Isotopic labelling for the characterisation of HNE-sequestering agents in plant-based extracts and its application for the identification of anthocyanidins in black rice with giant embryo

Article abstract of DOI:10.1080/10715762.2018.1490735

Reactive carbonyl species (RCS) are cytotoxic molecules that originate from lipid peroxidation and sugar oxidation. Natural derivatives can be an attractive source of potential RCS scavenger. However, the lack of analytical methods to screen and identify bioactive compounds contained in complex matrices has hindered their identification. The sequestering actions of various rice extracts on RCS have been determined using ubiquitin and 4-hydroxy-2-nonenal (HNE) as a protein and RCS model, respectively. Black rice with giant embryo extract was found to be the most effective among various rice varieties. The identification of bioactive compounds was then carried out by an isotopic signature profile method using the characteristic isotopic ion cluster generated by the mixture of HNE: ²H₅-HNE mixed at a 1:1 stoichiometric ratio. An in-house database was used to obtain the structures of the possible bioactive components. The identified compounds were further confirmed as HNE sequestering agents through HPLC-UV analysis.

Full text of DOI:10.1080/10715762.2018.1490735

Free Radical Research
ISSN: 1071-5762 (Print) 1029-2470 (Online) Journal homepage: http://www.tandfonline.com/loi/ifra20
Isotopic labelling for the characterisation of
HNE-sequestering agents in plant-based extracts
and its application for the identification of
anthocyanidins in black rice with giant embryo
Mara Colzani, Luca Regazzoni, Angela Criscuolo, Giovanna Baron, Marina
Carini, Giulio Vistoli, Yoon-Mi Lee, Sang-Ik Han, Giancarlo Aldini & Kyung-Jin
Yeum
To cite this article: Mara Colzani, Luca Regazzoni, Angela Criscuolo, Giovanna Baron, Marina
Carini, Giulio Vistoli, Yoon-Mi Lee, Sang-Ik Han, Giancarlo Aldini & Kyung-Jin Yeum (2018):
Isotopic labelling for the characterisation of HNE-sequestering agents in plant-based extracts and
its application for the identification of anthocyanidins in black rice with giant embryo, Free Radical
Research, DOI: 10.1080/10715762.2018.1490735
To link to this article: https://doi.org/10.1080/10715762.2018.1490735
Published online: 23 Jul 2018.
Submit your article to this journal
Article views: 15
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ifra20

FREE RADICAL RESEARCH
https://doi.org/10.1080/10715762.2018.1490735
ARTICLE
Isotopic labelling for the characterisation of HNE-sequestering agents
in plant-based extracts and its application for the identification of
anthocyanidins in black rice with giant embryo
a
a
a
a
a
a
Mara Colzani , Luca Regazzoni , Angela Criscuolo , Giovanna Baron , Marina Carini , Giulio Vistoli ,
b
c
a
b
ꢀ
ꢀ
Yoon-Mi Lee , Sang-Ik Han , Giancarlo Aldini and Kyung-Jin Yeum
a
b
Department of Pharmaceutical Sciences, University of Milan, Milan, Italy; Division of Food Bioscience, College of Biomedical and
Health Sciences, Konkuk University, Chungju-si, South Korea; National Institute of Crop Science, Rural Development Administration,
Suwon-si, South Korea
c
ABSTRACT
ARTICLE HISTORY
Received 7 April 2018
Revised 13 June 2018
Accepted 14 June 2018
Reactive carbonyl species (RCS) are cytotoxic molecules that originate from lipid peroxidation
and sugar oxidation. Natural derivatives can be an attractive source of potential RCS scavenger.
However, the lack of analytical methods to screen and identify bioactive compounds contained
in complex matrices has hindered their identification. The sequestering actions of various rice
extracts on RCS have been determined using ubiquitin and 4-hydroxy-2-nonenal (HNE) as a
protein and RCS model, respectively. Black rice with giant embryo extract was found to be the
most effective among various rice varieties. The identification of bioactive compounds was then
KEYWORDS
Black rice with giant
embryo; HNE sequestering
agents; mass spectrometry
carried out by an isotopic signature profile method using the characteristic isotopic ion cluster
generated by the mixture of HNE: H
2
5
-HNE mixed at a 1:1 stoichiometric ratio. An in-house data-
base was used to obtain the structures of the possible bioactive components. The identified
compounds were further confirmed as HNE sequestering agents through HPLC-UV analysis.
Introduction
human diseases [4,5], RCS have been considered not
only as biomarkers but also as potential drug targets.
Among the damaging RCS, 4-hydroxy-2-nonenal (HNE)
is one of the most studied since its discovery by
Hermann Esterbauer in 1964, due to its abundance,
reactivity, and biological effects [6,7]. The strict link
between elevated HNE tissue/blood levels and some
human diseases suggests that HNE contributes towards
pathophysiology of these diseases [8]. Several studies
based on cell signalling and protein covalent modifica-
tion have since reported the molecular mechanisms
involved in the cytotoxic effect of HNE [9,10].
Reactive carbonyl species (RCS) are a class of break-
down products arising from the oxidation of sugars and
lipids which can be formed either exogenously, for
instance during food processing or cooking, and
endogenously, via different pathophysiological condi-
tions [1]. RCS are chemically quite heterogeneous,
belonging to different classes, including di-aldehydes,
keto-aldehydes, and a,b-unsaturated aldehydes [2]. RCS
have the common property of covalently reacting with
nucleophilic substrates such as proteins. Covalent
adducts originating from lipids are named advanced
lipoxidation end-products (ALEs), whereas RCS originat-
ing from sugars are known as advanced glycoxidation
end-products (AGEs) [1]. Since the 1980s, RCS and their
corresponding reaction products have been widely
used as markers of oxidative stress and several analyt-
ical methods for their measurements have been devel-
oped and widely applied [3]. More recently, due to a
growing body of evidence reporting the involvement of
AGEs and ALEs in the onset and propagation of some
Different molecular approaches for reducing the
overproduction of HNE and in general of RCS have
been reported and summarised in some recent reviews
[4,11,12]. Among the proposed approaches, the most
promising is based on small nucleophilic molecules
(RCS sequestering agents) which covalently react with
RCS, forming unreactive adducts which are then metab-
olised and excreted. Several RCS sequestering agents
have been proposed such as aminoguanidine, pyridox-
amine, hydralazine, and carnosine [13,14]. Their
CONTACT Giancarlo Aldini
giancarlo.aldini@unimi.it
Department of Pharmaceutical Sciences, University of Milan, Milan, Italy; Kyung-Jin Yeum
kyeum@kku.ac.kr
These authors contributed equally to this work.
Division of Food Bioscience, College of Biomedical and Health Sciences, Konkuk University, Chungju-si, South Korea
ꢀ
ß 2018 Informa UK Limited, trading as Taylor & Francis Group

