Quantitation and Taste Contribution of Sensory Active Molecules in Oat (Avena sativa L.)

Article abstract of DOI:10.1021/acs.jafc.0c04022

A total of 59 taste-active molecules were quantitated and then rated for their individual taste impact on the basis of dose-over-threshold factors in oat flour (Avena sativa L.). A sensitive high-performance liquid chromatography-tandem mass spectrometry

Full text of DOI:10.1021/acs.jafc.0c04022

Subscriber access provided by CMU Libraries - http://library.cmich.edu
Chemistry and Biology of Aroma and Taste
Quantitation and taste contribution of sensory
active molecules in oat (Avena sativa L.)
Kirsten Günther-Jordanland, Corinna Dawid, and Thomas Hofmann
J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.0c04022 • Publication Date (Web): 11 Aug 2020
Downloaded from pubs.acs.org on August 18, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 42
Journal of Agricultural and Food Chemistry
1
Quantitation and taste contribution of sensory
2
3
active molecules in oat (Avena sativa L.)
§
§
&
§
4
5
Kirsten Günther-Jordanland , Corinna Dawid and Thomas Hofmann *
6
7
§Chair for Food Chemistry and Molecular Sensory Science, Technische Universität
München, Lise-Meitner-Straße 34, D-85354 Freising, Germany,
8
9
&Bavarian Center for Biomolecular Mass Spectrometry, Gregor-Mendel-Straße 4,
85354 Freising, Germany.
10
11
12
13
*
14
To whom correspondence should be addressed
15
16
PHONE
+49-8161/71-2902
FAX
+49-8161/71-2949
17
18
E-MAIL
thomas.hofmann@tum.de
1
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 2 of 42
20 ABSTRACT
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
A total of 59 taste active molecules were quantitated, and then rated for their individual
taste impact on the basis of dose-over-threshold factors (DoT) in oat flour (Avena
sativa L.). A sensitive HPLC-MS/MS-method was developed to quantitate bitter-tasting
steroidal and furostanol-saponins as well as avenanthramides. Four monoglycerides,
five free fatty acids and four saponins, were confirmed for the first time to be major
contributors to the bitter off-taste of oats, among them 1-linoleoyl-rac-glycerol,
1-stearoyl-rac-glycerol, 1-oleoyl-rac-glycerol, 1-palmitoyl-rac-glycerol, linoleic acid,
linolenic acid, oleic acid, palmitic acid and stearic acid, as well as avenacoside A and
B and the recently identified furostanosides 3-(O-α-L-rhamnopyranosyl(1→2)-[β-D-
glucopyranosyl(1→3)-β-D-glucopyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-
glucopyranosyl-(25R)-furost-5-ene-3β,22,26-triol
and
3-(O-α-L-
rhamnopyranosyl(1→2)-[β-D-glucopyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-
glucopyranosyl-(25R)-furost-5-ene-3β,22,26-triol. By means of a Stable Isotope
Dilution Assay (SIDA) quantitated avenanthramides 2c, 2p, 2f, 1p, 1c, 1f, and 3f were
found in concentrations below their thresholds and, therefore, did not contribute to the
bitter sensation of the tested oat flour.
37
38
39
KEYWORDS: oat, Avena sativa, saponins, avenanthramides, avenacosides, free
fatty acids, monoglycerides, taste, bitter, astringent, SIDA, LC-MS/MS.
40
2
ACS Paragon Plus Environment

Page 3 of 42
Journal of Agricultural and Food Chemistry
42 INTRODUCTION
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
In addition to its characteristic attractive aroma, the sensory profile of oats is
driven by its typical astringent and bitter off-taste1. Whereas, multiple attempts have
been made to correlate this off-taste with lipid-derived bitter metabolites, such as free
fatty acids2, 9-hydroxy-10(E),12(Z)-octadecadienoic acid (9-HODE), 13(S)-hydroxy-
9(Z),11(E)-octadecadienoic
acid
(13-HODE)3,
9,10,13-trihydroxy-trans-11-
octadecenoic acid3 and oxidized monoglycerides4, few studies have reached beyond
these compounds. Although oat saponins, such as avenacosides A and B and their
corresponding monodesmosides 26-desglucoavencoside A and B, have been studied
for many years5–10, there is only little knowledge on their sensory power, while other
chemical groups in oats are rather poorly investigated.
In order to identify which non-volatile, key taste molecules impart the bitterness
and astringency of oats, we recently applied the so-called taste dilution analysis
(TDA)11 on fractions isolated from oat flour after sequential solvent fractionation and
separation by means of solid phase extraction followed by high performance liquid
chromatography1. This activity-guided fractionation led to the identification of four
steroidal and furostanol saponins, as well as the three avenanthramides 2c (5), 2p (6),
2f (7) as key bitter and astringent phytochemicals in oats. Sensory evaluation of these
saponins and avenanthramides showed potent bitter sensation with taste recognition
thresholds between
4
and 170 µmol/L, with the saponin 3-(O-α-L-
rhamnopyranosyl(1→2)-[β-D-glucopyranosyl(1→3)-β-D-glucopyranosyl(1→4)]-β-D-
glucopyranosid)-26-O-β-D-glucopyranosyl-(25R)-furost-5-ene-3β,22,26-triol
exhibiting by far the lowest bitter threshold1.
(1)
To the best of our knowledge, a comprehensive mapping of key taste molecules
of oats linking human sensory recognition thresholds has not been made so far.
3
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 4 of 42
67
68
69
70
71
Therefore, this study aimed to position and quantitate previously identified taste active
metabolites and minerals followed by dose-activity-relationships to positioning these
compounds according to their orosensory impact, and validate previous findings by
taste re-engineering studies of the non-volatile sensometabolome of oat flour.
72 MATERIALS AND METHODS
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
Chemicals and Materials. All chemicals were acquired commercially from Merck
(Darmstadt, Germany) and Sigma-Aldrich (Steinheim, Germany), if not stated
elsewise. Sucrose, o-coumaric acid, p-hydroxybenzoic acid and protocatechuic acid
were purchased from Carl Roth (Karlsruhe, Germany), p-coumaric and 4,5-dimethoxy-
2-nitro-benzoic acid from Alfa Aesar (Karlsruhe, Germany), palladium on carbon from
EGA Chemie (Steinheim, Germany), and lauric acid from Fluka (Neu-Ulm, Germany).
As published recently1, saponin reference materials were isolated from oat flour and
avenanthramide standards were synthesized according to the published synthesis
protocol1. Stable-isotope labelled amino acids were acquired from Cambridge Isotope
Laboratories (Andover, USA), (+)-dihydrorobinetine and ginsenoside Rg1 and Rb1 were
obtained from Extrasynthese (Genay, France). Deuterated solvents were supplied by
Euriso-Top (St. Aubin, France). Solvents for HPLC applications were of HPLC grade
(Merck, Darmstadt, Germany), solvents for LC-MS uses were of LC-MS grade
(J.T.Baker, Deventer, Netherlands) whereas solvents used for extraction (Merck,
Darmstadt, Germany) were distilled prior to use. Water for HPLC separation was
purified by means of a Milli-Q water advantage A 10 water system (Millipore, Molsheim,
France). Bottled water (Evian, low mineralization: 405 mg/L) was adjusted to pH 6.4
with aqueous formic acid prior to gustatory analysis. Ethanol (absolute, Merck,
Darmstadt, Germany) was utilized for sensory analysis. Ground oat flour from heat
4
ACS Paragon Plus Environment

