3.2. Sample pre-Treatment and Storage

Leaves from 10 randomly selected plants per replicate (52 accessions 3 replicates) were
harvested eight weeks after transplanting and on the same day. They were washed, weighed to assess
Molecules 2017, 22, 851 10 of 15
3.2. Sample Pre-Treatment and Storage

Leaves from 10 randomly selected plants per replicate (52 accessions 3 replicates) were harvested
eight weeks after transplanting and on the same day. They were washed, weighed to assess their
biomass, and placed in Ziploc-type freezer bags at 20 C for post-harvest storage. The samples were
freeze-dried and ground using a pestle and mortar.
3.3. GLs Analysis by Liquid Chromatography with Ultraviolet Absorbance Detection (LC-UV)
A hundred mg of freeze-dried sample was heated at 75 C for 15 min in 2.5 mL of 70:30
methanolwater and 200 L of 10 mM sinigrin as an external standard (sinigrin hydrate, 85440 Fluka,
St. Louis, MO, USA) according to the ISO norm (ISO 9167-1, 1992). A second extraction was applied
after centrifugation (5 min, 5 103 g) using 2 mL of 70:30 methanol-water. The combined GLs extracts
were pipetted (1 mL) onto the top of an ion-exchange column containing 1 mL of Sephadex DEAE-A25
(40-125 m bead size, 30000 Da exclusion limit). Desulfation was carried out by addition of 75 L
of purified sulfatase (EC 3.1.6.1, type H-1 from Helix pomatia) (Sigma-Aldrich, St. Louis, MO, USA)
solution. Desulfated GLs were eluted with 2.5 mL of Milli-Q (Millipore, Bedford, MA, USA) ultrapure
water and analyzed with a 600 HPLC instrument (Waters) equipped with a model 486 UV tunable
absorbance detector (Waters, Milford, MA, USA) fixed at a wavelength of 229 nm. Separation was
carried out using a Lichrospher 100 RP-18 in Lichrocart column (125 mm 4 mm i.d., 5 m particle
size, Merck, Darmstadt, Germany). The HPLC chromatogram was compared to the desulfo-GL profile
provided by three certified reference materials recommended by the U.E. and ISO (CRMs 366, 190
and 367) (Commission of the European Communities, report EUR 13339 EN, 1-75) [52].
3.4. Determination of the Total Phenolic Fraction

The concentration of total phenolic compounds (TPC) was estimated by a modified version of
the FolinCiocalteu method [53], using gallic acid as standard, for which a calibration curve was run
with solutions of 50, 100, 200, 300, 400, 500 and 600 mg/L of this compound. A 0.06 mL aliquot of
extract 1.58 mL of distilled water, 0.1 mL of Folin-Ciocalteu reagent and 0.3 mL of Na2 CO3 (20% w/v)
were mixed and heated at 50 C for 5 min. After 30 min, the absorbance was measured at 765 nm
against a blank similarly prepared, but containing 70:30 ethanolwater mixture (pH 3.2) instead of
extract. Sodium carbonate (Panreac, Spain), FolinCiocalteu reagent (FCR) and gallic acid (both from
Sigma-Aldrich) were used to determine the total phenol fraction. The absorbance was measured with
a ThermoSpectronic UV-visible Spectrometer (Thermo Fisher Scientific, Waltham, MA, USA).
3.5. NIRS Analysis Calibration and Validation Development

