- Email: [email protected]
Hydroxyl radical oxidation of feruloylated arabinoxylan
Carbohydrate Polymers 152 (2016) 263–270
Contents lists available at ScienceDirect
Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol
Hydroxyl radical oxidation of feruloylated arabinoxylan Attila Bagdi a,b , Sándor Tömösközi b , Laura Nyström a,∗ a
Institute of Food, Nutrition and Health, ETH Zurich, Schmelzbergstrasse 9, CH-8092 Zurich, Switzerland Department of Applied Biochemistry and Food Science, Budapest University of Technology and Economics (BME), Muegyetem rkp. 3, H-1111 Budapest, ˝ Hungary b
a r t i c l e
i n f o
Article history: Received 17 May 2016 Received in revised form 26 June 2016 Accepted 27 June 2016 Available online 28 June 2016 Keywords: Arabinoxylan Hydroxyl radical Oxidation EPR Viscosity Gelation
a b s t r a c t Feruloylated arabinoxylan (AX) has a unique capacity to form covalent gels in the presence of certain oxidizing agents. The present study demonstrates that hydroxyl radical oxidation does not provoke ferulic acid dimerization and thus oxidative gelation. We studied the hydroxyl radical mediated oxidation of an alkali-extracted AX preparation (purity: 92 g/100 g dry matter) that showed gel-forming capability upon peroxidase/hydrogen peroxide treatment. Hydroxyl radicals were produced with ascorbate-driven Fenton reaction and the radical formation was monitored with electron paramagnetic resonance, using a POBN/EtOH spin trapping system. Oxidation was carried out at different catalytic concentrations of iron (50 and 100 M) and at different temperatures (20 ◦ C, 50 ◦ C, and 80 ◦ C). It was demonstrated that hydroxyl radical oxidation does not provoke gel formation, but viscosity decrease in AX solution, which suggests polymer degradation. Furthermore, it was demonstrated that hydroxyl radical formation in AX solution can be initiated merely by increasing temperature. © 2016 Published by Elsevier Ltd.
1. Introduction Hydroxyl radical (• OH) oxidation has been used for the modification of technological and nutritional properties of different biopolymers, but so far it has not been applied to feruloylated arabinoxylan (AX), which is one of the most important cereal polysaccharides. AX is an important dietary fibre component possessing both health-promoting (e.g. cholesterol lowering activity, prebiotic effect) as well as functional properties (e.g. interaction with gluten proteins) (Mendis & Simsek, 2014; Noort, van Haaster, Hemery, Schols, & Hamer, 2010). Composed of -1,4 linked d-xylopyranosyl backbone, AX contains intermittent ␣-larabinofuranoside monomeric units substituted at C (O)-3 and/or C (O)-2 positions of the xylose units. Additionally, esterification of trans-ferulic acid at the C (O)-5 position of arabinose residues is characteristic of cereal arabinoxylan. These ferulic acid moieties can form covalent linkages with other molecules, such as proteins or lignin (Kiszonas, Fuerst, & Morris, 2013; Lam & Stone, 1994; Stone & Morell, 2009).
Abbreviations: AX, feruloylated arabinoxylan; NOX, non-oxidized; OX1, oxidation level 1; OX2, oxidation level 2; • OH, hydroxyl radical; EPR, electron paramagnetic resonance. ∗ Corresponding author. E-mail addresses: [email protected] (A. Bagdi), [email protected] (S. Tömösközi), [email protected] (L. Nyström). http://dx.doi.org/10.1016/j.carbpol.2016.06.105 0144-8617/© 2016 Published by Elsevier Ltd.
Several attempts have been made to modify the structure of AX in order to customize the properties of the biopolymer for specific applications. A common approach is to enzymatically degrade AX in order to achieve better functionality in dough development (Courtin & Delcour, 2002). Chemical modifications of AX (e.g. esterification, carboxymethylation, ethylation) have also been studied by a number of researchers in order to modify the properties of the polymer for alternative applications (e.g. pharmaceutical, composites) (Buchanan et al., 2003; Saghir, Iqbal, Hussain, Koschella, & Heinze, 2008; Saghir, Iqbal, Koschella, & Heinze, 2009). Oxidative crosslinking of AX in the presence of free radical generating agents is one of the best-studied AX modifications within the literature. The crosslinking occurs through phenoxy radical-mediated oxida˜ tive dimerization of feruloyl units (Nino-Medina et al., 2009; Stone & Morell, 2009), which can be achieved chemically (Izydorczyk, Biliaderis, & Bushuk, 1990) as well as enzymatically (Izydorczyk et al., 1990; Martínez-López et al., 2011). The crosslinking results in the formation of a gel, giving AX, for example, the potential to be used in controlled protein delivery or wound management aid (Lloyd, Kennedy, Methacanon, Paterson, & Knill, 1998; MartínezLópez et al., 2013). • OHs might present an alternative mode of oxidative modification of AX. • OHs are highly reactive free radical species, which can be produced in aqueous systems by the Fenton-reaction (Wardman & Candeias, 1996). This reaction can be enhanced with ascorbate that acts as pro-oxidant, thus providing reduced catalytic metals
264
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
for the Fenton reaction (a.k.a. ascorbate-driven Fenton reaction) (Burkitt, 2003). • OH oxidation of cereal beta-glucan has been examined comprehensively, and it has been found that • OHs provoke polymer degradation resulting in the decrease of the viscosity of aqueous beta-glucan solution (Faure, Sánchez-Ferrer, Zabara, Andersen, & Nyström, 2014; Kivelä, Henniges, Sontag-Strohm, & Potthast, 2012). Additionally, it was also suggested that oxidation provokes other chemical changes (the formation of carbonyl groups), (Faure et al., 2014; Kivelä et al., 2012), technological changes (de Moura et al., 2011), as well as improved health-related properties (increased bile-acid binding capacity) of beta-glucan (de Moura et al., 2011). To the best of our knowledge direct oxidation of AX using • OH has not been examined until now. Since the structure of AX (branched backbone, the presence of ferulic acid esters) is considerably different from the structure of cereal beta-glucan (linear polymer without substitution), AX can be anticipated to behave differently upon • OH oxidation compared to beta-glucan. The aim of this study was to evaluate the effect of hydroxyl radical exposure of AX on its degradation using EPR to monitor the radicals formed in the Fenton reaction. AX was extracted from wheat aleuronerich flour and its degradation together with gelling behaviour was evaluated using viscosity measurements. 2. Materials and methods 2.1. Materials Aleurone-rich flour that was used as raw material for AX extraction was provided by Gyermelyi Zrt., Hungary. The composition of the flour was published earlier; its AX content is 12 g/100 g d.m. (Bagdi et al., 2016). ␣-Amylase (Termamyl® 300 L, from Bacillus licheniformis, 500 U/ml; Novozymes A/S, Bagsvaerd, Denmark), protease (Subtilisin A from Bacillus licheniformis, 300 U/ml; Megazyme, Bray, Ireland), and amyloglucosidase (from Aspergillus niger; 200 U/ml; Megazyme, Bray, Ireland) enzyme solutions were used as provided by their manufacturers. They were stored at 4 ◦ C. 1 ml lichenase (from Bacillus subtilis, 1000 U/ml; Megazyme, Bray, Ireland) enzyme solution was diluted to 20.0 ml with 20 mM sodium phosphate buffer (pH 6.5). 10 mg of peroxidase (from horseradish, 150 U/mg, Sigma–Aldrich, Buchs, Switzerland) enzyme preparation was dissolved in 5 ml of 50 mM sodium phosphate buffer (pH 6). Lichenase and peroxidase solutions were stored at −20 ◦ C. POBN (␣-(4-Pyridyl N-oxide)-N-tert-butylnitrone; 99%), TEMPO (2,2,6,6-Tetramethy-piperidin-1-yl)oxyl, sublimed, 99%)), hydrogen peroxide (30 wt.% in H2 O), iron (II) sulphate heptahydrate (≥99.0%), d-(+)-xylose (≥99%), l-(+)-arabinose (≥99%), ferulic acid (99 +%), o-coumaric acid (≥97%), acetic acid (≥99.8%), sodium acetate (≥99.0%), acetonitrile (≥99.9%), and hydrochloric acid (37%) were purchased from Sigma–Aldrich (Buchs, Switzerland). Ascorbic acid (99.5%), d-(+) glucose (≥99.5%), 2-Deoxy-d-glucose (≥98.0%), d-(+) galactose (≥99%), and ethanol (puriss. p.a., ACS reagent, absolute alcohol, without additive, 99.8%) were purchased from Fluka (Buchs, Switzerland). Sodium hydroxide (99.8%) was obtained from Fischer Chemicals AG (Zürich, Switzerland). Ultrapure water was used for all of the experiments. 2.2. Arabinoxylan extraction AX extraction method was developed based on earlier published procedures (Escarnot, Aguedo, Agneessens, Wathelet, & Paquot, 2011; Kale, Hamaker, & Campanella, 2013; Sárossy, Tenkanen, Pitkänen, Bjerre, & Plackett, 2013). 300 g of aleurone-rich flour
