1
Contribution of peptides and polyphenols from olive water to acrylamide formation in
2
sterilized table olives
3
4
Francisco Javier Casado*,a,b, Alfredo Montañob and Reinhold Carlea
5
6
7
a
8
University, Garbenstrasse 25, D-70593 Stuttgart, Germany
9
b
10
Institute of Food Science and Biotechnology, Chair Plant Foodstuff Technology, Hohenheim
Instituto de la Grasa (C.S.I.C.), Apartado 1078, 41012 Seville, Spain
*Tel.: +49(0)711-45923125, fax: +49(0)711-45924110, e-mail: javierhebrard@ig.csic.es
11
12
Running title: Peptides and polyphenols from olive water are acrylamide precursors
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1
27
Abstract
28
29
To confirm the role of peptides as principal precursors of acrylamide formation in sterilized
30
table olives, peptides from olive water were fractionated. After their partial fractionation by
31
solid phase extraction (SPE) and ultrafiltration (< 10,000 Da), respectively, small peptides
32
from olive water were isolated by size-exclusion chromatography (SEC). In the fractions
33
collected, peptides and polyphenolic compounds were determined colorimetrically, and
34
acrylamide was quantitated by LC-MS/MS after heating of the samples. Subsequently,
35
peptides
36
time of flight mass spectrometry (MALDI-TOF/TOF-MS), and polyphenols were analyzed by
37
LC-MS in the respective fractions. Finally, peptides containing fractions were purified on a
38
polymeric resin (Amberlite XAD 16HP) to remove unbound phenolic compounds by
39
adsorption. The results of the different experiments performed in complete absence of free
40
asparagine and reducing sugars strongly support small peptides bound to polyphenols to be
41
the principal precursors of acrylamide in sterilized table olives.
were
characterized
by
matrix-assisted
laser
desorption/ionization-tandem
42
43
44
45
Keywords: Acrylamide · olives · peptides · fractionation · size-exclusion chromatography ·
46
peptide-polyphenol complexes
47
48
49
50
51
52
2
53
1. Introduction
54
55
Acrylamide, a chemical compound classified as “probably carcinogenic to humans” by
56
the International Agency for Research on Cancer (IARC, 1994), was first detected in heated
57
carbohydrates-rich foods in 2002 (Tareke, Rydberg, Karlsson, Eriksson & Tornqvist, 2002).
58
The Maillard reaction between free asparagine (Asn) and reducing sugars and further
59
carbonyl sources, respectively, has been confirmed as the major pathway of acrylamide
60
formation in foodstuffs (Mottram, Wedzicha & Dodson, 2002; Stadler et al., 2002; Gökmen
61
& Palakzaglu, 2008). In general, the highest levels of acrylamide have been found in
62
carbohydrate-rich foods from potato, wheat and other cereals, and coffee (Friedman, 2003).
63
High acrylamide levels (from 200 to 2000 μg/kg) have been detected in black ripe
64
olives, one of the main types of table olive commercialized worldwide in which sterilization
65
treatment is compulsory (FDA, 2006; Casado & Montaño, 2008). Nevertheless, it has been
66
demonstrated that acrylamide formation in sterilized olives follows a different pathway, since
67
acrylamide levels of olives did not correlate with the contents of free Asn and any of the
68
reducing sugars determined prior to sterilization (Amrein, Andres, Escher & Amadò, 2007;
69
Casado & Montaño, 2008). Compared with potato and cereal products, little information is
70
available on acrylamide in sterilized black ripe and green ripe olives. Up to now, the
71
mechanism of acrylamide formation in olives is still unknown.
72
A detailed study dealing with heated model peptides and olive water fractions has
73
been reported previously (Casado, Montaño, Spitzner & Carle, 2013). According to our
74
findings, the role of free Asn and glucose, being the main reducing sugar in olives, as
75
acrylamide precursors in these fruits has been ruled out, suggesting peptides smaller than
76
10,000 Da to be the principal precursors of acrylamide formation in heated table olives. In the
77
absence of free Asn, only fractions containing peptides/proteins obtained by partial
78
fractionation of olive water by solid phase extraction (SPE) and precipitation with cold
3
79
acetone, respectively, generated significant amounts of acrylamide upon heating at usual
80
sterilization conditions (121 ºC for 30 min).
81
The aim of the present work was to fractionate and isolate peptides from olive water to
82
confirm their putative role as precursors of acrylamide in sterilized olives. Due to the complex
83
fruit matrix, isolation of peptides was to be expected very challenging, considering the
84
multiple interactions of peptides or proteins with phenolic compounds. Consequently, olive
85
water from untreated green olives was subjected to size-exclusion chromatography (SEC)
86
after partial fractionation by SPE and ultrafiltration. Peptides and polyphenolic compounds of
87
the fractions collected were measured colorimetrically. For the quantitation of acrylamide in
88
the heated samples, liquid chromatography-tandem mass spectrometry (LC-MS/MS) was
89
applied. Subsequently, peptides were further characterized by matrix-assisted laser
90
desorption/ionization-tandem time of flight mass spectrometry (MALDI-TOF/TOF-MS), and
91
polyphenols were determined by liquid chromatography-mass spectrometry (LC-MS) in the
92
respective fractions. Finally, phenolic compounds were removed from the peptides containing
93
fractions using a polymeric resin to elucidate the mechanism of acrylamide formation in table
94
olives.