2
M. COLZANI ET AL.
protective effects have been confirmed in animal mod-
els by several independent labs, further supporting the
promising therapeutic efficacy of targeting HNE
We have recently reported an in vitro high-resolution
MS method to test the ability of compounds, mixtures,
and extracts to trap RCS by measuring their efficacy in
inhibiting ubiquitin carbonylation induced by RCSs
including HNE [28]. Such a method was found suitable
to evaluate the overall quenching activity of extracts,
however it does not permit the identification of the
active components.
[8,13,15]. Carnosine has recently been reported as the
most efficient and selective sequestering agent of HNE
and its promising activity has been demonstrated in
animal models and most recently in obese subjects
[16–18]. However, the application of carnosine as an
HNE sequestering agent in humans is limited because
of its poor bioavailability, due to the presence of serum
carnosinases, which catalyse the hydrolytic cleavage of
the dipeptide [19]. Nowadays there is great interest in
the discovery of novel HNE sequestering agents such as
carnosine derivatives resistant to carnosinases [20,21].
Among others, an interesting discovery approach to
identify novel HNE sequestering agents is based on
searching for bioactive compounds in plant extracts
In this article, we report an LC-ESI-MS method based
on an isotopic labelling procedure which permits the
identification of the HNE sequestering agents contained
in a crude mixture. Together with the ubiquitin method,
it was then applied to search for HNE sequestering
agents contained in extracts prepared from different
varieties of rice.
Materials and methods
[22]. Plants are also subject to RCS stress, in particular
Chemicals
those mediated by a,b-unsaturated aldehydes with HNE
being the most abundant and toxic carbonyl com-
pound formed in stress conditions [23]. Because plants
are subject to RCS and protein carbonylation damage
Formic acid (HCOOH), sodium dihydrogen phosphate
(
NaH PO ꢁH O), sodium hydrogen phosphate (Na
2
4
2
2
HPO ꢁ2H O), phenylalanine, gamma-aminobutyric acid,
4
2
[24], they have developed efficient enzymatic and non-
histidine, tricin-5-O-glucoside, lyophilised ubiquitin
from bovine erythrocytes (BioUltra, ꢂ98%), and LC–MS
grade solvents were purchased from Sigma–Aldrich
enzymatic detoxification pathways. Hence, besides
detoxifying enzymes, it is reasonable to consider
the presence of secondary metabolites acting as
RCS-sequestering agents. Such a hypothesis has been
partially confirmed by previous studies which have
reported that EGCG and other polyphenols are effective
as sequestering agents of HNE [25] of other RCS, includ-
ing acrolein and methylglyoxal [26]. These results
prompted scientists to screen the sequestering activity
of different plant metabolites and, in particular, the
polyphenol class. As an example, Zhu et al. [27] com-
pared 21 natural polyphenols with diverse structural
characteristics for their ACR-/HNE-trapping capacities
under simulated physiological conditions.
To the best of our knowledge, the search for a
potential RCS natural sequestering agent has been car-
ried out by using isolated molecules or highly purified
extracts and not by applying off-target methods involv-
ing crude extracts. One reason is the lack of analytical
methods able to test the sequestering activity of com-
plex matrices and the inherent difficulties associated
with fishing out compounds that are bioactive. As an
example, one of the most used methods to evaluate
HNE sequestering efficacy is based on a HPLC-UV
method for measuring the residual amount of HNE.
Such an approach is clearly suitable for testing single
molecules and not mixtures, which would interfere with
HNE analysis.
(
Milan, Italy). Peonidin-3-O-glucoside, malvidin, peoni-
din, and pelargonidin were from Extrasynthese
Genay, France). Carnosine (b-alanyl-L-histidine) was a
(
generous gift from Flamma SpA (Chignolo d’Isola,
Bergamo, Italy). LC-grade H O (18 MX cm) was pre-
2
pared with a Milli-Q H O purification system (Millipore,
2
Bedford, MA). All other reagents were of analytical
grade. 4-Hydroxy-2-nonenal diethylacetal (HNE-DEA)
2
2
and H ꢃ 4-hydroxy-2-nonenal diethylacetal ( H -HNE-
5
5
DEA) were synthesised according to the literature [29]
ꢄ
and stored at ꢃ20 C. For each experiment, fresh HNE
was prepared starting from stored HNE-DEA, which was
evaporated under nitrogen stream and hydrolysed with
1
mM HCl, pH 3 for 1 h at room temperature to obtain
HNE. The concentration of HNE was estimated by meas-
uring the absorbance at k ¼ 224 nm (molar extinction
ꢃ1
ꢃ1
coefficient ¼ 13,750 M ꢅ cm ) [30].
Plant material
Black rice with giant embryo and white rice were
provided by the National Institute of Crop Science,
Rural Development Administration, Republic of Korea.
A detailed method for the development and culture of
black rice with giant embryo has been previously
reported [31].

