Content uploaded by Mariateresa Maldini
Author content
All content in this area was uploaded by Mariateresa Maldini on Feb 16, 2018
Content may be subject to copyright.
‘Moringa oleifera: study of phenolics and
glucosinolates by mass spectrometry’
†
Mariateresa Maldini,
a
Salwa A. Maksoud,
b
Fausta Natella,
c
Paola Montoro,
d
Giacomo Luigi Petretto,
a
Marzia Foddai,
a
Gina Rosalinda De Nicola,
e
Mario Chessa
a
and Giorgio Pintore
a
*
Moringa oleifera is a medicinal plant and an excellent dietary source of micronutrients (vitamins and minerals) and health-
promoting phytochemicals (phenolic compounds, glucosinolates and isothiocyanates). Glucosinolates and isothiocyanates are
known to possess anti-carcinogenic and antioxidant effects and have attracted great interest from both toxicological and
pharmacological points of view, as they are able to induce phase 2 detoxification enzymes and to inhibit phase 1 activation
enzymes. Phenolic compounds possess antioxidant properties and may exert a preventative effect in regards to the development
of chronic degenerative diseases.
The aim of this work was to assess the profile and the level of bioactive compounds in all parts of M. oleifera seedlings, by using
different MS approaches. First, flow injection electrospray ionization mass spectrometry (FI-ESI-MS) fingerprinting techniques and
chemometrics (PCA) were used to achieve the characterization of the different plant’s organs in terms of profile of phenolic com-
pounds and glucosinolates. Second, LC-MS and LC-MS/MS qualitative and quantitative methods were used for the identification
and/or determination of phenolics and glucosinolates in M. oleifera. Copyright © 2014 John Wiley & Sons, Ltd.
Keywords: Moringa oleifera; glucosinolates; PCA; LC-MS; MRM
Introduction
Moringa oleifera Lam. (synonim: M. ptreygosperma Gaertn.)
(Moringaceae) is a tree native of India, Pakistan, Bangladesh and
Afghanistan, widely distributed in tropical and sub-tropical areas
of the world.
[1,2]
Moringa, the sole genus in the family Moringaceae,
consists of 14 species, among which M. oleifera is the best known
and most widely distributed and naturalized. M. oleifera, also called
‘Miracle Vegetable’, is a multiuse plant used for human nutrition as
functional food, animal feeding and for medicinal purposes.
[3]
In
fact, a wide variety of nutritional and medicinal virtues have been
attributed to its roots, bark, leaves, flowers, fruits and seeds. All
these parts are used in folk medicine for the treatment of various
ailments including the treatment of inflammation and infectious
diseases along with cardiovascular, gastrointestinal, haematological
and hepatorenal disorders. In addition, extracts of various Moringa
tissues have been used for anti-bacterial and anti-cancer activity
(M. oleifera seeds),
[4]
anti-inflammatory and hepatoprotective
agents (M. oleifera fruits and bark),
[5]
while leaf extracts have been
shown to regulate thyroids status and cholesterol levels in rats.
[6]
Recently, this plant has attracted great interest as an important
food commodity because of its high nutritional value. Leaves,
flowers and green pods are used traditionally as vegetable, whereas
the seed can be consumed fresh, fried, roasted or ground to meal.
The seed is also the source of a high quality edible oil known as
moringa oil, or ben oil, that can be used inseveral applications such
as cooking, cosmetics and as a lubricant.
M. oleifera is an excellent dietary source of micronutrients, vita-
mins and minerals, and health-promoting phytochemicals such as
glucosinolates and phenolic compounds.
[7]
The stem bark has also
been reported to contain alkaloids (moringinine and moringine).
[8]
In particular, this plant genus contains unusual sugar-substituted
hydroxy-aromatic glucosinolates.
[9,10]
Generally, glucosinolates
are structurally characterized by a β-D-glucopyranosyl unit and an
O-sulfated anomeric (Z)-thiohydroximate function connected to a
variable side chain depending on the amino acid metabolism of
the plant species.
[11]
M. oleifera contains several uncommon mem-
bers of the glucosinolates family with peculiar characteristics given
by the presence in their structure of a second saccharide residue in
the aglycon side chain.
[12–14]
The predominant glucosinolate in this
species is 4-(α-L-rhamnopyranosiloxy)benzyl glucosinolate, known
as glucomoringin. Due to its atypical structure, this compound
could display biological properties distinctly different from those
of others GLs.
[14]
Glucosinolates and isothiocyanates are known to
possess anti-carcinogenic and antioxidant effects and have
* Correspondence to: Giorgio Pintore, University of Sassari, Department of
Chemistry and Pharmacy via F. Muroni, 23/b, 07100 Sassari, Italy. E-mail:
pintore@uniss.it
†
This article is part of the Journal of Mass Spectrometry special issue entitled “3rd
MS Food Day”edited by Gianluca Giorgi.
aUniversityof Sassari, Department of Chemistry and Pharmacy via F. Muroni, 23/b,
07100, Sassari, Italy
bUniversity of Cairo, Faculty of Science, Department of Botany, Egypt
cConsiglio per la Ricerca e la sperimentazione in Agricoltura, CRA-NUT, via
Ardeatina, 546, 00178, Roma, Italy
dUniversity of Salerno, Department of Pharmacy, via Giovanni Paolo II, 84084,
Fisciano, Sa, Italy
eConsiglio per la Ricerca e la sperimentazione in Agricoltura, CRA-CIN, via di
Corticella, 133, 40128, Bologna, Italy
J. Mass Spectrom. 2014, 49, 900–910 Copyright © 2014 John Wiley & Sons, Ltd.
Research article
Received: 21 February 2014 Revised: 1 July 2014 Accepted: 8 July 2014 Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/jms.3437
900
attracted great interest from both toxicological and pharmacologi-
cal points of view, as they are able to induce phase 2 detoxification
enzymes, and to inhibit phase 1 activation enzymes.
[15]
Since glucosinolates and their derivatives possess different rel-
evant biological activities, their identification and quantification
in plant tissues have become of great importance.