95
96
2. Materials and methods
97
98
2.1. Chemical and Materials
99
100
Acrylamide (99%) was purchased from ICN Biomedicals (Eschwege, Germany),
101
2,3,3-D3-labeled acrylamide (98%) was from Cambridge Isotope Laboratories (Andover, MA,
102
USA). Formic acid and acetonitrile (ACN), both gradient grade, ammonium acetate (99%),
103
ethanol
104
hexacyanoferrate (Carrez I) and zinc acetate (Carrez II) were provided by VWR (Darmstadt,
(98%),
acetone,
ethyl
acetate,
methanol,
4
hydrochloric
acid,
potassium
105
Germany). Glacial acetic acid and trifluoroacetic acid (TFA) were from Merck (Darmstadt,
106
Germany). 2,5-dihydroxybenzoic acid (DHB) and α-Cyano-4-hydroxycinnamic acid (HCCA),
107
were obtained from Bruker Daltonics (Bremen, Germany). Deionized water (Sartorius Arium
108
611 Ultrapure water system) was used throughout. Solid phase extraction cartridges (Isolute
109
Multimode, 1000 mg) were obtained from IST (Hengoed, Mid Glamorgan, UK). Silica-based
110
bonded-phase cartridges (Sep Pak Vac 20 cc/5g C18) were purchased from Waters (Milford,
111
MA, USA). Chem Elut cartridges for solid phase supported liquid-liquid extraction and
112
Hydromatrix diatomaceous earth were from Varian (Darmstadt, Germany).
113
114
2.2. Olive water extraction
115
116
Green olives (5 kg) of cv. ‘Hojiblanca’ (Seville, Spain) from a local processor were
117
used. Olives were pitted and homogenized using a mixer. The olive water was obtained by
118
pressing the mash using an hydraulic laboratory press (Hafico, HP-2, Düsseldorf, Germany)
119
and twofold centrifugation at 20,000g for 20 min at 20 ºC to remove the olive oil. The olive
120
water was stored at - 20 ºC until the different experiments were performed. After extraction,
121
the pH of the olive water was 4.4.
122
123
2.3. Partial fractionation of olive water
124
125
Peptides from olive water were fractionated according to the scheme showed in Fig 1.
126
To the partial fractionation of olive water by SPE, Sep Pak Vac C18 cartridges were
127
conditioned with methanol and water. After pH adjustment to 3.0 using HCl, aliquots (10 mL)
128
of olive water were filtered through a 0.45 µm syringe filter, loaded, and slowly (1 mL/min)
129
passed through the cartridges. Two different fractions were recovered: (1) fraction A, an
130
aqueous fraction consisting of the eluate, and the retentate obtained after elution from the
5
131
sorbent with 15 mL of water; and (2) fraction B, comprising the less polar fraction eluted
132
from the sorbent with 15 mL of methanol. Whereas fraction A was discarded, fraction B was
133
evaporated to dryness in a rotary evaporator, resuspended in 10 mL of water, and the pH was
134
adjusted to the initial value (≈ 4.4). Fraction B was used for the fractionation of peptides by
135
SEC after ultrafiltration.
136
137
2.4. Ultrafiltration
138
139
Fraction B (see section 2.3.) was re-circulated through a stirred ultrafiltration cell
140
model 8003 containing a cellulose membrane with a molecular weight cut-off of Mr 10,000
141
(Millipore, Beverly, MA, USA). Aliquots of the retentate (fraction C) and permeate (fraction
142
D), resuspended with water, were adjusted to pH 4.4. Fraction D was used for the
143
fractionation of peptides by SEC.
144
145
2.5. Fractionation of peptides by size-exclusion chromatography
146
147
A 5 mL of fraction D (permeate) was concentrated to 1 mL under reduced pressure.
148
The concentrate was applied to a glass column (Superformance, 600 x 26 mm, Merck,
149
Darmstadt, Germany) filled with Toyopearl HW-55F (Tosoh, Stuttgart, Germany). Elution
150
was performed at room temperature using 0.3 mol/L ammonium acetate buffer, pH 4.0. The
151
flow-rate of 2 mL/min was produced using a HPLC compact pump (Bischoff, Leonberg,
152
Germany). The absorbance at 280 nm was measured with a SPD-10AV UV/Vis detector
153
(Shimadzu, Duisburg, Germany). The eluate was collected in 40 mL fractions. The
154
fractionation was performed in sextuplicate. Appropriate fractions collected after six runs
155
were pooled, freeze-dried, and resuspended in water. The pH of the fractions ranged from 5.7
156
to 6.0.