The freeze-dried rocket samples were scanned using the FNS-6500 scanning instrument (FOSS
NIRSystems, Silver Spring, MD, USA) in a small ring cup (3.75 cm ) using the spinning sample
module. Spectra were collected on all samples in the reflectance mode, acquiring their spectra over a
wavelength range from 400 to 2500 nm (visible and near infrared regions). Samples were scanned in
duplicate and the average spectrum was used to develop the multivariate models. Reflectance data
were stored as log (1/R) (R = reflectance) at 2 nm intervals (1050 data points).
Before developing NIRS calibrations, the structure and spectral variability of the sample
population was determined using the CENTER algorithm included in the WinISI software. This
programme performs an initial principal component analysis (PCA) to calculate the centre of the
population and the distance of samples (spectra) from that centre in an n-dimensional space using the
Mahalanobis distance (GH); samples with a statistical value greater than three were considered outliers
or anomalous spectra [54]. Thus, having ordered the sample set by spectral distance (from smallest
to greatest distance to the centre), the 30 samples forming the validation set were selected by taking
one of every 5 samples in the final 156 sample set; the calibration set thus comprised the remaining
126 samples.
Molecules 2017, 22, 851 11 of 15
Calibration equations for physico-chemical were developed using the programme GLOBAL v. 1.50
(WINISI II, Infrasoft International, LLC, Port Matilda, PA, USA). Calibration equations were computed
using raw optical data (log 1/R, where R is reflectance), or first or second derivatives of the log 1/R
data, with several combinations of derivative (gap) sizes and smoothing [i.e., (0, 0, 1, 1; derivative
order and segment , first smooth, second smooth); (1, 4, 4, 1); (1, 10, 10, 1); (2, 5, 5, 2); (2, 20, 20, 2)]. The
regression method employed to correlate spectral information and quality compounds in the samples
was modified partial least squares (MPLS). This regression method is a soft-modelling method [41,42]
for constructing predictive models when the factors are many and highly collinear and allows a model
to be calculated that was tested on external samples observing its prediction ability. The final objective
of the mathematical procedure is to reduce the high number of spectral data points (absorbance values
from 400 to 2500 nm every 2 nm, i.e. 1050 data) and to eliminate the correlation of absorbance values
presented by neighbouring wavelengths [55]. Standard normal variate and detrend transformations
(SNV-DT) were used to correct baseline offset due to scattering effects (differences in particle size
among samples) [56]. Cross-validation was performed on the calibration set for determining the best
number of terms to use in the equation, as well as to determine the ability of each equation to make
predictions on unknown samples [57].
An external validation procedure in 30 independent samples was carried out to determine the
accuracy and precision of the equations obtained in the calibration for each quality component.
The coefficient of determination (R2 ) and standard error (SE) were calculated for both cross-validation
and external validation. The predictive ability of the mathematical models was assessed in the external
validation from the R2 , the RPD, which is the ratio of the standard deviation for the validation samples
to the standard error of prediction (performance) (SEP), and the RER, which is the ratio of the range in
the reference data (validation set) to the SEP.
Depending on the R2 value from the external validation, NIR models can be classified [37] as:
models with a low correlation (0.26 < R2 VAL < 0.49); models that can be used to discriminate between
low and high values of the samples (0.50 < R2 VAL < 0.64); models that can be used for rough predictions
of samples (0.65 < R2 VAL < 0.81); models with good correlations (0.82 < R2 VAL < 0.90); and models
with excellent precision (R2 VAL > 0.90).
The guideline used for setting performance calibrations stated that an RPD value of more than
3 is desirable for excellent calibration equations, whereas equations with an RPD of less than 1.5 are
unusable [46]. In relation to the range error ratio (RER), it should ideally be at least 10 [47].
The mathematical expressions of these statistics are as follows:
*" ! #1/2 +1
n
RPD = SD (yi yi ) 2
( N K 1) 1
i =1
where yi = lab reference value for the ith sample; y = NIR measured value; N = number of samples, K =
number of wavelengths used in an equation; SD = standard deviation.
*" ! #1/2 +1
n
RER = range (yi yi ) 2
( N K 1) 1
i =1
where yi = lab reference value for the ith sample; y = NIR measured value; N = number of samples, K =
number of wavelengths used in an equation.
3.6. Statistical Analysis