[16]
Several published indirect methods are based on the determina-
tion of their enzymatic volatile breakdown products by GC-MS
analysis, but unfortunately some glucosinolate breakdowns
products are unstable in the conditions applied, and they cannot
be correctly detected. Therefore, the current tendency is to
determine intact glucosinolates or desulfoglucosinolates by more
accurate and robust LC-MS methods.
[17]
LC coupled with tandem
mass spectrometry (LC-MS/MS) is an important tool that can be
used for both qualitative and quantitative analysis, especially in
the case of the characterization of GLs composition in less investi-
gated species.
[16,18]
Phenolics are a large class of secondary metabolites widely
distributed in plant kingdom. Previous phytochemical studies on
different tissues of M. oleifera reported the highest level of phenolic
compounds in leaves extracts, mainly flavonoids and caffeic acid
derivatives.
[9]
Crypto-chlorogenic acid, caffeoylquinic acids (5- and
3-isomers), isoquercetin and astragalin were detected to be the
major compounds isolated and identified in the leaves of
M. oleifera.
[9,19–21]
Concerning flavonoids, their profiles were
found more complex and characterized by flavonol glycosides
(kaempferol 3-O-rutinoside, kaempferol 3-O-glucoside, quercetin
3-O-glucoside and rutin). Amaglo et al.
[15]
reported leaves as the
tissue with the highest and most complex flavonoid contents, while
in roots or seeds any phenolic compounds are detected. Phenolics
and flavonoids are active antioxidant components in the leaves of
M. oleifera, responsible also for anti-inflammatory, atherosclerotic
and anti-diabetics activities.
[8,22–24]
The aim of the present study was to determine the profile of
glucosinolates and phenolics, as well as to assess the glucosinolate
content in pulp seed, seed coat, leaves and roots of M. oleifera 12-
day old seedlings.
Furthermore, we investigated whether ESI-MS coupled to
statistical analysis (principal components analysis (PCA)) could be
used as a simple and reliable technique to distinguish extracts from
different tissues. Direct infusion ESI-MS offers several advantages
since it is a very fast, versatile, reproducible and sensitive
technique,
[25]
which requires little or no sample preparation and
provides almost instantaneous information. On the other hand,
PCA is a chemometric approach that combines mathematical,
statistical and computing methods which allows obtaining the
maximum information from chemical data analyses.
[26]
First, on the basis of ESI-MS and ESI-MS/MS results, we described
the profiles of glucosinolates, along with proanthocyanidins and
phenolic compounds, in M. oleifera tissues. Second, the occurrence
and level of glucosinolates were determined in pulp and coat of
dried seeds, as well as in roots and leaves of 12-day-old seedlings,
by LC-MS/MS analysis and a rapid and sensitive LC-MS/MS
(MRM) method.
Experimental
Materials
Solvents used for extraction, HPLC grade methanol, acetonitrile and
formic acid were from Sigma-Aldrich Chemical Company (St Louis,
MO). HPLC grade water (18 mΩ) was prepared by using a Millipore
(Bedford, MA, USA) Milli-Q purification system. Glucoraphanin
potassium salt, glucoiberin potassium salt, glucotropaeolin potas-
sium salt, glucosinalbin potassium salt, glucobarbarin potassium
salt and glucoraphenin potassium salt were purchased from
PhytoLab GmbH & Co. KG (Vestenbergsgreuth, Germany).
Glucomoringin purification
Glucomoringin was isolated from M. oleifera Lam. (fam. Moringaceae)
seeds (cake powder PKM2 provided by Indena India Pvt. Ltd.;
Bangalore, India) according to a previously described method.
[14]
In brief, glucomoringin was purified by two sequential chromato-
graphic steps: isolation through anion exchange chromatography,
followed by gel filtration to achieve purification to homogeneity.
Glucomoringin was unambiguously characterized by
1
H- and
13
C-NMR spectrometry, and the purity was assessed by HPLC
analysis of the desulfo-derivative according to the ISO 9167-1
method,
[27]
yielding glucomoringin with a purity of 99% based
on peak area value and 95% on a weight basis.
Plant material and growth conditions
M. oleifera Lam. seeds werekindly provided from Moringa society of
Egypt. A part of dried seeds was separated in coat and pulp,
whereas another part was kept 2days in water for imbibition and
then germinated in field for 12 days. Afterwards, young leaves
and roots of 12-day-old seedlings were rapidly and gently collected,
immediately frozen in liquid nitrogen and then stored at 80 °C
prior analysis.
Extraction and sample preparation
Each sample of M. oleifera (pulp seed, seed coat, leaves and roots)
was ground to a fine powder and extracted with methanol:water
(70:30 v/v; sample to solvent ratio 1:25 w/v)at70°Cfor30minunder
vortex mixing to facilitate the extraction. The samples were succes-
sively centrifuged (4000 rpm, 30 min, 4 °C), the supernatants were
collected and the solvent was completely removed using a rotary
evaporator under vacuum at 40 °C. The dried samples were
dissolved in ultrapure water with the same volume of extraction
and filtered through 0.20-μm syringe PVDF filters (Whatmann
International Ltd., UK).
ESI-MS and ESI-MS/MS analyses
Full scan ESI-MS and collision-induced dissociation (CID) ESI-MS/MS
analyses ofsamples were performed on an ABSciex API2000 (Foster
City, CA, USA) spectrometer. The analytical parameters were opti-
mized by infusing a standard solution of glucomoringin (1μgml
1
in methanol 50%) into the source at a flow rate of 10 μlmin
1
.The
optimized parameters were declustering potential 40 eV, focus-
ing potential 400 eV and entrance potential 10 eV. Data were
acquired in the negative ion MS and MS/MS modes.
Full scan ESI-MS, MS/MS and MS
3
analyses of standards and
samples were performed on an ABSciex API32000 Q-Trap (Foster
City, CA, USA) spectrometer. The analytical parameters were opti-
mized by infusing a standard solution of glucomoringin (1μgml
1
in methanol 50%) into the source at a flow rate of 10 μlmin
1
.The
optimized parameters were declustering potential 73.6 eV,
entrance potential 4 eV, collision energy 39 eV and collision cell
exit potential 5 eV. Data were acquired in the negative ion MS and
MS/MS modes.