FREE RADICAL RESEARCH
3
2
Rice extract preparation: the seeds were manually
dehulled with a wooden rice dehuller and ground to a
powder by using a mortar and pestle. The milled rice
powders were stored at ꢃ80 C until further use. Water
H5-HNE. LC-HRMS were run in positive and negative
ion mode.
ꢄ
extract of black rice with giant embryo was provided by
National Institute of Crop Science, Rural Development
Administration, Republic of Korea. Two hundred
grammes of black rice with giant embryo extract (BRGE)
was extracted with 200 mL of water for 5 days at room
temperature, and the extracts were filtered and evapo-
rated under vacuum, followed by drying. The yield of
extraction was 4.32%.
Chromatographic conditions
Chromatographic separation was performed on a
reversed-phase Agilent Zorbax SB-C18 column
(150 ꢅ 2.1 mm, i.d. 3.5 mm, CPS analitica, Milan, Italy),
protected by an Agilent Zorbax guard column, by an
UltiMate 3000 system (Dionex) equipped with an auto-
ꢄ
sampler kept at 4 C working at a constant flow rate
Rice extract filtration: for the ubiquitin assay as well
as for the isotopic labelling procedure, rice powder
extracts were then dissolved in water and ultrafiltered
by using a 3 kDa ultrafilters so to remove proteins and
high MW components. The yield of the low molecular
fraction after 3 kDa ultrafiltration was ꢆ10% of the start-
ing sample for all the rice varieties tested.
(
200 mL/min). Ten microlitres of sample was injected
into the column and the analytes were eluted with a
5-min multistep gradient of phase A H O-HCOOH
4
2
(100:0.1, v/v) and phase B CH CN-HCOOH (100:0.1, v/v):
3
0
–5 min, isocratic of 1% B; 5–15 min, from 1% B to 25%
B; 15–30 min, from 25% B to 65% B; 30–32 min, from
6
3
5% B to 80% B; 32–36 min, isocratic of 80% B;
6–36.1 min, from 80% B to 1% B, and then 36.1–45 min
of isocratic 1% B.
Ubiquitin carbonylation assay
The effect of rice extracts on HNE-induced protein car-
bonylation was tested by using the ubiquitin assay as
previously reported [28]. Briefly, ubiquitin (10 mM final
concentration) was chosen as the protein target of
HNE-induced protein adduction. Ubiquitin was incu-
MS conditions
The acquisitions were performed on a LTQ-Orbitrap XL
mass spectrometer (Thermo Fisher Scientific, San Jose)
using an ESI source, acquiring in positive and in nega-
tive ion mode. A real-time mass calibration was
obtained by using a list of 20 background ions [32]. The
source parameters used for the positive mode are:
spray voltage 3 kV, capillary temperature 275 C, capil-
lary voltage 49 V, sheath gas flow 15 units, auxiliary gas
flow 5 units, tube lens offset 130 V; for the negative ion
ꢄ
bated for 24 h at 37 C in 10 mM phosphate buffer in
the presence of 500 mM HNE together with rice extracts
of increasing concentrations (10, 25, and 50 mg/mL). To
remove high molecular components, which could inter-
fere with the assay, the rice extracts were filtered by
using Amicon YM3 filters (Millipore, Milan, Italy) and the
filtered fractions were then used for the assay. After an
incubation time of 24 h, the reactions were stopped by
centrifugation using Amicon YM3 filters and the extent
of ubiquitin carbonylation was determined by MS intact
protein analysis using the microflow automated loop
injection procedure as already described [28].
ꢄ
mode: spray voltage 3.5 kV, capillary temperature
ꢄ
2
75 C, capillary voltage ꢃ17 V, sheath gas flow 15
units, auxiliary gas flow 5 units, tube lens offset ꢃ100 V.
Full MS spectra were acquired in profile mode by the
FT analyser in a scan range of 100–1000 m/z, using AGC
5
scan target 5 ꢅ 10 and resolution 100,000 FWHM at
Identifying the HNE sequestering agents by an
isotopic labelling procedure
m/z 400. Tandem mass spectra were acquired by the
linear ion trap (LTQ) that was set up to fragment the
3
three most intense ions exceeding 1 ꢅ 10 counts. Mass
After filtration by using Amicon YM3 filters, the black
rice giant embryo extract (BRGE) was dissolved in PBS
at a final concentration of 50 mg/mL and was then incu-
acquisition settings were: centroid mode, AGC scan tar-
4
get 1 ꢅ 10 , precursor ion isolation width of m/z 2.5,
bated for 24 hours in presence of HNE or deuterium
and collision energy (CID) of 30 eV. Dynamic exclusion
was enabled to reduce redundant spectra acquisition:
two repeat counts, 20 s repeat duration, 30 s of exclu-
sion duration. Instrument control and spectra analysis
were provided by the software Xcalibur 2.0.7 and
Chromeleon Xpress 6.80.
2
labelled HNE ( H -HNE). Control samples were incu-
5
bated neither with HNE nor with deuterium labelled
HNE. After 24 h, the samples were analysed individually
by LC-MS together with a sample prepared by mixing
in a 1:1 v/v ratio the samples incubated with HNE and