was suspended in 2100 ml of water and stirred for 10 min with a magnetic stirrer. The pH was adjusted to 6.0 with 2 M HCl and 240 l thermostable amylase was added to degrade starch. Amylase treatment was carried out at 95 ◦ C for 60 min. Afterwards, the slurry was cooled to room temperature and 4 M NaOH was added to reach a final concentration of 0.5 M and was stirred for 30 min. NaOH treatment aimed the solubilisation of AX that was covalently linked to insoluble fibre molecules. The suspension was neutralized with 4 M HCl to pH 7 and the slurry was centrifuged (3220g, 20 ◦ C, 20 min). The pellet, which contained all the nonwater-soluble material, was discarded. The pH of the supernatant, which contained the solubilized AX, was adjusted to 6.0 and the amylase treatment was repeated (95 ◦ C, 60 min) in order to achieve complete starch degradation. Afterwards, the pH was increased to 6.5 with NaOH and 200 l lichanase solution was added to degrade beta-glucan. The sample was stirred at 60 ◦ C for 60 min. Subsequently, the sample was autoclaved at 121 ◦ C for 20 min to deactivate the applied enzymes and to denature proteins. The precipitated material was separated by centrifugation (3220g, 20 ◦ C, 20 min). The pellet was discarded. The supernatant was mixed with equal volume of ethanol and stored at 4 ◦ C overnight. Next day the precipitate was separated by centrifugation (4 ◦ C, 3220g, 40 min). The supernatant was discarded and the pellet was dissolved by continuous stirring for 90 min in 300 ml water at 80 ◦ C, pH 4.5. 360 l amyloglucosidase solution was added and the sample was incubated at 60 ◦ C for 60 min. Afterwards, the pH was increased to 8.0 and 300 l protease enzyme solution was added. The sample was incubated at 60 ◦ C for 180 min. At the end of the incubation, the sample was boiled for 10 min to deactivate the enzymes. After cooling, the sample was ultra-centrifuged (48000g, 20 min, 20 ◦ C) and the pellet was discarded. The supernatant was dialyzed in 100 fold amount of water for 48 h. The dialyzed sample was mixed with equal volume of ethanol to precipitate AX and kept at 4 ◦ C overnight. Next day the sample was ultra-centrifuged (48000g, 40 min, 4 ◦ C). The supernatant was discarded and the pellet was collected and freeze-dried. The dried product was pulverized in a ball mill equipped with a 15 ml grinding bowl and two grinding balls with a diameter of 15 mm (Mini-Mill Pulverisette 23, Fritsch, Idar-Oberstein, Germany). The final AX preparation product was stored in a glass container at room temperature.
2.3. Chemical composition AX preparation was analysed for moisture and protein (Nx6.25) content, according to international standards (AACC International, 1999). Arabinoxylan content was determined using a hydrochloric acid hydrolysis – anion-exchange chromatography method with small modifications (Houben, de Ruijter, & Brunt, 1997). 20 mg of the AX preparation was dispersed in 2 ml water and 2 ml 4.0 M HCl, and was hydrolysed for 45 min at 100 ◦ C in a 10 ml screw cap tube. During hydrolysis, the sample was mixed frequently to aid dispersion. Following hydrolysis, the sample was cooled down and subsequently neutralised by the addition of 2 ml 4.0 M NaOH. Finally, the sample was diluted in order to reach an analyte concentration of 5–20 mg/ml and the internal standard (2-deoxy-dglucose) was added to reach a final concentration of 10 mg/ml. The separation and the detection were carried out with a Dionex CarboPac PA1 analytical column, using 1 M sodium hydroxide as the mobile phase following the procedure of Houben et al. (1997). The instrument consisted of Dionex AS50 Autosampler, Dionex AS50 Pneumatic controller, Dionex AS50 Thermal Compartment, Dionex ED50 Gradient Pump and a Dionex PED1 detector (Dionex, Sunnyvale, USA). Arabinoxylan content was calculated as 0.88 × (g/100 g arabinose + g/100 g xylose).
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270 Table 1 Experimental concentration of feruloylated arabinoxylan (AX), ferrous ion (Fe2+ ), ascorbic acid (AH2 ), and hydrogen peroxide in oxidation experiments. Name of treatment
AX (g/ml)
Fe 2+ (M)
AH2 (M)
H2O2 (M)
Non-oxidized (NOX) Oxidation level 1 (OX1) Oxidation level 2 (OX2)
0.02 0.02 0.02
0 50 100
0 500 500
0 500 500
The ferulic acid content was determined with an RP-HPLC method (Rouau et al., 2003). An Agilent 1100 HPLC with an xBridge Shield RP18 3.5 um (3.0 × 150 mm) column from Waters was used. O-coumaric acid was applied as internal standard. The measurements were carried out in triplicates. For the measurement of the solid-state AX isolate, 5 mg sample was used. For the measurement of AX solutions, 250 l solution with 0.02 g/ml concentration was tested. Iron and copper content of the AX preparation was determined by atomic absorption methods. Samples were mineralized by microwave-assisted digestion under nitrogen atmosphere in a TurboWave Digestion System (MLS GmbH, Germany). The concentrations were measured by graphite furnace atomic absorption spectrometry (AA240Z, Agilent, USA). 2.4. Oxidation of AX AX solution was prepared by dispersing 132 mg AX preparation in water to reach a final volume of 5 ml. The dispersion was heated and vortexed frequently in a water bath at 80 ◦ C until the solid material was entirely dissolved. AX solution was prepared freshly every day. • OH oxidized samples were prepared by adding H O, Fe 2+ , AH , 2 2 and H2 O2 in this order to the AX solution to reach the given final concentrations (Table 1). In order to monitor the viscosity change and the evaluation of the EPR signal simultaneously, the samples were split into two. The sample pairs (one for viscosity, one for EPR measurement) were incubated together and the measurements were carried out at the same time. Experiments were carried out at different temperatures, i.e. 20 ◦ C, 50 ◦ C, and 80 ◦ C. The EPR spectra and the viscosity were recorded at different times of incubation (0 min, 30 min, 60 min, 90 min, 120 min, and 180 min). Ascorbic acid, iron (II) sulphate, and hydrogen peroxide solutions were prepared freshly every day, 15 min before the beginning of the tests to avoid changes during storage. 2.5. Rheological tests Rheological measurements were carried out using an AR-2000 rheometer (TA Instruments, New Castle, DE, USA) equipped with a cone and plate geometry (radius: 20 mm; cone angle: 2◦ ). 59 m gap between the cone and plate was used. The formation of AX gel was monitored with a small amplitude oscillatory method as follows: 0.026% g/ml AX solution was mixed with water, peroxidase enzyme solution and H2 O2 to reach a final concentration of 0.02 g/ml AX, 20 M H2 O2 and 30 U/ml peroxidase activity. The solutions were kept at 4 ◦ C before mixing. After mixing, the solution was immediately placed on the rheometer plate, of which temperature was set to 4 ◦ C. A solvent trap cover was used with silicone oil to prevent water evaporation. AX gelation was initiated by an increase in temperature from 4 ◦ C to 20 ◦ C. Measurements were carried out at 20 ◦ C by recording the storage modulus (G ), the loss modulus (G ), and the phase angle ␦ (G /G ). Time sweep measurement was carried out at a frequency of 0.25 Hz, 20 ◦ C, and 5% strain in the linear region of the viscoelastic behaviour.