Moringa oleifera, glucosinolates, PCA, LC-MS
J. Mass Spectrom. 2014, 49, 900–910 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms
901
HPLC-ESI-MS and HPLC-ESI-MS/MS analyses
Qualitative on-line HPLC-ESI-MS/MS analysis of extracts was per-
formed using an HPLC system interfaced to an Applied Biosystems
(Foster City, CA, USA) API3200 Q-Trap instrument in ion trap mode.
LC analyses were conducted using a system equipped with a 200
binary pump (Perkin-Elmer, USA). Samples were injected (10 μl) into
aLunaC
18
column (Phenomenex, USA) (150 × 2.1 mm i.d., 5 μmd)
and eluted at a flow rate of 0.3 ml min
1
. Mobile phase A was
H
2
O containing 0.1% formic acid, while mobile phase B was aceto-
nitrile containing 0.1% formic acid. Elution was carried out using a
gradient commencing at 100% A (gradient:1) and changing to
20:80 (A:B) in 55min (gradient: 4), then from 20:80 (A:B) to 100%:B
in 5 min (gradient: 1). The column was kept at 25 °C, using a Peltier
Column Oven Series 200 (Perkin Elmer). The flow from the chro-
matograph was injected directly into the ESI source. Qualitative
analysis of the compounds was performed using IDA (information
dependent acquisition). The IDA method created included an IDA
criteria (specify the charge state, mass range), enhanced MS scan,
enhanced resolution, enhanced product ion scan or MS/MS scan.
The source temperature was held at 450°C, and MS parameters
were those optimized for the ESI-MS and ESI-MS/MS analyses with
ion spray voltage at 4500. MS data were acquired using the
software provided by the manufacturer (Analyst software 1.5.1),
and extracted ion fragmentograms (XIC) were elaborated in order
to identify glucosinolates from their deprotonated molecular ions
and retention time.
Quantitative on-line HPLC-ESI-MS/MS analyses were performed
using the same LC-ESI-MS/MS equipment, but the mass spectrom-
eter worked with triple quadrupole analyzer in Multiple Reaction
Monitoring (MRM) mode. Elution was carried out using a gradient
commencing at 98% A (gradient:1) and changing to 88:12 (A:B) in
5 min (gradient: 4), then from 88:12 (A:B) to 75:25 (A:B) in 21min
(gradient: 1).
The API 3200 ES source was operated in negative ion mode
and was tuned by infusing solutions of standards (1 μgμl
1
in
methanol 50%) into the source at a flow rate of 10 μlmin
1
.
The optimized parameters, fragmentation reactions selected
for each compound, dwell time and retention times were
reported in Table 1. The voltage applied was 4500. Data
acquisition and processing were performed using Analyst
software 1.5.1.
Principal component analysis procedure
A m × n matrix (where m is the number of samples, and n is
the number of variables) was used in PCA. For the flow injection
ESI-MS matrix construction, the mass spectra were expressed as
the intensities of the individual [M H]
ions (variables) of the
most intense ions in the fingerprint of each sample. The data were
autoscaled and PCA was run. Thus, quantitative data of each chem-
ical marker were used to define a data set with 12 observation and
545 variables. The resulting metabolomics data were processed
using SIMCA P+ software 12.0 (Umetrix AB, Umea, Sweden).
Calibration and quantification of glucosinolates
In order to prepare the calibration plot, a sample (1 mg) of each
standard was weighted accurately into a 1-ml volumetric flask,
dissolved in methanol 50% (v/v) and the volume made up to the
mark with methanol. The resulting stock solution was diluted with
methanol in order to obtain reference solutions containing 0.5, 1,
5, 10, 25, 50, 100 and 200 μgml
1
of external standards.
A suitable amount of Internal Standard (IS), namely glucobar-
barin for glucomoringin, glucosinalbin and glucotropaeolin deter-
mination, and glucoraphenin for glucoraphanin and glucoiberin
analysis, was added to each reference solution to give a final
concentration of 20 μgml
1
and 5 μgml
1
for glucobarbarin and
glucoraphenin, respectively. Calibration curves were constructed
by analyzing reference/IS solutions in triplicate at each concentra-
tion level. The ratios of the peak areas of the external standard to
those of the IS were calculated and plotted against the correspond-
ing standard concentration using weighted linear regression to
generate standard curves. All quantitative data were elaborated
with the aid of Analyst software (Applied Biosystems).
Method validation
LC-MS/MS method was validated according to the European Med-
icines Agency (EMEA) guidelines relating to the validation of
Table 1. LC-MS/MS conditions for quantitation of glucosinolates by negative ion MRM
Compound t
R
Precursor ion
[M H]
Product ion
[A H]
Dwell
time (ms)
Declustering
potential
Entrance
potential
Collision
energy
Collision cell
exit potential
Glucomoringin 11 570 97 60 73.6 6.1 56.4 1.13
3-Hydroxy-4-(α-L-rhamnopyranosyloxy)benzyl
glucosinolate
10.74 586 97 60 78 657 4
4-(2′-O-Acetyl-α-L-rhamnopyranosyloxy)benzyl
glucosinolate or4-(3′-O-Acetyl-α-L-
rhamnopyranosyloxy)benzyl glucosinolate
or4-(4′-O-Acetyl-α-L-rhamnopyranosyloxy)
benzyl glucosinolate
15.45 612 259 100 76 11 41 5.6
16.01
19.85
unknown (m/z 912) 10.8 912 570 200 49 4.1 30 8
11.3
Glucotropaeolin 17.6 408 328 60 44 420 9
Glucosinalbin 11.6 424 97 60 61.5 4.34 37 1.8
Glucoraphanin 9.3 436 178 60 51 537.8 4
Glucoiberin 8.8 422 358 60 48.5 426 11.6
Glucobarbarin (I.S.) 16.9 438 97 60 55 7.7 40 1.7
Glucoraphenin (I.S.) 9.6 434 259 60 56 4.8 35.1 6.5
M. Maldini et al.
wileyonlinelibrary.com/journal/jms Copyright © 2014 John Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 900–910
902
analytical methods.