6
157
2.6. Adsorptive removal of phenolic compounds from olive water fractions using polymeric
158
resin
159
160
Phenolic compounds from appropriate olive water fractions were removed by
161
adsorption with a polymeric resin, Amberlite XAD 16 HP (Rohm and Haas, Philadelphia, PA,
162
USA), following a method developed by Weisz, Schneider, Schweiggert, Kammerer and
163
Carle (2010) with several modifications.
164
The resin (15 g) was activated overnight by soaking in 96% EtOH (5 mL/g). The same
165
volume of deionized water at 50 ºC was used to remove the alcohol from the resin beads prior
166
to the experiments. Subsequently, the adsorbent material was conditioned by purging with 3
167
mL/g of NaCl solution (1.3 mol/L, pH 6.0).
168
For each experiment, 15 mL olive water fraction adjusted to pH 6.0 was combined
169
with 1 g of pre-treated resin in a 50 mL screw cap flasks. To prevent polyphenol oxidation,
170
the flask was flushed with nitrogen. The suspension was stirred using a magnetic stirrer at 150
171
rpm in a thermostatted water bath at 20 ºC during 120 min. After separation of the solution
172
(fraction E), MeOH (20 mL) was added to the colored resin and the suspension was stirred at
173
room temperature for 30 min to elute the retained compounds. Subsequently, the colored
174
solution was evaporated to dryness in a rotary evaporator (30 ºC, 40 mbar), and resuspended
175
in 15 mL of water (fraction F). The pH of fraction F was adjusted to 6.0.
176
177
2.7. Heat treatment of model systems
178
179
Heat treatments were performed using 5 mL of the corresponding sample (aliquots of
180
fraction B after ultrafiltration, SEC or adsorption treatment with resin) filled in a stainless
181
tubular steel reactor (internal diameter 1 cm, length 8.5 cm). The reactor was sealed, and then
182
heated in a thermostatted oil bath at 121 ºC (±1 ºC) for 30 min in all experiments to mimic
7
183
sterilization conditions. After heating, the sample was immediately cooled in ice water for 3
184
min to stop the reaction. All heating experiments were performed in triplicate.
185
186
2.8. Sample preparation for acrylamide determination
187
188
For acrylamide determination, samples were prepared according to a method recently
189
published by Casado, Montaño, Spitzner and Carle (2013) using D3-acrylamide as internal
190
standard, and analysed by LC-MS/MS analysis.
191
192
2.9. Determination of acrylamide by LC-MS/MS
193
194
Acrylamide was determined by liquid chromatography-tandem mass spectrometry
195
(LC-MS/MS) according to the method developed by Claus, Weisz, Kammerer, Carle and
196
Schieber (2005). Analyses were carried out on an Agilent 1100 series HPLC system (Agilent,
197
Waldbronn, Germany). A Bruker Esquire 3000+ ion trap mass spectrometer (Bremen,
198
Germany) fitted with an electrospray ionisation (ESI) source was coupled on-line to the LC
199
system. Esquire Control software was used for data acquisition and processing.
200
Chromatographic separation of acrylamide was carried out on a Hypercarb column (100 x 2.1
201
mm; 5 µm) (Thermo Hypersil, Dreieich, Germany), equipped with a C18 guard column (4.0 x
202
3.0 mm) (Phenomenex, Torrance, CA, USA) at 30 ºC. The sample (20 µL) was separated
203
isocratically using a mobile phase composed of water/ACN/formic acid (99 mL/1 mL/0.05
204
mL) in water at a flow rate of 0.2 mL/min, within a total run time of 10 min. The mass
205
spectrometer was operated in the positive electrospray ionization mode (ESI+). Characteristic
206
fragmentation transitions m/z 72.3 > m/z 55.5 for acrylamide and m/z 75.3 > m/z 58.5 for D3-
207
acrylamide were recorded using multiple reaction monitoring (MRM). The signals at m/z 55.5
8
208
and m/z 58.5 were used for the quantitation of acrylamide and D3-acrylamide, respectively,
209
while the signals at m/z 44.5 and m/z 45.5 were used for qualification.
210
211
2.10. Colorimetric determination of peptides and polyphenols
212
213
Peptide contents of the fractions obtained by SEC were measured according to
214
Bradford (1976). Bovine serum albumin from Merck (Darmstadt, Germany) was used as a
215
standard. The absorbance at 595 nm was recorded against the reagent blank using a Cary 100
216
Cone spectrophotometer (Varian, Palo Alto, CA, USA). All determinations were performed in
217
triplicate.
218
The quantification of phenolic compounds was based on the method of Singleton and
219
Rossi (1965). Folin-Ciocalteu reagent was provided by Merck (Darmstadt, Germany). Caffeic
220
acid (≥ 95%, Roth, Karlsruhe, Germany) was used as a reference. The absorbance was read at
221
760 nm. All determinations were performed in triplicate.