4
M. COLZANI ET AL.
HPLC-UV HNE sequestering assay
6–31 G basis set as implemented by the GAMESS soft-
ware [34].
The compounds identified as putative HNE binders with
the isotopic labelling procedure were purchased as
pure standards and tested individually for their HNE
sequestering ability. The assay was a 3-hour time course
Results and discussion
Overview of the approach
ꢄ
experiment at pH 7.4 and 37 C, in the presence of 50
RCS are well-known pathogenetic factors of various oxi-
dative stress associated diseases [5]. Recently, molecular
approaches based on detoxification of RCS have been
found to be effective in preventing the onset and pro-
gression of some diseases in different animal models
mM HNE and 1 mM of the tested compound. The
residual HNE was measured hourly within a 3-hour reac-
tion time by using the HPLC-UV assay described by
Vistoli et al. [33] with some minor modifications. Briefly,
HNE concentration was determined by a Surveyor HPLC
platform (Thermo Scientific, Milan, Italy) setting the UV
[11]. Among such efforts, those based on small nucleo-
philic compounds capable of trapping and detoxifying
RCS are of particular interest. Several synthetic and nat-
ural RCS sequestering agents have so far been reported
to be effective in several animal models, but for many
of these compounds, clinical application is limited due
to their ill-defined activity, lack of selectivity, and of
suitable bioavailability. Hence, a discovery approach
aimed at identifying novel RCS sequestering agents is
necessary to overcome this gap. In this study, we report
a discovery approach aimed at identifying RCS seques-
tering agents in crude mixtures such as natural extracts.
Natural sources and, in particular, plant extracts, repre-
sent an attractive source of RCS sequestering agents
due to the fact that plants are also subject to RCS
stress, in particular mediated by a,b-unsaturated alde-
hydes, and that they have developed efficient enzym-
atic and nonenzymatic detoxification pathways [23,35].
Many nonenzymatic pathways and in particular those
based on small nucleophilic agents are still unexplored
and they would represent an interesting potential
source of novel bioactive compounds. However, search-
ing for RCS sequestering agents in a crude mixture is
quite a challenging process because they are usually
present as minor components with respect to the com-
plex matrix, making their identification and character-
isation analytically challenging. Moreover, a robust
method capable of measuring the overall RCS seques-
tering activity is needed in order to select and identify
the most effective plant extracts.
detector at 224 nm. HNE elution was performed in
ꢄ
6
minutes by reverse phase chromatography at 37 C
and flow rate of 0.3 mL/min A Kinetex column
(
75 ꢅ 2.1 mm i.d., 2.6-mm particle size, 100-Å porosity,
Phenomenex, Castel Maggiore, Italy) was used as
stationary phase, while mobile phase consisted of H O-
2
CH CN-HCOOH, 80:20:0.1 (v/v/v).
3
The amount of reacted HNE was calculated by the
following formula:
½
HNEꢇ ꢃ½HNEꢇ
t0
t3
reacted HNE ð%Þ ¼
ꢀ 100
½
HNEꢇ
t0
[
HNE] being the concentration of HNE at the begin-
t0
ning of the incubation and [HNE] being the concentra-
t3
tion of HNE at the end of the incubation. Carnosine,
which is a HNE sequestering agent not containing a
thiol moiety, was used as positive control to assess the
assay reproducibility. To demonstrate that HNE con-
sumption over time is not caused by auto-oxidation, an
incubation in pure phosphate buffer was performed as
negative control.
The activity of each compound was also given as car-
nosine units (i.e. the ratio between the % of HNE
reacted with the compound at a given time and the %
of HNE reacted with carnosine in the same span
of time).
Computational studies
Here we report a new method to overcome these
challenges that is based on the following steps: (i) the
RCS sequestering activities of the tested extracts are
first characterised by using the ubiquitin assay; (ii) the
most active extracts are then selected in order to iden-
tify sequestering agents by using an innovative isotopic
labelling procedure, and (iii) the RCS sequestering activ-
ities of the identified compounds are then validated by
using standard compounds and a HPLC-UV method.
The method proposed has been applied by using
HNE as a target aldehyde and eight rice extracts as
For simplicity, attention was focussed on two anthocya-
nidins chosen as representative of active (pelargonidin)
and inactive (malvidin) compounds. Given the tested
physiological pH, the compounds were simulated in
their quinoidal neutral form by considering three major
tautomers. In order to investigate their stereo-electronic
properties, these structures were minimised at their
ground state and at gas phase by density functional
theory (DFT) using the Becke three-parameter hybrid
function with LYP correlation (DFT/B3LYP) and with the

FREE RADICAL RESEARCH
5
potential sources of RCS sequestering agents. HNE is an
a,b-unsaturated aldehyde generated endogenously
and exogenously (it is present in food) by the radical-
mediated peroxidation of x-6 polyunsaturated fatty
acids [36]. Since its identification, 4-HNE, an autoxida-
tion product of unsaturated fats and oils (at first errone-
ously described as 4-hydroxy-octenal), has attracted
great scientific interest, as demonstrated by the publi-
cation of more than 4200 papers since 1980. Compared
to other 4-hydroxyalkenals, HNE is the most extensively
studied and reviewed as it was the first one discovered
the proposed method to test the RCS sequestering cap-
ability of rice and so we evaluated the water extracts of
white, brown, and black rice.
HNE sequestering activity of rice extracts
The overall HNE sequestering activity of eight rice
extracts was studied using the ubiquitin assay. The
method consists of incubating ubiquitin as protein tar-
get with HNE at a specific molar ratio and incubation
time, thus inducing ubiquitin covalent modification of
almost 50%. The protein carbonylation was monitored
by measuring the ratio of the areas of two peaks: one
at m/z 779.61239 relative to the 11-charged peak
(named z11) of native ubiquitin and the second at m/z
793.80415 corresponding to the 11-charged peak of
ubiquitin covalently modified by HNE. Figure 1 shows
the MS spectra of ubiquitin incubated in the absence
and in the presence of HNE. When ubiquitin incubated
with HNE was spiked with rice extracts, a dose-
dependent reduction of the peak area relative at m/z
793.80415 was observed. The IC50 value, which is the
rice extract concentration able to decrease by 50% the
level of protein carbonylation, was determined. Table 1
summarises the IC50 of the tested rice extracts. White
rice extract was found less effective in respect to col-
oured rice and of these, black rice with giant embryo
was the most effective and for this reason it was
[37], it represents the main 4-hydroxyalkenal formed
during the autoxidation of unsaturated fatty acids, it is
highly reactive and numerous biological effects have
been demonstrated [6,38], leading HNE being consid-
ered as a promising drug target [8].
Rice (Oryza sativa) is one of the five different cereal
grains, which are the most commonly consumed
throughout the world and are functionally reported to
have antioxidant activity. Bioactive components of rice,
have demonstrated antioxidant and anti-inflammatory
activities in cells and animals [39,40].
However, although the direct and indirect antioxi-
dant activity of rice has been demonstrated in different
in vitro and also in vivo models, no data to our know-
ledge have so far been reported regarding the efficacy
of rice towards protein carbonylation and, in particular,
its ability to sequester RCS. We then decided to apply
Figure 1. 11-charged MS peaks of ubiquitin incubated in the absence and in the presence of HNE and increasing concentration
of two extract: white rice extract (panel on the left) and BRGE (panel on the right).