265
Table 2 Composition of the arabinoxylan isolate. Means of triplicates ± standard deviation is presented. Moisture (g/100 g) Protein (Nx6.25) (g/100 g) Arabinoxylan (g/100 g) Arabinose/Xylose Glucose (g/100 g) Galactose (g/100 g) Ferulic acid (mg/g) Fe (g/g) Cu (g/g)
4.2 ± 0.03 4.2 ± 0.02 88.5 ± 2.71 0.37 ± 0.01 0.26 ± 0.02 0.46 ± 0.04 0.98 ± 0.005 36.5 ± 7.87 25.1 ± 0.7
Frequency sweep (0.01–10 Hz) was carried out at the end of the network formation (30 min) at 5% strain and 20 ◦ C. To measure the apparent viscosity of the • OH oxidized AX solutions, aliquots were collected from the incubated AX solutions with micropipettes (110 l) at different incubation times (see 2.4) and were placed on the plate of the rheometer. The temperature of the measurements was 20 ◦ C. The flow curves were obtained in rotational mode over a shear rate range of 10–1500 1/s. Comparison of the results was done at 40 1/s. Due to the high deviation of the absolute values of viscosity, remaining viscosity was calculated according to Eq. (1). remainingviscosityat401/s (%) =
apparentviscosityat40 1s atgivenoftheoxidation apparentviscosityat40 1s atthebeginningoftheoxidation
∗ 100 (1)
2.6. Spin trapping and electron spin resonance measurement The formation of • OH was indirectly monitored with EPR measurement using POBN-EtOH spin trapping with initial addition of the spin trap according to Faure, Andersen, and Nyström (2012). In this method, the spin trap POBN was added together with EtOH to the sample of interest right after the beginning of the oxidation (see 2.4) to reach a final concentration of 2% (v/v) and 50 mM respectively. Afterwards, the mixture was vortexed and incubated. Sample aliquots were collected (50 l micropipettes (Brand GmbH, Wertheim, Germany)) and EPR signal was measured at different incubation times (see 2.4). The EPR spectra were recorded with a High Sensitive Benchtop EPR Spectrometer MiniScope MS300 (Magnettech, Berlin, Germany) at room temperature using the following settings: B0-field: 3350 G; sweep width: 100 G; sweep time: 30 s; steps: 4096; number of passes: 3; modulation frequency: 1000 mG; microwave attenuation: 10 dB. Relative EPR signal corresponding to the content of spin adducts was obtained by calculating the ratio between the peak-to-peak- amplitude of the first doublet in the EPR signal of POBN-hydroxyethyl radical adduct and the peak-to-peak-amplitude of the first singlet in the EPR signal of a TEMPO ((2,2,6,6-tetramethylpiperidin-1-yl)oxidanyl) reference solution (2 M in H2 O). The TEMPO solution was measured in triplicate on each day. 3. Results and discussion 3.1. Characterisation of the AX preparation 3.1.1. Composition The AX preparation produced had outstandingly high arabinoxylan content (Table 2). This purity (92.3 g/100 g in dry matter) is comparable to that of commercially available arabinoxylan preparations. Yet, it must be mentioned that the high purity was achieved
266
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
Fig. 1. Gel development of 0.02 g/ml AX solution upon oxidation with peroxidase/H2 O2 at 20 ◦ C, 0.25 Hz, and 5% strain (A) and the mechanical spectrum of AX (0.02 g/ml) gel at 5% strain (B). Phase angle (␦ (G /G ) ), and loss modulus (G ) are presented.
), storage modulus (G
at the expense of the yield. The yield by raw material dry basis was 1 g AX/100 g aleurone-rich flour. Other substances present in the AX preparation were mainly proteins and moisture. The low amounts of glucose and galactose demonstrate that negligible amount of other carbohydrates were present in the sample. The ferulic acid content of the AX preparation was 1.11 mg/ g AX. The feruloylation level of extracted AX samples reported in the literature is quite variable. For instance, Carvajal-Millan, Guigliarelli, Belle, Rouau, and Micard (2005) reported a ferulic acid content of 2.3 mg/g AX, extracted from a wheat milling fraction (endosperm of wheat kernel), while Martínez-López et al. (2011) reported 0.25 mg/g AX, produced from nixtamalized maize bran. The feruloylation level of AX preparations is highly dependent on the applied extraction method, in particular on the conditions of the alkali treatment (Kale et al., 2013). 3.1.2. Gel-forming capacity Oxidative gelation of AX with a treatment of peroxidase/H2 O2 is well documented, (Carvajal-Millan et al., 2005; Dervilly-Pinel, Rimsten, Saulnier, Andersson, & Åman, 2001; Vansteenkiste, Babot, Rouau, & Micard, 2004), and also, the AX sample used in this study exhibited a weak gel behaviour typical to AX gels after such treatment (Fig. 1). At low frequencies, G was obviously higher than G indicating pronounced elastic behaviour (Fig. 1B). At higher frequency range (10 Hz), these values equalized. While G was highly frequency-dependent indicating AX chain mobility in the gel, G increased only slightly with frequency (from 19.9 to 25.6 Pa). The development of AX gel took place rapidly; a visco-elastic material had already been formed when the time-sweep measurement started, in spite of the low temperature applied (Fig. 1A). At the beginning of the measurement the values of ␦ (G /G ), G , and G were 14.4◦ , 16.6 and 6.1 Pa respectively. After the introduction of the peroxidase and H2 O2 , gel development took place in the first 12 min of the measurement, followed by a plateau region that
Fig. 2. EPR signal of POBN-hydroxyethyl radical spin adduct generated by ascorbatedriven Fenton-reaction in AX solution at 20 ◦ C, 60 min after the beginning of the oxidative reaction, at three different oxidation levels (non-oxidized: NOX, 0.02 g/ml AX; oxidation level 1: OX1, 0.02 g/ml AX, 50 M Fe 2+ , 500 M AH2 and H2 O2 ; oxidation level 2: OX2, 0.02 g/ml AX, 100 M Fe 2+ , 500 M AH2 and H2 O2 ).
stayed stable until the end of the measurement. The values of ␦ (G /G ), G , and G at the plateau region were 4.6◦ , 21 and 1.7 Pa, respectively. The ferulic acid content of AX has a great impact on its gelforming capability since AX gelling occurs as a result of ferulic acid dimerization (Kale et al., 2013). The presented gelling behaviour along with the measured ferulic acid content (1.11 mg/g AX d.m.) is in accordance with the findings of Carvajal-Millan et al. (2005); who reported similar gel-forming behaviour for a ferulic acid content of 1.40 mg/g AX. Overall these results demonstrate that the examined AX preparation contains a sufficient amount of ferulic acid for oxidative gelation. 3.2. Formation of • OH in AX solutions The formation of hydroxyl radicals generated in the Fenton reaction was monitored using EPR spin trapping. The EPR signal increased in the course of time, which shows that stable spin adducts were formed and accumulated to detectable levels. The recorded EPR spectra (Fig. 2) showed a characteristic signal of the POBN-hydroxyethyl radical spin adduct (hyperfine coupling constants: aN = 15.5 G and aH = 2.6 G) (Buettner, 1987), which reflect the amount of hydroxyl radicals trapped by the POBN-ethanol system, and thereby the radicals present in the reaction mixture. The oxidation levels used in this work were chosen based on preliminary studies at room temperature in order to reach different levels of • OH development (data not shown). Three different conditions with varying levels of • OH formation were selected (Table 1).