222
223
2.11. Characterization of olive water peptides by MALDI-TOF/TOF-MS
224
225
Protein spectra, peptide mass fingerprints and MS/MS spectra of appropriate olive
226
water fractions were acquired on an Autoflex III MALDI-TOF/TOF mass spectrometer
227
(Bruker Daltonics, Bremen, Germany). The instrument was operated in the positive ion mode,
228
and externally calibrated using protein mass or peptide calibration standards (Bruker
229
Daltonics, Bremen, Germany), respectively. Samples were desalted using either C 4- or C18-
230
ZipTips (Millipore, Schwalbach, Germany) following the manufacturer’s protocols. Peptides
231
and proteins were eluted directly onto a stainless steel target using HCCA (5 mg/mL in 50%
232
ACN/ 50% 0.1% TFA) or DHB (20 mg/mL in 50% ACN/ 50% 0.1% TFA) matrix solutions,
233
respectively. Protein molecular weights in the range of 5-50 kDa were obtained in the linear
9
234
mode using an accelerating voltage of 20 kV. Peptide mass fingerprint data and MS/MS
235
spectra of olive water fractions were recorded in the reflector mode using an accelerating
236
voltage of 21 kV. Peptide mass fingerprint spectra were acquired in the m/z range between
237
700 and 5,000 Da. 2,000 laser shots per sample were acquired to ensure good signal/noise
238
quality for precursor ion selection. MS/MS analysis was done with a varying number of laser
239
shots. Flex Analysis 3.0 and Bio-Tools 3.0 software (Bruker Daltonics, Bremen, Germany)
240
were used for data processing.
241
242
2.12. Determination of polyphenols by LC-MS
243
244
LC-MS analyses were carried out using an Agilent HPLC series 1100 (Agilent,
245
Waldbronn, Germany). Polyphenols from appropriate olive water fractions were separated by
246
HPLC according to the method described by Desportes, Charpentier, Duteurtre, Maujean and
247
Duchiron (2000) with slight modifications. The separation was performed using a porous
248
graphitic carbon (PGC) Hypercarb column (100 x 2.1 mm; 5 µm) (Thermo Hypersil,
249
Dreieich, Germany), equipped with a C18 guard column (4.0 x 3.0 mm) (Phenomenex,
250
Torrance, CA, USA). The column was kept at 30 ºC. The volume injection was 50 µL. Eluent
251
A was 0.1 mL TFA/100 mL in deionized water, and eluent B was 0.1 mL TFA/100 mL in
252
ACN. The following gradient was used: 0-5 min, 0% B; 5-20 min, 0-10% B; 20-25 min,
253
10%B; 25-40 min, 10-30% B; 40-45 min, 30-50% B; 45-55 min, 50-100% B; 55-60 min, 100-
254
0% B, 60-70 min, 0% B. The flow rate was 0.8 mL/min. MS conditions were adopted from
255
Kaiser, Carle and Kammerer (2013). A Bruker Esquire 3000+ ion trap mass spectrometer
256
(Bremen, Germany) fitted with an ESI source was coupled on-line to the HPLC system Data
257
acquisition and processing were performed using Esquire Control software. Mass
258
spectrometer was operated in the negative ion mode. Mass spectra of the column eluate were
259
recorded in the range of m/z 50-1500 at a scan speed of 13,000 Th/s.
10
260
261
3. Results and Discussion
262
263
3.1. Partial fractionation by SPE and ultrafiltration of olive water
264
265
Olive water from untreated green olives (for olive water extraction see section 2.2.)
266
was partially fractionated by SPE. Two different fractions were recovered: fraction A,
267
comprising the permeate and the compounds eluted from the sorbent with water, and fraction
268
B, containing the retained compounds. Previous experiments carried out with the same olive
269
water in similar conditions demonstrated that: (1) after heat treatment (121ºC for 30 min) of
270
both fractions, significant amounts of acrylamide was formed in fraction B, while only traces
271
were found in fraction A, (2) peptides and/or proteins were presented in fraction B, and (3)
272
free Asn was in complete absence in fraction B (Casado, Montaño, Spitzner & Carle, 2012).
273
After discarding fraction A, fraction B was ultrafiltered to remove compounds having Mr
274
>10,000. The resulting fraction C and fraction D, resuspending with water, both with pH
275
adjusted to 4.4, were heated (121 ºC, 30 min) in the tubular reactor. Acrylamide was detected
276
in both, heated fraction D (1,217 ± 145 µg/L) and heated fraction C (147 ± 15 µg/L).
277
Acrylamide content in heated ultrafiltrate represented 89% of total. Although larger peptides
278
and/or proteins also contribute in minor way, these results suggested that the main acrylamide
279
precursors in sterilized fresh olive are peptides displaying Mr < 10,000.
280
281
3.2. Fractionation of peptides by SEC
282
283
A Toyopearl HW-55F column was used to fractionate peptides from fraction B. This
284
resin is a hydroxylated methacrylic polymer for SEC of proteins and peptides, having Mr >
11
285
1,000. Figure 2 shows a typical elution profile of the fraction obtained after partial
286
fractionation by SPE, and ultrafiltration of olive water on Toyopearl HW-55F.