Page 5 of 42
Journal of Agricultural and Food Chemistry
93
94
treated oat grains was used and stored at −20 °C until use.
Syntheses of Avenanthramide 3f (13) and Labelled Internal Standards [2H3]-
95
96
Avenanthramide 1f and 2f (16,17) (c.f. Figure 1). Synthesis of avenanthramide 3f
was performed in a stepwise approach following the protocols described in the
literature12–14, with some modifications. In order to generate the educt for the final
condensation reaction of the monomalonate with vanilline, 5-hydroxy-4-methoxy-
anthranilic acid was synthesized from 4,5-dimethoxy-2-nitro-benzoic acid in a two-step
approach. An amount (20 mmol) of 4,5-dimethoxy-2-nitro-benzoic acid was dissolved
in 20 mL of NaOH (6 mol) and heated to 100 °C for 3 h14. After cooling to room
temperature, the solution was dripped into a mixture of concentrated hydrochloric acid
and crushed ice (pH 2). To obtain 5-hydroxy-4-methoxy-2-nitro-benzoic acid as an
orange powder (18.1 mmol, 91% yield) the icy slurry was extracted with ethylacetate,
washed with a saturated saline solution, dried over MgSO4, filtered, freed from solvent
under vacuum, and finally dried at 110 °C.
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
In a second step, reduction of the amino-function was conducted using a catalyst
consisting of palladium on activated charcoal12. The catalyst (10 wt %) and 33 mL of
methanol (dried with molecular sieve 4 Å) were introduced into a three-necked glass
with 12.5 mmol of 5-hydroxy-4-methoxy-2-nitro-benzoic acid. Reduction was
performed for 4 h with gaseous hydrogen from a reservoir balloon. The solid was
filtered off, washed with methanol, and removed from solvent to give 5-hydroxy-
4-methoxy-anthranilic acid (5.6 mmol, 45% yield).
As reported recently13, 5-hydroxy-4-methoxy-anthranilic acid (6 mmol) was mixed in
equimolar ratio with meldrums acid, dissolved in dried toluene (15 mL) and refluxed for
4 h. The mixture was allowed to cool down to room temperature, and treated with
saturated NaHCO3 (15 mL) solution, followed by a stepwise addition of concentrated
5
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 6 of 42
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
HCl (10 mL). The precipitate was filtered off, washed with water and dried at 100 °C to
obtain 5-hydroxy-4-methoxy-2-(2-carboxyacetyl)aminobenzoic acid (2.3 mmol, 36 %
yield). In a final step, 5-hydroxy-4-methoxy-2-(2-carboxyacetyl)aminobenzoic acid
(0.5 mmol), vanilline (0.5 mmol) and a catalytic amount of β-alanine were dissolved in
pyridine (3 mL) and refluxed for 110 min at 115 °C13. The solution was cooled with ice
and acidified with concentrated HCl (3 mL). The resulting solid was filtered, washed
with water and dried in a lab oven at 100 °C. Crude synthesis was purified by means
of preparative HPLC on a Phenyl-Hexyl Luna column (250 x 21.0 mm, 5 µm,
Phenomenex, Aschaffenburg, Germany). Monitoring the effluent at 280 and 340 nm,
chromatography was performed using 0.1% formic acid in water (effluent A) and
acetonitrile (effluent B) at a flow rate of 19 mL/min and isocratic conditions of 52% of
effluent B, respectively. Collected fractions containing the target compounds were
combined and separated from solvent under vacuum. After lyophilisation,
avenanthramide 3f (13) was obtained as white, amorphous powder in purity of more
than 98 % (HPLC-DAD, 1H NMR). The structure was verified by means of LC-MS and
NMR spectroscopy. Isotope labelled standards were synthesized as described above,
using deuterated [2H3]-vanilline for final condensation reaction with the monomalonate.
135
136
137
138
139
140
141
142
143
N-(4‘-Hydroxy-3‘-methoxy-(E)-cinnamoyl)-5-hydroxy-4-methoxy-anthranilic acid
(avenanthramide 3f), 13 (Figure 1): UV/vis (ACN/0.1% HCOOH, 75/25 v/v) λmax
=
248 nm, 348 nm; LC-MS (ESI−) m/z 358.1 ([C18H18NO7−H]−), 314.1 ([C17H17NO5−H]−),
298.1 ([C17H17NO4−H]−); MS/MS (declustering potential (DP)= −90 V) m/z (%) 358.0
(100), 314.0 (13), 298.0 (24), 282.6 (8), 176.0 (36), 149.8 (34); LC-TOF-MS m/z
358.0928 ([M−H]−, measured), calculated for [C18H17NO7−H]− m/z 358.0927; 1H-NMR
(500 MHz, MeOD-d4) δ 3.92* [s, 3H, H-C(10‘)], 3.94* [s, 3H, H-C(8)], 6.57 [d, 1H,
J=15.7 Hz, H-C(8‘)], 6.82 [d, 1H, H-C(5‘)], 7.10 [dd, 1H, J=8.2, 1.8 Hz, H-C(6‘)], 7.23
[d, 1H, J=1.7 Hz, H-C(2‘)], 7.53 [s, 1H, H-C(6)], 7.56 [d, 1H, J=15.7 Hz, H-C(7‘)], 8.42
6
ACS Paragon Plus Environment

Page 7 of 42
Journal of Agricultural and Food Chemistry
144
145
146
147
148
[s, 1H, H-C(3)]; 13C-NMR [125 MHz, MeOD-d4) δ 56.3/56.5 [C(10‘/8)], 104.6 [C(3)],
111.6 [C(2‘)], 113.8 [C(1)], 116.5 [C(5‘)], 118.4 [C(6)], 120.2 [C(8‘)], 123.8 [C(6‘)], 128.2
[C(1‘)], 136.3 [C(2)], 142.7/142.8 [C(7‘/5)], 149.4 [C(3‘)], 150.2 [C(4‘)], 152.3 [C(4)],
166.7 [C(9‘)], 172.7 [C(7)]; *maybe interchangeable.
[10‘-2H3]-N-(4‘-hydroxy-3‘-methoxy-(E)-cinnamoyl)-anthranilic
acid
([2H3]-
2
149
150
151
152
153
154
155
156
157
158
159
160
161
avenanthramide 1f), 16 (Figure 1): LC-MS (ESI−): m/z 315.0 ([C17H12 H3NO5-H]−),
271.1 ([C16H12 H3NO3-H]−), 162.0 ([C8H5NO3-H]−), 133.2 ([C9H10O-H]−); MS/MS (DP=
2
−85 V): m/z (%) 315.0 (100), 271.1 (32), 252.0 (68), 177.0 (10), 133.2 (37); LC-TOF-
MS m/z 315.1067 ([M−H]−, measured), calculated for [C17H12 H3NO5-H]− m/z
2
1
315.1107; H-NMR (400 MHz, MeOD-d4) δ/ppm: δ 6.59 [d, 1H, J=15.6 Hz, H-C(8‘)],
6.83 [d, 1H, J=8.2 Hz, H-C(5‘)], 7,11 [dd, 1H, J=8.2, 1.9 Hz, H-C(6‘)], 7.15 [t, 1H, J=7.6,
0.9 Hz, H-C(5)], 7.24 [d, 1H, J=1.9 Hz, H-C(2‘)], 7.57 [t, 1H, J=8.0, 1.6 Hz, H-C(4)],
7.60 [d, 1H, J=15.5 Hz, H-C(7‘)], 8.11 [dd, 1H, J=8.0, 1.5 Hz, H-C(6)], 8.69 [dd, 1H,
J=8.4, 0.7 Hz, H-C(3)]; 13C-NMR [100 MHz, MeOD-d4) δ/ppm: δ 111.63 [C(2‘)], 116.42
[C(5‘)], 117.65 [C(1)], 119.52 [C(8‘)], 121.58 [C(3)], 123.93/123.97 [C(5/6‘)], 127.96
[C(1‘)], 132.64 [C(6)], 135.18 [C(4)], 142.74 [C(2)], 144.03 [C(7‘)], 149.40 [C(3‘)],
150.39 [C(4‘)], 167.16 [C(9‘)], 171.60 [C(7)].
[10‘-2H3]-N-(4‘-hydroxy-3‘-methoxy-(E)-cinnamoyl)-5-hydroxyanthranilic
acid
2
162
([2H3]-avenanthramide 2f), 17 (Figure 1): LC-MS (ESI−) m/z 331.0 ([C17H12 H3NO6-H]−),
2
163
164
165
166
167
168
169
287.0 ([C16H12 H3NO4-H]−), 178.0 ([C8H5NO4-H]−), 133.2 ([C9H10O-H]−); MS/MS (DP=
−85 V): m/z (%) 331.0 (100), 287.0 (41), 268.0 (65), 178.0 (7), 160.0 (32), 133.2 (52);
LC-TOF-MS m/z 331.1018 ([M−H]−, measured), calculated for [C17H12 H3NO6-H]− m/z
2
1
331.1056; H-NMR (400 MHz, DMSO-d6) δ/ppm: δ 6.56 [d, 1H, J=15.6 Hz, H-C(8‘)],
6.95 [d, 1H, J=8.2 Hz, H-C(5‘)], 7.02 [dd, 1H, J=3.0, 9.0 Hz, H-C(4)], 7.10 [dd, 1H,
J=2.0, 7.9 Hz, H-C(6‘)], 7.11 [d, 1H, J=1.9 Hz, H-C(2‘)], 7.37 [d, 1H, J=2.9 Hz, H-C(6)],
7.42 [d, 1H, J=15.4 Hz, H-C(7‘)], 8.32 [d, 1H, J=9.0 Hz, H-C(3)]; 13C-NMR (101 MHz,
7
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 8 of 42
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
DMSO-d6) δ/ppm: δ 111.19 [C(5‘)], 113.77 [C(2‘)], 116.40 [C(6)], 118.72 [C(1)], 119.75
[C(8‘)], 120.60 [C(6‘)], 120.71 [C(4)], 122.40 [C(3)], 127.31 [C(1‘)], 132.65 [C(2)],
140.50 [C(7‘)], 146.55 [C(3‘)], 149.44 [C(4‘)], 152.44 [C(5)], 163.37 [C(9‘)], 168.97
[C(7)].
Quantitation of Avenanthramides (5−13) by means of Stable Isotope
Dilution Assay (SIDA). Sample Preparation. A sample weight (7 g) of oat flour was
placed into a conical flask, spiked with an internal standard solution (150 µL) of [2H3]-
avenanthramide 1f (16, 174 mg/L) and [2H3]-avenanthramide 2f (17, 220 mg/L),
dissolved in 70% acetonitrile and equilibrated for 30 min. In a first step, samples were
defatted with n-hexane (60 mL) and centrifuged. After decantation, the residue was
taken up in 140 mL of methanol/water (70/30, v/v, aqueous phase adjusted to pH 4.0)
and extracted for 2 h under stirring. The liquid phase was freed from solvent under
vacuum and lyophilized to give the methanol/water extract. The extract was taken up
in a mixture of acetonitrile/0.1 % formic acid (40/60, v/v), membrane filtered and
analyzed by means of HPLC-MS/MS.
LC-MS/MS Analysis. An aliquot (10 µL) of the prepared sample was injected into an
LC-MS/MS system combined with a Luna C18 PhenylHexyl (150 × 2.0 mm i.d., 5 µm,
Phenomenex, Aschaffenburg, Germany) equipped with a guard column of the same
type. Eluent A was composed of 0.1 % formic acid in acetonitrile, eluent B of 0.1 %
formic acid in water. Using a flow rate of 300 µL/min, the system was operated at 25 °C
starting with 40 % A under isocratic conditions for 1 min, then increasing content A to
70 % within 1 min (holding for 2 min), followed by an increase to 100 % A within 2 min
and keeping isocratic conditions for 3 min. Eluent was pumped down again to 40 % A
within 2 min and held isocratically for further 3 min. Analysis was performed in ESI−
mode using the mass transition and declustering potential (DP in V), collision energy
(CE in V), cell exit potential (CXP in V), and entrance potential (EP in V) as follows:
8
ACS Paragon Plus Environment

