Research article

Received: 7 September 2012 Revised: 17 December 2012 Accepted: 20 December 2012 Published online in Wiley Online Library

([Link]) DOI 10.1002/jms.3160

Structural characterization of flavonoid

glycosides from leaves of wheat (Triticum
aestivum L.) using LC/MS/MS profiling of the
target compounds
Anna Wojakowska,a Juliusz Perkowski,b Tomasz Góralc and
Maciej Stobieckia*

The aim of this study was to present integrated mass spectrometric methods for the structural characterization and identification
of flavonoid glycoconjugates. During the liquid chromatography/mass spectrometry analyses, TriVersa NanoMate chip-based
system with nanoelectrospray ionization and fraction collection was combined to a quadrupole time-of-flight mass spectrometer.
In the extract samples prepared from green leaves of wheat plantlets, 41 flavonoid derivatives were recognized. Part of the target
natural products had the full structure being characterized after the registration of mass spectra, where m/z values for protonated
[M + H]+ and deprotonated molecules [M
H]
were annotated. MS2 and pseudo-MS3 experiments were performed for [M + H]+ or
[M
H]
and aglycone ions (Y+/
3
0 -type), respectively. It should be underlined that pseudo-MS mass spectra were registered for
aglycone product ions in the mass spectra of O-glycosides present in the extract samples. In many cases, only tentative structural
identification of aglycones was possible, mainly because of the presence of numerous C-monoglycoside or C-diglycoside in the
samples. Acylation of the sugar moiety and/or methylation of the aglycone in the flavonoid glycosides under study was observed.
The existence of isobaric and/or isomeric compounds was demonstrated in the extract studied. The collision-induced dissociation
mass spectra registered for C,O-diglycosides and C,C-diglycosides did not permit to draw complete structural conclusions about
the compounds studied. For the investigated class of natural products, unambiguous classification of sugar moieties linked to
the aglycones from the recorded mass spectra was not possible. Registration of the positive and negative ion mass spectra
did not lead to the precise conclusion about the glycosylation position at C-6 or C-8, and O-40 or O-7 atoms. It was possible, on
the basis of the collected MS2 spectra, to differentiate between O-glycosides and C-glycosides present in the samples analyzed.
Copyright © 2013 John Wiley & Sons, Ltd.
Supporting information may be found in the version of this article.

Keywords: collision-induced dissociation; electrospray ionization; flavonoid C-glycoconjugates and/or O-glycoconjugates; liquid
chromatography/mass spectrometry; secondary metabolite profiling; structural characterization; wheat

Introduction data. Information about the identification of the flavonoid

glycosides based on the registered CID mass spectra of product
Flavonoids and their glycoconjugates constitute an important ions is widely described.[13–22] The published papers presented
class of secondary metabolites present in the plant kingdom. mainly examples of structural MS analyses concerned with the size
There are over 8000 known different natural products belonging of sugar rings and substitution pattern of sugars on an aglycone or
to this class of metabolites.[1] Interestingly, polyphenols and their placement of an interglycosidic bond.[12–15] Additionally, in LC/MS
derivatives create a group of natural products representing a systems with a triple quadruple or quadrupole time-of-flight
wide range of biological roles or activities.[2–8] (q-ToF) analyzer, it is possible to obtain pseudo-MS3 spectra for
Possibilities of profiling and structural analysis of flavonoid aglycone ions in the glycosides and comparison of the product
conjugates were reported in various reviews.[1,9–15] Experimental
papers described applications of different liquid chromatography/
mass spectrometry systems (LC/MS), where various methodical * Correspondence to: Maciej Stobiecki, Institute of Bioorganic Chemistry,
approaches were used for gaining structural information about Polish Academy of Sciences, Noskowskiego 12/14, 61-704 Poznań, Poland.
E-mail: mackis@[Link]
target natural products during the analyses of extract samples with
the complicated composition. When electrospray ionization (ESI) of a Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego
the LC eluate is applied, the principal methodology is based on the 12/14, 61-704 Poznań, Poland
application of collision-induced dissociation (CID) in tandem MS
experiments for obtaining information about masses of aglycone b Department of Chemistry, Poznań University of Life Sciences, Wojska Polskiego
75, 60-625 Poznań, Poland
and sugars. In well-defined ionization conditions and properly
chosen energy of CID experiments, it is possible to obtain good- c Department of Plant Pathology, Plant Breeding and Acclimatization Institute – NRI,
329

quality product ion mass spectra that bear interesting structural Radzików, 05-870 Błonie, Poland