287
288
Eleven different fractions were collected. Subsequently, all 11 fractions were
subjected to polyphenols and peptides determination.
289
Results of polyphenols and peptides determination are displayed in Figure 3. As can
290
be seen, phenolic compounds were detected in all fractions with a maximum (0.075 g caffeic
291
acid/100 mL) in fraction 5, corresponding to the second highest peak of the elution profile.
292
Concerning peptides, fractions 2 to 4 exhibited the greatest concentrations (38, 36 and 38 mg
293
peptides/L, respectively). In general, a decreasing trend was observed from fraction 4 to 11.
294
Since the highest peptide concentrations were detected at the beginning of the elution profile,
295
most of the peptides contained in the ultrafiltrate were supposed to exhibit a Mr close to the
296
cut off of the ultrafiltration membrane (10,000 Da).
297
Figure 3 also displays the acrylamide contents of each fraction after heating of the
298
samples (121 ºC for 30 min). Different amounts of acrylamide were found in fractions 1 to 5,
299
whereas the compound was not detected in fractions 6 to 11. These results demonstrated that
300
fractions 2 and 3 generated the largest amounts of acrylamide (528 ± 31, and 416 ± 13 µg/L,
301
respectively). Taking into account all these findings, acrylamide precursors were mainly
302
accumulated in fraction 2 and 3. Therefore, both fractions were exclusively used for the
303
following experiments.
304
305
3.3. Characterization of olive water peptides by MALDI-TOF/TOF-MS
306
307
In the past decade, MALDI-TOF MS has become one of the most powerful tools for
308
the analysis of biomolecules including proteins and peptides. This analytical technique has
309
demonstrated increasing utility in the molecular weight determination and identification of
310
peptides. In an attempt to characterize the peptides present in fractions 2 and 3, both fractions
12
311
were subjected to MALDI-TOF/TOF-MS analysis. Surprisingly, masses detected in both
312
fractions were below 600 Da. Since these fractions were collected at the beginning of the
313
elution profile of ultrafiltered olive water (10,000 Da cut off membrane), peptides present in
314
both fractions should have a bigger size. One possible explanation for the unsuccessful
315
detection of peptides is that these compounds were bound to oxidized polyphenols, thus
316
forming high-molecular complexes. MALDI-TOF/TOF-MS has proven to be very efficient
317
for structural analysis of peptides, but limitations in the evaluation of the molecular weight
318
distributions of peptide-polyphenol mixtures are obvious. The successful application of
319
MALDI-TOF/TOF-MS to such complex mixtures is mainly restricted by peptide-polyphenol
320
interactions, and depends on the properties and molecular weight of both the peptide and the
321
polyphenol. The higher the molecular size of the polyphenol, the greater the tendency to form
322
complexes with peptides, and the worse the resolution due to problems with the
323
fragmentation. Mané et al. (2007) found that complexes resulting from interaction between
324
proteins from grape seed and tannins such as β-lactoglobulin, β-casein or myoglobin were not
325
detected by MALDI-TOF-MS. From the results obtained in section 3.4., and 3.5. it is
326
suggested that peptides related to acrylamide formation in olives were covalently bound to
327
polyphenols, thus forming larger complexes.
328
According to the findings of Konno et al. (1999), by enzymatic conversion of
329
oleuropein, a naturally-occurring secoiridoid polyphenol in olives, into a glutaraldehyde-like
330
α,ß-unsaturated highly reactive dialdehyde, proteins are covalently cross-linked. The resultant
331
oleuropein aglycone has been proposed as a cross-linking agent for collagen. Therefore, it can
332
be assumed, that the activated form of oleuropein interacts with olive peptides as the
333
aldehydic component. Furthermore, in the above mentioned study, polyphenol oxidase
334
activity was shown to activate dihydroxyphenolic moieties through the formation of o-
335
quinones to exert protein-binding activities. Briante et al. (2000) and Antunes et al. (2013)
13
336
confirmed that the aglycone form of oleuropein has a strong affinity for nucleophilic binding
337
sites such as amino and thiol groups of proteins, thus forming protein-polyphenol-adducts.
338
339
3.4. Determination of polyphenols by LC-MS
340
341
Polyphenols profile in table olives has been extensively studied. Olive fruits are rich
342
sources of hydroxytyrosol, verbascoside, ligustroside, salidroside, rutin, luteolin 7-glucoside
343
and many other phenolics including secoiridoid derivatives like oleuropein (Ryan et al., 2002;
344
Romero et al, 2004; Cardoso et al., 2005). For the determination of polyphenols in fractions 2
345
and 3 by LC-MS, a PGC column was used. This stationary phase was shown to be suited to
346
separate peptides from phenolics compounds by Desportes, Charpentier, Duteurtre, Maujean
347
and Duchiron (2000) using HPLC-fractionation of small peptides from wine. Surprisingly,
348
despite the positive Folin-Ciocalteu reaction demonstrating the presence of polyphenols in
349
fractions 2 and 3, both fractions exhibited dark coloration providing further evidence of
350
polyphenols being involved in both fractions. However, none of the fragment ions typical of
351
olive polyphenols was detected by ESI-MS. This fact further supported our hypothesis of
352
polyphenols in both fractions being covalently bound to peptides.