[28]
Precision was evaluated at three concentra-
tion levels for each compound through triplicate intra-day assays
and inter-day assays over 3 days;the intra-day precision (coefficient
of variance) was within 8%, while the inter-day was within 9% for all
analytes (Table 2). Specificity was defined as the non-interference
by other analytes detected in the region of interest. As regards
the LC-MS/MS method, which was developed on the basis of the
characteristic fragmentation of detected glucosinolates, no other
peaks interfered with the analytes in the MS/MS detection mode.
Accuracy of the analytical procedure was evaluated using the
recovery test. Pulp seed samples were added with three different
amounts of the eight standards, and recoveries were calculated
from the difference between the amount of analytes measured
before and after standards addition. The mean recoveries for each
standard and each concentration level are reported in Table 2.
The calibration graphs, obtained by plotting the area ratio between
the external and internal standards against the known concentra-
tion of external standards, were linear in the range used for the
analysis of all glucosinolates. The sensitivity of the method was
determined with respect to limit of quantification (LOD) and limit
of detection (LOQ). The LOQ (equivalent to sensitivity of the
quantitative method), defined as the lowest concentration of ana-
lyte that could be quantified with acceptable accuracy and
precision, was estimated by injecting a series of increasingly dilute
standard solutions until the signal-to-noise ratio was reduced to 10.
The LOD (equivalent to sensitivity of the qualitative method),
defined as the lowest concentration of analyte that could be
detected, was estimated by injecting a series of increasingly dilute
standard solutions until the signal-to-noise ratio was reduced to 2.
Linearity (calibration curves equations and regression), together
with LOQ and LOD for each of the five compounds analyzed, is
reported in Table 3.
Results
ESI-MS, ESI-MS/MS analyses and PCA.
Direct infusion electrospray ionization mass spectrometry in the
negative ion mode was initially used to obtain the fingerprints of
the extracts of the following different tissues: seed pulp and coat,
and leaves and root of 12-day-old seedlings of M. oleifera.Thefull
spectrum of each sample was recorded in triplicate with the aim
to rapidly provide a visual and statistical evaluation of similarities
and differences of secondary metabolites among tissues. The ESI-
MS fingerprints of samples were very characteristic, thus showing
distinctive sets of polar markers for each different tissue. Consider-
ing the large amount of data set obtained by negative ion ESI-MS
fingerprints of extracts under investigation, a chemometric ap-
proach was performed using PCA, to characterize the different
plant’s tissue and to evaluate differences in terms of metabolites.
The PCA, using SIMCA-P
+
Software, was applied to the matrix
obtained as described in the experimental section, resulting in the in-
tensity of each compared signal which shows the quantitative level
of each marker compound in each of the overall 12 samples
(4 tissues × 3 replicates). PCA is an unsupervised method and was used
to reduce the dataset in order to obtain the maximum variation
between the samples. Pareto scaling was chosen for scaling data.
Figure 1 shows the 2D projection plot of the 12 M. oleifera sam-
ples. The first component (R1X) explained the 57% of variance
whilst, andthe second (R2X) the 16%. Principal component’s choice
was done on the basis of the fitting (R
2
X) and predictive (Q
2
X)
values for the PCA model. In our case, the second component gave
the closest value to 1 for both of them. Variance was evaluated by
significance level for Hotelling’sT2.
The 2D diagram showed confined cluster areas, representing
each a link to a different part of the plant extracted; in fact, we
can observe separated regions related to samples of pulp seed,
seed coat, root and leaves.
In order to evaluate the influence of each variable on the clas-
sification of the samples, the loading plot obtained for the same
dataset was then studied, and it is reported in Fig. 2. For each
region of the 2D space, the loading plot shows the m/z values
corresponding to the peak observed in the specificsamples.In
Table 2. Accuracy and precision of eight analytes at three concentra-
tion levels
Compound Concentration Accuracy Precision
Intra-day
(CV%)
Precision
Inter-day
(CV%)
(μg/ml) (% recovery)
Glucomoringin 5 112 1.8 3
50 101 1.2 4.3
100 99 4.7 3.6
Glucotro paeolin 1 87 8 8
10 99 2.4 1
50 98 4.4 2.2
Glucosinalbin 1 112 5.3 6.4
10 113 2.7 3.2
50 98 3.1 0.3
Glucoraphanin 1 109 6.3 4.6
10 110 5.1 2.2
50 101 3.6 9.8
Glucoiberin 1 106 4.1 8.2
10 110 3.5 6
50 100 5.4 1.9
Precision and accuracy were evaluated at three concentration levels for
each compound through triplicate intra-day assays and inter-day assays
over 3 days
Table 3. Linearity, LOQ and LOD of LC-ESI-QqQ-MS/MS MRM method for the analysis of standard compounds
Compound Calibration curve R
2
LOQ LOD
equation (μgml
1
)(μgml
1
)
Glucomoringin y = 0.0757x + 0.705 0.993 0.02 0.005
Glucotropaeolin y = 0.0241x + 0.0156 0.999 0.05 0. 0015
Glucosinalbin y = 0.149x0.00236 0.999 0.011 0.005
Glucoraphanin y = 0.883x+ 1.72 0.998 0.06 0.009
Glucoiberin y = 0.565x + 1.07 0.994 0.0019 0.05
Moringa oleifera, glucosinolates, PCA, LC-MS
J. Mass Spectrom. 2014, 49, 900–910 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms
903
Figure 1. Principal component analysis score plot.
Figure 2. Principal component analysis loading plot.
M. Maldini et al.
wileyonlinelibrary.com/journal/jms Copyright © 2014 John Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 900–910
904
particular, the variables that contribute most to the differentia-
tion of the samples and to their location in a specific area of
the space can be highlighted. Interesting, values like m/z 288.9
and m/z 577.0 are localized in the area corresponding to the
seed coat in the score plot, whereas values like m/z 570, m/z
612, m/z 407.9 and m/z 912, are confined in the area corre-
sponding to pulp seed, and leaves and roots of 12-day-old seed-
lings in the score plot.