Page 9 of 42
Journal of Agricultural and Food Chemistry
196
197
198
199
200
201
202
203
204
avenanthramide 2c (5) (m/z 314.2 → 178.0; -65/-16/-11/-10); avenanthramide 2p (6)
(m/z 298.1 → 253.9; -50/-22/-15/-10); avenanthramide 2f (7) (m/z 328.0 → 283.8;
-95/-24/-19/-10); avenanthramide 1p (8) (m/z 282.0 → 238.0; -85/-24/-7/-10);
avenanthramide 1c (9) (m/z 298.0 → 161.7; -65/-16/-11/-10); avenanthramide 1f (10)
(m/z 312.0 → 251.7; -90/-34/-17/-10); avenanthramide 1s (11) (m/z 342.0 → 298.1;
-85/-24/-9/-10); avenanthramide 2s (12) (m/z 358.1 → 314.1; -80/-24/-9/-10);
avenanthramide 3f (13) (m/z 358.1 → 175.8; -90/-28/-11/-10); [2H3]-avenanthramide 2f
(16) (m/z 331.0 → 286.8; -85/-24/-19/-10); [2H3]-avenanthramide 1f (17) (m/z 315.0 →
271.1; -85/-24/-7/-10).
205
206
207
208
209
210
Calibration. Internal standards [2H3]-avenanthramide 2f (16) and [2H3]-avenanthramide
1f (17) were mixed with analytes 5−13 in molar ratios from 0.03 to 6.8 keeping the
concentration of the internal standard constant. Calibration curves were prepared by
plotting peak area ratios of analyte/internal standard against concentration ratios of
analyte/internal standard. Application of linear regression led to linear responses
showing correlation coefficients of >0.99 each.
211
212
213
214
215
216
217
218
219
Method Validation. To investigate the repeatability (intraday precision) of oat samples,
four independent repetitions were analyzed using three replicates each. Precision,
given as relative standard deviation (RSD), was confirmed to be between ±3% (5) and
±9% (9 and 10) according to the individual avenanthramide. Linearity was determined
in the range between 0.0025−0.4753 mg/L. To determine recovery, oat flour samples
were spiked with methanolic solutions containing known amounts of 5−10 (c.f. Table
1) at increasing concentrations and analyzed in duplicates as described before. In
order to calculate recovery, baseline samples were defined from a 4-fold determination
of unspiked oat flour.
220
Quantitation of Saponins (1−4). Sample Preparation. A sample weight (7 g) of
9
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 10 of 42
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
oat flour was placed into a conical flask, spiked with a methanolic solution of internal
standards ginsenoside Rg1 (14, 400 µL of 577 mg/L) and ginsenoside Rb1 (15, 450 µL
of 487 mg/L) and equilibrated for 30 min. Samples were extracted twice under stirring
for 10 h using 250 mL of a methanol/water mixture (70/30, v/v; aqueous phase adjusted
to pH 4.0) each. Combined liquid phases were separated from solvent under vacuum,
freeze-dried and taken up to 10 mL of an acetonitrile/0.1 % formic acid mixture (25/75,
v/v). After membrane filtration and, if appropriate dilution, samples were analyzed by
means of HPLC-MS/MS.
LC-MS/MS Analysis. An aliquot (5 µL) of the prepared sample was injected into an LC-
MS/MS system combined with a 150 × 2.0 mm i.d., 3 µm, Nucleodur C18 Pyramid
(Machery-Nagel, Düren, Germany) equipped with a guard column of the same type.
Eluent A consisted of 0.1 % formic acid in acetonitrile, eluent B of 0.1 % formic acid in
water. Using a flow rate of 300 µL/min, the system was operated at 40 °C starting with
25 % A under isocratic conditions for 1 min, then increasing content A to 65 % within
5 min, followed by an increase to 100 % A within 2 min and keeping isocratic conditions
for 2 min. Eluent was pumped down to 25 % A within 2 min and held isocratically for
further 4 min. Analysis was performed in ESI− mode using the mass transition and
declustering potential (DP in V), collision energy (CE in V), cell exit potential (CXP in
V), and entrance potential (EP in V) as follows: 3-(O-α-L-rhamnopyranosyl(1→2)-[β-D-
glucopyranosyl(1→3)-β-D-glucopyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-
glucopyranosyl-(25R)-furost-5-ene-3β,22,26-triol (1) (m/z 1225.7 → 901.2qn=quantifier
;
-225/-74/-55/-10; 1225.7 → 900.9ql=qualifier; -225/-76/-55/-10); 3-(O-α-L-rhamnopyra-
nosyl(1→2)-[β-D-gluco-pyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-gluco-
pyranosyl-(25R)-furost-5-ene-3β,22,26-triol
(2)
(m/z
1064.7
→
902.5qn;
-185/-62/-51/-10; 1064.7 → 901.9ql; -185/-62/-51/-10) were quantitated using
ginsenoside Rg1 (14) (m/z 799.6 → 89.0qn; -200/-50/-5/-10; 799.5 → 637.5ql;
10
ACS Paragon Plus Environment