Correlation analysis was assessed by the Pearson test among TGLs, GRA, GLSAT and TPC to
determine significant relationships. Statistical analyses were performed using SPSS 13.0 (SPSS Inc.,
Chicago, IL, USA).
Molecules 2017, 22, 851 12 of 15
4. Conclusions
Results reported in this work show that NIRS is able to predict the glucosinolate and phenolic
contents in the leaf rocket, with sufficient accuracy for screening purposes. Each sample that
we analyzed by using the NIRS method took us approximately 1.5 min, and prediction results
for the individual and total glucosinolate and phenolic contents were monitored instantaneously.
The equations shown in this work have been recently applied to the evaluation of both compounds
in 3000 individual plants of rocket. This has allowed the rapid identification and selection of the
genotypes of interest, which would not have been possible by using the reference method. NIRS is
thus ideal for mass screening programmes in large-scale plant monitoring and quality control.
Acknowledgments: The authors wish to express their thanks to the Project entitled Innovacin sostenible en
horticultura protegida (PP.AVA.AVA201601.7), FEDER y FSE (Programa Operativo FSE de Andalucia 2007-2013
Andaluca se mueve con Europa) for the funding of this research. We thank Nicholas Davies for his help in the
grammatical revision of the manuscript.
Author Contributions: M.D.R.C. and R.F. conceived and designed the experiments; M.V.P. performed the field
trials; S.O.C. and A.D.H.B. determined the glucosinolate content; E.M.T.M. developed the NIRS calibrations and
wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Zeven, A.C.; de Wet, J.M.J. Dictionary of Cultivated Plants and Their Regions of Diversity, 2nd ed.; Centre for
Agricultural Publishing and Documentation: Wageningen, The Netherlands, 1982; p. 107.
2. Warwick, S.I. Guide to the wild germplasm of brassica and allied crops. Pt. V: Life history and geographical
data for wild species in the tribe Brassiceae (Cruciferae). Agric. Can. Tech. Bull. 1994, 2E, 61.
3. Yaniv, Z.; Schafferman, D.; Amar, Z. Tradition, uses, and biodiversity of rocket (Eruca. sativa) in Israel.
Econ. Bot. 1998, 52, 394400. [CrossRef]
4. Villatoro-Pulido, M.; Priego-Capote, F.; lvarez-Snchez, B.; Saha, S.; Philo, M.; Obregn-Cano, S.;
De Haro-Bailn, A.; Font, R.; Del Ro-Celestino, M. An approach to the phytochemical profiling of rocket
[Eruca. sativa (Mill.) Thell]. J. Sci. Food Agric. 2013, 93, 38093819. [CrossRef] [PubMed]
5. Windsor, A.J.; Reichelt, M.; Figuth, A.; Svatos, A.; Kroymann, J.; Kliebenstein, D.J.; Gershenzon, J.;
Mitchell-Olds, T. Geographic and evolutionary diversification of glucosinolates among near relatives of
Arabidopsis thaliana (Brassicaceae.). Phytochemistry 2005, 66, 13211333. [CrossRef] [PubMed]
6. Louda, S.; Mole, S. Glucosinolates: chemistry and ecology. In Herbivores: Their Interactions with Secondary
Plant Metabolites; Rosenthal, G.A., Berembaum, M.R., Eds.; Academic Press: New York, NY, USA, 1991;
Volume 1, pp. 123164.
7. Bell, L.; Wagstaff, C. Glucosinolates, myrosinase hydrolysis products, and flavonols found in rocket