6
M. COLZANI ET AL.
Table 1. HNE Sequestering activity of rice extracts.
Rice extract
Average IC50 (mg/mL)
SD
CV%
2.7
Black rice giant embryo (BRGE)
Black rice giant embryo bran
Miryang282
White rice (brown)
Black rice
Black rice giant embryo 2 (BRGE2)
Red rice
White rice (brown)
29.10
30.63
39.32
45.46
52.04
65.92
71.32
151.00
0.79
3.70 12.4
2.48
2.29
8.34 16
2.13
4.04
17.70 11.7
6.3
5
3.2
5.7
Activities are reported as IC50 which represents the concentration of the
tested compound able to reduce by 50% HNE-induced ubiquitin
carbonylation.
Δ mass
5.03088 Da
Rice
extract
HNE 2H5-HNE
Molecule
identity
Database
incubation
HRMS
Figure 2. Graphic representation of the method. The first
step consists of spiking the plant extract with HNE
2
(
156.11502 Da) and H -HNE (161.07590 Da) at the same molar
5
ratio. After incubation, the HNE adducts are identified based
on the typical isotopic pattern and the identity of the HNE
scavenger determined by matching the experimental accurate
masses and MS/MS fragment ions with those contained in
a database.
Figure 3. Workflow’s nodes used in Compound Discoverer.
selected for the next step aimed at identifying the com-
pounds responsible for the sequestering effect of
the extract.
mass of 5.03088 Da, a chromatogram of the peak pairs
characterised by the set delta mass was obtained.
Elution times of the most abundant peaks were then
extracted and the m/z values of each peak pair were
then obtained by the total ion chromatogram (TIC).
HNE adducts were then confirmed by manually inspect-
ing the characteristic relative abundance and mass shift
of the isotopic pattern.
Further confirmation of the identity of the HNE
adducts was given by the Unknown detector node which
cheques for the absence of the detected peaks as
above reported in nonspiked samples.
As a final adduct confirmation, the MS/MS spectra of
each ion of the isotopic pattern (nondeuterated and
deuterated ions) were checked in order to verify the
overlap between the spectra, thus confirming the same
structures and avoiding false positives. The MW of the
bioactive compounds were then calculated by subtract-
ing the MW of the HNE molecule (ꢆ156 Da) from the
masses of the identified compounds, and assuming
that the detected ions are attributed to the Michael
adduct between HNE and the bioactive molecule, as
reported for a series of bioactive molecules able to
quench HNE [41].
Set-up of an isotopic labelling procedure for the
identification of HNE-sequestering agents
An off-target approach based on isotopic signature was
then set-up to identify HNE adducted compounds
when present in complex matrices such as plant
extracts. As summarised in Figure 2, the method con-
sists of the following steps: first the plant extract is
2
spiked with HNE (156.11502 Da) and
H5-HNE
(161.07590 Da) mixed at the same molar ratio.
Compounds which are reactive towards HNE, react with
both the isotopes at the same rate forming reaction
products which have a specific isotopic pattern. A pair
of peaks with a similar intensity spaced by the delta
mass corresponding to the incremental mass of
2
H5-HNE with respect to HNE (5.03088 Da ±10 ppm) is
generated. The second step consists of extracting the
m/z values relative to the ions of the HNE adducts. This
TM
is automatically carried out by Compound Discoverer
on the basis of the specific isotopic pattern character-
ised by the two peaks with the same intensity and with
a delta mass corresponding to five deuterium atoms.
Figure 3 shows in more detail the specific workflow
generated to extract the peak lists from the mass spec-
tra. By setting the “Pattern Tracer” node with a delta
The method was firstly validated by using gluta-
thione (GSH) and carnosine as known sequestering
agents. Figure 4 shows the Full MS and MS/MS
2
spectra of the GSH spiked with HNE and H5-HNE
mixed at the same molar ratio. The GSH-HNE and

FREE RADICAL RESEARCH
7
(
a)
464.20629
z=1
469.23712
100
z=1
80
60
40
20
0
4
46.19599
z=1
451.22736
z=1
445
450
455
460
m/z
465
470
(
b)
(c)
446.06
451.09
100
100
8
6
4
2
0
0
0
0
0
80
60
40
20
0
3
08.12
308.11
300
300
350
400
450
350
400
450
m/z
m/z
2
Figure 4. Full MS and MS/MS spectra of GSH spiked with HNE and H -HNE mixed at the same molar ratio. (a) Full MS spectrum
5
þ
þ
showing the ions at m/z 464.20629 and at m/z 446.19599 relative to the [M þ H] and the [M þ H-H O] of GSH-HNE, and the
2
þ
þ
2
ions at m/z 469.23712 and m/z 451.22736 referred to the [M þ H] and the [M þ H-H O] of GSH- H -HNE. (b) MS/MS spectrum
2
5
of GSH-HNE adduct: the ion at m/z 446.06 derives from the loss of H O and that at m/z 308.12 from the neutral loss of HNE.
2
2
(
c) MS/MS spectrum of GSH- H -HNE adduct: the ion at m/z 451.09 derives from the loss of H O and the ion at m/z 308.11 which
5
2
2
is the GSH moiety deriving from the neutral loss of H -HNE.
5
Figure 5. The “pattern tracer” chromatogram relative to BRGE incubated in the presence of HNE and 2H5-HNE mixed at the
same molar ratio. The chromatogram reports only peak pairs characterised by the delta mass of 5.03088 Da.
2
GSH- H -HNE adducts are present as single charged
5
Identification of sequestering components in black
rice with giant embryo
ions at m/z 464.20629 and m/z 469.23712. They are
þ
also present as [M þ H-H O] ions as shown in
2
Figure 4(a). The fragmentation patterns are character-
Figure 5 shows the pattern trace of BRGE incubated in
2
2
ised by the H O loss and the HNE and H -HNE neu-
the presence of HNE and H
-HNE mixed at the same
5
2
5
tral loss (Figure 4(b,c)), giving the ion at m/z 308.1
that is the GSH moiety.
molar ratio. Eleven peaks over the threshold intensity
were easily identified and their HNE adducted nature