Page 11 of 42
Journal of Agricultural and Food Chemistry
247
248
249
250
251
252
-180/-36/-31/-10); avenacoside B (3) (m/z 1223.5 → 899.5qn; -200/-72/-41/-10; 1223.5
→ 1061.6ql; -200/-72/-53/-10; avenacoside A (4) (m/z 1062.5 → 899.9qn; -
235/-60/-43/-10; 1062.5 → 754.3ql; -235/-64/-35/-10); protodioscin 1047.5 → 901.4qn;
-175/-126/-9/-10; 1047.5 → 901.4ql; -175/-126/-9/-10) were quantitated using
ginsenoside Rb1 (15) (m/z 1107.6 → 783.6qn; -210/-64/-37/-10; 1107.6 → 945.4ql; -
210/-60/-43/-10) as internal standard (c.f. Figure 2).
253
254
255
256
257
258
259
260
261
262
263
264
265
266
Calibration. Internal standards of ginsenoside Rg1 (14, 577 mg/L) and
ginsenoside Rb1 (15, 487 mg/L) were diluted 1:10 using acetonitrile/0.1 % formic acid
(25/75, v/v) and then added to the analytes 1−4 in molar ratios of 0.3 to 8, keeping the
concentration of the internal standard constant. Calibration curves were prepared by
plotting peak area ratios of analyte/internal standard against concentration ratios of
analyte/internal standard using linear regression (R2 > 0.99).
Method Validation. To investigate the repeatability (intraday precision) of oat samples,
three independent repetitions were analyzed using three replicates each. Method
precision, given as relative standard deviation (%), was confirmed as follows: 1 (6%),
2 (8%), 3 (5%), 4 (11%). Linearity was determined in the range between
0.2−51.1 mg/L. To determine recovery, oat flour samples were spiked with methanolic
solutions containing known amounts of 1−4 (c.f. Table 2) at increasing concentrations
and analyzed in duplicates as described before. In order to calculate recovery, baseline
samples were defined from a 3-fold determination of unspiked oat flour.
267
268
269
270
271
Quantitation of Free Fatty Acids (18−24) by Means of SIDA. Sample
Preparation. In order to quantitate free fatty acids, a literature procedure using a stable
isotope dilution analysis was followed15. An amount of oat flour (12 g) was placed into
a conical flask, spiked with 100 µL each of an isopropanolic solution of internal
standards [13C]-palmitic acid (25, 480 mg/L) and [2H2]-9,10-oleic acid (26, 720 mg/L)
11
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 12 of 42
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
and equilibrated for 30 min. Samples were extracted three times with 300 mL of
methanol for 2 h under stirring. Liquid phases were combined, freed from solvent under
vacuum, and separated from di- and triglycerides by means of solid phase extraction
(SPE). The residue was dissolved in 5 mL of methanol and applied onto a RP-18
material (Strata C18-E Giga Tubes, 10 g/60 mL, Phenomenex, Aschaffenburg,
Germany). Elution of fatty acids was carried out using 60 mL of methanol. The resulting
fraction was narrowed down under vacuum, transferred into a 10 mL volumetric flask
and filled up with starting conditions of the HPLC gradient. After membrane filtration
and dilution, free fatty acids were determined by means of HPLC-MS/MS using isotope
labelled standards.
LC-MS/MS Analysis. An aliquot (5 µL) of the prepared sample was injected into
an LC-MS/MS system combined with a 150 × 2.0 mm i.d., 4 µm, Polar-RP ODS C18
(Phenomenex, Aschaffenburg, Germany) equipped with a guard column of the same
type. Eluent A consisted of isopropanol: acetonitrile/ 0.1 % HCOOH (v/v, 60:40), eluent
B of 0.1 % formic acid in water. Using a flow rate of 400 µL/min, the system was
operated at 40 °C starting with 40 % A under isocratic conditions for 2 min, then
increasing content A to 65 % within 8 min, continuing to 100 % A within 2 min and
keeping isocratic conditions for 4 min, pumping down to 40 % A within 2 min and
keeping constant for further 3 min. Analysis was performed in negative ESI mode using
the mass transition and declustering potential (DP in V), collision energy (CE in V), cell
exit potential (CXP in V), and entrance potential (EP in V) as follows: lauric acid (18)
(m/z 199.0 → 199.1; -70/-14/-1/-10); myristic acid (19) (m/z 227.1 → 227.0; -
85/-12/-15/-10); palmitic acid (20) (m/z 255.2 → 255.2; -80/-12/-7/-10); stearic acid (21)
(m/z 283.2 → 283.2; -80/-14/-17/-10) were quantitated using [13C]-palmitic acid (25)
(256.1 → 256.2; -95/-16/-7/-10); oleic acid (22) (m/z 281.2 → 281.4; -100/-14/-7/-10);
linoleic acid (23) (m/z 279.2 → 279.2; -85/-14/-17/-10) and linolenic acid (24) (m/z
12
ACS Paragon Plus Environment

Page 13 of 42
Journal of Agricultural and Food Chemistry
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
277.2 → 277.1; -80/-14/-7/-10) were quantitated using [2H2]-9,10-oleic acid (26) (m/z
283.2 → 283.1; -100/-14/-7/-10) as internal standard (c.f. Figure 3). An ionisation of
the fatty acid molecules by negative electrospray ionisation could be achieved during
optimisation of MS-parameters, however a fragmentation was neither achieved by ESI
nor by APCI ionisation. Nevertheless, the free fatty acids were present in sufficient
amounts in the oat flour and separated well given the selectivity of the HPLC method,
therefore, the detection of a free fatty acid was obtained by its unfragmented molecular
ion.
Calibration. Internal standards [13C]-palmitic acid (25) and [2H2]-9,10-oleic acid
(26) were mixed with analytes 18−24 in molar ratios of 0.03 to 6.8 keeping the
concentration of the internal standard constant. Calibration curves were set up by
plotting peak area ratios of analyte to internal standard against concentration ratios of
analyte to internal standard. Execution of linear regression led to linear responses
showing correlation coefficients of R2 > 0.99 each.
Quantitation of Monoglycerides and oxidized fatty acids by means of LC-
MS/MS. Sample Preparation. According to a literature protocol15 monoglycerides and
oxidized fatty acids were quantitated as follows with some modifications: 10 g of oat
flour was weighed into a conical flask and extracted three times with methanol (3 ×
250 mL) for 1 h under stirring. Extracts were combined and freed from solvent under
vacuum. To separate from di- and triglycerides, a solid phase extraction (SPE) was
carried out on a RP-18 material (Strata C18-E Giga Tubes, 10 g/60 mL, Phenomenex,
Aschaffenburg, Germany). Therefore, the residue was taken up in 5 mL of methanol,
applied onto the cartridge, and the fatty acids were eluted with 60 mL of methanol. The
resulting fraction was concentrated under vacuum, transferred into a 10 mL volumetric
flask and filled up with 40% of eluent A. After membrane filtration and dilution, free fatty
acids were determined by means of HPLC-MS/MS using external calibration.
13
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 14 of 42
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
LC-MS/MS Analysis. An aliquot (2 µL) of the prepared sample was injected into
an LC-MS/MS system combined with a Polar-RP ODS C18 (150 × 2.0 mm i.d., 4 µm,
Phenomenex, Aschaffenburg, Germany) equipped with a guard column of the same
type. Eluent A consisted of isopropanol: acetonitrile/ 0.1 % HCOOH (v/v, 60:40), eluent
B of 0.1 % formic acid in water. A flow rate of 400 µL/min was used for both analyte
groups.
For analysis of monoglycerides (29-33) the system was operated at 40 °C starting
at 40 % A increasing to 70 % within 2 min, rising to 100 % A within 5 min and keeping
isocratic conditions for a further 5 min, pumping down again to 40 % A within 3 min
and holding constant conditions for 3 min. Operation of HPLC gradient for analysis of
oxidized fatty acids (27/28) was run at same temperature starting at 40 % A and kept
at constant conditions for 2 min, then increasing to 65 % within 8 min, rising further to
100 % A within 2 min and holding isocratic conditions for 4 min before going back to
40 % A within 2 min and keeping for further 3 min. The oxidized fatty acids 9-HODE
(27) and 13-HODE (28) were not chromatografically separated and therefore
integrated together (27/28). Analysis was performed in positive ESI mode for
monoglycerides and negative mode for oxidized fatty acids using the mass transition
and declustering potential (DP in V), collision energy (CE in V), cell exit potential (CXP
in V), and entrance potential (EP in V) as follows: 1-myristoyl-rac-glycerol (29) (m/z
303.2 → 284.9; 91/13/8/10); 1-palmitoyl-rac-glycerol (30) (m/z 331.2 → 312.9;
86/13/10/10); 1-stearoyl-rac-glycerol (31) (m/z 359.2 → 341.3; 71/17/10/10); 1-oleoyl-
rac-glycerol (32) (m/z 357.2 → 339.0; 51/17/12/10); 1-linoleoyl-rac-glycerol (33) (m/z
355.0 → 262.8; 41/13/8/10); 9-HODE/ 13-HODE (27/28) (m/z 295.2 → 295.0;
-100/-12/-19/-10); 13-HpODE (m/z 311.3 → 112.9; -95/-22/-1/-10) (c.f. Figure 4).
Calibration. Quantification was performed using external calibration of monoglyceride
and oxidized fatty acid reference standards. Monoglyceride calibration standards were
14
ACS Paragon Plus Environment