J. Mass Spectrom. 2013, 48, 329–339 Copyright © 2013 John Wiley & Sons, Ltd.
 A. Wojakowska et al.

ion spectra with the standards.[20,21] It should be underlined Dried extract samples were dissolved in water (1.5 ml) and
that these data are accessible with ion trap and orbitrap mass subjected to SPE on RP C-18 silica gel (300 mg) from Baker (Omaha,
spectrometers in regular MS3 experiments.[22] Moreover, post- NE, USA). Next SPE column was washed with 2 ml of H2O, and
column addition of cations to the LC eluate prior to ESI ionization the adsorbed target natural products were eluted with 2 ml of
helps gain an amount of structural information because of the methanol. The sample solution after SPE was evaporated to
appearance of new product ions in mass spectra.[23,24] The application dryness (vacuum concentrator) and dissolved in 600 ml of 80%
of medium-resolution analyzers [about 15 000 full width at half MeOH and immediately subjected to the LC/MS analyses.
maximum (FWHM)] in the mass spectrometers enables the registra-
tion of m/z values of ions with accuracy up to the fourth decimal
Liquid chromatography/mass spectrometry
point, and it allows to achieve, for masses of protonated or depro-
tonated molecules above 200 Da, an accuracy better than 5 ppm. LC/MS analyses of plant extracts were performed with Waters
On the basis of accurate m/z values, it is possible to differentiate Acquity UPLC system (Milford, MA, USA) connected to a mass
isobaric compounds present in the studied extract samples. For spectrometer Model micrOTOF-Q from Bruker Daltonics (Bremen,
example, these were acylated with substituted aromatic acid Germany) combined with TriVersa NanoMate system from
molecule flavone diglycosides and flavone triglycoside.[25,26] Advion (Ithaca NY, USA).
Profiling of flavonoid glycoconjugates and other phenolic The analyses were carried out using Poroshell 120 EC-C18
natural products with LC/MS methods in the samples obtained columns (2.1  100 mm, granulation of 2.7 mm, Agilent). The
from wheat plant leaves or grains was reported earlier.[27–31] injection volume to the LC/MS system of the sample solution
We were able to detect O-glycosides or C-glycosides, mixed O, was 5 ml. Chromatographic separation was performed at a
C-diglycosides and C,C-diglycosides in the studied samples from 0.6 ml/min flow rate by using mixtures of two solvents: A
wheat plants. In this paper, we describe our strategy, where (99.5% H2O/0.5% formic acid v/v) and B (99.5% acetonitrile/
CID-MS2 and pseudo-MS3 experiments combined with information 0.5% formic acid v/v) with a 2 : 1 split of the column effluent,
about the accurate mass of protonated/deprotonated molecules so 0.2 ml/min was delivered to the ESI ion source. The elution
permits to conduct profiling and, to some extent, the structural steps were as follows: 0–5 min linear gradient from 10% to 30% of
analysis of flavonoid glycoconjugates present in a complicated B, 5–12 min isocratic at 30% of B, 12–13 min linear gradient from
mixture. The attempts were undertaken to obtain maximum 30–95% of B and 13–15 min isocratic at 95% of B. After returning
structural information about sugar classes and sizes and their to the initial conditions, the equilibration was achieved after 4 min.
substitution on aglycones. The recognition of acylated flavonoid The ESI source of the micrOToF-Q mass spectrometer operated
glycosides with aromatic acids present in the samples of wheat at a voltage of +4.5 kV in the positive ion mode and
4.5 kV in
leaf extracts was achieved. Pre-concentration of the wheat the negative ion mode, with an ion transfer energy of 7 eV,
extract samples with solid-phase extraction (SPE) on reversed- during MS and MS/MS experiments run in both polarization.
phase (RP) C18 column was applied for the elimination of matrix Nebulization with nitrogen was performed at 1.2 bar and dry
compounds present in the samples in order to improve the gas flow of 8.0 l/min at a temperature of 220  C. The instrument
signal-to-noise ratio. was operated using the micrOTOF control program ver. 2.3, and
data were analyzed using the Bruker Data Analysis ver. 4 package.
The system was calibrated externally using a calibration mixture
Experimental containing sodium formate clusters. Additional internal calibration
was performed for every run by injection of the calibration mixture
Plant material
by using the diverter valve during the LC separation. All calcula-
Plant material was obtained from winter wheat (Triticum aestivum tions of m/z values were performed with the enhanced quadratic
L. var. Legenda); this variety is present on the Polish National List algorithm with the accuracy of at least 5 ppm.
of Varieties and is grown in Poland in commercial fields. Flavonoid conjugates were detected by comparison of the exact
Plant material was collected from young seedlings. Pre- molecular masses and fragmentation pathways of the precursor
germinated healthy kernels were sown in germination trays. [M + H]+ or [M
H]
ions recorded during CID MS/MS experiments
Seedlings were cultivated in the growth chamber with a 16-h with those obtained for standards and data from the literature.
photoperiod at 15  C for 10 days. Next, seedling leaves were The targeted MS/MS experiments were performed using a collision
collected to 4-ml cryogenic vials, and about ten leaves were cut energy ramp of 15 to 30 eV (positive ion mode) and 25 to 30 eV
for single samples. Immediately, the plant material was frozen in (negative ion), unless otherwise stated. For the pseudo-MS3 experi-
liquid nitrogen and next kept in a
80  C freezer for analysis. ments, ionization energy [in-source CID (ISCID)] was increased
from 0 to 80 or 85 eV in the positive and negative ion modes,
respectively. During the experiments, the collision energy was
Extraction of phenolic secondary metabolites from plant tissues
between 20 and 25 eV, in order to achieve high fragmentation of
For the structural elucidation of flavonoid glycoconjugates, the protonated and deprotonated molecules ([M + H]+ and [M
H]
).
frozen plant material (600 mg fresh weight) was homogenized The instrument was operated at a resolution higher than 15 000
in the ball mill from Retsch, model 400 (Dusseldorf, Germany). FWHM. The spectra were recorded in the targeted mode within
The plant material was suspended in 4 ml of 80% MeOH, and the m/z range of 50–1100, and metabolite profiles were registered
the suspension was placed in an ultrasonic bath for 30 min. The in the positive and negative ion modes. For identification of the
extract was centrifuged, and the supernatant was transferred compounds, the instrument operated in the MS, CID MS/MS and
to new screw-cupped tubes. The solvent was evaporated in a pseudo-MS3 modes and single-ion chromatograms for exact
vacuum concentrator, Savant SPD 121P (Thermo Electron masses of [M + H]+ or [M
H]
ions (0.005 Da) were recorded; this
Corporation, Waltham, MA, USA); during this procedure, the tubes accuracy corresponds – in the mass range between 400 and
330

were placed in a tray rotating in vacuum at room temperature. 1000 amu – between 12 and 5 ppm.

[Link]/journal/jms Copyright © 2013 John Wiley & Sons, Ltd. J. Mass Spectrom. 2013, 48, 329–339
Secondary metabolite profiles in wheat leaves

A chip-based nano-ESI source (TriVersa NanoMate system) Findings about glycosylation pattern and presence of the
coupled to the q-ToF analyzer was used in the LC/MS fraction O-glycosides or C-glycosides could be drawn after analysis of the
collection mode for a more effective separation of the isomeric fragmentation pathways of protonated [M + H]+ or deprotonated
flavonoid compounds. The NanoMate source was operated in molecules [M
H]
in the registered MS2 spectra. The presence
the positive ionization mode with an HD_A_384 chip with a spray of characteristic product ions after the elimination of sugar moiety
voltage of 1.7 kV. A spray sensing was used in order to prevent (Yn ions) in the case of O-glycosides and disruption of water
blockage of the nozzle. In the cases when the spray current molecules (En ions) or cleavage of the sugar ring ([M  H]+/
– 90
dropped for more than 30 s below 2 nA or was above 7000 nA, or 120)+/
for C-glycosides can be concluded (Scheme 1). However,
the robot was moved to the next nozzle (maximum two times the exact placement of glycosidic bond position for each detected
for injection). The eluent flow was split in front of the NanoMate, compound was difficult to elucidate.
at 86.7 ml/min to the fraction collector and 0.099 ml/min to the
nano-ESI source. The LC fractions were collected every 10 s into
Flavone O-glycosides
384 well plates ([Link] Eppendorf) and cooled at 4  C. After
fractionation of the UPLC eluate, the plate with the collected In the case of tricetin trimethyl ether, conjugates were identified
fractions was sealed to prevent evaporation. Off-line injection as O-glycosides (32, 34, 36 and 38). They were tentatively
of the fractions was performed with the aforementioned assigned to the O-7 hydroxyl group, because of the lack of free
parameters of q-ToF. hydroxyls in the B ring of the flavonoid molecule, and the
hydroxyl groups at the C-5 position in ring A was not considered
as a place of sugar substitution (Fig. 1). The same conclusion
Results and discussion would be valid for tricin glycosides because the presence of
two methyl groups on hydroxyls at C-30 and C-50 in ring B autho-
During UPLC/MS analyses with ESI and hybrid analyzer (q-ToF), rized a spatial hindrance for sugar substitution at the hydroxyl
attention was directed to the identification of flavonoid glyco- group present at C-40 carbon atom (17, 22, 24, 26, 28, 30 and 33).
conjugates present in the extract samples from wheat plantlet Important structural characterization of the flavone aglycones could
leaves. The structural characterization of the target compounds in be obtained after registering the pseudo-MS3 spectra only for
the eluates of the LC column relied on the following data obtained the O-glycoconjugates. This technique relies on more intense
during the registration of mass spectra in three independent fragmentation of protonated or deprotonated molecules. This
analyses, run in the MS, MS2 and pseudo-MS3 modes targeted effect occurred after the ESI ionization, mainly as a result of
for the aglycone ions. The CID MS2 and pseudo-MS3 mass spectra frequent ions and molecules collisions because of the high pressure
were registered in the positive and negative ion modes. During existing after entering the mass analyzer.[19–21] Because of the
the analyses conducted, sensitivity achieved with the CID MS/MS collisions and increased in-source dissociation of protonated or
spectra registered in the positive ion mode was better than deprotonated molecules occurring at the mass spectrometer ana-
that obtained for the negative mode. In the positive ion mode, lyzer entrance system (ISCID) of the q-ToF instrument, further CID
41 flavonoid glycoconjugates were tentatively recognized, and in experiments were performed on more abundant aglycone ions
the negative ion mass spectra, only 30 phenolic metabolites were (Y+0 /Y