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
Fig. 3. Flow curve of 0.02 g/ml arabinoxylan (AX) solutions at different oxidation levels, 180 min after the beginning of the oxidation. Non-oxidized sample (NOX, , 0.02 g/ml AX, 50 M Fe2+ , 500 M AH2 and H2 O2 ) at 20 ◦ C, and oxidation level 1 at 80 ◦ C ( ) are presented.. AX) and oxidation level 1 (OX1
267
, 0.02 g/ml
Fig. 4. Relationship between the relative EPR signal and the remaining viscosity (%) in hydroxyl radical (• OH) oxidized feruloylated arabinoxylan (AX) solution (0.02 g/ml). Squared Pearson’s correlation coefficient is presented.
The NOX (non-oxidized) sample at 20 ◦ C did not show EPR signal, which shows that there were no detectable radical formation, while OX2 (oxidation level 2) sample showed higher relative EPR signal than OX1 (oxidation level 1) (see discussion in 3.5). Since the applied spin trapping system has been shown to be selective to • OH (Pou et al., 1994), it can be concluded that the applied oxidizing levels produced • OH at different speeds, which was successfully detected.
3.3. Apparent viscosity of AX solution with and without oxidation Within the examined range of shear rate the apparent viscosity of NOX solution showed a pseudoplastic behaviour. At low shear rate the viscosity showed moderate shear rate dependency, while at higher shear rate substantial shear thinning was observed (Fig. 3). This behaviour is characteristic of linear polymer solutions and has been presented with AX solutions (Izydorczyk & Biliaderis, 1992; Shelat et al., 2010). The flow curve of the oxidized solutions changed compared to NOX sample (see discussion in 3.4). The apparent viscosity (40 1/s) of the freshly dissolved AX preparation (0.02 g/ml) varied between 0.08 and 0.11 Pa s among batches prepared for different measurements. Due to this high deviation we calculated the remaining viscosity in percentage and compared the values during oxidation to the initial viscosities to present the effect of oxidation. The measured values are in accordance with ear-
lier published results (Izydorczyk & Biliaderis, 1992; Shelat et al., 2010).
3.4. Relationship between relative EPR signal and the viscosity change of AX solution A strong negative correlation was found between relative EPR signal of oxidized AX solutions and remaining viscosity (Fig. 4). Since the 0 min viscosity values were arbitrarily set to 100%, they were excluded from the analysis. This result suggests that • OH formation is almost exclusively responsible for the viscosity loss of AX solution. Not only the remaining viscosity, but also the shape of the flow curve changed with increasing EPR signal: the pseudoplastic behaviour decreased in samples with more hydroxyl radicals and therefore higher EPR signal (Fig. 3). This observation along with the decreasing viscosity suggests a lower molecular weight in the solution and therefore indicates that • OH oxidation causes the depolymerisation of AX (Izydorczyk & Biliaderis, 1992). Based on these findings it can be stated that • OH oxidation did not induce AX gel formation, but resulted in polymer degradation, although at lower extent of viscosity decrease other attributes (i.e. conformational changes) might also have played a role (see 3.5). The lack of gel formation does not necessarily mean that ferulic acid dimerization did not happen; it is also possible that ferulic
268
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
Fig. 5. The effect of hydroxyl radical (• OH) oxidation on the remaining viscosity (%) and the relative EPR signal of feruloylated arabinoxylan (AX) solution at different temperatures and at different oxidation levels. I.A: viscosity change at 20 ◦ C; I.B: relative EPR signal at 20 ◦ C; II.A: viscosity change at 50 ◦ C; II.B: relative EPR signal at 50 ◦ C; , 0.02 g/ml AX), oxidation level 1 (OX1, , 0.02 g/ml AX, 50 M Fe2+ , 500 M AH2 III.A: viscosity change at 80 ◦ C; III.B: relative EPR signal at 80 ◦ C. Non-oxidized (NOX, and H2 O2 ), and oxidation level 2 (OX2, , 0.02 g/ml AX, 100 M Fe2+ , 500 M AH2 and H2 O2 ) samples are presented. Means of triplicates and 95% confidence intervals are presented.
acid dimerization occurred, but the gel development was inhibited by the simultaneous depolymerisation of the xylan chain. In order to clarify this point we examined the ferulic acid content of • OH oxidized AX in comparison with the ferulic acid content of NOX sample. The results showed that the ferulic acid content did not change upon hydroxyl radical oxidation (NOX, 20 ◦ C, 180 min: 0.99 ± 0.01 mg/g; OX1, 20 ◦ C, 180 min: 0.99 ± 0.01 mg/g; OX1 80 ◦ C, 180 min: 0.98 ± 0.03 mg/g), confirming that • OH oxidation did not induce ferulic acid dimerization. Depolymerisation through • OH radicals was also likely the cause of AX degradation when subjected to ␥-irradiation in the study of Micard, Surget, Raffi, and Rouau (2003) since ␥-radiation is known to produce • OHs. Carvajal-Millan et al. (2005) also reported radicalmediated AX degradation working with laccase-induced AX gels. In their study laccase-induced phenoxy radicals caused gel-hardness reduction, which was shown to be caused by the depolymerisation of AX chains.
Viscosity decrease provoked by • OHs was observed also in the case of beta-glucan oxidation (Faure et al., 2012; Fry, 1998). Based on other studies working with beta-glucan, it can be reasonably assumed that apart from polymer degradation, • OH oxidation leads to other chemical changes in AX as well, such as new carbonyl and carboxylic groups on the polymer backbone (Faure et al., 2014; Kivelä et al., 2012). 3.5. The effect of temperature on • OH oxidation of AX The viscosity changes were shown to be dependent on the amount of radicals formed at different temperatures (Fig. 5). In the NOX sample with no added oxidizing agents at 20 ◦ C no EPR signal was detected (Fig. 5I.B) and no viscosity decrease occurred (Fig. 5I.A). At 50 ◦ C there was a small, but detectable radical formation (Fig. 5II.B), however, viscosity decrease was not observed (Fig. 5IIA) in the NOX samples. At 80 ◦ C a remarkable increase in
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
relative EPR signal (Fig. 5III.B) and a viscosity decrease up to 73% (Fig. 5III.A) was observed. Since no oxidative chemicals were added to NOX sample, this result indicates that at higher temperature there were suitable conditions in the AX solution for • OH formation. Similar behaviour was observed earlier in the case of beta glucan (Faure, Knüsel, & Nyström, 2013; Faure, Koppenol, & Nyström, 2015). In these studies, it was assumed that the endogenous iron content and the supposed reducing capacity of the carbohydrate led to Fenton reaction and provoked • OH formation. We suggest that the same processes are responsible for • OH formation in AX solution at 80 ◦ C as well: the endogenous Fe2+ oxidizes to Fe3+ in the presence of oxygen, causing • OH formation, while the reducing carbohydrate regenerates Fe2+ from Fe3+ resulting in a self-maintaining process. The AX preparation contains a considerable amount of metal elements (Fe, Cu) that can take part in Fenton reaction (Table 2). These metals either can be complexed metal residues that could not be removed during dialysis or can originate from the ball-milling process, as contamination. The final concentration of the endogenous Cu and Fe in the experimental AX solution was calculated to 7.8 M and 13.2 M respectively, which concentrations are comparable to the concentration of the Fe2+ (50 M) that was added to the oxidized samples. These endogenous metals are likely to participate in the oxidative reaction, and oxidized metals can be reduced by AX. The metal-reducing effect of AX was shown previously with Au3+ and Ag+ ions (Amin et al., 2013). The fact that this behaviour was observed only at 80 ◦ C is in accordance with earlier observations reporting that polysaccharides have reducing