Page 15 of 42
Journal of Agricultural and Food Chemistry
350
351
352
353
354
355
356
prepared at concentrations of 0.2−335 mg/L using 1-myristoyl-rac-glycerol (29), 1-
palmitoyl-rac-glycerol (30), 1-stearoyl-rac-glycerol (31), 1-oleoyl-rac-glycerol (32) and
1-linoleoyl-rac-glycerol (33). For oxidized fatty acids calibration standards were made
up from 9-HODE/ 13-HODE (27/28) and 13-hydroperoxy-9Z,11E-octadecadienoic acid
(13-HpODE) at concentrations of 0.2−165 mg/L. Calibration curves were prepared by
plotting peak areas of the analyte against the analyte concentration. Correlation
coefficients of linear regression revealed values greater than 0.99 each.
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
Quantitation of Free Amino Acids by Means of SIDA. Free amino acids were
quantitated by means of HILIC-MS/MS using a stable isotope dilution analysis (SIDA)
following a literature protocol16. Sample Preparation. A sample weight (10 g) of oat
flour was placed into a conical flask, spiked with 250 µL of an internal standard solution
containing the isotope labelled amino acids in acetonitrile (0.1 mg/L) and equilibrated
for 30 min. Samples were defatted with 200 mL of n-hexane under stirring followed by
centrifugation for 25 min at 4 °C (Avanti J-E, Beckman Coulter, Krefeld, Germany) and
decantation of the solvent. Then, samples were extracted two times with 200 mL of
methanol/water (50/50, v/v; aqueous phase adjusted to pH 4.0) for 2 h under stirring.
Combined extracts were treated on a rotary evaporator at 40 °C to remove solvents
and lyophilized afterwards. A weight of 60 mg of the resulting extract was transferred
into a volumetric flask (5 mL), filled up with 85% of eluent A, filtered and diluted if
required. Quantification of amino acids was done by means of HPLC-MS/MS and
stable isotope dilution assay.
LC-MS/MS Analysis. An aliquot (2 µL) of the prepared sample was injected into
an LC-MS/MS system combined with a TSK-gel Amide 80 HILIC column (150 × 2.0
mm i.d., 3 µm, Tosoh Bioscience, Stuttgart, Germany) equipped with a guard column
of the same type. Eluent A consisted of 5 mmol/L ammonium acetate buffer (pH 3.0,
adjusted with acetic acid) in 95 % acetonitrile, eluent B of 5 mmol/L ammonium acetate
15
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 16 of 42
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
buffer (pH 3.0, adjusted with acetic acid) in water. Using a flow rate of 200 µL/min, the
system was operated at 40 °C starting with 85 % A under isocratic conditions for 3 min,
then increasing content A to 45 % within 7 min, continuing to 100 % A within 4 min and
keeping isocratic conditions for 4 min, pumping back to initial 85 % A within 3 min and
keeping constant for further 5 min. Analysis was performed in positive ESI mode using
the mass transition and declustering potential (DP in V), collision energy (CE in V), cell
exit potential (CXP in V), and entrance potential (EP in V) as follows: L-alanine (m/z
90.1 → 90.0; 26/5/6/10); L-alanine-[13C3] (m/z 93.0 → 93.0; 41/5/5/10); L-arginine (m/z
175.0 → 70.1; 41/35/4/10); L-arginine-[13C6] (m/z 181.0 → 74.0; 78/36/5/10); L-
asparagine (m/z 132.9 → 73.9; 1/19/6/10); L-asparagine-[15N2]
(m/z 135.0
→
89.0; 39/14/5/10); L-aspartic acid (m/z 134.0 → 74.0; 46/19/6/10); L-aspartic acid-[13C4-
15N] (m/z 139.1 → 92.0; 39/14/10/10); L-glutamine (m/z 147.0 → 84.0; 46/23/6/10);
L-glutamine-[13C5] (m/z 152.0 → 88.0; 43/23/10/10); L-glutamic acid (m/z 148.0 → 84.0;
31/23/6/10) was quantitated over L-glutamine-[13C5]; L-histidine (m/z 156.1 → 110.0;
41/21/8/10); L-histidine-[13C6] (m/z 162.0 → 115.0; 46/21/10/10); L-isoleucine (m/z
132.0 → 86.0; 41/15/6/10); L-isoleucine [13C6] (m/z 138.0 → 91.0; 47/14/10/10); L-
leucine (m/z 132.0 → 86.0;
41/15/6/10); L-leucine-[13C2] (m/z 134.1 → 87.2;
44/14/10/10); L-lysine (m/z 147.0 → 84.0;
46/23/6/10);
L-lysine-[13C6-15N2] (m/z 155.0 → 90.0; 44/23/10/10); L-methionine (m/z 150.0
→ 104.0;
31/15/8/10); L-methionine-[2H3] (m/z 153.1 → 107.0; 50/14/10/10);
L-phenylalanine (m/z 166.0 → 120.0; 51/19/10/10); L-phenylalanine-[2H5] (m/z 171.1
→ 125.0;
48/19/10/10); L-proline (m/z 116.0 → 70.0;
21/21/4/10);
L-proline-
[13C5-15N] (m/z 122.0 → 75.0;
73/25/5/10); L-serine (m/z 106.0
→
60.0;
26/17/4/10); L-serine-[13C3] (m/z 109.0 → 62.0;
38/16/5/10); L-threonine
(m/z 120.0 → 55.9; 36/25/4/10); L-threonine-[13C4-15N] (m/z 125.0 → 78.0;
32/14/5/10); L-tryptophan (m/z 205.1 → 146.0;
41/25/12/10);
16
ACS Paragon Plus Environment

Page 17 of 42
Journal of Agricultural and Food Chemistry
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
L-tryptophan-[2H5] (m/z 210.0 → 150.0;
40/37/10/10); L-tyrosine (m/z 182.1
→ 91.0;
26/41/6/10); L-tyrosine-[2H4]
186.0 → 140.0;
38/19/10/10);
L-valine
(m/z 118.0 → 72.1; 21/15/6/10); L-valine-[13C5-15N] (m/z 124.0
→
77.0; 64/31/ 10/10);
Calibration. Internal standards were mixed with analytes in molar ratios of 0.01 to
7, keeping the concentration of the internal standard constant. Calibration curves were
set up by plotting peak area ratios of analyte to internal standard against concentration
ratios of analyte to internal standard. Execution of linear regression led to linear
responses showing correlation coefficients of >0.99 each.
Quantitation of Polyphenols by means of SIDA. Sample Preparation. For the
quantification of polyphenols a modified literature protocol was used17. Oat flour (15 g)
was placed into a conical flask, spiked with an internal standard solution (250 µL, 0.015
mg/L) and equilibrated for 30 min. Samples were defatted with n-hexane (300 mL) for
30 min in an ultrasonic bath and centrifuged. After decantation, the residue was taken
up two times in 250 mL of methanol/water (70/30, v/v, aqueous phase adjusted to pH
4.0) and extracted for 30 min in an ultrasonic bath. The solvent was removed from the
liquid phase under vacuum and lyophilized to give the methanol/water extract. The
extract was taken up in starting conditions of the HPLC gradient, filtered and analyzed
by means of HPLC-MS/MS. Quantification was performed by unlabelled and labelled
internal standards.
LC-MS/MS Analysis. An aliquot (5 µL) of the prepared sample was injected into an LC-
MS/MS system combined with a 150 × 2.0 mm i.d., 3 µm, Luna C18 (Phenomenex,
Aschaffenburg, Germany) equipped with a guard column of the same type. Eluent A
was composed of 0.1 % formic acid in acetonitrile, eluent B of 0.1 % formic acid in
water. Using a flow rate of 300 µL/min, the system was operated at 40 °C starting from
40 % A to 70 % A within 2 min. Effluent was increased to 100 % within 5 min (holding
17
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 18 of 42
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
for 5 min), followed by a pump down to 40 % A within 3 min and holding conditions for
a further 3 min. Analysis was performed in negative ESI mode using the mass transition
and declustering potential (DP in V), collision energy (CE in V), cell exit potential (CXP
in V), and entrance potential (EP in V) as follows: p-coumaric acid (m/z 162.98 →
119.0; -55/-20/-1/-10); caffeic acid (m/z 178.96 → 135.0;
-50/-22/-9/-10) and ferulic
acid (m/z 193.01 → 133.8; -55/-22/-7/-10) were quantitated using o-coumaric acid
(m/z 162.98 → 119.1; -55/-20/-1/-10); protocatechuic acid (m/z 152.96 → 108.8;
-20/-20/-5/-10) and gallic acid (m/z 168.96 → 124.8; -50/-22/-7/-10) were quantitated
using [2H5]-gallic acid ethyl ester (m/z 201.92 → 123.9; -85/-32/-7/-10); vanilline (m/z
151.03 → 135.8; -50/-18/-9/-10); p-hydroxybenzoic acid (m/z 136.94 → 92.8;
-50/-22/-7/-10); p-hydroxybenzaldehyde (m/z 121.00 → 91.80; -90/-34/-5/-10);
dihydrorobinetine (m/z 302.97 → 284.8; -60/-14/-17/-10) and quercetine (m/z 301.02
→ 150.8; -80/-30/-9/-10) were quantitated using [2H3]-methoxybenzoic acid (m/z
154.04 → 91.8;
-65/-30/-5/-10) while sinapinic acid (m/z 223.02 → 163.9;
-60/-20/-9/-10) quantification was done using [2H5]-caffeic acid ethyl ester (m/z 211.98
→ 133.9; -80/-28/-7/-10).
444
445
446
447
448
Calibration. Internal standards were mixed with analytes in molar ratios from 0.01 to
2.3 keeping the concentration of the internal standard constant. Calibration curves
were prepared by plotting peak area ratios of analyte/internal standard against
concentration ratios of analyte/internal standard. Application of linear regression led to
linear responses showing correlation coefficients of >0.99 each.
449
450
451
452
Quantitation of Minerals by means of HPIC. Sample Preparation. A defined amount
of oat flour (1.5 g) was placed into a conical tube (50 mL, Carl Roth, Karlsruhe,
Germany), filled with 15 mL deionized water and homogenized for 75 sec at 9000 rpm
by means of an Ultra-Turrax T25 digital (Ika Labortechnik, Staufen, Germany).
18
ACS Paragon Plus Environment