0 -type). In the case of O-glycosides, proper identification of
distinguished in the extract samples from the wheat leaf tissue the flavones present in the given molecule was feasible after
(Table 1). The literature reports concerned with the analyses of comparison of pseudo-MS3 mass spectra of the defined aglycone
biological material from wheat plants showed that both ionization with those registered for the flavone standards. In the case of
modes were applied during the mass spectrometric analyses of C-glycosides, it was not possible to observe enough intense
the samples (27–31). appropriate aglycone product ions in the MS spectra, so further
CID experiments were not performed. Their structures were
proposed only tentatively from m/z values noted for the glyco-
Interpretation of product ions mass spectra registered with the
conjugates and elemental composition proposed. An analogy
UPLC/MS/MS system
to molecular masses of known O-glycosides present in wheat
All mass spectra performed in the MS mode were registered with a leaf extracts was also considered. The nomenclature of product
high resolution, better than 15 000 (FWHM). The registration of ions created after cleavage of the glycosidic bonds follows that
mass spectra in the MS mode for the wheat leaf extract sample proposed by Domon and Costello for O-glycosides[34] and by
allowed to achieve the accuracy better than 5 ppm for all Waridel and others for C-glycosides[12] (Scheme 1).
protonated [M + H]+ and deprotonated [M
H]
molecules of Among the aglycones characterized in the LC/MS analysis of the
flavone glycoconjugates (1–41) recognized during a single analysis. O-glycoconjugates, we identified the following flavones: apigenin,
Twenty-four natural products included in Table 1 were not reported luteolin, chrysoeriol, tricetin, tricin and tricetin trimethyl ether (Fig. 1);
earlier in the wheat plant leaves; errors are defined for protonated four of them were methylated on the hydroxyl groups of the flavone
molecules only. The achieved accuracy was sufficient for an B ring. Earlier published data in the literature confirmed the presence
unambiguous elucidation of the elemental composition of ions with of all these flavones and partially their glycosides in the wheat plant
m/z values annotated to the fourth decimal point for protonated or tissue or grains.[27–31] In the aforementioned papers, the number of
deprotonated molecules in the MS spectra. Differentiation on this recognized flavonoid glycosides was lower than those recognized
basis of isobaric compounds containing sugar rings or acyl groups during our investigation. Structures of six aglycones were
was possible with the hybrid q-ToF analyzer applied. Existence unambiguously defined for recognized O-glycosides after com-
of some compounds in the wheat plant was not reported earlier, paring the mass spectra registered in the pseudo-MS3 mode with
especially when they were acylated with ferulic or caffeic acid those of standard flavones. In the case of O-monoglycosides or O-
flavone glycosides 13, 17, 18, 20 and 23 or malonylated tricetin glycosyl-O-glycosides, it was possible to perform pseudo-MS3
331

trimethyl ether glycoside 38. experiments on Y+0 or Y
0 fragment ions (created after the cleavage

J. Mass Spectrom. 2013, 48, 329–339 Copyright © 2013 John Wiley & Sons, Ltd. [Link]/journal/jms
 332

Table 1. Flavones and their glycoconjugates identified in wheat (Triticum aestivum L.) leaf extract.

No Compound Rt MW Elemental [M + H]+ m/z Error [M + H]+ observed/product [M
H]
observed/product Metabolite
(min) composition calculated (ppm) ions (m/z) ions (m/z) identification
levela

1. Luteolin O,C,-trihexoside (I)b,c 0.9 772 C33H40O21 773.2135
0.1 773.2136/611, 593, 575, 545, 491, 449 771.1971/609, 519, 489 3
2. Luteolin O,C,C-trihexoside (II)b,c,d 1.2 772 C33H40O21 773.2135
0.5 773.2139/611, 593, 491, 431, 381, 209 771.1974/609, 383 3
3. Luteolin 6-C-hexosyl-O-hexosided,e 1.2 610 C27H30O16 611.1607 0.0 611.1607/593, 575, 545, 515, 491, 461, 609.1492/447, 429, 357, 327 3
431, 413, 395, 383, 353, 329, 299
4. Luteolin C-hexosyl-O-hexoside O-deoxyhexosideb,c,d 1.3 756 C33H40O20 757.2186 0.4 757.2183/611, 593, 491, 449, 431, nr 3
413, 383, 353, 329

[Link]/journal/jms
5. Luteolin 6-C-hexoside O-hexosideb,d,e 1.5 610 C27H30O16 611.1607 1.1 611.1600/449, 431, 413, 395, 383, 609.1471/447, 429, 411, 393, 387, 2
353, 329, 299, 287 369, 357, 327, 297
6. Tricetin 6-C-hexosided,e 1.5 464 C21H20O12 465.1028
0.4 465.1030/447, 429, 411, 399, 393, 463.0871/385, 373, 355, 343, 301, 2
381, 369, 345, 315 193, 165, 149
7. Luteolin C-hexoside, C-pentosideb,d,f 1.7 580 C26H28O15 581.1501
0.5 581.1504/563, 545, 527, 515, 509, 579.1334/561, 543, 519, 501, 489, 3
497, 485, 473, 463, 461, 455, 449, 471, 459, 441, 429, 411, 399, 369,
443, 437, 425, 413, 407, 395, 365, 339, 327
353, 341, 323, 311
8. Luteolin O-hexoside C-hexosideb,d 1.8 610 C27H30O16 611.1607 0.5 611.1604/449, 431, 395, 383, 377, 609.1492/591, 531, 489, 429, 411, 3
365, 353, 329, 299 393, 369, 357, 351, 327 593.1508/
473, 431
9. Luteolin C-hexoside O-deoxyhexosideb,d 1.9 594 C27H30O15 595.1657 1.0 595.1651/449, 431, 413, 395, 383, 563, 1408/545, 503, 485, 473, 455, 2
353, 329, 299 443, 425, 413, 383, 353
10. Apigenin C-hexoside C-pentosideb,d,e 2.0 564 C26H28O14 565.1552 0.7 565.1548/547, 529, 511 623.1619/503, 461, 371, 341 2
11. Chrysoeriol C-hexoside O-hexosideb,d 2.0 624 C28H32O16 625.1763 1.9 625.1751/607, 589, 559, 529, 505, 463, 447.0933/429, 399, 369, 357, 339, 2
445, 427, 409, 397, 367, 343, 313 327, 311, 297
12. Luteolin 6-C-glucosided,e,f 2.1 448 C21H20O11 449.1078
0.4 449.1080/431, 413, 395, 377, 365, 771.1757/651, 609, 429, 401 2

Copyright © 2013 John Wiley & Sons, Ltd.

353, 329, 299
13. Luteolin C-[(O-caffeoyl-hexosyl)-O-hexoside]b,c,d 2.2 772 C36H36O19 773.1924
0.1 773.1925/611, 449, 431, 413, 395, 593.1504/575, 503, 485, 473, 443, 2
383, 365, 353, 329, 163 413, 383, 357
14. luteolin C-deoxyhexosyl-O-hexosideb,c,d 2.2 594 C27H30O15 595.1657 2.9 595.1640/577, 559, 541, 511, 499, 623.1621/443, 383, 341, 323 nr 2
487, 475, 457, 433, 415, 397, 367,
355, 337, 313
15 Chrysoeriol O-hexoside C-hexosideb,d,f 2.5 624 C28H32O16 625.1763 0.5 625.1760/463, 445, 427, 409, 397, 813.2090/483, 329 2
367, 343, 313
16. Apigenin 6-C-glucosided,f 2.5 432 C21H20O10 433.1129 0.7 433.1126/415, 397, 379, 349, 337, 741.1669/579, 429 2
313, 283
17. Tricin 7-O-[(O-feruoyl-deoxyhexosyl)-O-hexoside] c,d,g 2.6 814 C39H42O19 815.2393 2.5 815.2373/669, 493, 331, 177 743.1814/ 607.1657/487, 461, 443, 2
581, 449, 431
18. Luteolin O-(O-caffeoyl-pentoside) C-hexosideb,c,d 2.7 742 C35H34O18 743.1818 0.5 395, 383, 353, 329, 295, 163 428, 407, 383, 371, 365, 353, 341, 323 2
19. Chrysoeriol O-deoxyhexoside-C-hexosideb,c,d 2.7 608 C28H32O15 609.1814 0.0 609.1814/463, 445, 427, 409, 397, 785.1932/665, 609, 429 461.1660/ 2
391, 379, 367.353, 343, 313 383, 371, 353, 341, 311, 298
A. Wojakowska et al.