effect only at high temperatures (Amin et al., 2013; Faure et al., 2013). An obvious explanation for the observed viscosity decrease would be that • OH-provoked polymer degradation was responsible for it. However, in earlier studies (Andrewartha, Phillips, & Stone, 1979; Carvajalmillan, Guigliarelli, Belle, Rouau, & Micard, 2005) it was found that although viscosity decrease at high temperature (boiled for 15 min) occurred, molecular weight reduction could not be observed. Andrewartha et al. (1979) suggested that instead of degradation conformational changes in WEAX chains are responsible for the observed viscosity decrease, although the reason for this behaviour could not be explained. Whether indeed conformational changes occurred under the present experimental conditions and if • OHs played a role in these conformational changes need further examinations. In the case of the oxidized samples OX1 and OX2, a pronounced increase in relative EPR signal along with considerable viscosity loss was observed. At 20 ◦ C the radical formation (Fig. 5IB) and the viscosity change (Fig. 5IA) occurred gradually throughout the duration of measurement: OX2 samples showed faster • OH formation and slightly larger viscosity loss than OX1. Since these samples differed only in Fe2+ concentration, this result shows that Fe2+ concentration was rate-determining in • OH formation at 20 ◦ C. Increased temperature accelerated the formation of • OH, resulting in larger and faster viscosity loss. It was reported in the case of beta-glucan, that increasing temperature enhances the polymer degradation effect of • OH oxidation (Faure et al., 2013). At 50 ◦ C the relative EPR signal increased gradually (Fig. 5II.B), while the remaining viscosity reached its minimum within 30 min and stayed stable until the end of the measurement (Fig. 5II.A). At 80 ◦ C, the relative EPR signal reached its maximum within 30 min and remained stable (Fig. 5III.B), in accordance with the remaining viscosity, which decreased rapidly as well (Fig. 5III.A). At 50 ◦ C both samples showed a maximum of 55% viscosity loss (Fig. 5II.A), while at 80 ◦ C the viscosity decreased to 70% of the starting value (Fig. 5III.A). It is remarkable that OX1 and OX2 behaved differently only at 20 ◦ C (Fig. 5I.A and I.B), while at higher temperatures these samples did not show any differences (Fig. 5II.A, II.B, III.A, and III.B). This result shows that at high temperatures (50 and 80 ◦ C) Fe2+
269
concentration is not rate determining at the applied concentration range, unlike at 20 ◦ C. 4. Conclusion • OH
oxidation did not induce ferulic acid dimerization and thus gelation processes in aqueous AX solution, although the examined AX sample was capable of gel formation upon enzymatic oxidation. Instead of gelation, • OH oxidation provoked viscosity decrease, which probably originated from the oxidative degradation of the polymer. It was demonstrated that the degradation can be enhanced with a higher concentration of oxidizing agents and by increased temperature. It was observed that high temperature induces • OH formation in AX solution, also without the application of any oxidizing agents. This probably occurs due to the presence of endogenous transition metals and the reducing capacity of AX at high temperature, which together provoke radical-generating Fenton reaction. Merely by increasing the temperature, viscosity decrease of AX solution was also induced, however, it is not clear if depolymerisation or conformational changes were responsible for this behaviour. The exact changes that occur upon • OH oxidation in the polymer structure are unknown and should be the subject of future studies. Acknowledgements The study was carried out within the framework of a Scientific Exchange Programme NMS-CH project (code: 13.026) and was funded by ETH Zurich, Switzerland. We thank Gyermelyi Zrt., Hungary for providing aleurone-rich flour and Diana Góngora Jurado for her technical assistance. We thank Prof. Raffaele Mezzenga (Laboratory of Food and Soft Materials, ETH Zurich) for kindly allowing us to use the rheometer and Christophe Zeder for carrying out the AAS measurements. References AACC International. (1999). Approved Methods of Analysis (Methods 44-15.02, 46-11.02). Amin, M., Iram, F., Iqbal, M. S., Saeed, M. Z., Raza, M., & Alam, S. (2013). Arabinoxylan-mediated synthesis of gold and silver nanoparticles having exceptional high stability. Carbohydrate Polymers, 92(2), 1896–1900. http://dx. doi.org/10.1016/j.carbpol.2012.11.056 Andrewartha, K. A., Phillips, D. R., & Stone, B. A. (1979). Solution properties of wheat-flour arabinoxylans and enzymically modified arabinoxylans. Carbohydrate Research, 77, 191–204. Bagdi, A., Tóth, B., L"orincz, R., Szendi, S., Gere, A., Kókai, Z., . . . & Tömösközi, S. (2016). Effect of aleurone-rich flour on composition, baking, textural, and sensory properties of bread. LWT—Food Science and Technology, 65, 762–769. http://dx.doi.org/10.1016/j.lwt.2015.08.073 Buchanan, C. M., Buchanan, N. L., Debenham, J. S., Gatenholm, P., Jacobsson, M., Shelton, M. C., . . . & Wood, M. D. (2003). Preparation and characterization of arabinoxylan esters and arabinoxylan ester/cellulose ester polymer blends. Carbohydrate Polymers, 52(4), 345–357. http://dx.doi.org/10.1016/S01448617(02)00290-4 Buettner, G. R. (1987). Spin trapping: ESR parameters of spin adducts. Free Radical Biology and Medicine, 3(4), 259–303. http://dx.doi.org/10.1016/S08915849(87)80033-3 Burkitt, M. J. (2003). Chemical, biological and medical controversies surrounding the Fenton reaction. Progress in Reaction Kinetics and Mechanism, 28(1), 75–103. http://dx.doi.org/10.3184/007967403103165468 Carvajal-Millan, E., Landillon, V., Morel, M.-H., Rouau, X., Doublier, J.-L., & Micard, V. (2005). Arabinoxylan gels: impact of the feruloylation degree on their structure and properties. Biomacromolecules, 6(1), 309–317. http://dx.doi.org/ 10.1021/bm049629a Carvajalmillan, E., Guigliarelli, B., Belle, V., Rouau, X., & Micard, V. (2005). Storage stability of laccase induced arabinoxylan gels. Carbohydrate Polymers, 59(2), 181–188. http://dx.doi.org/10.1016/j.carbpol.2004.09.008 Courtin, C. M., & Delcour, J. A. (2002). Arabinoxylans and endoxylanases in wheat flour bread-making. Journal of Cereal Science, 35(3), 225–243. http://dx.doi.org/ 10.1006/jcrs.2001.0433 Dervilly-Pinel, G., Rimsten, L., Saulnier, L., Andersson, R., & Åman, P. (2001). Water-extractable arabinoxylan from pearled flours of wheat, barley, rye and triticale. Evidence for the presence of ferulic acid dimers and their involvement
270
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
in gel formation. Journal of Cereal Science, 34(2), 207–214. http://dx.doi.org/10. 