Page 19 of 42
Journal of Agricultural and Food Chemistry
453
454
455
456
457
458
459
460
461
462
463
464
465
466
Samples were centrifuged for 30 min at 4 °C and 4700 rpm using a Heraeus Multifuge
X3 FR Zentrifuge (Thermo Fisher Scientific, Waltham, USA). The aqueous supernatant
was transferred into a volumetric flask, filled up to the mark with deionized water and,
finally, filtered. Analysis of the samples was performed by means of high performance
ion chromatography (HPIC), following a modified literature procedure18, using an
Dionex ICS-2000 Ion Chromatography System (Thermo Fisher Scientific, Waltham,
USA) consisting of a DS 6 heated conductivity cell detector, a dual piston pump, a
CSRS 300 suppressor cell, an eluent generator RFIC EluGen Cartridge; EGC II MSA
and an AS-AP autosampler. Using 5 mM methanesulfonic acid as mobile phase and a
flow rate of 300 µL/min the system was operated at 40 °C under isocratic conditions
for 20 min (injection volume 10 µL). Quantitative data were generated by means of
external calibration at concentration levels between 0.8 and 250 mg/L. To quantitate
cations, reference standards of potassium chloride, sodium chloride, magnesium
chloride, and calcium chloride were used.
467
468
469
470
471
472
473
474
475
476
Quantitation of Carbohydrates by means of Enzymatic Assay. Sample
Preparation. Quantification of D-(+)-glucose, D-(+)-fructose und sucrose was
performed using an enzymatic assay kit for glucose/fructose/sucrose (R-Biopharm,
Darmstadt, Germany). An amount of 10 g oat flour was placed into a volumetric flask
(100 mL), and filled up with 60 mL deionized water. After the addition of 5 mL of
Carrez I and II solutions for protein precipitation, the flask was filled up to the mark with
deionized water and then filtered. Samples were treated according to the protocol of
the assay kit and measurements were recorded on a UV-2401 PC UV-Vis
spectrophotometer (Shimadzu, Kyoto, Japan) at 340 nm. The quantitative data are
given as the mean of three replicates (RSD for each data point <4.1 %).
477
19
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 20 of 42
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
Sensory Experiments. Panel Training. According to literature19 and in order to
familiarize the panelists (six female, six male; 22-35 years in age) with the taste
language, assessors were trained on different qualities of oral sensation. The
panelists, who had no history of known taste disorders and who gave informed consent
to participate in the sensory studies, took part in weekly sessions for at least two years.
Aqueous solutions (2 mL; pH 6.4) of the following standard taste components were
used for training: sucrose (50 mM) for sweet taste, L-lactic acid (20 mM) for sour taste,
NaCl (20 mM) for salty taste, caffeine (1 mM) for bitter taste, and sodium glutamate
(3 mM) for umami taste. For the astringent oral sensation, tannic acid (0.05 %) was
used to train the panelists using the so-called half-tongue test.19,20 All sensory studies
were performed in a sensory panel room at 22−25 °C in three different sessions and
using nose clips to avoid cross-modal interactions with odorants.
Pretreatment of Food Fractions and Taste Compounds for Sensory Analysis.
Prior to sensory analysis, fractions or compounds, either commercially available or
isolated from synthesis, were suspended in water and freeze dried twice, after
1
removing the volatiles in high vacuum (<5 mPa). H NMR and ion chromatography
revealed that fractions treated by that procedure are essentially free of solvents and
buffer compounds.
Taste Profile Analysis. An amount of methanol/water fraction II (2.5 g) isolated
from oat flour as reported recently1, was suspended in water (25 mL) containing 3%
ethanol and adjusted to pH 6.4. After a dilution of 1:10, the aqueous solution was
presented to the trained sensory panel to evaluate the taste sensations sweet, sour,
umami, salty, bitter, and astringent on a linear intensity scale from 0 (not detectable)
up to 5 (strongly detectable).
Taste Recognition Threshold Concentrations. Following the protocol reported
recently21, commercial or synthesized standard compounds were first proven to be
20
ACS Paragon Plus Environment

Page 21 of 42
Journal of Agricultural and Food Chemistry
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
pure by means of 1H-NMR spectroscopy and HPLC-MS and, then, taken up in bottled
water (pH 6.4, 2% ethanol) to determine threshold concentrations for sweet, sour,
salty, bitter and umami, as well as astringent taste sensation. Taste recognition
thresholds were recorded by means of the half-tongue19,22 test in order of increasing
concentration. Taste thresholds concentrations were determined as the geometric
mean across all panelists in three individual sessions. The values between individual
and separate sessions differed by not more than plus or minus one dilution step,
meaning that a threshold concentration of, for example, 81 µmol/L for avenanthramide
3f (13) represents a range of 40.5−162 µmol/L.
Preparation of Taste Recombinants in Aqueous Solution for Comparative Taste
Analysis. A basic taste recombinant OR was prepared by dissolving the tastants from
group I−V (see Table 3) in their “natural” concentrations (as determined in oat flour).
Tastants were taken up in bottled water with a pH adjusted to 6.4 using 1 % formic acid
and an ethanol content of 3 %. A 1:10 dilution of this stock solution was rated against
a reference of the methanol-water fraction II (pH 6.4, 3 % ethanol, dilution 1:10)
isolated from oat flour. Taste sensations sweet, sour, umami, salty, bitter, and
astringent were then evaluated on a linear scale from 0 (not detectable) up to 5
(strongly detectable).
In the same way, a second recombinant (OR≥1) was prepared consisting of the
tastants from group I−V showing DoT-factors solely greater than 1.0 (Table 3) in their
“natural” concentrations (as determined in oat flour). This recombinant was then rated
against a reference of the basic taste recombinant OR (pH 6.4, 3 % ethanol, dilution
1:10) as described before. The overall comparative taste profiles of the recombinants
were evaluated by means of a taste profile analysis using nose-clips and ranked on a
linear scale from 0 to 5. Data are based on the average of three individual panel
sessions.
21
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 22 of 42
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
High Performance Liquid Chromatography (HPLC). Preparative analysis of
crude synthesis was performed on a HPLC apparatus (Jasco, Groß-Umstadt,
Germany) consisting of two PU-2087 Plus pumps, a MD 2010 Plus photodiode array
detector and a Rh 7725i type Rheodyne injection valve (Rheodyne, Bensheim,
Germany). Data acquisition was executed by means of Chrompass 1.9. (Jasco, Groß-
Umstadt, Germany).
High-Performance Liquid Chromatography/Tandem Mass Spectrometry
(LC-MS/MS). Mass spectral analysis was performed on an API 4000 Q-Trap
LC-MS/MS (AB Sciex. Darmstadt, Germany) connected to a Dionex UHPLC UltiMate-
system 3000 (Dionex, Idstein, Germany). For electrospray ionization, the ion spray
voltage was set at −4500 V in the negative mode and at 5500 V in the positive mode.
Both quadrupoles operated at unit mass resolution, and nitrogen served as both curtain
gas (25 psi) and turbo gas (425 °C). Fragmentation of the pseudo molecular ions
[M+H]+ or [M−H]− into specific product ions was induced by collision with nitrogen (4.5
× 10−5 Torr). Data acquisition and instrumental control were performed with Analyst
1.5.1 (AB Sciex, Darmstadt, Germany).
Nuclear Magnetic Resonance Spectroscopy (NMR). One- and two-
dimensional 1H and 13C NMR spectra were acquired on a 500 MHz Bruker Avance III
equipped with a triple resolution cryo probe (TCI) (Bruker, Rheinstetten, Germany).
Samples were dissolved in methanol-d4 containing 0.03 % trimethylsilane (TMS) or
pyridine-d5. The chemical shifts are referenced to the TMS or the solvent signal
1
(pyridine-d5: H: 7.22 ppm; 13C: 123.87 ppm). TOPSPIN version 2.1 (Bruker,
Rheinstetten, Germany) was used for data processing. 1H-, 13C-, COSY-, HSQC- and
HMBC-spectroscopy were recorded using standard pulse sequences of the Bruker
library. Interpretation of the obtained spectra was performed with MestReNova 8.1.0-
11315. (Mestrelab Research, Santiago de Compostela, Spain).
22
ACS Paragon Plus Environment