J. Mass Spectrom. 2013, 48, 329–339

 [M + H]+ observed/product [M
H]
Metabolite identification
ions (m/z) observed/ levela
product
ions (m/z)

20. Chrysoeriol C-[(O-caffeoyl-hexosyl)-O-hexoside]b,c,d 2.8 786 C37H38O19 787.2080
0.5 787.2084/625, 463, 445, 427, 409, 667.1504/329 2
397, 367, 343, 163
21. Chrysoeriol 6-C-glucosidec,d,f 2.8 462 C22H22O11 463.1235
0.2 463.1236/445, 427, 409, 397, 391, 771.1773/609, 357, 327, 299 2
379, 367, 353, 343, 313
22. Tricin 7-O-glucuronyl-O-hexosideb,c,d,g 2.8 668 C29H32O18 669.1661
0.7 669.1666/493, 331 653.1700/329 2
23. Luteolin O-(O-caffeoyl-hexoside) C-hexosideb,c,d 2.9 772 C36H36O19 773.1924 0.0 773.1924/449, 431, 413, 395, 383, 755.1804/635, 593, 579, 447, 429, 2
353, 329, 163 411, 369, 357
24. Tricin 7-O-hexosyl-O-hexosidec,d,f,g 2.9 654 C29H34O17 655.1869
0.9 655.1875/493, 331 nr 2
25. Luteolin C-[pentosyl-O-(feruoyl-O-hexoside)]b,c,d 3.1 756 C36H36O18 757.1974 0.8 757.1968/449, 431, 413, 383, 353, 607.1657/299 2
329, 309, 177

J. Mass Spectrom. 2013, 48, 329–339

26. Tricin O-pentoside O-dihexosidec,d 3.1 800 C35H44O21 801.2448
1.7 801.2462/655, 493, 331 nr 2
27. Chrysoeriol O-deoxy-hexosyl-O-hexosided,g 3.2 608 C28H32O15 609.1814 0.3 609.1812/463, 301 799.2204/623, 443 2
28. Tricin 7-O-deoxyhexosyl-O-hexosidec,g 3.3 638 C29H34O16 639.1920
0.3 639.1922/493, 331 nr 2
29. Chrysoeriol O-hexoside C-(O-feruoyl-hexoside)b,c,d 3.3 800 C38H40O19 801.2448
1.0 801.2456/639, 463, 445, 427, 177 461.1060/299 nr 2
30. Tricin 7-O-hexosidec,d,e,g 3.5 492 C23H24O12 493.1341 0.0 493.1341/331 nr 2
Secondary metabolite profiles in wheat leaves

31. Chrysoeriol O-hexoside 3.6 462 C22H22O11 463.1235 1.5 463.1228/301 2

32. Tricetin trimethyl ether 7-O-hexosyl-hexosidec,g 4.2 668 C30H36O17 669.2025 0.9 669.2031/507, 345 nr 2
33. Tricin 7-O-hexoside malonylatedc,g 4.3 578 C26H26O15 579.1344
1.0 579.1350/493, 331 285.0383/201, 175, 151, 133, 121, 107 2
34. Tricetin trimethyl ether 7-O-deoxyhexosyl- 4.7 652 C30H36O16 653.2076 0.2 653.2075/507, 345 nr 2
hexosidec,d,g
35. Luteolinc,d,h 4.7 286 C15H10O6 287.1550 1.7 287.0545/269, 241, 213, 179, 161, 269.0467/227, 183, 159, 151, 117, 107 1
153, 135, 117, 107, 89
36. Tricetin trimethyl ether 7-O-hexosidec,d,g,h 5.0 506 C24H26O12 507.1497 0.6 507.1494/345 nr 2
37. Apigeninh 5.6 270 C15H10O5 271.0601 1.1 271.0598/243, 229, 197, 163, 153, 329.0629/314, 299, 285, 271, 227, 1
145, 131, 119, 91 203, 185, 161
38. Tricetin trimethyl ether 7-O-hexoside 5.7 592 C27H28O15 593.1501
0.2 593.1502/345 nr 2
malonylatedc,g
39. Tricinc,d,h 5.8 330 331.0812 0.6 331.0810/315, 301, 287, 270, 258, 343.0817/327, 313, 309, 285, 270 1

Copyright © 2013 John Wiley & Sons, Ltd.

C17H14O7
242, 153
40. Chrysoeriold,h 5.9 300 C16H12O6 301.0707 0.3 301.0706/286, 258, 229, 212, 201, 1
184, 153
41. Tricetin trimethyl etherd,h 10.2 344 C18H16O7 345.0969 0.3 345.0968/329, 315, 284, 255 1

nr, spectrum not registered.

a
Metabolite identification level according to Metabolomics Standards Initiative recommendation.[36]
b
Substitution of sugar molecule at C-6 or C-8 carbon atom on the aglycone was not estimated.
c
Not reported in Triticum aestivum.
d
Recognized also with TriVersa NanoMate system.
e
Compound reported in the literature of Cummins et al.[27]
f
Compound reported in the literature of Ioset et al.[28]
g
Placement of the sugar moieties proposed on the basis of spatial hindrance on B ring of the aglycone.
h
Compound reported in the literature of Dinelli.[30]

[Link]/journal/jms
333
 A. Wojakowska et al.

Scheme 1. Fragmentation nomenclature commonly used for flavonoid glycosides (illustrated on luteolin-7-O-rutinoside), according to Vukics and
Guttman[13] and Domon and Costello.[34]

intensities) registered for the Y+0 fragment ion for tricin 7-O-glucoside
(30) and tricetin trimethyl ether 7-O-glucoside (36) verified the iden-
tity of the flavones after comparison of the spectra with those for the
aglycone standard (Fig. 2).
From the mass spectrum of the metabolite 22, we were able to
propose the following structure: tricin 7-O-glucuronyl-O-hexoside
(MW = 668). In the mass spectra registered in the positive and
negative ion modes, cleavages of both O-glycosidic bonds were
Figure 1. Structure of flavones identified in wheat leaf. observed, and ions at m/z 493 and 331, or 491 and 329 (Y1 and
Y0 types) were created in the positive and negative ion spectra,
respectively. Hydroxylation and methoxylation patterns of tricin
of O-glycosidic bonds), and the structure of the flavone was eluci- suggested the proposed structure of diglycoside and the
dated. Mass spectra registered for compounds 30 and 36 are an fragmentation pathway from the C-7 carbon-substituted hydroxyl
example of the application of the pseudo-MS3 experiment for the group (Fig. 3).
evaluation structure of the aglycones. Comparison of the pseudo- In the case of O-monoglucosides, we proposed the substitution of
MS3 mass spectra (m/z values of the product ions and their relative sugar at the C-7 carbon atom of the flavone skeleton (30 and 36);

Figure 2. Mass spectra of tricin 7-O-hexoside 30 (MW = 492), in positive ion mode: (a) CID MS/MS of [M + H]+ ion; (b) pseudo-MS3 spectrum of the agly-
cone ion (Y+0 ); mass spectra of tricetin trimethyl ether 7-O-hexoside 36 (MW = 506), in positive ion mode: (c) CID MS/MS of [M + H]+ ion; (d) pseudo-MS3
334

spectrum of the aglycone ion (Y+0 ).