As evidenced by PCA, the ESI-MS fingerprint obtained for seed
coat extract showed the [M H]
ions at m/z values of 289, 577
and 865 corresponding at catechin/epicatechin and dimeric and tri-
meric procyanidins, respectively (Fig. 3A). We confirmed the
presence of these compounds by opportune ESI-MS/MS experi-
ments (data not shown). To our knowledge, this is the first study
showing the profiling of proanthocyanidins in seed of M. oleifera;
in fact, previous published works reported just the total
proanthocyanidin content.
[29]
The full negative mass spectrum recorded for seed pulp ex-
tract (Fig. 3B) revealed a major intense ion peak at m/z of 570,
relative to the glucosinolate glucomoringin together with 3-
hydroxy-4-(α-L-rhamnopyranosyloxy)benzyl glucosinolate (m/z 586)
and three glucosinolates with close structure similarity to
glucomoringin, except for the presence of an acetyl group in the
compound located at C-2′, C-3′and C-4′on the α-L-
rhamnopyranosyl unit (m/z 612). The identity of the revealed
glucosinolates was verified by the comparison of the MS
2
spectra
recorded for each compound with those of the standards and/or
with those reported in literature.
Anothermostabundantunknownionpeakwasevidenced
in this spectrum at m/z value of 912. Fragmentation experi-
ments were performed with the aim to individuate the nature
of this compound. MS
2
spectrum (Fig. 4A) evidenced only one
major ion peak at m/z value of 570 corresponding to the
deprotonated ion of glucomoringin. Sequential MS
3
spectrum
of m/z 570 (Fig. 4B) displayed most intense fragment ions at
values of m/z 424, 328, 275 and 259. The first two were
presumably generated by the subsequent loss of a
rhamnopyranosyloxyl moiety [M H146]
,followedbythe
loss of sulfate ion [M H146 96]
. The last two fragments
resulted to be characteristic diagnostic ions typical of frag-
mentation pattern of glucosinolates. Comparing these results
with the fragmentation pattern obtained for glucomoringin
standard in using ESI-QqQ-MS and ESI-QqQ-MS/MS (Fig. 4C),
we could hypothesize that the ion at m/z 912 is a glucosino-
late structurally correlated to glucomoringin. Further studies
are needed to clearly identify this compound.
Figure 3. ESI-MS (negative ion mode) fingerprints of coat (A) and pulp seed (B).
Moringa oleifera, glucosinolates, PCA, LC-MS
J. Mass Spectrom. 2014, 49, 900–910 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms
905
The negative ion ESI-MS spectra of leaves and roots extracts were
more complicated, suggesting the high complexity of the mixtures
analyzed (data not shown).
The deprotonated spectrum of roots extract showed over all a
most intense peak at m/z of 408 corresponding to glucosinolate
glucotropaeolin, along with minor ion relative to glucomoringin
(m/z 570). The comparison of the MS
2
spectra recorded for each
compound with those of the standards confirmed the nature of
revealed compounds.
The ESI-MS spectrum of leaves extract showed, along with
major ion peak of glucomoringin (m/z 570), the acetyl-α-L-
rhamnopyranosyloxy-benzyl glucosinolate (m/z 612), glucotro-
paeolin (m/z 408)andanothermostabundantionatm/z
value of 353, ascribable to chlorogenic acid. The identity of
this compound was confirmed by the comparison of MS/MS
spectrum with that of the standard. As evidenced by PCA
(Fig. 2), this variable contributes to the differentiation of
leaves sample.
Moreover, negative ESI-MS spectrum evidenced other minor ion
peaks at m/z value of 447, 463, 593 and 609 corresponding to
kaempferol 3-O-β-glucoside, quercetin 3-O-β-glucoside, kaempferol
3-O-rutinoside and rutin, respectively. The identity of the revealed
phenolic compounds was verified by the comparison of the MS
2
spectra with those of the corresponding standards and/or with
those reported in literature.
[9,15,19–21]
By using IDA software, phenolics found in M. oleifera leaves were
characterized according to their retention time and their MS and
MS/MS spectra and by comparison with standard reference com-
pounds, when available (data not shown). Besides confirming the
presence of the revealed compounds, LC-ESI-MS/MS analysis
allowed to identify three chlorogenic acid isomers (m/z 353).
Focusing on glucosinolates, an opportune IDA method with
EMS survey scans, ER and EPI scans was developed to clearly
identify these compounds by comparison of both their MS
2
and
retention times with those observed for the analytical standards
in LC-ESI-MS/MS analyses (data not shown). Besides
glucomoringin, glucotropaeolin (benzyl glucosinolate), 3-hydroxy-
4-(α-L-rhamnopyranosyloxy)benzyl glucosinolate, 4-(-2′-O-acetyl-α-
L-rhamnopyranosyloxy)benzyl glucosinolate, 4-(-3′-O-acetyl-α-L-
rhamnopyranosyloxy)benzyl glucosinolate and 4-(-4′-O-acetyl-α-L-
rhamnopyranosyloxy)benzyl glucosinolate, LC-ESI-MS/MS analyses
allowed to evidence the presence of other glucosinolates at
m/z values of 424, 422 and 436. The comparison of MS
2
spectra and the retention times with those of reference
standards allowed us to identify these compounds as
glucosinalbin (4-hydroxybenzyl glucosinolate), glucoiberin
Figure 4. ESI-IT-MS/MS (A), ESI-IT-MS
3
(B) spectra of compound at m/z 912 and ESI-IT-MS/MS (C) of glucomoringin.
M. Maldini et al.
wileyonlinelibrary.com/journal/jms Copyright © 2014 John Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 900–910
906
(3-methylsulfinylpropyl glucosinolate) and glucoraphanin
(4-methylsulfinylbutyl glucosinolate), respectively. Except for
glucosinalbin, these latter glucosinolates have never been
previously reported in M. oleifera.
In order to obtain accurate data concerning the amounts of
revealed glucosinolates in different tissue extracts, a quantitative
LC-ESI/MS (MRM) analysis was performed using a method previ-
ously described with the opportune modifications.
[18]
MS/MS spectra of glucomoringin and 3-hydroxy-4-(α-L-rhamno-
pyranosyloxy)benzyl glucosinolate showed the most intense
peak at specificproductionatm/z value of 97 corresponding to
the [SO
4
H]
ion, while the MS
2
spectra of 4-(O-acetyl-α-L-
rhamnopyranosyloxy)benzyl glucosinolate and glucosinalbin
showed a major ion peak at m/z value of 259 due to a sulfated
glucose moiety. For the unknown compounds at m/z 912, the only
intense product ion generated was the ion at m/z 570.