(Eruca. sativa and Diplotaxis. tenuifolia). J. Agric. Food Chem. 2014, 62, 44814492. [CrossRef] [PubMed]
8. Srensen, H. Glucosinolates: Structure, properties, function. In Canola and Rapeseed. Production, Chemistry,
Nutrition and Processing Technology; Shahidi, F., Ed.; Springer: Berlin, Germany, 1990; pp. 149172.
9. Rosa, E.A.S.; Heaney, R.K.; Fenwick., G.R.; Portas, C.A.M. Glucosinolates in crop plants. Hortic. Rev. 1997,
19, 99225.
10. Agerbirk, N.; Olsen, C.E. Glucosinolate structures in evolution. Phytochemistry 2012, 77, 1645. [CrossRef]
[PubMed]
11. Kim, H.J.; Lee, M.J.; Jeong, M.H.; Kim, J.E. Identification and quantification of glucosinolates in kimchi by
liquid chromatography-electrospray tandem mass spectrometry. Int. J. Anal. Chem. 2017, 2017, 6753481.
[CrossRef] [PubMed]
12. Fahey, J.W.; Zalcmann, A.T.; Talalay, P. The chemical diversity and distribution of glucosinolates and
isothiocyanates among plants. Phytochemistry 2001, 56, 551. [CrossRef]
13. Pasini, F.; Verardo, V.; Caboni, M.F.; D0 Antuono, L.F. Determination of glucosinolates and phenolic
compounds in rocket salad by HPLC-DADMS: Evaluation of Eruca. sativa Mill and Diplotaxis. tenuifolia L.
genetic resources. Food Chem. 2012, 133, 10251033. [CrossRef]
Molecules 2017, 22, 851 13 of 15
14. Angus, J.; Gardner, P.; Kirkegaard, J.; Desmarchlier, J. Biofumigation: Isothiocyanates released from Brassica
roots inhibit growth of take-all fungus. Plant. Soil 1994, 162, 107112. [CrossRef]
15. Fahey, J.W.; Haristoy, X.; Dolan, P.M.; Kensler, T.W.; Scholtus, I.; Stephenson, K.K.; Talalay, P.; Lozniewski, A.
Sulforaphane inhibits extracellular, intracellular, and antibiotic-resistant strains of Helicobacter pylori and
prevents benzo[a]pyrene-induced stomach tumors. Proc. Natl. Acad. Sci. USA 2002, 99, 76107615. [CrossRef]
[PubMed]
16. Chen, S.; Andreasson, E. Update on glucosinolate metabolism and transport. Plant. Physio. Bioch. 2001, 39,
743758. [CrossRef]
17. Mithen, R. Glucosinolates-biochemistry, genetics and biological activity. Plant. Growth Regul. 2001, 34,
91103. [CrossRef]
18. Juge, N.; Mithen, R.F.; Traka, M. Molecular basis of chemoprevention by sulforaphane: A comprehensive
review. Cell. Mol. Life Sci. 2007, 64, 11051127. [CrossRef] [PubMed]
19. Keum, Y.S.; Jeong, W.S.; Kong, A.N. Chemoprevention by isothiocyanates and their underlying molecular
signaling mechanisms. Mutat. Res. 2004, 555, 191202. [CrossRef] [PubMed]
20. Iori, R.; Bernardi, R.; Gueyrard, D.; Rollin, P.; Palmieri, S. Formation of glucoraphanin by chemoselective
oxidation of natural glucoerucin: A chemoenzymatic route to sulforaphane. Bioorg. Med. Chem. Lett. 1999, 9,
10471048. [CrossRef]
21. Fimognari, C.; Nusse, M.; Berti, F.; Iori, R.; Cantelli-Forti, G.; Hrelia, P. Isothiocyanates as novel cytotoxic and
cytostatic agents: Molecular pathway on human transformed and nontransformed cells. Biochem. Pharmacol.
2004, 68, 11331138. [CrossRef] [PubMed]
22. Fimognari, C.; Nusse, M.; Iori, R.; Cantelli-Forti, G.; Hrelia, P. The new isothiocyanate 4-(methylthio)butyl
isothiocyanate selectively affects cell-cycle progression and apoptosis induction of human leukemia cells.