[Link]/journal/jms Copyright © 2013 John Wiley & Sons, Ltd. J. Mass Spectrom. 2013, 48, 329–339
Secondary metabolite profiles in wheat leaves

Figure 3. CID MS/MS spectra of tricin 7-O-glucuronyl-hexoside 22 (MW = 668): (a) positive ion mode; (b) negative ion mode.

linkage of the sugar group on the B ring together with the methyl compound consisted of chrysoeriol as an aglycone and two sugars,
group is rather unlikely. where hexose and deoxyhexose (probably glucose and rhamnose)
were distinguished; hexose was linked through the C-bond to the
aglycone (Fig. 4). The relative intensities of the En-type product ions
Flavone C,O-diglycosides and C-glycosylglycosides
in the positive mode at m/z 445 and 427 may suggest substitution
During the analysis of mass spectra, we also tried to distinguish of glucose at carbon C-8.[12] On the other hand, in the negative ion
these compounds originating from O,C-diglycosides and mass spectra, the basic ion at an m/z of 443 (Z1-type ion) and/or
C-glycosylglycosides (Fig. S1). Another problem was the evaluation m/z 461 – H2O (E1-type ion) and a small relative intensity of
of interglycosidic bond; it was impossible to conclude about the ([M
H]
– Rha) ion at an m/z of 461 suggested a direct O-glycosidic
exact linkage pattern between the sugars or substitution of linkage of rhamnose to the hydroxyl group at the C-7 carbon atom
aromatic acyl groups only from m/z values of the registered product of flavone instead of substitution as hexosyl-O-hexoside directly at
ions and their relative intensities in the obtained mass spectra. the C-8 carbon atom.[17]
Because of lack of glycoside standards, it was not possible to On the basis of fragmentation, pathways induced during
estimate this structural feature. However, Ferreres and coworkers, CID MS2 experiments and from accurate m/z values of product
and earlier Cuykens and Clayes, reported possible differentiation ions, it was possible to classify glycosides as O-glycosides and
of the glycosidic linkage placement on the basis of relative intensi- C-glycosides or C,O-glycosides (Fig. S1a, b). The estimation of
ties of the product ions.[17,32,33] exact placement of C-glycosidic or O-glycosidic bonds between
Some phenolic glycoconjugates were C,O-dihexosides or the sugar moiety and an aglycone was difficult and, in many
C-hexosyl-O-hexosides; an example of this class of metabolites is cases without NMR analyses, impossible. Some conclusions could
compound 19. From the m/z values of the product ions and the ac- be drawn after comparing the mass spectra registered in the
curate mass value of the [M + H]+ ion, it was known that the positive and negative ion modes followed by the analysis of the

335

Figure 4. CID MS/MS spectra of chrysoeriol O-deoxyhexoside C-hexoside 19 (MW = 608): (a) positive ion mode; (b) negative ion mode.

J. Mass Spectrom. 2013, 48, 329–339 Copyright © 2013 John Wiley & Sons, Ltd. [Link]/journal/jms
 A. Wojakowska et al.

literature data.[10,12,14,22,27–33] The information concerned with with the proper interpretation of the product ion mass spectra
the structural features of the target compounds drawn from because of lack of intense signals from deprotonated mole-
positive and negative mode mass spectra were complementary. cules, so careful control of collision energy during experiments
On the other hand, in the case of C-glycosides of flavonoids, is necessary.
better fragmentation was observed because of the presence of In some cases, isobaric compounds containing an aromatic
product ions of En-type and ions created after cleavage of the acid as an acyl group substituted in the sugar ring were recog-
sugar ring (Scheme 1). The sugar part of the glycoconjugates nized. In the registered profiles, monoglycosides, diglycosides
constituted hexose, deoxyhexose or pentose. and triglycosides of flavones were detected. For example, luteolin
In the single-ion chromatogram of the accurate mass of diglycoside (13) was acylated with the aromatic acid molecule;
611.1607 Da corresponding to the luteolin dihexosides (elemental this was an isobaric natural product, which corresponds to the
composition – C27H31O16), recorded in the positive and negative molecular masses of triglycosides (1, 2). It was possible to differ-
ion modes, three peaks were observed, corresponding to entiate the presence of the aromatic group, instead of the sugar
compounds 3, 5 and 8 (Fig. S1a, b). In three CID MS/MS spectra moiety, on the basis of elemental composition of protonated or
registered in the positive ion mode (Fig. S1a2–a3), there was unclear deprotonated molecules – 13, 17, 18, 20, 23, 25 and 29
evidence about the substitution pattern of both hexoses in the (Figs S2–S5). The rule observed in the positive ion mode CID
luteolin molecule. From all three mass spectra registered in the mass spectra of the analyzed compound acylated with phenyl-
positive ion mode, it was obvious that a minimum one sugar ring propenoid acid was registration in the MS2 mass spectrum of
was linked through the C-glycosidic bond, because of the presence the product ion with m/z value corresponding to the mass of
Y+1 ion (m/z 449) and (m/z 449 – En) product ions at m/z 431, 413 and the acid residue; there were ions at m/z 163 and 177. These ions
395 produced from [M + H]+. However, information obtained from originated from caffeic and ferulic acid residues, respectively. In
the negative ion mass spectra was that, in the first step, O-linked the mass spectra registered in the negative polarity, the product
sugars were eliminated and Y