On the basis of the results, the specific transitions from
deprotonated molecular ions to the corresponding fragment
ions for each compound were selected to develop an
MRM method. IS (internal standards) were characterized by
MRM through the transitions from precursor ion m/z 438.0
to product ion m/z 97.0 for glucobarbarin and from precursor
ion m/z 434.0 to product ion m/z 259.0 for glucoraphenin.
The structures of detected compounds and selected IS are
reported in Fig. 5. The calibration curves, obtained by
plotting the area ratios between the external standards
and internal standards against the known concentration of
each compound, were linear in the range of 0.5–200 μgml
1
with r
2
values of >0.993 (Table 3). Retention times and
selected transitions for analyzed compounds are reported in
Table 1.
The method based on the characteristic fragmentation reac-
tions of glucosinolates was highly specific with no other peaks
interfering at the retention times of the marker compounds in
the MRM chromatograms (Fig. 6).
The quantitative analysis results are summarized in Table 4. It is
possible to observe that glucomoringin is the most abundant glu-
cosinolate in all parts of M. oleifera, particularly in the pulp seed,
followed by 3-hydroxy-4-(α-L-rhamnopyranosyloxy)benzyl gluco-
sinolate in the pulp seed and seed coat, and by glucotropaeolin
in the leaves. In the pulp, we can also evidence, for the first time
in M. oleifera, the presence of glucoraphanin and glucosinalbin
and minor quantities of 4-(-2′-O-acetyl-α-L-rhamnopyranosyloxy)
benzyl glucosinolate, 4-(-3′-O-acetyl-α-L-rhamnopyranosyloxy)ben-
zyl glucosinolate, 4-(-4′-O-acetyl-α-L-rhamnopyranosyloxy)benzyl
glucosinolate, glucotropaeolin, glucoiberin and in addition two un-
identified compounds, most likely glucosinolates.
To our knowledge, this is the first paper showing a direct
quantitative determination of predominant glucomoringin
and the other glucosinolates in seeds and different tissues
of M. oleifera 12-day-old seedlings. Previously, Bennett
et al.
[9,17]
reported the identification (by a LC-MS method)
and the indirect quantification (by LC-UV method) of the
major glucosinolates present in seeds
[16]
and in tissues of 1-
year-old plants of M. oleifera.
[9]
In another study, performed
by Amaglo et al.
[15]
a direct quantitative analyses of
glucomoringin and congeners in 100-, 320- and 380-day-old
plants is reported. Furthermore, the study of Bellostas et al.
[30]
measured the total glucosinolate level in leaves of 2–3-year-
old plants of three different Moringa species.
Despite showing similar patterns, the levels of glucosinolates ob-
tained in our study are different from those already reported.
[15,30]
These differences could be attributed to several reasons. In
addition to the different analytical methods used, difference in
the varieties examined, the growth conditions, as well as the plant
health and nutrition could represent important factors.
Figure 5. Molecular structure of glucosinolates.
Moringa oleifera, glucosinolates, PCA, LC-MS
J. Mass Spectrom. 2014, 49, 900–910 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms
907
Figure 6. LC/ESI(QqQ)/MS/MS XICs (extracted ion chromatograms) of MRM analysis of glucosinolates in M. oleifera.
M. Maldini et al.
wileyonlinelibrary.com/journal/jms Copyright © 2014 John Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 900–910
908
Our data demonstrated that seed and seedlings of M. oleifera
can represent a good source of glucomoringin. This com-
pound, as well as its acetyl derivatives and hydrolyzed products,
can exert a broad biological activity, from antimicrobial to
antiproliferative properties,
[5,31–36]
as well as an effective
anticarcinogenic activity.
[14,36–39]
Conclusion
This paper describes for the first time the MS profiling of
proanthocyanidins in seed coat and the simultaneous determina-
tion of 11 glucosinolates in different M. oleifera tissues by using
mass spectrometric approaches.
ESI-MS and ESI-MS/MS fingerprints of seed coat have not been
previously performed and allowed us to emphasize the presence
of dimeric and trimeric proanthocyanidins together with the
related monomer (catechin/epicatechin).
Furthermore, the use of the full ESI-MS spectra along with the
PCA approach proved to be a potentially useful and effective tool
to rapidly provide both visual and statistical evaluation of similari-
ties and differences in M. oleifera tissues.
The LC-ESI-MS/MS IDA method allowed us to individuate two
glucosinolates (glucoiberin and glucoraphanin) never reported in
M. oleifera before.Moreover, an LC–MS/MS MRM method allowed
us to quantify all the identified glucosinolates in different tissues
of M. oleifera.
M. oleifera pulp resulted a very rich source of glucosinolates, in
particular of glucomoringin, an uncommon member of the
glucosinolate family with promising antimicrobial and anti-
carcinogenic properties. Finding of phenolic compounds is
interesting because they are active antioxidant components of
M. oleifera responsible for its anti-inflammatory, atherosclerotic
and antidiabetic activities.
Acknowledgements
This work was supported by grants P.O.R. SARDEGNA F.S.E.
2007–2013—Obiettivo competitività regionale e occupazione,
Asse IV Capitale umano, Linea di Attività l.3.1., Fondazione Banco
di Sardegna and EMAP (Edible Medicinal Aromatic Plants) IRSES
PROGRAMME—Call ID ‘FP7-PEOPLE-IRSES-2009’.
We thank Dr. Renato Iori (Consiglio per la Ricerca e la
sperimentazione in Agricoltura, CRA-CIN, Bologna, Italy) for his kind
support and all the volunteers for their participation.
References
[1] L. O. Manguro, P. Lemmen. Phenolics of Moringa oleifera leaves. Nat.
Prod. Res. 2007,21, 56.
[2] Y. Kashiwada, F. A. Ahmed, S. Kurimoto, S. Y. Kim, H. Shibata, T. Fujioka,
Y. Takaishi. New α-glucosides of caffeoyl quinic acid from the leaves of
Moringa oleifera Lam. J. Nat. Med. 2012,66, 217.