8
M. COLZANI ET AL.
Table 2. HNE Quenchers identified in rice extract BRGE by
using the isotopic labelling procedure.
Table 3. HNE Sequestering activity of standard compounds
identified in rice extract BRGE. Activity was determined by
measuring free-HNE by HPLC-UV analysis.
Quencher
monoisotopic MW
Calculated MW
Dppm
Possible structure
Activity
HNE
1
1
1
2
3
3
3
3
3
03.06336
55.06950
65.07922
71.06133
01.07236
17.06864
17.12164
30.07634
31.08361
103.06333
155.06948
165.07898
271.06064
301.07121
317.06612
317.12229
330.07395
331.08177
0.304
0.151
1.462
2.514
3.809
7.923
ꢃ2.082
7.232
5.533
c-Aminobutyric acid
Histidine
Phenylalanine
Pelargonidin
Peonidin
Petunidin
Asp–Ser–Pro
Tricine
(carnosine units)
consumption (%)
1
h
3 h
1 h
3 h
Reference compounds
Buffer
Carnosine
Polyphenols
Peonidin-3-O-glucoside
Malvidin
<0.10
<0.10
0.17 ± 0.31
0.71 ± 0.67
1.00
1.00 13.78 ± 2.34 36.06 ± 4.98
0.29
<0.10
<0.10
0.57
0.20
<0.10
0.12
0.35
0.20
4.00 ± 1.39
1.15 ± 1.10
1.02 ± 0.47
7.27 ± 2.42
1.87 ± 1.35
4.31 ± 0.92
Malvidin
Peonidin
Pelargonidin
Tricin-5-O-glucoside
Amino acids
Phenylalanine
GABA
7.81 ± 0.49 12.60 ± 1.99
0.30
5.03 ± 2.58
8.31 ± 2.91
was confirmed by manual inspection taking into
account the mass shift between the ion pairs, their rela-
tive abundances and the overlapping MS/MS fragmen-
tation pattern. The MW of the sequestering agents
were then determined by subtracting the adduct MW
from that of HNE. The identity of the parent com-
pounds was then assigned by matching the experimen-
tal accurate masses and MS/MS fragment ions with
those contained in a database of compounds compiled
by including the chemicals contained in rice [42–44].
The database contained 112 entries (32 anthocyanins,
<0.10
<0.10
0.12
<0.10
<0.10
0.17
0.84 ± 0.78
0.34 ± 0.59
1.77 ± 1.82
1.28 ± 0.47
0.35 ± 0.62
6.26 ± 2.69
Histidine
However, no histidine peptides were detected as HNE
adducts in the rice extract. The HNE sequestering activ-
ity of GABA and phenylalanine, which is more likely due
to their primary amino group, is very weak and much
lower in respect to that of carnosine (the HNE seques-
tering activity was under the detection limit as deter-
mined by using the HPLC-UV assay). However, even if
these two compounds are characterised by a very low
potency as HNE sequestering agents in comparison to
that of carnosine, they act as HNE sequestering agents
in the crude mixture. This apparent contradiction could
be explained by considering that they are probably
contained in high concentrations in the extract or could
be explained by the specific milieu which catalyses their
reactions. In other words, it should be considered that
the overall sequestering efficacy of compounds in plant
extracts is not only due to their intrinsic nucleophilic
reactivity but also to their relative concentration as well
as to the matrix which could affect the intrinsic activ-
ity itself.
A set of polyphenols were then identified as HNE
sequestering agents, four anthocyanidins, and one fla-
vone (tricine). All of the identified compounds were
found to be less effective than carnosine but character-
ised in the HPLC-UV assay by a well detectable activity.
Among these polyphenols, pelargonidin was the most
potent, followed by tricine, peonidin, and malvidin. The
activity of anthocyanindins as HNE sequestering agents
which are characteristic components of black rice can
also explain the higher overall activity of black rice in
respect to white rice and to the other rice varieties.
1
4 phenolic acids, 22 between flavones and flavonols,
nine tocopherols, 20 ferulates (c-oryzanol), four carote-
noids, seven sterols, and four triterpenoids). A list of
nine compounds with a MW lower than 350 Da was
then obtained and listed in Table 2. Five out of nine
compounds were flavonoids, four of them belonging to
the anthocyanidins class, three compounds were amino
acids, two of them proteogenic, while one a peptide.
HNE sequestering activity
The commercially available compounds identified by
the isotopic labelling procedure were then tested as
HNE sequestering agent by using an HPLC-UV test. The
test is based on measuring free-HNE after incubating
the tested compound with HNE. Table 3 reports the
percentage of HNE sequestered by the active com-
pounds identified in BRGE extract. Results are reported
either as % of HNE consumption at 1 and 3 hours and
as carnosine units at 1 and 3 hours, setting as 1 the
value of carnosine activity for both the time points.
Histidine, phenylalanine, and aminobutyric acid were
the amino acids identified as having HNE sequestering
activity. The reactivity for histidine is due to the imid-
azole ring as has already been reported, together with
the full elucidation of the reaction adduct by nuclear
magnetic resonance [45]. The activity of histidine is
increased when bonded to specific amino acids such as
b-alanine, as in the case of carnosine, or analogues,
which are histidine peptides found in mammals [46].
Anthocyanidins as HNE sequestering agents:
molecular modelling studies
Several studies investigating various adducts on antho-
cyanidins discovered that the C8 atom is the most

FREE RADICAL RESEARCH
9
Figure 6. DFT-based MEP as projected on solvent accessible surface for the three major pelargonidin tautomers. The brown
arrow indicates the C8 electrophilic attack centre in the most nucleophilic tautomer.
Table 4. DFT-based properties of the three major tautomer of malvidin and pelargonidin (DE values are
expressed in kJ/mol; nucleophilicity descriptors are based on the parr’s indices as nucleo ¼1/x [44].
0
0
DE_N4
Compound
Nucleo_N5
Nucleo _N7
Nucleo _N4
DE_N5
DE_N7
Malvidin
Pelargonidin
5.18
5.15
5.08
5.29
4.86
4.60
0.00
þ7.78
þ9.59
þ15.63
þ18.27
0.00
prone centre for the electrophilic attack [47]. As exem-
plified by the pelargonidine tautomers in Figure 6, DFT
calculations showed that the electron density at C8
vastly differs in the three simulated major tautomers.
These differences, which are similarly detected in both
considered compounds, revealed that the N7 tautomers
are the most electron rich one at C8 due to the effect
of the vicinal deprotonation in seven, while the
complex matrices for their overall activity as HNE
sequestering agents and the identification of small
molecules responsible for the sequestering effects.
This approach could be applied to screen a variety of
plant extracts with the final aim of identifying
novel and potent HNE sequestering agents. These
methods can also be easily adapted to other cytotoxic
RCS compounds such as MDA, glyoxal, and methyl-
glyoxal for which potent sequestering agents are yet to
be found.
0
N4 tautomers are the least electron rich one at C8 due
to the electron-drawing effect exerted by the C4’
deprotonation. While showing comparable stereo-elec-
tronic features, Table 4 reveals that the tautomers of
pelargonidin and malvidin greatly differ in their relative
stability. In detail, Table 4 shows that the N7 tautomer
is the most stable for pelargonidin, while the N5 tauto-
By using this assay, we also found that rice, and in
particular black rice with giant embryo is a valuable
source of detoxifying agents of HNE which is known to
be formed during food digestion and has a role in the
promotion of colon cancer [48]. HNE derived from food
0
mer is the preferred one for malvidin and the C4 tauto-
ꢃ1
is estimated to be 16 mg day as daily exposure in
mer is the least stable form for both anthocyanidins.
Table 4 also evidences that the nucleophilicity differen-
ces, despite being very limited, parallel the tautomer
stability, the pelargonidin N7 tautomer being the most
nucleophilic species.
Korean diet for a 60-kg Korean adult [49]. Based on the
fact that 29.1 mg, which is corresponding to 673-mg
black rice, inhibits 50% of protein carbonylation
ꢃ1
induced by 78.0 mg mL of HNE, we can conclude that
ꢃ1
the HNE daily exposure of 16 mg day should be easily
These results are in agreement with similar calcula-
tions performed by Estevez et al. [44] and suggest that
the observed differences in the quenching activity
between pelargonidin and malvidin are ascribable to
different relative abundance of the reactive N7 tauto-
mer. Differences in the stereo-electronic properties, des-
pite the high nucleophilicity of the pelargonidin N7
tautomer, appear to be too limited to explain alone the
difference in reactivity towards HNE. Taken together,
these results confirm that the relative tautomer stability
is markedly influenced by the substituents on the B
ring and suggest that the stability of the reactive N7
tautomer is lowered by electron-donating groups on
the B ring, a finding which may explain the intermedi-
ate reactivity of peonidin.
detoxified in the gastro-intestinal lumen by a serving
portion of 75 g. However, we do not know whether the
g.i. microbiota changes the black rice components thus
affecting their HNE sequestering activity (an increase or
decrease of activity are both plausible), an aspect which
should be further investigated for better understanding
the in vivo effect.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
In conclusion, we report a novel approach, which
permits the screening of plant extracts and in general
This work was supported in part by Rural Development
Administration, Republic of Korea [grant number: PJ010059].