1 and Z1 ions of different relative ions created from aromatic acids were not observed.
intensities were observed (Fig. S1b). Only for isomer 3 (Rt = 1.2 min), In the figures presented in the supporting information, there
the CID MS/MS spectrum showed the presence of a series of are mass spectra of four compounds containing an aromatic acyl
product ions obtained after the elimination of the water molecule group (13, 20, 25 and 29). Differentiation between the acyl
(En-type ions) from the [M + H]+ ion (Fig. S1a2). In this situation, a and sugar moieties was, in every case, possible on the basis of
series of ions at m/z 593, 575 and 557 were created after consecutive accurate masses of the registered protonated or deprotonated
water molecule elimination from [M + H]+, and next the dehydrated molecules (Figs S2–S5). The mass spectra of acylated compounds
hexose from these E+n ions was possible to cleave (ions at m/z 431, in the supporting information were registered for metabolites
413 and 395). Moreover, the product ion of low intensity at with one sugar linked to the aglycone moiety with the
m/z 449 ([M + H]+ – 162) was also observed. This one was probably C-glycosidic bond and the second with the O-glycosidic bond
created after the cleavage of the O-glycosidic bond between two via the C-substituted sugar moiety 13, 20 and 25. There was
hexoses. On the other hand, in the negative product ion, the mass one exception: For compound 29 in C-monoglycoside, the second
spectra recorded an ion at m/z 447 with a high abundance only for sugar molecule was substituted directly to the hydroxyl group on
one isomeric compound 3 (Fig. S1b2). This ion was created after the flavone skeleton at the C-7 carbon atom (Fig. S5). The relative
the rupture of O-glycosidic bond between hexoses. In the spectra, intensities of product ions registered in CID mass spectra of all
the isomers 3, 5 and 8, and Z1-type ions were also observed at compounds in the positive mode were quite similar. On the other
m/z 429 ([M
H]
– 180). Their relative intensities were different hand, in the negative mode, they varied substantially, and the
only in the spectrum of compound 8; it was the main ion in the mass main reason was different glycosylation patterns and various
spectrum (Fig. S1b4). Placement of the C-glycosidic bond on the substitutions of acyl group. The common feature of CID mass
basis of data available was not authorized. Compounds 5 and 8 were spectra registered in the positive ion mode was that product ions
probably glycosylated C-6 or C-8, respectively. The second sugar unit characteristic for cleavage bonds in C-bonded sugar (En-type ions)
was attached to the hydroxyl group, most probably at the C-7, not appeared in the second step of fragmentation after the elimina-
C-40 , carbon atom. On the basis of the information obtained from tion of O-linked sugar moiety. In the case of the C-substituted
the mass spectra, it was not possible to characterize structurally both sugars, the relative intensities of the product ions created after
compounds studied. the cleavage of one sugar moiety and acyl group did not suggest
In wheat, the presence of glucose and galactose or glucuronic substitution of hexose at the C-6 or C-8 position. Placement of
acid, rhamnose or arabinose was reported.[27–31] Moreover, these the second sugar on the aglycone moiety at the C-7 carbon or
luteolin triglycosides were isomeric compounds with different O-bonded to the sugar linked to the aglycone moiety with the
patterns of sugar molecules substitution. From the product ions C-glucosidic bond is not clear. The Z1-type product ion (at m/z 443
observed in the mass spectra, we suggested that, in compound or 429 for negative ions and m/z 445 or 431 for positive ions)
1, one sugar is bonded to the hydroxyl group at the C-7 carbon presented different relative intensities; it may suggest presence of
atom, and two others are substituted as hexosyl-O-hexoside to the O-glycosidic bond in the studied molecules.[15] Data obtained
the carbon atom at the C-6 or C-8 position. In the second from positive and negative CID mass spectra did not permit to draw
compound 2, all sugars are substituted to different carbons in consistent conclusions about the complete substitution pattern for
the molecule C-6, C-8 carbons and C-7 via oxygen atom. all acylated natural products (Figs S2–S5). For compound 13, it was
difficult to propose the placement of the C-bonded sugar. The
relative intensity of the Z1 product ion (negative ion mode) provided
Acylated flavone glycosides
evidence of the linkage between both hexoses (Fig. S2). This
In the negative spectra of malonylated glycosides, the intense compound had the following proposed structure: luteolin
elimination of carbon dioxide from protonated molecules C-[(O-caffeoyl-hexosyl)-O-hexoside]. In the first step of fragmen-
336

[M
H–CO2]
was observed. This effect might create problems tation in the positive ion mode, cleavage of aromatic acyl was

[Link]/journal/jms Copyright © 2013 John Wiley & Sons, Ltd. J. Mass Spectrom. 2013, 48, 329–339
Secondary metabolite profiles in wheat leaves

observed on the basis of mass difference, and it was estimated or caffeic acid residue was followed by rupture of hexose
as the caffeoyl group. In this spectrum, ion at m/z 163 was also and creation of the product ion at m/z 443; relative intensities
observed, which corresponded to the elemental composition of these product ions were different. We concluded from the
(C9H9O4), which suggested the protonated caffeoyl residue. Other positive and negative product ion spectra that the caffeoyl
two ions of high intensities at m/z 449 (Y1 ion) and m/z 431 group in the compound 20 was attached to the sugar linked at
(Z1 ion) were also observed. However, these ions at m/z 431 and the non-reducing end of the C-diglycoside. Additionally, the
next at m/z 413 or 395 would also be created after consecutive low relative intensity of the Z1 type ion (m/z 443) provided
elimination of water molecules (En-type ion) from the Y+1 ion at evidence of the presence of the diglycoside at the C-carbon
m/z 449. On the other hand, in the negative ion mode, among atom (Fig. S4). In the fourth acylated compound, it can be
others, strong ions were recognized at m/z 357 and 327 because speculated that the ferulic acid molecule is substituted to
of the rupture of hexose ring (glucose), which is substituted with hexose at the C-carbon atom. However, in the negative product
the C-glycosidic bond linked directly to the aglycone moiety. In ion mass spectra, in the first step, elimination of the feruoyl
the negative ion mode, the fragments created after the cleavage group was observed, and in the second step, cleavage of
of the O-glycosidic bond were not observed at m/z 447, the neutral hexose resulted in the creation of the Z1-type ion
which corresponded to the ion created after the cleavage of at m/z 443, as it is a basic ion in the mass spectrum (Fig. S5).
the glycosidic bond between sugars (Fig. S2). The mass spectra
of the natural product acylated with ferulic acid was luteolin-C,
Flavone C,C-diglycosides
O-diglycoside (hexose and pentose, MW = 756) 25; differences
between the m/z values of product ions permitted to elucidate In the case of C,C-diglycosides, from the elemental composition of
the size of sugars and the acyl group; the relative intensities of [M + H]+/[M
H]
ions in the MS spectra, we proposed presence
product ions did not give structural information. The positive of luteolin and apigenin in the molecules of the conjugates 7 and
product ion MS2 spectrum demonstrated that the feruloyl group 10, respectively. Hexoses and pentoses were recognized in the
is probably substituted at the hexose ring linked through the C sugar parts of these compounds. On the other hand, in MS/MS
bond to the aglycone moiety (Fig. S3). However, in the negative spectra, consecutive elimination of water molecules from the sugar
ion mass spectra, the proposed structure was not confirmed. rings ([M
H]
–nH2O)
or ([M + H]+–nH2O)+ En-type product ions
In the negative ion spectrum, the Z1-type ion was observed at in the negative and positive ion modes were observed (Figs 5 and
m/z 429. It was created after an earlier elimination of the feruoyl S6). The precise placement of pentose (apiose, arabinose or xylose)
group and rupture of the neutral pentose. According to the and hexose (glucose or galactose) at the C-6 or C-8 carbon atom on
information from the literature, we suggest substitution of the basis of m/z values of the En ions and relative intensities
pentose with the O-glycosidic bond to hexose and later through recorded in the MS/MS spectra was not possible.[12] In the negative
the C-carbon atom of the aglycone skeleton.[15] Two other meta- ion spectra of compounds 7 and 10, the elimination of 60, 90 and
bolites acylated with caffeic and ferulic acid had various patterns 120 Da neutral molecules from [M
H]
ions was observed,
of glycosylation; these were chrysoeriol C-[(O-caffeoyl-hexosyl)- but they did not provide evidence for the presence of pentose or
O-hexoside] (20) and chrysoeriol O-hexoside C-(O-feruoyl- hexose at the C-8 or C-6 carbon. From the m/z values and their
hexoside) (29). The mass spectra registered in the positive relative intensities annotated for the product ions in the mass
ion mode for 20 and 29 differed in the presence of the ion at spectra, it was not possible to infer about the structural features
m/z 639 created after the elimination in the first step of the of both natural products. Rules proposed by Waridel and coworkers
hexose residue (Figs S4 and S5). A different order of the frag- for the discrimination of C-6 and C-8 glycosides[12] permitted to
mentation pathway was also allowed; see the ions at m/z 625 estimate the substitution of sugar rings in C-monoglycosides at
and 445 created after the elimination of the feruoyl group and the C-6 carbon for compounds 6, 12, 16 and 21. In earlier studies
later, hexose. In the negative mode, the elimination of the ferulic with fast atom bombardment ionization (or liquid secondary ion

337

Figure 5. CID MS/MS spectra of luteolin C-hexoside C-pentoside 7 (MW = 580): (a) positive ion mode; (b) negative ion mode.