[3] S. C. Verma, R. Banerji, G. Misra, S. K. Nigam. Nutritional value of
Moringa.Curr. Sci. 1976,45, 769.
[4] O. A. Oluduro, T. O. Idowu, B. I. Aderiye, O. Famurewa, O. O. Omoboye.
Evaluation of Antibacterial Potential of Crude Extract of Moringa
oleifera seed on Orthopaedics Wound Isolates and Characterization of
Phenylmethanamine and Benzyl Isothiocyanate Derivatives. Res. J.
Med. Plant. 2012,6,383.
[5] S. Cheenpracha, E. J. Park, W. Y. Yoshida, C. Barit, M. Wall, J. M. Pezzuto,
L. C. Chang. Potential anti-inflammatory phenolic glycosides from the
medicinal plant Moringa oleifera fruits. Bioorg. Med. Chem. 2010,18,
6598.
[6] F. Anwar, S. Latif, M. Ashraf, A. H. Gilani. Moringa oleifera: a food plant
with multiple medicinal uses. Phytother. Res. 2007,21, 17.
[7] M. Mbikay. Therapeutic Potential of Moringa oleifera Leaves in
Chronic Hyperglycemia and Dyslipidemia: A Review. Front
Pharmacol. 2012,3, 24.
[8] S. Sreelatha, P. R. Padma. Antioxidant activity and total phenolic
content of Moringa oleifera leaves in two stages of maturity. Plant
Foods Hum. Nutr. 2009,64, 303.
[9] R. N. Bennett, F. A. Mellon, N. Foidl, J. H. Pratt, M. S. Dupont, L. Perkins,
P. A. Kroon. Profiling glucosinolates and phenolics in vegetative and
reproductive tissues of the multi-purpose trees Moringa oleifera L.
(horseradish tree) and Moringa stenopetala L. J. Agric. Food Chem.
2003,51, 3546.
[10] S. Faizi, B. S. Siddiqui, R. Saleem, F. Noor, S. Husnain. Isolation and
structure elucidation of a novel glycoside niazidin from the pods of
Moringa oleifera,J. Nat. Prod. 1997,60, 1317.
[11] N. Agerbirk, C. E. Olsen. Glucosinolatestructures in evolution. Phytoche.
m2012,77, 16.
[12] A. Smiechowska, A. Bartoszek, J. Namiesnik. Determination of
glucosinolates and their decomposition products-indoles and
isothiocyanates in cruciferous vegetables. Crit. Rev. Anal. Chem. 2010,
40, 202.
[13] M. Jahangir, I. B. Abdel-Farid, H. K. Kim, Y. H.Choi, R. Verpoorte. Healthy
and unhealthy plants: The effect of stress on the metabolism of
Brassicaceae. Environ. Exp. Bot. 2009,67, 23.
[14] D. Brunelli, M. Tavecchio, C. Falcioni, R. Frapolli, E. Erba, R. Iori, P. Rollin,
J. Barillari, C. Manzotti, P. Morazzoni, M. D’Incalci. The isothiocyanate
Table 4. Quantitative results for glucosinolates detected in extracts of seed pulp, seed coat, roots and leaves of 12-day-old seedlings of Moringa oleifera
Glucosinolates (mg/100 g)
Compound Pulp Coat Roots Leaves
Glucomoringin 8619.44 ± 573.20 28.27 ± 0.6 3.99 ± 0.47 77.7 ± 8.07
3-Hydroxy-4-(α-L-rhamnopyranosyloxy)benzyl glucosinolate
a
65.75 ± 10.85 3.3 ± 0.02 0.02 ± 0.01 0.45 ± 0.05
4-(2′-O-Acetyl-α-L-rhamnopyranosyloxy)benzyl glucosinolate
or 4-(3′-O-acetyl-α-L-rhamnopyranosyloxy)benzyl glucosinolate
or 4-(4′-O-acetyl-α-L-rhamnopyranosyloxy)benzyl glucosinolate
a
(t
R
15.45) 0.29 ± 0.03 ND 0.01 ± 0.01 0.50 ± 0.01
(t
R
16.01) 0.16 ± 0.01 ND 0.01 ± 0.01 0.35 ± 0.05
(t
R
19.85) 0.39 ± 0.11 ND 0.02 ± 0.01 2.22 ± 0.26
Unknown (m/z 912)
a
(t
R
10.8) 0. 14 ± 0.11 ND ND ND
(t
R
11.3) 0. 14 ± 0.05 ND ND ND
Glucotropaeolin ND ND 0.27 ± 0. 03 15.66± 1. 04
Glucosinalbin 3.17 ± 0.67 ND 0.05 ± 0.01 0.84 ± 0.19
Glucoraphanin 3.57 ± 0.32 0.86 ± 0.22 0.58 ± 0.03 2.2 ± 0.27
Glucoiberin 0.09± 0.02 ND 0.02 ± 0.01 0.05 ± 0.02
Each data is the mean of three replicates (mean ± SD)
ND: not detected
a
Quantified as equivalent of glucomoringin
Moringa oleifera, glucosinolates, PCA, LC-MS
J. Mass Spectrom. 2014, 49, 900–910 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms
909
produced from glucomoringin inhibits NF-kB and reduces myeloma
growth in nude mice in vivo. Biochem. Pharmacol. 2010,79, 1141.
[15] N.K.Amaglo,R.N.Bennett,R.B.LoCurto,E.A.S.Rosa,V.LoTurco,
A. Giuffrida, A. Lo Curto, F. Crea, G. M. Timpo. Profiling selected
phytochemicals and nutrients in different tissues of the multipurpose
tree Moringa oleifera L., grown in Ghana. Food Chem. 2010,122, 1047.
[16] B. Kusznierewicz, R. Iori, A. Piekarska, J. Namieśnik, A. Bartoszek.
Convenient identification of desulfoglucosinolates on the basis of
mass spectra obtained during liquid chromatography-diode array-
electrospray ionisation mass spectrometry analysis: method
verification for sprouts of different Brassicaceae species extracts.