Page 23 of 42
Journal of Agricultural and Food Chemistry
556
557
558 RESULTS AND DISCUSSION
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
As reported recently1, sequential application of solvent extraction, solid phase
extraction (SPE), and isolation by means of RP-HPLC followed by a taste dilution
analysis disclosed that, besides steroidal saponins avenacoside A and B, two novel
furostanol saponins as well as, unique oat constituents of avenanthramides were
identified as bitter and astringent taste active molecules in oat flour1. In order to
evaluate the sensory impact of previous findings, LC-MS/MS methods were developed
to quantitate saponins and avenanthramides in oat flour samples. Furthermore,
additional known taste active compounds were screened and analysed by means of
LC-MS/MS, such as free fatty acids15, monoglycerides15, free amino acids16,
polyphenols17, whereas soluble carbohydrates17 were determined by means of an
enzymatic assay, and minerals16 were quantitated by means of HPIC.
HPLC-MS/MS Method Development and Quantification of Saponins 1−4 in
Oats. A Selected Ion Monitoring method by means of LC-MS/MS was established with
the aim of determining the concentrations in oats of recently1 identified bitter and
astringent tasting saponins, namely, 3-(O-α-L-rhamnopyranosyl(1→2)-[β-D-
glucopyranosyl(1→3)-β-D-glucopyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-
glucopyranosyl-(25R)-furost-5-ene-3β,22,26-triol
(1),
3-(O-α-L-
rhamnopyranosyl(1→2)-[β-D-glucopyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-
glucopyranosyl-(25R)-furost-5-ene-3β,22,26-triol (2), avenacoside
avenacoside B (3).
A
(4) and
To quantitate exact concentrations, commercially available ginsenosides Rb1
(14) and Rb1 (15) were utilized as internal calibration standards. For method validation,
oat flour measurements were run in three independent analyses, each injected into the
23
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 24 of 42
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
LC-MS/MS system in triplicate. Resulting measurement precision, expressed as
relative standard deviation, ranged between 5% (3) and 11% (4). Recovery studies
were performed by spiking known concentrations of saponins 1−4 into oat flour at two
different concentrations. A comparison to those amounts found in unspiked oat flour
samples resulted in average recovery rates of 101.0% (4), 104.0% (3), 126.0% (2),
and 140.5% (1).
Quantitative analysis revealed avenacoside A (4, 231.8 µmol/kg) and
avenacoside B (3, 178.9 µmol/kg) as to be the most abundant saponins in oats, which
is well in line with literature reported earlier10. The so far unknown furostanosides 1
and 2 were quantitated for the first time in oats and were found at 40-fold lower
concentrations, with 6.0 and 12.3 µmol/kg, respectively.
HPLC-MS/MS Method Development and Quantification of Avenanthramides
5−10 and 13 in Oat by Means of SIDA. With the aim to investigate the recently1
identified bitter and astringent tasting avenanthramides, namely, 2c (5), 2p (6), 2f (7),
1p (8), 1c (9), 1f (10), 1s (11), 2s (12), and the newly synthesized 3f (13), a LC-MS/MS
method using Selected Ion Monitoring mode was set up to screen their presence in
oats. As reported recently1, the presence of 1s (11) and 2s (12) has not been confirmed
in oat flour so far, whereas bismethoxylated avenanthramide 3f (13) was screened and
successfully confirmed23 to be present in oats within this study after tuning and
optimisation of LC-MS/MS parameters using synthesized avenanthramide compounds
in combination with a screening of oat flour samples and co-chromatography
experiments.
To quantitate exact concentrations by means of SIDA, isotope labelled standards
of [2H3]-avenanthramide 2f (16) and [2H3]-avenanthramide 1f (17) were synthesized by
condensation reaction with labelled [2H3]-vanillin. Incorporation of the labelled OCD3-
function was confirmed by means of NMR and LC-MS/MS spectra. 1H-NMR spectra of
24
ACS Paragon Plus Environment

Page 25 of 42
Journal of Agricultural and Food Chemistry
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
unlabelled avenanthramide 1f (10) revealed a singlet at δ 3.93 ppm, demonstrating an
intensity of three protons, for the methoxylation at position C-H(10’) whereas for the
labelled [2H3]-avenanthramide 1f (17) no proton signal could be detected at C-H(10’).
In addition, mass spectrometry measurements of 10 revealed an exact mass of m/z
312.0868 for [M−H]−, while the labelled [2H3]-avenanthramide 1f 17 gave a mass of
m/z 315.1067 for [M−H]−, showing an increase of three daltons for the incorporated
three deuterium molecules.
To validate the method, oat flour measurements were run in four independent
analysis, each injected into LC-MS/MS in three replicates. Resulting measurement
precision ranged between 3% (5) and 9% (9 and 10). Recovery studies were performed
by spiking known concentrations of avenanthramides 5−10 to oat flour samples at
three different concentrations. A comparison to those amounts found in unspiked oat
flour samples resulted in average recovery of 53.7% (9), 56.3% (8), 78.5% (10), 90.3%
(5), 95.5% (6), and 98.3% (7).
Quantitative analysis revealed avenanthramide 3f (13, 35.9 µmol/L), 2f (7,
33.5 µmol/L), 2p (6, 23.0 µmol/L), 2c (5, 14.8 µmol/L) as the major abundant
avenanthramides, whereas 1p (8), 1c (9), 1f (10), were found in lower quantities,
between 0.3 (10) and 1.3 µmol/L (9). These results are well in line with data reported
in previous studies23-25
.
Concentration and DoT Factors of Taste Compounds. With the aim of
evaluating the relative contribution of individual taste molecules to the overall off-taste
in oats, a total of 59 compounds were quantitated in oat flour, among them the
saponins 1−4, avenanthramides 5−10 and 13, free fatty acids 18−24 and 27/28,
monoglycerides 29−33, and in addition, ten polyphenols, eighteen amino acids, four
minerals, and three soluble carbohydrates. Taking their individual taste recognition
thresholds into consideration, quantitated molecules of oat taste were grouped into five
25
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 26 of 42
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
clusters according to their different taste sensations (see Table 3). In order to map out
the key molecules driving the overall oat taste sensation, Dose-over-Threshold (DoT)
factors were calculated for each individual compound.
Representing compounds of bitter and astringent oro-sensations, saponins 1−4,
avenanthramides 5−10 and 13, free fatty acids 18−24 and 27/28, monoglycerides
29−33 as well as magnesium, calcium, L-tryptophan, L-arginine, L-lysine, L-valine,
L-histidine, L-phenylalanine, L-isoleucine, L-tyrosine, L-leucine were classified into
group I (c.f. Table 3).
Calculation of the DoT factors for bitter taste showed an extremely high value of
305.1 for 1-linoleoyl-rac-glycerol (33). Furthermore, 1-stearoyl-rac-glycerol (31), 1-
oleoyl-rac-glycerol (32), linoleic acid (23), avenacoside A (4), avenacoside B (3), 1-
palmitoyl-rac-glycerol (30), linolenic acid (24), oleic acid (22) and palmitic acid (20)
were rated with slightly lower DoT factors between 4.6 and 42.9 for bitterness, while 3-
(O-α-L-rhamnopyranosyl(1→2)-[β-D-glucopyranosyl(1→3)-β-D-
glucopyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-glucopyranosyl-(25R)-furost-5-
ene-3β,22,26-triol
(1)
and
3-(O-α-L-rhamnopyranosyl(1→2)-[β-D-gluco-
pyranosyl(1→4)]-β-D-glucopyranosid)-26-O-β-D-glucopyranosyl-(25R)-furost-5-ene-
3β,22,26-triol (2) and stearic acid (21) only slightly exceeded their bitter thresholds
showing DoT factors between 1 and 1.5. Although lipid derived compounds are well
known to induce a bitter off-taste3,26-29 this study positions monoglycerides, free fatty
acids and the furostanol saponins (1) and (2) for the first time as to be the major bitter
taste stimuli in oats.
In contrast, twenty-one out of the 34 bitter and astringent compounds where
found at concentrations below their respective taste threshold among them the bitter
amino acids, 1-myristoyl-rac-glycerol (29), the bivalent cations magnesium and
calcium, the fatty acids 9/13-hydroxy-9Z,11E-octadecadienoic acid (27/28) myristic
26
ACS Paragon Plus Environment