J. Mass Spectrom. 2013, 48, 329–339 Copyright © 2013 John Wiley & Sons, Ltd. [Link]/journal/jms
 A. Wojakowska et al.

MS), where mass spectrometers with electromagnetic sector obtain better product ions annotation accuracy for product
analyzers and high-energy collisions were used, more complete ions registered during MS2 experiments. When analyses were
structural information would be obtained.[35] Relative intensities of performed in positive and negative ion modes, the distinction of
fragments obtained after cleavage of sugar moieties [M + H – 90]+ C-glycosides and O-glycosides was possible. From the mass differ-
and [M + H – 120]+, which were observed in the mass spectra ences between the Yn-type ions, the size of sugar rings was
registered with q-ToF analyzers, did not bring information about estimated, but the identification of type of hexose, pentose or
sugar substitution. deoxyhexose was not allowed. Unambiguous conclusions about
the sugar substitution pattern at the defined carbon atom of the
flavonoid molecule were difficult from the data collected in the
Interpretation of product ion mass spectra registered with chip-
mass spectra registered. Co-elution of many target metabolites
based nanoelectrospray system (TriVersa NanoMate system)
and their differences in abundances preclude the effective applica-
TriVersa NanoMate system from Advion was used in two different tion of the MS mode of analysis only for the identification
ways. In the first case, it was used as the nanoelectrospray source procedure of the natural products studied. It was necessary to use
with split flow from normal UPLC column applied for separation the CID MS/MS techniques for the differentiation of the natural
components of the analyzed sample and registration of the CID products present.
mass spectra; the split flow delivered to the nano-ESI source Acylated with aliphatic or aromatic acids, flavonoid glycoconju-
was 0.099 ml/min. No differences in the relative intensities of gates is better to analyze in the positive ion mode because of
product ions registered in the MS2 mass spectra recorded after possible intensive fragmentation in the negative ions mass spectra
splitting LC eluate to nano-ESI source were registered. Among of the malonyl group in deprotonated molecules of the compounds
others, it was observed for tricin 7-O-hexosyl-O-hexoside 24 or lack of product ions originating from the aromatic residue
(Fig. S7a, b) and luteolin O-(O-caffeoyl-hexoside) C-hexoside 23 created after cleavage of the group from the [M
H]
ions.
(Fig. S8a, b). During the analysis, the same collision energy was During this project, 24 flavonoid C-glycoconjugates and
applied. In the second approach, during the UPLC/MS run, O-glycoconjugates were recognized the first time in the wheat
additional fractionation of the LC column eluate was performed. plant leaf (Table 1). Full structural characterization of these natural
Fractions of 10 ml volume each were collected. In the next step, products directed to sugar identification and placement of their
nano-ESI CID MS/MS experiments with controlled variable collision substitution was impossible on the basis of structural features
energies (from 15 to 35 eV) in the positive ion mode on the consec- recorded in CID MS2 or MS3 spectra. This type of analysis requires
utive fractions were performed, with a flow of 0.099 ml/min; registration of MS2 and MS3 spectra in positive and negative ion
fragmentation was reproducible. The recorded mass spectra were modes. However, it should be underlined that in positive ion mode,
averaged from 60 to about 100 scans. The spray achieved in the we were able to recognize higher number of targeted natural
positive ion mode for the solutions obtained after fractionation in products. On the other hand, spectra registered in both polarities
TriVersa NanoMate was stable without the addition of isopropanol. were submitting supplementary data about structural features of
The content of acetonitrile in the liquid phase was at least 10% after the compounds studied.
fractionation procedure of the LC column eluate into the wells. The elemental composition of the analyzed molecules permitted
After UPLC separation of the studied sample, the relative intensi- to make some conclusion about structures of studied molecules.
ties registered for the product ions were different from those for the For structural characterization of phenolic glycoconjugates, the
mass spectra obtained with normal and nano-ESI source (Figs S7c targeted mode of analyses was necessary, and control of energetic
and S8c). The information about C-glycosylation or O-glycosylation parameters during CID experiments was also very important.
was achieved, also presence of the acyl group was possible to Ramping of the collision energy enables the registration in CID
elucidate from m/z values of the product ions recorded. However, mass spectra product ions of proper relative intensities. After full
signal-to-noise ratio due to the presence of matrix substances was NMR, structural characterization of the consecutive natural
lower in the case of the nano-ESI source than in the mass spectra products and registration of collision-induced mass spectra with
registered during direct analysis of LC column eluate with normal controlled energetic parameters will make it possible to post the
ESI source. In the mass spectrometer equipped with a q-ToF mass spectra in one of existing databases of natural products
analyzer, only MS2 spectra for protonated ions could be recorded, mass spectra. This proposed strategy would be compatible with
whereas in depth spectral trees, which can be obtained with ion that described by the Chemical Analysis Working Group in the
trap–orbitrap instruments, these could not be generated.[36] During Metabolomics Standards Initiative.[37]
our studies, the mass spectra registered with Triversa NanoMate
were only recorded in the positive ion mode. Acknowledgements
This work was partly supported by the National Science Center
Conclusions in Poland (project no. UMO2011/01/N/NZ2/00025 and NCN
2704/B/P01/2011/40) and a research grant from the Institute of
Monocotyledonous plants possess very complicated profiles of Bioorganic Chemistry PAS (no. 501 0901). A. W. is grateful for
flavonoid glycoconjugates, where O-glycosylated and C-glycosylated the scholarship funded within the project from Sub-measure
flavonoids are present. During LC/MS analyses of extracts from 8.2.2 Human Capital Operational Program, co-financed by the
wheat plant leaves, it was possible to propose the elemental European Union Fund.
composition for molecules of the recognized compounds after
the application of mass analyzer with a resolution better than
Supporting information
15 000 FWHM. It enabled the inferences concerning the presence
in the molecules of aromatic acyl groups instead of sugar rings. Supporting information may be found in the version of this
338

However, increased resolution of at least 40 000 would permit to article.

[Link]/journal/jms Copyright © 2013 John Wiley & Sons, Ltd. J. Mass Spectrom. 2013, 48, 329–339
Secondary metabolite profiles in wheat leaves