Invest. New Drugs 2004, 22, 119129. [CrossRef] [PubMed]
23. Kim, S.J.; Kawaharada, C.; Jin, S.; Hashimoto, M.; Ishii, G.; Yamauchi, H. Structural elucidation of
4-(cysteine-S-yl)butyl glucosinolate from the leaves of Eruca. sativa. Biosci. Biotech. Bioch. 2007, 71, 114121.
[CrossRef] [PubMed]
24. Cataldi, T.R.I.; Rubino, A.; Lelario, F.; Bufo, S.A. Naturally occurring glucosinolates in plant extracts of
rocket salad (Eruca. sativa L.) identified by liquid chromatography coupled with negative ion electrospray
ionization and quadrupole ion-trap mass spectrometry. Rapid Commun. Mas. Spectrom. 2007, 21, 23742388.
[CrossRef] [PubMed]
25. Kim, S.J.; Ishii, G. Glucosinolate profiles in the seeds, leaves and roots of rocket salad (Eruca. sativa Mill.) and
anti-oxidative activities of intact plant powder and purified 4-methoxyglucobrassicin. Soil Sci. Plant. Nutr.
2006, 52, 394400. [CrossRef]
26. Weckerle, B.; Michel, K.; Balzs, B.; Schreier, P.; Tth, G. Quercetin 3,30 ,40 -tri-O--D-glucopyranosides from
leaves of Eruca. sativa (Mill.). Phytochemistry. 2001, 57, 547551. [CrossRef]
27. Vallejo, F.; Toms-Barbern, F.A.; Garca-Viguera, C. Potential bioactive compounds in health promotion
from broccoli cultivars grown in Spain. J. Sci. Food Agric. 2002, 82, 12931297. [CrossRef]
28. Danielak, R.; Borkowski, B. Biologically active compounds in seeds of crucifers Part III. Chromatographical
search for glucosinolates. Dissert. Pharm. Pharma. 1969, 2, 563575.
29. Velasco, L.; Becker, H.C. Analysis of total glucosinolate content and individual glucosinolates in Brassica spp.
by near-infrared reflectance spectroscopy. Plant. Breeding 1998, 117, 97102. [CrossRef]
30. Villatoro-Pulido, M.M.; Moreno-Rojas, R.; Muoz-Serrano, A.; Cardeosa, V.; Amaro-Lpez, M.A.; Font, R.;
Del Ro-Celestino, M. Characterization and prediction by near-infrared reflectance of mineral composition
of rocket (Eruca. vesicaria subsp. sativa and Eruca. vesicaria subsp. vesicaria). J. Sci. Food Agric. 2012, 92,
13311340. [CrossRef] [PubMed]
31. Biston, R.; Dardenne, P.; Cwikowski, M.; Marlier, M.; Severin, M.; Wathelet, J.P. Fast Analysis of Rapeseed
Glucosinolates by near-Infrared Reflectance Spectroscopy. J. Am. Oil Chem. Soc. 1988, 65, 15991600.
[CrossRef]
32. Daun, J.K.; Clear, K.M.; Williams, P. Comparison of three whole seed near-infrared analyzers for measuring
quality components of canola seed. J. Am. Oil Chem. Soc. 1994, 71, 10631608. [CrossRef]
33. Font, R.; Del Ro, M.; Fernndez-Martnez, J.M.; De Haro, A. Use of near-infrared spectroscopy for screening
the individual and total glucosinolate content in Indian mustard seed (Brassica juncea L. Czern. & Coss.).
J. Agric. Food Chem. 2004, 52, 35633569. [CrossRef]
Molecules 2017, 22, 851 14 of 15
34. Font, R.; Del Ro-Celestino, M.; Cartea, M.; De Haro-Bailn, A. Quantification of glucosinolates in leaves
of leaf rape Brassica napus ssp. pabularia) by near-infrared spectroscopy. Phytochemistry 2005, 66, 175185.
[CrossRef] [PubMed]
35. Font, R.; Del Ro-Celestino, M.; Rosa, E.; Aires, A.; De Haro-Bailn, A. Glucosinolate assessment in Brassica