353
354
3.5. Adsorptive removal of phenolic compounds from olive water fractions using polymeric
355
resin
356
357
To isolate peptides from fractions 2 and 3, and to confirm the presence of peptide-
358
polyphenol complexes, both fractions were subjected to adsorptive removal of polyphenolics
359
using a polymeric resin. Amberlite XAD 16HP resin has been successfully used by Weisz,
360
Schneider, Schweiggert, Kammerer and Carle (2010) to decolorize sunflower protein extracts.
14
361
They observed initial phenolic contents to be decreased by 86% upon addition of the resin,
362
whereas protein loss was below 5%.
363
From each fraction, two different fractions were obtained: (1) fraction E, a colorless
364
solution containing the compounds not retained on the polymeric adsorber resin, and (2)
365
fraction F, a pigmented solution containing the compounds retained on the resin after their
366
subsequent recovery by elution as explained in section 2.6. In fraction E, the initial phenolic
367
concentrations in fractions 2 and 3 amounting to 0.040 and 0.043 g caffeic acid/100 mL,
368
respectively, were decreased by 85 and 87%, respectively, upon the addition of the resin.
369
These losses were comparable to those reported by Weisz, Schneider, Schweiggert,
370
Kammerer and Carle (2010). When analyzing the phenolic compounds of fraction F from
371
both fractions, the proportion of polyphenols retained on the resin represented 80 and 83% of
372
the initial concentration in fractions 2 and 3, respectively. Regarding peptides, their
373
adsorption to the resin was also found to be high in both fractions, since their initial peptide
374
contents in fractions 2 and 3 (34 and 35 mg peptides/L, respectively) was reduced by 90 and
375
83%, respectively. In contrast, fraction F from fractions 2 and 3 still contained 85 and 72% of
376
the initial peptide content, respectively. Four commercial food-grade adsorbent and three ion
377
exchange resins, including Amberlite XAD 16HP, using different conditions were tested in
378
sunflower extracts by Weisz, Schneider, Schweiggert, Kammerer and Carle (2010), and
379
maximum protein loss was 31% for Amberlite XAD 1180N resin, while protein losses were
380
generally below 5% for most other resins. Therefore, substantial peptides losses in fractions 2
381
and 3 also suggested peptides to form complexes with polyphenols.
382
Fraction E and F from fractions 2 and 3 were subjected to heat treatment (121 ºC for
383
30 min). While only trace amounts of acrylamide were detected in fraction E (containing the
384
compounds not retained on the resin) from fraction 2, acrylamide was not detected in the
385
same solution from fraction 3. However, heating fractions F (containing the compounds
386
retained on the resin after their subsequent recovery by elution) yielded significant amounts of
15
387
acrylamide in both fractions (240 ± 13, and 130 ± 4 µg/L in fractions 2 and 3, respectively).
388
As earlier reported, peptide concentrations in fraction F from fractions 2 and 3 were 85 and
389
72% of the original fraction, respectively. Again, these findings confirmed the role of peptides
390
smaller than 10,000 Da to be the principle precursors of acrylamide formation in heated olives
391
being completely devoid of reducing sugars and free Asn.
392
393
4. Conclusion
394
395
Although the exact mechanism of acrylamide formation in sterilized table olives still
396
remains to be elucidated, the results of the present study confirm small peptides (< 10,000 Da)
397
to be the principal precursors of acrylamide following heat treatment of olives mimicking
398
sterilization (121 ºC, 30 min). Therefore, in the complete absence of reducing sugars and free
399
Asn, after removal of low molecular compounds by SPE, olive water from green olives and
400
peptide containing fractions isolated by different methods (ultrafiltration, SEC and adsorption
401
with polymeric resin), respectively, generated significant amounts of acrylamide after heat
402
treatment. Upon heating to 121 ºC for 30 min, the low molecular fractions obtained by SEC
403
containing the highest concentrations of peptides, yielded the highest amounts of acrylamide.
404
Taking together the results of this study, covalent binding of olive polyphenols to
405
peptides is assumed to be responsible for acrylamide formation. As previously discussed,
406
different authors (Konno et al., 1999; Briante et al., 2000; Antunes et al., 2013) have
407
demonstrated adduct formation of enzymatically activated oleuropein and proteins.
408
Oleuropein hydrolyzed by β-glucosidase had even stronger protein-crosslinking and lysine-
409
alkylating activities than glutaraldehyde. In our previous report (Casado, Montaño, Spitzner &
410
Carle, 2013), formation of acrylamide from Ala, Asp and Met containing model peptides at
411
200°C for 180 min was abundant without the involvement of any carbonyl component.