[20] G.C. Justin, C.M. Borges, M.N.H. Florencio. Electrospray ionization

References tandem mass spectrometry fragmentation of protonated flavone
[1] O.M. Andersen, K.R. Markham. Flavonoids Chemistry, Biochemistry and and flavonol aglycones: a re-examination. Rapid Commun. Mass
Applications. CRC Press Taylor & Francis Group: Boca Raton, 2006. Spectrom. 2009, 23, 237–248.
[2] L.G. Laundry, C.C. Chapple, R.L. Last. Arabidopsis mutants lacking [21] L. Abranko, J.F. Garcia-Reye, A. Molina-Diaz. In-source fragmentation
phenolic sun screens exhibit enhanced ultraviolet-B injury oxidative and accurate mass analysis of multiclass flavonoid conjugates be
damage. Plant Physiol. 1995, 109, 1159–1166. electrospray ionization time of flight mass spectrometry. J. Mass
[3] R.A. Dixon, N.L. Paiva. Stress-induced phenylpropanoid metabolism. Spectrom. 2011, 46, 478–488.
Plant Cell 1995, 7, 1085–1097. [22] J.J.J. van der Hooft, J. Vervoort, R.J. Bino, J. Beekwilder, R.C.H. de Vos.
[4] D. Treutter. Significance of flavonoids in plant resistance: a review. Polyphenol identification based on systematic and robust high-resolu-
Environ. Chem. Lett. 2006, 4, 147–157. tion accurate mass spectrometry fragmentation. Anal. Chem. 2011, 83,
[5] N. Hounsome, B. Hounsom, D. Tomos, G. Edwards-Jones. Plant 409–416.
metabolites and nutritional quality of vegetables. J. Food Sci. 2008, [23] M. Pikulski, J.S. Brodbelt. Differentiation of flavonoid glycoside
73, R48–R65. isomers by using metal complexation and electrospray ionization
[6] D.F. Birt, S. Hendric, W. Wang. Dietary agents in cancer prevention: mass spectrometry. J. Am. Soc. Mass. Spectrom. 2003, 14, 1437–1453.
flavonoids and isoflavonoids. Pharmacol. Ther. 2001, 90, 157–77. [24] M. Stobiecki, A. Skirycz, L. Kerhoas, P. Kachlicki, D. Muth, J. Einhorn,
[7] L. Korkina, V. Kostyuk, C. De Luca, S. Pastore. Plant phenylpropanoids B. Mueller-Roeber. Profiling of phenolic glycosidic conjugates in leaves
as emerging anti-inflammatory agents. Mini-Rev. Med. Chem. 2011, of Arabidopsis thaliana using LC/MS. Metabolomics 2006, 2, 197–219.
11, 823–835. [25] Ł. Marczak, M. Stobiecki, M. Jasińsk, W. Oleszek, P. Kachlicki. Frag-
[8] Q. Zhao, R.A. Dixon. Transcriptional networks for lignin biosynthesis: mentation pathways of acylated flavonoid diglucuronides from
more complex than we thought? Trends Plant Sci. 2011, 16, 227–233. leaves of Medicago truncatula. Phytochem. Anal. 2010, 21, 224–233.
[9] J.K. Prasain, C.C. Wang, S. Barnes. Mass spectrometric methods for [26] A. Staszków, B. Swarcewicz, J. Banasiak, D. Muth, M. Jasiński,
the determination of flavonoids in biological samples. Free Radical M. Stobiecki. LC/MS profiling of flavonoid glycoconjugates
Biol. Med. 2004, 37, 1324–1350. isolated from hairy roots, suspension root cell cultures and seedling
[10] M. Stobiecki, P. Kachlicki. In The Science of Flavonoids, E. Grotevold roots of Medicago truncatula. Metabolomics 2011, 7, 604–613.
(Eds). Springer Science and Business Media, Inc.: New York, 2006, [27] I. Cummins, M. Brazier-Hicks, M. Stobiecki, R. Frański, R. Edwards.
47–69. Selective disruption of wheat secondary metabolism by herbicide
[11] R. March, J. Brodbelt. Analysis of flavonoids: tandem mass spectrom- safeners. Phytochemistry 2006, 67, 1722–1730.
etry, computational methods, and NMR. J. Mass Spectrom. 2008, 43, [28] J.R. Ioset, B. Urbaniak, K. Ndjoko-Ioset, J. Wirth, F. Martin,
1581–1617. W. Gruissem, K. Hostettmann, C. Sautter. Flavonoid profiling
[12] P. Waridel, J.L. Wolfender, K. Ndjoko, K.R. Hobby, H.J. Major, among wild type and related GM wheat varieties. Plant Mol. Biol.
K. Hostettmann. Evaluation of quadrupole time of flight tandem 2007, 65, 645–654.
mass spectrometry and ion trap multiple-stage mass spectrometry [29] Q. Liu, Y. Qiu, T. Beta. Comparison of antioxidant activities of different
for differentiation of C-glycosidic flavonoid isomers. J. Chromatog. A colored wheat grains and analysis of phenolic compounds. J. Agr. Food
2001, 926, 29–41. Chem. 2010, 58, 9235–9241.
[13] V. Vukics, A. Guttman. Structural characterization of flavonoid [30] G. Dinelli, A. Segura-Carretero, R. Di Silvestro, I. Marotti, D. Arraez-Roman,
glycosides by multi-stage mass spectrometry. Mass Spectrom. Rev. S. Benedettelli, L. Ghiselli, A. Fernadez-Gutierrez. Profiles of phenolic
2010, 29, 1–16. compounds in modern and old common wheat varieties determined
[14] B. Abad-García, S. Garmón-Lobato, L.A. Berrueta, B. Galloand, by liquid chromatography coupled with time-of-flight mass spectrome-
F. Vicente. New features on the fragmentation and differentiation try. J. Chromatogr. A 2011, 1218, 7670–7681.
of C-glycosidic flavone isomers by positive electrospray ionization [31] A. Moheb, R.K. Ibrahim, R. Roy, F. Sarhan. Changes in wheat leaf
and triple quadrupole mass spectrometry. Rapid Commun. Mass phenolome in response to cold acclimation. Phytochemistry 2011,
Spectrom. 2008, 22, 1834–1842. 72, 2294–2307.
[15] B. Abad-García, S. Garmón-Lobato, L.A. Berrueta, B. Galloand, [32] F. Cuyckens, M. Claeys. Mass spectrometry in structural analysis of
F. Vicente. Practical guidelines for characterization of O-diglycosyl flavonoids. J. Mass Spectrom. 2004, 39, 1–15.
flavonoid isomers by triple quadrupole MS and their applications [33] Y.L. Ma, F. Cuyckens, H. van den Heuvel, M. Claeys. Mass spectrometric
for identification of some fruit juices flavonoids. J. Mass. Spectrom. methods for the characterisation and differentiation of isomeric
2009, 44, 1017–1025. O-diglycosyl flavonoids. Phytochem. Anal. 2001, 12, 159–165.
[16] R.J. Hughes, T.R. Croley, C.D. Metcalfe, R.E. March. A tandem mass [34] B. Domon, C.E. Costello. A systematic nomenclature for carbohydrate
spectrometric study of selected characteristic flavonoids. Int. J. Mass fragmentations in FAB MS/MS spectra of glycoconjugates. Glycoconj.
Spectrom. 2001, 210, 371–385. J. 1988, 5, 397–409.
[17] F. Ferreres, R. Llorach, A. Gil-Izquierdo. Characterization of the [35] M. Becchi, D. Fraisse. Fast atom bombardment and fast atom
interglycosidic linkage in di-, tri-, tetra- and pentaglycosylated bombardment collision-activated dissociation/mass-analysed ion
flavonoids and differentiation of positional isomers by liquid kinetic energy analysis of C-glycosidic flavonoids. Biol. Mass.
chromatography/electrospray ionization tandem mass spectrometry. Spectrom. 1989, 18, 122–130.
J. Mass Spectrom. 2004, 39, 312–321. [36] J.J.J. van der Hooft, J. Vervoort, R.J. Bino, R.C.H. de Vos. Spectral trees
[18] R.E. March, X.S. Miao, C.D. Metcalfe, M. Stobiecki, Ł. Marczak. A frag- as a robust annotation tool in LC/MS/MS based metabolomics.
mentation study of an isoflavone glycoside, genistein-7-O-glucoside, Metabolomics 2012, 8, 691–703.
using quadrupole time of flight mass spectrometry at high mass [37] L.W. Sumner, A. Amberg, D. Barett, M.H. Bealle, R. Beger, C.A.
resolution. Int. J. Mass Spectrom. 2004, 232, 171–183. Daykin, T.W-M. Fan, O. Fiehn, R. Goodacre, J.L. Griffin, T.
[19] P. Kachlicki, J. Einhorn, D. Muth, L. Kerhoas, M. Stobiecki. Evaluation Hankemeier, N. Hardy, J. Harnly, R. Higashi, J. Kopka, A.N. Lane, J.C.
of glycosylation and malonylation patterns in flavonoid glycosides Lindon, P. Mariott, A.W. Nicholls, M.D. Reilly, J.J. Thaden, M.R. Viant.
during LC/MS/MS metabolite profiling. J. Mass Spectrom. 2008, 43, Proposed minimum reporting standards for chemical analysis.
572–586. Metabolomics 2007, 3, 211–221.
339

J. Mass Spectrom. 2013, 48, 329–339 Copyright © 2013 John Wiley & Sons, Ltd. [Link]/journal/jms