oleracea leaves by Near-Infrared Spectroscopy. J. Agric. Sci. 2005, 143, 6573. [CrossRef]
36. Hernndez-Hierro, J.M.; Valverde, J.; Villacreces, S.; Reilly, K.; Gaffney, M.; Gonzalez-Miret, M.L.; Heredia, F.J.;
Downey, G. Feasibility Study on the Use of Visible-Near-Infrared Spectroscopy for the Screening of Individual
and Total Glucosinolate Contents in Broccoli. J. Agric. Food Chem. 2012, 60, 73527358. [CrossRef] [PubMed]
37. Bennett, R.N.; Rosa, E.A.S.; Mellon, F.A.; Kroon, P.A. Ontogenic Profiling of Glucosinolates, Flavonoids, and
Other Secondary Metabolites in Eruca. sativa (Salad Rocket), Diplotaxis. erucoides (Wall Rocket), Diplotaxis.
tenuifolia (Wild Rocket), and Bunias. orientalis (Turkish Rocket). J. Agric. Food Chem. 2006, 54, 40054015.
[CrossRef] [PubMed]
38. Bennett, R.N.; Carvalho, R.; Mellon, F.A.; Eagles, J.; Rosa, E.A.S. Identification and quantification of
glucosinolates in sprouts derived from deeds of wild Eruca. sativa L. (salad rocket) and Diplotaxis. tenuifolia L.
(Wild Rocket) from diverse geographical locations. J. Agric. Food Chem. 2007, 55, 6774. [CrossRef] [PubMed]
39. Perez-Balibrea, S.; Moreno, D.A.; Garca-Viguera, C. Genotypic effects on the phytochemical quality of seeds
and sprouts from commercial broccoli cultivars. Food Chem. 2011, 125, 348354. [CrossRef]
40. Bell, L.; Yahya, H.N.; Oloyede, O.O.; Methven, L.; Wagstaff, C. Changes in rocket salad phytochemicals
within the commercial supply chain: Glucosinolates, isothiocyanates, amino acids and bacterial load increase
significantly after processing. Food Chem. 2017, 221, 521534. [CrossRef] [PubMed]
41. Naes, T.; Isaksson, T.; Fearn, T.; Davies, T. A User-Friendly Guide to Multivariate Calibration and
Classification. NIR Publication: Chichester, UK, 2002.
42. Wold, S.; Sjostrom, M.; Eriksson, L. PLS-regression: A basic tool of chemometrics. Chemometr. Intel. Lab. 2001,
58, 109130. [CrossRef]
43. Osborne, B.G.; Fearn, T.; Hindle, P. Practical NIR Spectroscopy with Applications in Food and Beverage Analysis,
2nd ed.; Addison-Wesley Longman Ltd.: Harlow, UK, 1993.
44. Murray, I.; Williams, P.C. Chemical principles of near-infrared technology. In Near-Infrared Technology in the
Agricultural and Food Industries; Williams, P., Norris, K., Eds.; The American Association of Cereal Chemists:
St. Paul, MN, USA, 1987; pp. 1734.
45. Petisco, C.; Garca-Criado, B.; Vzquez-de-Aldana, B.R.; de Haro, A.; Garca-Ciudad, A. Measurement of
quality parameters in intact seeds of Brassica species using visible and near-infrared spectroscopy. Ind. Crops
Prod. 2010, 32, 139146. [CrossRef]
46. Williams, P.; Norris, K. Near-Infrared Technology in the Agricultural and Food Industries, 2nd ed.; The American
Association of Cereal Chemists: St. Paul, MN, USA, 2001; pp. 171185.
47. Williams, P.C.; Sobering, D.C. How do we do it: A brief summary of the methods we use in developing near
infrared calibrations. In Near Infrared Spectroscopy: The Future Waves; Davies, A.M.C., Williams, P.C., Eds.;
NIR Publications: Chichester, UK, 1996; pp. 185188.
48. Mika, V.; Tillmann, P.; Koprna, R.; Nerusil, P.; Kucera, V. Fast prediction of quality parameters in whole seeds