1
0
M. COLZANI ET AL.
function in Zucker obese rats. J Cell Mol Med.
011;15:1339–1354.
17] Albrecht T, Schilperoort M, Zhang S, et al. Carnosine
attenuates the development of both type 2 diabetes
and diabetic nephropathy in BTBR ob/ob mice. Sci
Rep. 2017;7:44492.
18] Regazzoni L, de Courten B, Garzon D, et al. A carno-
sine intervention study in overweight human volun-
teers: bioavailability and reactive carbonyl species
sequestering effect. Sci Rep. 2016;6:27224.
19] Yeum K-J, Orioli M, Regazzoni L, et al. Profiling histi-
dine dipeptides in plasma and urine after ingesting
beef, chicken or chicken broth in humans. Amino
Acids. 2010;38:847–858.
20] Vistoli G, Colzani M, Mazzolari A, et al. Computational
approaches in the rational design of improved car-
bonyl quenchers: focus on histidine containing dipep-
tides. Future Med Chem. 2016;8:1721–1737.
21] Vistoli G, De Maddis D, Straniero V, et al. Exploring
the space of histidine containing dipeptides in search
of novel efficient RCS sequestering agents. Eur J Med
Chem. 2013;66:153–160.
[22] Hwang S, Lee Y-M, Aldini G, Yeum K-J. Targeting
reactive carbonyl species with natural sequestering
agents. Molecules. 2016;21:280.
23] Mano J, Nagata M, Okamura S, et al. Identification of
oxidatively modified proteins in salt-stressed
Arabidopsis: a carbonyl-targeted proteomics approach.
Plant Cell Physiol. 2014;55:1233–1244.
References
2
[
1] Vistoli G, De Maddis D, Cipak A, et al. Advanced glyco-
xidation and lipoxidation end products (AGEs and
ALEs): an overview of their mechanisms of formation.
Free Radic Res. 2013;47:3–27.
[
[
[
[
[
[
2] Sousa BC, Pitt AR, Spickett CM. Chemistry and analysis
of HNE and other prominent carbonyl-containing lipid
oxidation compounds. Free Radic Biol Med. 2017;111:
294–308.
[
3] Mano J, Biswas MS. Analysis of reactive carbonyl spe-
cies generated under oxidative stress. Methods Mol
Biol. 2018;1743:117–124.
[
4] Aldini G, Dalle-Donne I, Facino RM, et al. Intervention
strategies to inhibit protein carbonylation by lipoxida-
tion-derived reactive carbonyls. Med Res Rev.
2007;27:817–868.
[
[
5] Dalle-Donne I, Aldini G, Carini M, et al. Protein carbon-
ylation, cellular dysfunction, and disease progression.
J Cell Mol Med. 2006;10:389–406.
6] Poli G, Schaur RJ, Siems WG, Leonarduzzi G. 4-hydrox-
ynonenal: a membrane lipid oxidation product of
medicinal interest. Med Res Rev. 2008;28:569–631.
7] Poli G, Schaur RJ. 4-hydroxynonenal in the pathome-
chanisms of oxidative stress. Tbmb. 2001;50:315–321.
8] Mol M, Regazzoni L, Altomare A, et al. Enzymatic and
non-enzymatic detoxification of 4-hydroxynonenal:
methodological aspects and biological consequences.
Free Radic Biol Med. 2017;111:328–344.
[
[
[
[
9] N ꢀe gre-Salvayre A, Garoby-Salom S, Swiader A, et al.
[24] Biswas MS, Mano J. Reactive carbonyl species
activate caspase-3-like protease to initiate pro-
grammed cell death in plants. Plant Cell Physiol.
Proatherogenic effects of 4-hydroxynonenal. Free
Radic Biol Med. 2017;111:127–139.
[
[
10] Xiao M, Zhong H, Xia L, et al. Pathophysiology of
mitochondrial lipid oxidation: role of 4-hydroxynone-
nal (4-HNE) and other bioactive lipids in mitochondria.
Free Radic Biol Med. 2017;111:316–327.
11] Aldini G, Vistoli G, Stefek M, et al. Molecular strategies
to prevent, inhibit, and degrade advanced glycoxida-
tion and advanced lipoxidation end products. Free
Radic Res. 2013;47:93–137.
2
016;57:1432–1442.
[25] Beretta G, Furlanetto S, Regazzoni L, et al.. Quenching
of alpha, beta-unsaturated aldehydes by green tea
polyphenols: HPLC-ESI-MS/MS studies.
Biomed Anal. 2008;448:606–611.
J
Pharm
[26] Yeh W-J, Hsia S-M, Lee W-H, Wu C-H. Polyphenols
with antiglycation activity and mechanisms of action:
a review of recent findings. J Food Drug Anal.
2017;25:84–92.
[27] Zhu Q, Zheng Z-P, Cheng K-W, et al. Natural polyphe-
nols as direct trapping agents of lipid peroxidation-
derived acrolein and 4-hydroxy-trans-2-nonenal. Chem
Res Toxicol. 2009;22:1721–1727.
[28] Colzani M, Criscuolo A, De Maddis D, et al. A novel
high resolution MS approach for the screening of 4-
hydroxy-trans-2-nonenal sequestering agents. J Pharm
Biomed Anal. 2014;91:108–118.
[
[
12] Davies SS, Zhang LS. Reactive carbonyl species
scavengers-novel therapeutic approaches for chronic
diseases. Curr Pharmacol Rep. 2017;3:51–67.
13] Colzani M, De Maddis D, Casali G, et al. Reactivity,
selectivity, and reaction mechanisms of aminoguani-
dine, hydralazine, pyridoxamine, and carnosine as
sequestering agents of reactive carbonyl species:
a
1
comparative study. ChemMedChem 2016;11:
778–1789.
[
[
[
14] Aldini G, Dalle-Donne I, Colombo R, et al.
Lipoxidation-derived reactive carbonyl species as
potential drug targets in preventing protein carbonyla-
tion and related cellular dysfunction. ChemMedChem
[29] Rees MS, van Kuijk FJGM, Siakotos AN, Mundy BP.
Improved synthesis of various isotope labeled
4-hydroxyalkenals and peroxidation intermediates.
Synth Commun. 1995;25:3225–3236.
[30] Aldini G, Granata P, Marinello C, et al. Effects of UVB
radiation on 4-hydroxy-2-trans-nonenal metabolism
and toxicity in human keratinocytes. Chem Res
Toxicol. 2007;20:416–423.
[31] Dasuri K, Ebenezer P, Fernandez-Kim SO, et al. Role of
physiological levels of 4-hydroxynonenal on adipocyte
biology: implications for obesity and metabolic syn-
drome. Free Radic Res. 2013;47:8–19.
2006;1:1045–1058.
15] Aldini G, Carini M, Yeum KJ, et al. Novel molecular
approaches for improving enzymatic and nonenzy-
matic detoxification of 4-hydroxynonenal: toward the
discovery of a novel class of bioactive compounds.
Free Radic Biol Med. 2014;69:145–156.
16] Aldini G, Orioli M, Rossoni G, et al. The carbonyl scav-
enger carnosine ameliorates dyslipidaemia and renal