1006/jcrs.2001.0392 Escarnot, E., Aguedo, M., Agneessens, R., Wathelet, B., & Paquot, M. (2011). Extraction and characterization of water-extractable and water-unextractable arabinoxylans from spelt bran: study of the hydrolysis conditions for monosaccharides analysis. Journal of Cereal Science, 53(1), 45–52. http://dx.doi. org/10.1016/j.jcs.2010.09.002 Faure, A. M., Andersen, M. L., & Nyström, L. (2012). Ascorbic acid induced degradation of beta-glucan: hydroxyl radicals as intermediates studied by spin trapping and electron spin resonance spectroscopy. Carbohydrate Polymers, 87(3), 2160–2168. http://dx.doi.org/10.1016/j.carbpol.2011.10.045 Faure, A. M., Knüsel, R., & Nyström, L. (2013). Effect of the temperature on the degradation of -glucan promoted by iron(II). Bioactive Carbohydrates and Dietary Fibre, 2(2), 99–107. http://dx.doi.org/10.1016/j.bcdf.2013.09.003 Faure, A. M., Sánchez-Ferrer, A., Zabara, A., Andersen, M. L., & Nyström, L. (2014). Modulating the structural properties of -d-glucan degradation products by alternative reaction pathways. Carbohydrate Polymers, 99, 679–686. http://dx. doi.org/10.1016/j.carbpol.2013.08.022 Faure, A. M., Koppenol, W. H., & Nyström, L. (2015). Iron(II) binding by cereal beta-glucan. Carbohydrate Polymers, 115, 739–743. http://dx.doi.org/10.1016/j. carbpol.2014.07.038 Fry, S. C. (1998). Oxidative scission of plant cell wall polysaccharides by ascorbate-induced hydroxyl radicals. Biochemical Journal, 332, 507–515. Houben, R., de Ruijter, C. F., & Brunt, K. (1997). Determination of the pentosan content of wheat products by hydrolysis, glucose oxidase treatment and analysis by HPAEC/PAD. Journal of Cereal Science, 26(1), 37–46. http://dx.doi. org/10.1006/jcrs.1996.0110 Izydorczyk, M. S., & Biliaderis, C. G. (1992). Influence of structure on the physicochemical properties of wheat arabinoxylan. Carbohydrate Polymers, 17(3), 237–247. http://dx.doi.org/10.1016/0144-8617(92)90010-N Izydorczyk, M. S., Biliaderis, C. G., & Bushuk, W. (1990). Oxidative gelation studies of water-soluble pentosans from wheat. Journal of Cereal Science, 11(2), 153–169. http://dx.doi.org/10.1016/S0733-5210(09)80117-3 Kale, M. S., Hamaker, B. R., & Campanella, O. H. (2013). Alkaline extraction conditions determine gelling properties of corn bran arabinoxylans. Food Hydrocolloids, 31(1), 121–126. http://dx.doi.org/10.1016/j.foodhyd.2012.09.011 Kiszonas, A., Fuerst, E., & Morris, C. (2013). Wheat arabinoxylan structure provides insight into function. Cereal Chemistry, 509. Retrieved from: http:// cerealchemistry.aaccnet.org/doi/abs/10.1094/CCHEM-02-13-0025-FI Kivelä, R., Henniges, U., Sontag-Strohm, T., & Potthast, A. (2012). Oxidation of oat -glucan in aqueous solutions during processing. Carbohydrate Polymers, 87(1), 589–597. http://dx.doi.org/10.1016/j.carbpol.2011.08.028 Lam, T. B., & Stone, B. A. (1994). Covalent cross-cinks in the cell wall, 315–320. Lloyd, L. L., Kennedy, J. F., Methacanon, P., Paterson, M., & Knill, C. J. (1998). Carbohydrate polymers as wound management aids. Carbohydrate Polymers, 37(3), 315–322. http://dx.doi.org/10.1016/S0144-8617(98)00077-0 Martínez-López, A. L., Carvajal-Millan, E., Lizardi-Mendoza, J., López-Franco, Y. L., ˜ Rascón-Chu, A., Salas-Munoz, E., . . . & Micard, V. (2011). The peroxidase/H2O2 system as a free radical-generating agent for gelling maize bran arabinoxylans: rheological and structural properties. Molecules (Basel, Switzerland), 16(10), 8410–8418. http://dx.doi.org/10.3390/molecules16108410 Martínez-López, A. L., Carvajal-Millan, E., Miki-Yoshida, M., Alvarez-Contreras, L., Rascón-Chu, A., Lizardi-Mendoza, J., & López-Franco, Y. (2013). Arabinoxylan microspheres: structural and textural characteristics. Molecules (Basel, Switzerland), 18(4), 4640–4650. http://dx.doi.org/10.3390/molecules18044640
Mendis, M., & Simsek, S. (2014). Arabinoxylans and human health. Food Hydrocolloids, 42, 239–243. http://dx.doi.org/10.1016/j.foodhyd.2013.07.022 Micard, V., Surget, A., Raffi, J., & Rouau, X. (2003). Feruloylated arabinoxylan and arabinoxylan-protein solutions do not gel upon ␥-irradiation. Food Hydrocolloids, 17(3), 297–304. http://dx.doi.org/10.1016/S0268005X(02)00080-2 de Moura, F. A., Pereira, J. M., da Silva, D. O., Zavareze, E. D. R., da Silveira Moreira, A., Helbig, E., & Dias, A. R. G. (2011). Effects of oxidative treatment on the physicochemical, rheological and functional properties of oat -glucan. Food Chemistry, 128(4), 982–987. http://dx.doi.org/10.1016/j.foodchem.2011.04.003 ˜ Nino-Medina, G., Carvajal-Millán, E., Rascon-Chu, A., Marquez-Escalante, J. A., ˜ Guerrero, V., & Salas-Munoz, E. (2009). Feruloylated arabinoxylans and arabinoxylan gels: structure, sources and applications. Phytochemistry Reviews, 9(1), 111–120. http://dx.doi.org/10.1007/s11101-009-9147-3 Noort, M. W. J., van Haaster, D., Hemery, Y., Schols, H. A., & Hamer, R. J. (2010). The effect of particle size of wheat bran fractions on bread quality-evidence for fibre-protein interactions. Journal of Cereal Science, 52(1), 59–64. http://dx.doi. org/10.1016/j.jcs.2010.03.003 Pou, S., Ramos, C. L., Gladwell, T., Renks, E., Centra, M., Young, D., . . . & Rosen, G. M. (1994). A kinetic approach to the selection of a sensitive spin trapping system for the detection of hydroxyl radical. Analytical Biochemistry, 217(1), 76–83. http://dx.doi.org/10.1006/abio.1994.1085 Rouau, X., Cheynier, V., Surget, A., Gloux, D., Barron, C., Meudec, E., . . . & Criton, M. (2003). A dehydrotrimer of ferulic acid from maize bran. Phytochemistry, 63(8), 899–903. http://dx.doi.org/10.1016/S0031-9422(03)00297-8 Sárossy, Z., Tenkanen, M., Pitkänen, L., Bjerre, A.-B., & Plackett, D. (2013). Extraction and chemical characterization of rye arabinoxylan and the effect of -glucan on the mechanical and barrier properties of cast arabinoxylan films. Food Hydrocolloids, 30(1), 206–216. http://dx.doi.org/10.1016/j.foodhyd.2012.05.022 Saghir, S., Iqbal, M. S., Hussain, M. A., Koschella, A., & Heinze, T. (2008). Structure characterization and carboxymethylation of arabinoxylan isolated from Ispaghula (Plantago ovata) seed husk. Carbohydrate Polymers, 74(2), 309–317. http://dx.doi.org/10.1016/j.carbpol.2008.02.019 Saghir, S., Iqbal, M. S., Koschella, A., & Heinze, T. (2009). Ethylation of arabinoxylan from Ispaghula (Plantago ovata) seed husk. Carbohydrate Polymers, 77(1), 125–130. http://dx.doi.org/10.1016/j.carbpol.2008.12.009 Shelat, K. J., Vilaplana, F., Nicholson, T. M., Wong, K. H., Gidley, M. J., & Gilbert, R. G. (2010). Diffusion and viscosity in arabinoxylan solutions: implications for nutrition. Carbohydrate Polymers, 82(1), 46–53. http://dx.doi.org/10.1016/j. carbpol.2010.04.019 Stone, B., & Morell, M. K. (2009). Carbohydrates. In K. Khan, & P. R. Shewry (Eds.), Wheat chemistry and technology (4th ed., pp. 322–332). St. Paul, Minnesota: AACC International, Inc. Vansteenkiste, E., Babot, C., Rouau, X., & Micard, V. (2004). Oxidative gelation of feruloylated arabinoxylan as affected by protein. Influence on protein enzymatic hydrolysis. Food Hydrocolloids, 18(4), 557–564. http://dx.doi.org/10. 1016/j.foodhyd.2003.09.004 Wardman, P., & Candeias, L. P. (1996). Fenton chemistry: an introduction. Radiation Research, 145(5), 523. http://dx.doi.org/10.2307/3579270
Contents lists available at ScienceDirect
Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol
Hydroxyl radical oxidation of feruloylated arabinoxylan Attila Bagdi a,b , Sándor Tömösközi b , Laura Nyström a,∗ a
Institute of Food, Nutrition and Health, ETH Zurich, Schmelzbergstrasse 9, CH-8092 Zurich, Switzerland Department of Applied Biochemistry and Food Science, Budapest University of Technology and Economics (BME), Muegyetem rkp. 3, H-1111 Budapest, ˝ Hungary b
a r t i c l e
i n f o
Article history: Received 17 May 2016 Received in revised form 26 June 2016 Accepted 27 June 2016 Available online 28 June 2016 Keywords: Arabinoxylan Hydroxyl radical Oxidation EPR Viscosity Gelation
a b s t r a c t Feruloylated arabinoxylan (AX) has a unique capacity to form covalent gels in the presence of certain oxidizing agents. The present study demonstrates that hydroxyl radical oxidation does not provoke ferulic acid dimerization and thus oxidative gelation. We studied the hydroxyl radical mediated oxidation of an alkali-extracted AX preparation (purity: 92 g/100 g dry matter) that showed gel-forming capability upon peroxidase/hydrogen peroxide treatment. Hydroxyl radicals were produced with ascorbate-driven Fenton reaction and the radical formation was monitored with electron paramagnetic resonance, using a POBN/EtOH spin trapping system. Oxidation was carried out at different catalytic concentrations of iron (50 and 100 M) and at different temperatures (20 ◦ C, 50 ◦ C, and 80 ◦ C). It was demonstrated that hydroxyl radical oxidation does not provoke gel formation, but viscosity decrease in AX solution, which suggests polymer degradation. Furthermore, it was demonstrated that hydroxyl radical formation in AX solution can be initiated merely by increasing temperature. © 2016 Published by Elsevier Ltd.