of oilseed rape (Brassica napus L.). Plant. Soil Environ. 2003, 49, 141145.
49. Bala, M.; Singh, M. Non-destructive estimation of total phenol and crude fiber content in intact seeds of
rapeseedmustard using FTNIR. Ind. Crops Prod. 2013, 42, 357362. [CrossRef]
50. Thomas, J.R.; Gausman, H.W. Leaf reflectance vs. leaf chlorophyll and carotenoid concentrations for eight
crops. Agron. J. 1977, 69, 799802. [CrossRef]
51. Cozzolino, D. The Role of Visible and Infrared Spectroscopy Combined with Chemometrics to Measure
Phenolic Compounds in Grape and Wine Samples. Molecules 2015, 20, 726737. [CrossRef] [PubMed]
52. Wathelet, J.P.; Wagstaffe, P.J.; Boenke, A. The certification of the total glucosinolate and sulphur contents of
three rapeseed (colza), CRMs 190, 366 and 367. Comission of the European Communities, report EUR 13339;
Office for Official Publications of the European Communities: rue Mercier, Luxembourg, 1991; pp. 175.
53. Singleton, V.L.; Rossi, J.A. Colorimetry of total phenolics with phosphomolybdic phosphotungstic acid
reagents. Am. J. Enol. Vitic. 1965, 16, 144158.
54. Shenk, J.S.; Westerhaus, M.O. Population structuring of near infrared spectra and modified partial least
squares regression. Crop. Sci. 1991, 31, 15481555. [CrossRef]
Molecules 2017, 22, 851 15 of 15
55. Shenk, J.S.; Westerhaus, M.O. The application of near infrared reflectance spectroscopy (NIRS) to forage
analysis. In Forage Quality, Evaluation and Utilization; Fahey, G.C., Jr., Collins, M., Mertens, D.R., Moser, L.E.,
Eds.; Crop Science Society of America, Soil Science Society of America: Madison, WI, USA, 1994; pp. 406450.
56. Barnes, R.J.; Dhanoa, M.S.; Lister, S.J. Standard normal variate transformation and de-trending of near
infrared diffuse reflectance spectra. Appl. Spectrosc. 1989, 43, 772777. [CrossRef]
57. Shenk, J.S.; Workman, J.; Westerhaus, M. Application of NIR spectroscopy to agricultural products. In Marcel
Dekker Handbook of Near Infrared Analysis; Burns, D.A., Ciurczac, E.W., Eds.; CRC press: Boca Raton, FL, USA,
2001; pp. 419474.
Sample Availability: Seeds of all the accessions are available from the authors.
2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

2017 Toledo Font Del Rio NIRS Glucos Eruca Molecules Publicado - 22-00851

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

2017 Toledo Font Del Rio NIRS Glucos Eruca Molecules Publicado - 22-00851

Caricato da

Copyright:

Formati disponibili

molecules

Academic Editor: Derek J. McPhee

Molecules 2017, 22, 851; doi:10.3390/molecules22050851 www.mdpi.com/journal/molecules

Molecules 2017, 22, 851 5-(Methylsulfinyl)pentyl Glucoalyssin 3 of 15

2. Results and Discussion

2.1. Reference Values

Parameter Range Mean SD 1 R 2C 2 SEC 3 R 2 CV 4 SECV 5 RPDcv 6

Parameter TGLs GRA GLSAT GER TPC

Determined by reference Determined by reference

2.2.3. Modified Partial Least Square Loadings

3.2. Sample pre-Treatment and Storage

3.2. Sample Pre-Treatment and Storage

3.4. Determination of the Total Phenolic Fraction

3.5. NIRS Analysis Calibration and Validation Development

3.6. Statistical Analysis

Potrebbero piacerti anche