FREE RADICAL RESEARCH
11
[
[
32] Keller BO, Sui J, Young AB, Whittal RM. Interferences
and contaminants encountered in modern mass spec-
trometry. Anal Chim Acta. 2008;627:71–81.
33] Vistoli G, Orioli M, Pedretti A, et al. Design, synthesis,
and evaluation of carnosine derivatives as selective
and efficient sequestering agents of cytotoxic reactive
carbonyl species. ChemMedChem. 2009;4:967–975.
34] Schmidt MW, Baldridge KK, Boatz JA, et al. General
atomic and molecular electronic structure system. J
Comput Chem. 1993;14:1347–1363..
carbonyl species: from structural studies to thera-
peutic perspectives. BioFactors 2005;24:77–87.
[42] Abdel-Aal e-SM, Young JC, Rabalski I. IAnthocyanin.
composition in black, blue, pink, purple, and red
cereal grains. J Agric Food Chem. 2006;54:4696–4704.
[43] Pereira-Caro G, Cros G, Yokota T, Crozier A.
Phytochemical profiles of black, red, brown, and white
rice from the Camargue region of France. J Agric
Food Chem. 2013;61:7976–7986.
[
[
[
44] Suttiarporn P, Chumpolsri W, Mahatheeranont S, et al.
Structures of phytosterols and triterpenoids
35] Mano J. Reactive carbonyl species: their production
from lipid peroxides, action in environmental stress,
and the detoxification mechanism. Plant Physiol
Biochem. 2012;59:90–97.
with potential anti-cancer activity in bran of black
non-glutinous rice. Nutrients. 2015;7:1672–1687.
[
45] Hashimoto M, Sibata T, Wasada H, et al. Structural
basis of protein-bound endogenous aldehydes.
Chemical and immunochemical characterizations of
configurational isomers of a 4-hydroxy-2-nonenal-histi-
dine adduct. J Biol Chem. 2003;278:5044–5051.
46] Aldini G, Carini M, Beretta G, et al. Carnosine is a
quencher of 4-hydroxy-nonenal: through what mech-
anism of reaction? Biochem Biophys Res Commun.
[
[
36] Schaur RJ. Basic aspects of the biochemical reactivity
of 4-hydroxynonenal. Mol Aspects Med. 2003;24:
149–159.
37] Esterbauer H, Schaur RJ, Zollner H. Chemistry and
biochemistry of 4-hydroxynonenal, malonaldehyde
and related aldehydes. Free Radic Biol Med. 1991;11:
[
81–128.
[
[
[
38] Schaur R, Siems W, Bresgen N, Eckl P. 4-hydroxy-
nonenal-A bioactive lipid peroxidation product.
Biomolecules. 2015;5:2247–2337.
39] Masisi K, Beta T, Moghadasian MH. Antioxidant prop-
erties of diverse cereal grains: a review on in vitro and
in vivo studies. Food Chem. 2016;1;196:90–97.
2
002;298:699–706.
[
[
47] He F, Pan Q-H, Shi Y, Duan C-Q. Chemical synthesis of
proanthocyanidins in vitro and their reactions in aging
wines. Molecules. 2008;13:3007–3032.
48] Bastide NM, Chenni F, Audebert M, et al. A central
role for heme iron in colon carcinogenesis associated
with red meat intake. Cancer Res. 2015;75:870–879.
40] Lee Y-M, Han S-I, Song BC, Yeum K-J. Bioactives in
commonly consumed cereal grains: implications for
oxidative stress and inflammation.
015;18:1179–1186.
41] Aldini G, Facino RM, Beretta G, Carini M. Carnosine
J
Med Food.
[49] Surh J, Kwon H. Estimation of daily exposure to 4-
hydroxy-2-alkenals in Korean foods containing n-3
and n-6 polyunsaturated fatty acids. Food Addit
Contam. 2005;22:701–708.
2
[
and related dipeptides as quenchers of reactive