412
Moreover, acrylamide was not formed following mild heating of the same model peptides at
16
413
121°C for 30 min, representing typical sterilization conditions for black ripe olives. Although
414
the addition of some genuine phenolic compounds present in olives at different levels, e. g.
415
hydroxytyrosol and 3,4-dihydroxyphenyl glycol, has been reported to insignificantly affect
416
acrylamide formation in heated olive water (Casado, Sánchez & Montaño, 2010), our findings
417
suggest that peptides need to be activated, e. g. by the formation of adducts with reactive
418
polyphenols (i. e. oleuropein aglycone and o-quinones, respectively) to form acrylamide in
419
table olives following sterilization without the involvement of free Asn and reducing sugars,
420
respectively. Our assumption is supported by our previous investigations of synthetic peptides
421
which produced acrylamide only when heated to 200°C without the involvement of carbonyl
422
components, whereas acrylamide formation was observed under less harsh heat conditions
423
(121°C) in the presence of the reactive polyphenols, thus suggesting activation by the
424
carbonyl adduct. Further investigations are still required to unravel the activating role of
425
polyphenols by the aldehydic component in the acrylamide formation mechanism of table
426
olives.
427
428
Acknowledgements
429
430
This work was supported in part by the European Union (FEDER funds) and the
431
Spanish government through Project AGL 2010-19178. The Spanish government is also
432
thanked for the fellowship of the first author.
433
434
References
435
436
Amrein, T. M., Andres, L., Escher, F., & Amadò, R. (2007). Occurrence of acrylamide in
437
selected foods and mitigation options. Food Additives and Contaminants, 24, 13-25.
17
438
Antunes, A. P. M., Attenburrow, G., Covington, A. D., & Ding, J. (2013). Utilisation of
439
oleuropein as crosslinking agent in collagenic films. Journal of Leather Science, 2, 17-23.
440
Bradford, M. M. (1976). A rapid sensitive method for the quantification of microgram
441
quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry,
442
72, 248-254.
443
Briante, R., La Cara, F., Febbraio, F., Barone, R., Piccialli, G., Carolla, R., Mainolfi, P., Da
444
Napoli, L., Patumi, M., Fontanazza, G., & Nucci, R. (2000). Hydrolysis of oleuropein by
445
recombinant β-glycosidase from hyperthermophilic archaeon Sulfolobus solfataricus
446
immobilised on chitosan matrix. Journal of Biotechnology, 77, 275-286.
447
Cardoso, S. M., Guyot, S., Marnet, N., Lopes-da-Silva, J. A., Renard, C. M. G. C., &
448
Coimbra, M. A. (2005). Characterisation of phenolic extracts from olive pulp and olive
449
pomace by electrospray mass spectrometry. Journal of the Science of Food and Agriculture,
450
85, 21-32.
451
Casado, F. J., & Montaño, A. (2008). Influence of processing conditions on acrylamide
452
content in black ripe olives. Journal of Agricultural and Food Chemistry, 56, 2021-2027.
453
Casado, F. J., Sánchez, A. H., & Montaño, A. (2010). Reduction of acrylamide content of ripe
454
olives by selected additives. Food Chemistry, 119, 161-166.
455
Casado, F. J., Montaño, A., Spitzner, D., & Carle, R. (2013). Investigation into acrylamide
456
precursors in sterilized table olives: Evidences of a peptic fraction being responsible for
457
acrylamide formation. Food Chemistry, 141, 1158-1165.
458
Desportes, C., Charpentier, M., Duteurtre, B., Maujean, A., & Duchiron, F. (2000). Liquid
459
chromatographic fractionation of small peptides from wine. Journal of Chromatography A,
460
893, 281-291.
461
Friedman, M. (2003). Chemistry, biochemistry, and safety of acrylamide. A review. Journal
462
of Agricultural and Food Chemistry, 51, 4504-4526.
18
463
FDA (US Food and
Drug
464
Individual Food Products. US Department of Health and Human Services. Center for Food
465
Safety
466
http://www.fda.gov/Food/FoodSafety/FoodContaminantsAdulteration/ChemicalContaminants
467
/Acrylamide/ucm053549.htm
468
Gökmen, V., & Palakzaglu, T. (2008). Acrylamide formation in foods during thermal
469
processing with a focus on frying. Food and Bioprocess Technology, 1, 35-42.
470
International Agency for Research on Cancer (1994). IARC Monographs on the evaluation of
471
carcinogenic risks to humans: some industrial chemicals. Vol. 60. Acrylamide. Lyon, France:
472
IARC 1994, pp 389-433.
473
Kaiser, A., Carle, R., & Kammerer, D. R. (2013). Effects of blanching on polyphenol stability
474
of innovative paste-like parsley (Petroselinum crispum (Mill.) Nym ex A. W. Hill) and
475
marjoram (Origanum majorana L.) products. Food Chemistry, 138, 1648-1656.