1. Introduction Hydroxyl radical (• OH) oxidation has been used for the modification of technological and nutritional properties of different biopolymers, but so far it has not been applied to feruloylated arabinoxylan (AX), which is one of the most important cereal polysaccharides. AX is an important dietary fibre component possessing both health-promoting (e.g. cholesterol lowering activity, prebiotic effect) as well as functional properties (e.g. interaction with gluten proteins) (Mendis & Simsek, 2014; Noort, van Haaster, Hemery, Schols, & Hamer, 2010). Composed of -1,4 linked d-xylopyranosyl backbone, AX contains intermittent ␣-larabinofuranoside monomeric units substituted at C (O)-3 and/or C (O)-2 positions of the xylose units. Additionally, esterification of trans-ferulic acid at the C (O)-5 position of arabinose residues is characteristic of cereal arabinoxylan. These ferulic acid moieties can form covalent linkages with other molecules, such as proteins or lignin (Kiszonas, Fuerst, & Morris, 2013; Lam & Stone, 1994; Stone & Morell, 2009).
Abbreviations: AX, feruloylated arabinoxylan; NOX, non-oxidized; OX1, oxidation level 1; OX2, oxidation level 2; • OH, hydroxyl radical; EPR, electron paramagnetic resonance. ∗ Corresponding author. E-mail addresses: [email protected] (A. Bagdi), [email protected] (S. Tömösközi), [email protected] (L. Nyström). http://dx.doi.org/10.1016/j.carbpol.2016.06.105 0144-8617/© 2016 Published by Elsevier Ltd.
Several attempts have been made to modify the structure of AX in order to customize the properties of the biopolymer for specific applications. A common approach is to enzymatically degrade AX in order to achieve better functionality in dough development (Courtin & Delcour, 2002). Chemical modifications of AX (e.g. esterification, carboxymethylation, ethylation) have also been studied by a number of researchers in order to modify the properties of the polymer for alternative applications (e.g. pharmaceutical, composites) (Buchanan et al., 2003; Saghir, Iqbal, Hussain, Koschella, & Heinze, 2008; Saghir, Iqbal, Koschella, & Heinze, 2009). Oxidative crosslinking of AX in the presence of free radical generating agents is one of the best-studied AX modifications within the literature. The crosslinking occurs through phenoxy radical-mediated oxida˜ tive dimerization of feruloyl units (Nino-Medina et al., 2009; Stone & Morell, 2009), which can be achieved chemically (Izydorczyk, Biliaderis, & Bushuk, 1990) as well as enzymatically (Izydorczyk et al., 1990; Martínez-López et al., 2011). The crosslinking results in the formation of a gel, giving AX, for example, the potential to be used in controlled protein delivery or wound management aid (Lloyd, Kennedy, Methacanon, Paterson, & Knill, 1998; MartínezLópez et al., 2013). • OHs might present an alternative mode of oxidative modification of AX. • OHs are highly reactive free radical species, which can be produced in aqueous systems by the Fenton-reaction (Wardman & Candeias, 1996). This reaction can be enhanced with ascorbate that acts as pro-oxidant, thus providing reduced catalytic metals
264
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
for the Fenton reaction (a.k.a. ascorbate-driven Fenton reaction) (Burkitt, 2003). • OH oxidation of cereal beta-glucan has been examined comprehensively, and it has been found that • OHs provoke polymer degradation resulting in the decrease of the viscosity of aqueous beta-glucan solution (Faure, Sánchez-Ferrer, Zabara, Andersen, & Nyström, 2014; Kivelä, Henniges, Sontag-Strohm, & Potthast, 2012). Additionally, it was also suggested that oxidation provokes other chemical changes (the formation of carbonyl groups), (Faure et al., 2014; Kivelä et al., 2012), technological changes (de Moura et al., 2011), as well as improved health-related properties (increased bile-acid binding capacity) of beta-glucan (de Moura et al., 2011). To the best of our knowledge direct oxidation of AX using • OH has not been examined until now. Since the structure of AX (branched backbone, the presence of ferulic acid esters) is considerably different from the structure of cereal beta-glucan (linear polymer without substitution), AX can be anticipated to behave differently upon • OH oxidation compared to beta-glucan. The aim of this study was to evaluate the effect of hydroxyl radical exposure of AX on its degradation using EPR to monitor the radicals formed in the Fenton reaction. AX was extracted from wheat aleuronerich flour and its degradation together with gelling behaviour was evaluated using viscosity measurements. 2. Materials and methods 2.1. Materials Aleurone-rich flour that was used as raw material for AX extraction was provided by Gyermelyi Zrt., Hungary. The composition of the flour was published earlier; its AX content is 12 g/100 g d.m. (Bagdi et al., 2016). ␣-Amylase (Termamyl® 300 L, from Bacillus licheniformis, 500 U/ml; Novozymes A/S, Bagsvaerd, Denmark), protease (Subtilisin A from Bacillus licheniformis, 300 U/ml; Megazyme, Bray, Ireland), and amyloglucosidase (from Aspergillus niger; 200 U/ml; Megazyme, Bray, Ireland) enzyme solutions were used as provided by their manufacturers. They were stored at 4 ◦ C. 1 ml lichenase (from Bacillus subtilis, 1000 U/ml; Megazyme, Bray, Ireland) enzyme solution was diluted to 20.0 ml with 20 mM sodium phosphate buffer (pH 6.5). 10 mg of peroxidase (from horseradish, 150 U/mg, Sigma–Aldrich, Buchs, Switzerland) enzyme preparation was dissolved in 5 ml of 50 mM sodium phosphate buffer (pH 6). Lichenase and peroxidase solutions were stored at −20 ◦ C. POBN (␣-(4-Pyridyl N-oxide)-N-tert-butylnitrone; 99%), TEMPO (2,2,6,6-Tetramethy-piperidin-1-yl)oxyl, sublimed, 99%)), hydrogen peroxide (30 wt.% in H2 O), iron (II) sulphate heptahydrate (≥99.0%), d-(+)-xylose (≥99%), l-(+)-arabinose (≥99%), ferulic acid (99 +%), o-coumaric acid (≥97%), acetic acid (≥99.8%), sodium acetate (≥99.0%), acetonitrile (≥99.9%), and hydrochloric acid (37%) were purchased from Sigma–Aldrich (Buchs, Switzerland). Ascorbic acid (99.5%), d-(+) glucose (≥99.5%), 2-Deoxy-d-glucose (≥98.0%), d-(+) galactose (≥99%), and ethanol (puriss. p.a., ACS reagent, absolute alcohol, without additive, 99.8%) were purchased from Fluka (Buchs, Switzerland). Sodium hydroxide (99.8%) was obtained from Fischer Chemicals AG (Zürich, Switzerland). Ultrapure water was used for all of the experiments. 2.2. Arabinoxylan extraction AX extraction method was developed based on earlier published procedures (Escarnot, Aguedo, Agneessens, Wathelet, & Paquot, 2011; Kale, Hamaker, & Campanella, 2013; Sárossy, Tenkanen, Pitkänen, Bjerre, & Plackett, 2013). 300 g of aleurone-rich flour