Page 27 of 42
Journal of Agricultural and Food Chemistry
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
acid (19), as well as the avenanthramides 5−10 and 13 showing DoT factors below 0.1
up to 0.6. Avenanthramides 1s (11), 2s (12) and lauric acid (18) were not detected by
means of LC-MS/MS in oat flour samples and considered not to be present in oat flour.
Among the astringent taste actives of group 2, p-hydroxybenzaldehyde was
found to be the most abundant polyphenol with a DoT factor of 104.0, followed by
ferulic acid, p-hydroxybenzoic acid, caffeic acid, p-coumaric acid and sinapinic acid
showing DoT values between 2.1 and 27.3 (see Table 3). Polyphenols of
protocatechuic acid, vanillin, and gallic acid were detected at concentrations below
their taste thresholds (DoT <1.0).
Group 3, consisting of sweet-tasting soluble carbohydrates and sweet amino
acids (Table 3), revealed sucrose as to be the key sweet parameter in oat flour with a
DoT factor of 1.9. In contrast, fructose, L-serine, glucose, L-proline, L-alanine,
L-threonine, and L-methionine did not reach DoT values greater than 0.2.
Potassium and sodium, both analyzed at concentrations above their recognition
thresholds (DoT 3.4 and 2.0), were classified into salty-tasting compounds of group 4.
Finally, the umami-tasting compounds of group 5 revealed L-aspartic acid as to
be the major umami active component, showing a DoT greater than 1, while for
L-glutamic acid, L-asparagine and L-glutamine the calculated DoT values were ≤0.1.
With the exception of sucrose and L-aspartic acid, we conclude that sweet- and umami-
tasting compounds are of minor importance to the characteristic taste of oats, taking
into consideration that the overall concentrations of group 3 and 5 were below their
taste recognition thresholds.
Taste Reconstitution Experiments. In order to confirm the results of
quantitation studies and to determine whether the compounds already identified can
create the characteristic taste of oats, we compiled an aqueous taste recombinant of
59 substances in “natural” concentrations as given in Table 3 and compared it against
27
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 28 of 42
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
a freshly prepared aqueous solution of the methanol water-extract isolated from oat
flour. The taste compounds were dissolved in bottled water and the pH value was
adjusted to those of the methanol-water fraction by addition of trace amounts of formic
acid. After dilution (1:10) of both aqueous solutions, trained panelists were asked to
judge taste quality intensity of the taste recombinant against the methanol-water
extract on a linear scale from 0 (not detectable) to 5 (strongly detectable). Averaged
intensities for bitter (3.4/2.7 ± 1.2), astringent (2.2/1.8 ± 0.6), sour (1.1/1.0 ± 0.3), sweet
(0.8/0.7 ± 0.3), scratchy (0.4/0.5 ± 0.4), given in Table 4, were close to those of the
reference. Descriptors for umami and salty-taste were rated below 0.5 and were
therefore excluded from further discussions. Assuming a symmetric distribution, results
were tested for significance using a Wilcoxon rank test30,31. Calculated test values for
bitterness (PGcalc 29; PGtab 21) and astringency (PGcalc 24; PGtab 21) exceeded
tabulated values, which confirmed the null hypothesis H0 to be true, resulting in a non-
significant difference between the taste recombinant and the methanol-water fraction.
Based on these findings it was concluded, that the compounds in group 1 and 2 are
sufficient to create the typical taste of an extract prepared from oat flour. For sweet
(PGcalc 19; PGtab 21), sour (PGcalc 21; PGtab 21) and scratchy (PGcalc 19; PGtab 21)
attributes, the alternative hypothesis H1 was accepted, leading to the conclusion, that
the extract and the recombinant are significantly different from each other. Due to the
fact that we focused, in our previous SENSOMICS study, on astringent and bitter
notes, further not-quantitated sweet, sour and scratchy taste stimuli may be present in
oats. While the major sweet tasting molecules are more or less screened, the presence
of oat organic acids e.g. such as free anthranilic acid or 5-hydroxyanthranilic acid could
be clarified in a further study. Furthermore, it could be tested if previously reported fatty
acids derivatives, such as 9,12,13-trihydroxy-trans-10-octadecenoic acid3, 9,10,13-
trihydroxy-trans-11-octadecenoic acid3, 9-hydroxy-trans,cis-10,12-octadecadienoic-1’-
28
ACS Paragon Plus Environment

Page 29 of 42
Journal of Agricultural and Food Chemistry
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
monoglyceride4 and 13-hydroxy-cis,trans-9,11-octadecadienoic-1’-monoglyceride4
impact not only the bitter but also the scratchy impression described by the sensory
panel.
To further investigate the importance of compounds showing a DoT factor below
1 to the overall taste profile, a partial taste recombinant OR≥1 was prepared from 23
analytes showing a DoT factor ≥ 1 (c.f. Table 4) and compared to the taste recombinant
OR containing all of the 59 analytes quantitated in oat flour. The resulting averaged
descriptors bitter (2.4/2.1 ± 0.4), astringent (1.8/1.6 ± 0.4), sour (0.8/0.9 ± 0.3), sweet
(0.5/0.6 ± 0.1), scratchy (0.9/0.6 ± 0.4) were very close to those of the completed
recombinant. A significance test, using a Wilcoxon rank test as discussed before,
revealed higher test values compared to tabulated values (PGcalc > PGtab ), leading to
a non-significant difference between the two recombinants: bitter (PGcalc 20; PGtab 8),
astringent (PGcalc 21; PGtab 8), sweet (PGcalc 25; PGtab 8), sour (PGcalc 10; PGtab 8) and
scratchy (PGcalc 24; PGtab 8). Therefore, it can be concluded that the overall sensory
profile of oat flour can be achieved solely from the 23 compounds used within the
partial taste recombinant OR≥1. Consequently, it can be derived that neither
avenanthramides 5−7 and 13 nor bitter amino acids, bivalent cations such as
magnesium and calcium or oxidized fatty acids 27/28 play an important role in oat
bitterness.
731
732
733
734
735
736
737
In summary, quantification of non-volatile taste active compounds in oat flour
and recombination experiments led to the overall identification of 59 taste active
components. Out of these, 23 compounds were identified as key taste contributors
showing a DoT factor above a threshold of 1.0. Lipid-derived taste molecules such as
monoglycerides and free fatty acids, namely, 1-linoleoyl-rac-glycerol (33), 1-stearoyl-
rac-glycerol (31), 1-oleoyl-rac-glycerol (32), 1-palmitoyl-rac-glycerol (30), linolenic acid
29
ACS Paragon Plus Environment