476
Konno, K., Hirayama, C., Yasui, H., & Nakamura, M. (1999). Enzymatic activation of
477
oleuropein: A protein crosslinker used as a chemical defense in the privet tree. Procedure of
478
the National Academy of Sciences, 96, 9159-9164.
479
Mané, C., Sommerer, N., Yalcin, T., Cheynier, V., Cole, R. B., & Fulcrand, H. (2007).
480
Assessment of the molecular weight distribution of tannin fractions through MALDI-TOF MS
481
analysis of protein-tannin complexes. Analytical Chemistry, 79, 2239-2248.
482
Mottram, D. S., Wedzicha, B. L., & Dodson, A. T. (2002). Acrylamide is formed in the
483
Maillard reaction. Nature, 419, 448-449.
484
Romero, C., Brenes, M., Yousfi, K., García, P., García, A., & Garrido, A. (2004). Effect of
485
cultivar and processing method on the contents of polyphenols in table olives. Journal of
486
Agricultural and Food Chemistry, 52, 479-484.
and
Administration) (2006). Survey Data on Acrylamide in Food:
Nutrition,
available
19
online
at
487
Ryan, D., Antolovich, M., Herlt, T., Prenzler, P. D., Lavee, S., & Robards, K. (2002).
488
Identification of phenolic compounds in tissues of novel olive cultivar Hardy’s Mammoth.
489
Journal of Agricultural and Food Chemistry, 50, 6716-6724.
490
Singleton, V. L., & Rossi, J. A. (1965). Colorimetry of total phenolics with
491
phosphomolybdic-phosphotungstic acid reagent. American Journal of Oenology and
492
Viticulture, 16, 144-158.
493
Stadler, R. H., Blank, I., Varga, N., Robert, F., Hau, J.; Guy, P. A., Robert, M.-C., &
494
Riediker, S. (2002). Acrylamide from Maillard reaction products. Nature, 419, 449-450.
495
Tareke, E., Rydberg, P., Karlsson, P., Eriksson, S., & Tornqvist, M. (2002). Analysis of
496
acrylamide, a carcinogen formed in heated foodstuffs. Journal of Agricultural and Food
497
Chemistry, 50, 4998-5006.
498
Weisz, G. M., Schneider, L., Schweiggert, U., Kammerer, D. R., & Carle, R. (2010).
499
Sustainable sunflower processing – I. Development of a process for the adsorptive
500
decolorization of sunflower [Helianthus annuus L.] protein extracts. Innnovative Food Science
501
and Emerging Technologies, 11, 733-741.
20
Figure captions
Figure 1. Scheme for the fractionation of peptides from olive water.
Figure 2. Elution profile of the fraction obtained after partial fractionation by SPE (Sep Pak
Vac C18 Cartridge) and ultrafiltration (cut off 10,000 Da) of olive water. Toyopearl HW-55F
in glass column (600 x 26 mm I.D.). Eluent: 0.3 mol/L ammonium acetate, pH 4.0. Flow: 2
mL/min. Absorbance at 280 nm. (1)-(11) Collected fraction numbers.
Figure 3. Polyphenols (g caffeic acid/100 mL x 10-3), peptides (mg/L) and acrylamide (µg/L
x 10-1) after heat treatment (121ºC for 30 min) in the 11 different fractions obtained after
SPE, ultrafiltration and SEC (Toyoupearl HW55-F, eluent: 0.3 mol/L ammonium acetate, pH
4.0, flow: 2 mL/min) of olive water. Data represent mean values (n = 3). Error bars indicate
95% confidence intervals.
21
Figure(s)
Partial fractionation of olive water by SPE
Fraction A (H2O)
Fraction B (MeOH)
Ultrafiltration
(Molecular weight cut-off of 10,000 Da)
Fraction C (retentate)
Analysis for acrylamide after
heat treatment (LC-MS/MS)
Fraction D (permeate)
Analysis for acrylamide after heat treatment
(LC-MS/MS)
Fractionation of peptides by SEC
(Glass column with Toyopearl HW-55F)
Analysis of the different fractions (fractions 1–11)
Analysis for peptides
(Bradford method)
Analysis for polyphenols
(Folin-Ciocalteu assay)
Characterization of peptides
(MALDI-TOF/TOF-MS)
Analysis for acrylamide
after heat treatment
(LC-MS/MS)
Determination
of polyphenols
(LC-MS)
Adsorptive removal of polyphenols
(Amberlite XAD 16 HP)
Fraction E
(Compounds not retained on the resin)
Analysis for peptides and polyphenols
Analysis for acrylamide after heat
treatment (LC-MS/MS)
22
Fraction F
(Compounds retained on the resin)
Analysis for peptides and polyphenols
Analysis for acrylamide after heat
treatment (LC-MS/MS)
Figure(s)
1
2
3
4
5
6
7
8
9
10
11
nm
Absorbance
at 280
Absorbancia
a 280nm
350 mV
80
120
160
200
240
80
160
240
320
400
23
280
480
min
mL
Figure(s)
Polyphenols
Peptides
Acrylamide
Fraction
24