J. Chromatogr. A 2013,1278, 108.
[17] R. N. Bennett, F. A. Mellon, P. A. Kroon, Screening crucifer seeds as
sources of specific intact glucosinolates using ion-pair high-
performance liquid chromatography negative ion electrospray mass
spectrometry, J. Agric. Food Chem. 2004,52, 428.
[18] M. Maldini, S. Baima, G. Morelli, C. Scaccini, F. Natella. A liquid
chromatography-mass spectrometry approach to study
"glucosinoloma" in broccoli sprouts. J. Mass Spectrom. 2012,47,1198.
[19] B. Vongsak, P. Sithisarn, W. Gritsanapan. Simultaneous HPLC
quantitative analysis of active compounds in leaves of Moringa
oleifera Lam. J. Chromatogr. Sci. 2013 4,1.
[20] L. O. A. Manguro, P. Lemmen. Phenolics of Moringa oleifera leaves. Nat.
Prod. Res. 2007,21, 56.
[21] Y. Kashiwada, F. A. Ahmed, S. Kurimoto, S.-Y. Kim, H. Shibata, T. Fujioka,
Y. Takaishi. New α-glucosides of caffeoyl quinic acid from the leaves of
Moringa oleifera Lam. J. Nat. Med. 2012,66, 217.
[22] A. R. Verma, M. Vijayakumar, C. S.Mathela, C.V. Rao. In vitro and in vivo
antioxidant properties of different fractions of Moringa oleifera leaves.
Food Chem. Toxicol. 2009,47, 2196.
[23] M. Ndong, M. Uehara, S. Katsumata, K. Suzuki. Effects of oral
administration of Moringa oleifera Lam. on glucose tolerance in
gotokakizaki and wistar rats. J Clin Biochem Nutr. 2007,40, 229.
[24] P. Chumark, P. Khunawat, Y. Sanvarinda, S. Phornchirasilp,
N. P. Morales, L. Phivthong-ngam, P. Ratanachamnong, S. Srisawat,
K. U. Pongrapeeporn. The in vitro and ex vivo antioxidant properties,
hypolipidaemic and antiatherosclerotic activities of water extract of
Moringa oleifera Lam. Leaves. J. Ethnopharmacol. 2008,116, 439.
[25] J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong, C. M. Whitehouse.
Electrospray Ionization for Mass-Spectrometry of Large Biomolecules.
Science 1989,246, 64.
[26] P. K. Hopke. The evolution of chemometrics. Anal. Chim. Acta 2003,
500, 65.
[27] European Economic Community, Commisssion Regulation, EEC No.
1864/90. Oilseeds—determination of glucosinolates high performance
liquid chromatography. Off. J. Eur. Comm. 1990,L170,27–34.
[28] EMEA. Quality guidelines: Validation of analytical procedures: Text and
methodology (ICH Q2). Available at: http://www.emea.europa. eu/pdfs/
human/ich/038195en.pdf.
[29] W. R. Compaoré, P. A. Nikièma, H. I. N. Bassolé, A. Savadogo,
J. Mouecoucou, D. J. Hounhouigan and S. A. Traoré. Chemical
Composition and Antioxidative Properties of Seeds of Moringa
oleifera and Pulps of Parkia biglobosa and Adansonia digitata
Commonly used in Food Fortification in Burkina Faso. Curr. Res. J. Biol.
Sci. 2011,3,64.
[30] N. Bellostas, J. C. Sørensen, A. Nikiema, H. Sørensen, D. Pasternak and
S. Kumar. Glucosinolates in leaves of Moringa species grown and
disseminated in Niger. Afr. J. Agric. Res. 2010,5, 1338.
[31] E. J. Park, S. Cheenpracha, L. C. Chang, T. P. Kondratyuk, J. M. Pezzuto.
Inhibition of lipopolysaccharide-induced cyclooxygenase-2 and
inducible nitric oxide synthase expression by 4-[(2′-O-acetyl-α-L-
rhamnosyloxy) benzyl] isothiocyanate from Moringa oleifera.Nutr.
Cancer 2011,63, 971.
[32] Y.M.Lee,M.R.Seon,H.J.Cho,J.S.Kim,J.H.Park.Benzylisothiocyanate
exhibits anti-inflammatory effects in murine macrophages and in
mouse skin. J. Mol. Med. 2009,87, 1251.
[33] N. Juge,R. F. Mithen,M. Traka. Molecular basis for chemoprevention by
sulforaphane: A comprehensive review. Cell. Mol. Life Sci. 2007,64,
1105.
[34] U. Leuck, H. Kunz. Synthesis of active principles from the leaves of
Moringa oleifera using S-pent-4-enyl thioglucosides. Carbohydr. Res.
1998,312,33.
[35] J. W. Fahey. Moringa oleifera: A review of the medical evidence for its
nutritional, Therapeutic, and prophylactic properties. Part 1. Trees for
Life J. www.TFLJournal.org, 2005.
[36] M. Galuppo, G. R. De Nicola, R. Iori, P. Dell’Utri, P. Bramanti,
E. Mazzon.Antibacterial activity of glucomoringin bioactivated
with myrosinase against two important pathogens affecting the
health of long-term patients in hospitals. Molecules 2013,18,
14340–14348.
[37] F. Anwar, S. Latif, M. Ashraf, A. H. Gilani. Moringa oleifera:a
food plant with multiple medicinal uses. Phytother. Res. 2007,
21, 17.
[38] A. P. Guevara, C. Vargas, H. Sakurai, Y. Fujiwara, K. Hashimoto, T. Maoka,
M. Kozuka, Y. Ito, H. Tokuda, H. Nishino. An antitumor promoter from
Moringa oleifera.Mutat. Res. 1999,440, 181.
[39] B. Dayal, V. R. Yannamreddy, R. Amin, M. A. Lea, A. B. Attygalle.
Bioactive Compounds in Moringa oleifera: isolation, structure
elucidation, and their antiproliferative properties. In Tropical and
Subtropical Fruits: Flavors, Color, and Health Benefits,B.Patil,etal.,
ACS Symposium Series; American Chemical Society: Washington,
DC, 2013.
M. Maldini et al.
wileyonlinelibrary.com/journal/jms Copyright © 2014 John Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 900–910
910