was suspended in 2100 ml of water and stirred for 10 min with a magnetic stirrer. The pH was adjusted to 6.0 with 2 M HCl and 240 l thermostable amylase was added to degrade starch. Amylase treatment was carried out at 95 ◦ C for 60 min. Afterwards, the slurry was cooled to room temperature and 4 M NaOH was added to reach a final concentration of 0.5 M and was stirred for 30 min. NaOH treatment aimed the solubilisation of AX that was covalently linked to insoluble fibre molecules. The suspension was neutralized with 4 M HCl to pH 7 and the slurry was centrifuged (3220g, 20 ◦ C, 20 min). The pellet, which contained all the nonwater-soluble material, was discarded. The pH of the supernatant, which contained the solubilized AX, was adjusted to 6.0 and the amylase treatment was repeated (95 ◦ C, 60 min) in order to achieve complete starch degradation. Afterwards, the pH was increased to 6.5 with NaOH and 200 l lichanase solution was added to degrade beta-glucan. The sample was stirred at 60 ◦ C for 60 min. Subsequently, the sample was autoclaved at 121 ◦ C for 20 min to deactivate the applied enzymes and to denature proteins. The precipitated material was separated by centrifugation (3220g, 20 ◦ C, 20 min). The pellet was discarded. The supernatant was mixed with equal volume of ethanol and stored at 4 ◦ C overnight. Next day the precipitate was separated by centrifugation (4 ◦ C, 3220g, 40 min). The supernatant was discarded and the pellet was dissolved by continuous stirring for 90 min in 300 ml water at 80 ◦ C, pH 4.5. 360 l amyloglucosidase solution was added and the sample was incubated at 60 ◦ C for 60 min. Afterwards, the pH was increased to 8.0 and 300 l protease enzyme solution was added. The sample was incubated at 60 ◦ C for 180 min. At the end of the incubation, the sample was boiled for 10 min to deactivate the enzymes. After cooling, the sample was ultra-centrifuged (48000g, 20 min, 20 ◦ C) and the pellet was discarded. The supernatant was dialyzed in 100 fold amount of water for 48 h. The dialyzed sample was mixed with equal volume of ethanol to precipitate AX and kept at 4 ◦ C overnight. Next day the sample was ultra-centrifuged (48000g, 40 min, 4 ◦ C). The supernatant was discarded and the pellet was collected and freeze-dried. The dried product was pulverized in a ball mill equipped with a 15 ml grinding bowl and two grinding balls with a diameter of 15 mm (Mini-Mill Pulverisette 23, Fritsch, Idar-Oberstein, Germany). The final AX preparation product was stored in a glass container at room temperature.
2.3. Chemical composition AX preparation was analysed for moisture and protein (Nx6.25) content, according to international standards (AACC International, 1999). Arabinoxylan content was determined using a hydrochloric acid hydrolysis – anion-exchange chromatography method with small modifications (Houben, de Ruijter, & Brunt, 1997). 20 mg of the AX preparation was dispersed in 2 ml water and 2 ml 4.0 M HCl, and was hydrolysed for 45 min at 100 ◦ C in a 10 ml screw cap tube. During hydrolysis, the sample was mixed frequently to aid dispersion. Following hydrolysis, the sample was cooled down and subsequently neutralised by the addition of 2 ml 4.0 M NaOH. Finally, the sample was diluted in order to reach an analyte concentration of 5–20 mg/ml and the internal standard (2-deoxy-dglucose) was added to reach a final concentration of 10 mg/ml. The separation and the detection were carried out with a Dionex CarboPac PA1 analytical column, using 1 M sodium hydroxide as the mobile phase following the procedure of Houben et al. (1997). The instrument consisted of Dionex AS50 Autosampler, Dionex AS50 Pneumatic controller, Dionex AS50 Thermal Compartment, Dionex ED50 Gradient Pump and a Dionex PED1 detector (Dionex, Sunnyvale, USA). Arabinoxylan content was calculated as 0.88 × (g/100 g arabinose + g/100 g xylose).
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270 Table 1 Experimental concentration of feruloylated arabinoxylan (AX), ferrous ion (Fe2+ ), ascorbic acid (AH2 ), and hydrogen peroxide in oxidation experiments. Name of treatment
AX (g/ml)
Fe 2+ (M)
AH2 (M)
H2O2 (M)
Non-oxidized (NOX) Oxidation level 1 (OX1) Oxidation level 2 (OX2)
0.02 0.02 0.02
0 50 100
0 500 500
0 500 500
The ferulic acid content was determined with an RP-HPLC method (Rouau et al., 2003). An Agilent 1100 HPLC with an xBridge Shield RP18 3.5 um (3.0 × 150 mm) column from Waters was used. O-coumaric acid was applied as internal standard. The measurements were carried out in triplicates. For the measurement of the solid-state AX isolate, 5 mg sample was used. For the measurement of AX solutions, 250 l solution with 0.02 g/ml concentration was tested. Iron and copper content of the AX preparation was determined by atomic absorption methods. Samples were mineralized by microwave-assisted digestion under nitrogen atmosphere in a TurboWave Digestion System (MLS GmbH, Germany). The concentrations were measured by graphite furnace atomic absorption spectrometry (AA240Z, Agilent, USA). 2.4. Oxidation of AX AX solution was prepared by dispersing 132 mg AX preparation in water to reach a final volume of 5 ml. The dispersion was heated and vortexed frequently in a water bath at 80 ◦ C until the solid material was entirely dissolved. AX solution was prepared freshly every day. • OH oxidized samples were prepared by adding H O, Fe 2+ , AH , 2 2 and H2 O2 in this order to the AX solution to reach the given final concentrations (Table 1). In order to monitor the viscosity change and the evaluation of the EPR signal simultaneously, the samples were split into two. The sample pairs (one for viscosity, one for EPR measurement) were incubated together and the measurements were carried out at the same time. Experiments were carried out at different temperatures, i.e. 20 ◦ C, 50 ◦ C, and 80 ◦ C. The EPR spectra and the viscosity were recorded at different times of incubation (0 min, 30 min, 60 min, 90 min, 120 min, and 180 min). Ascorbic acid, iron (II) sulphate, and hydrogen peroxide solutions were prepared freshly every day, 15 min before the beginning of the tests to avoid changes during storage. 2.5. Rheological tests Rheological measurements were carried out using an AR-2000 rheometer (TA Instruments, New Castle, DE, USA) equipped with a cone and plate geometry (radius: 20 mm; cone angle: 2◦ ). 59 m gap between the cone and plate was used. The formation of AX gel was monitored with a small amplitude oscillatory method as follows: 0.026% g/ml AX solution was mixed with water, peroxidase enzyme solution and H2 O2 to reach a final concentration of 0.02 g/ml AX, 20 M H2 O2 and 30 U/ml peroxidase activity. The solutions were kept at 4 ◦ C before mixing. After mixing, the solution was immediately placed on the rheometer plate, of which temperature was set to 4 ◦ C. A solvent trap cover was used with silicone oil to prevent water evaporation. AX gelation was initiated by an increase in temperature from 4 ◦ C to 20 ◦ C. Measurements were carried out at 20 ◦ C by recording the storage modulus (G ), the loss modulus (G ), and the phase angle ␦ (G /G ). Time sweep measurement was carried out at a frequency of 0.25 Hz, 20 ◦ C, and 5% strain in the linear region of the viscoelastic behaviour.
265
Table 2 Composition of the arabinoxylan isolate. Means of triplicates ± standard deviation is presented. Moisture (g/100 g) Protein (Nx6.25) (g/100 g) Arabinoxylan (g/100 g) Arabinose/Xylose Glucose (g/100 g) Galactose (g/100 g) Ferulic acid (mg/g) Fe (g/g) Cu (g/g)
4.2 ± 0.03 4.2 ± 0.02 88.5 ± 2.71 0.37 ± 0.01 0.26 ± 0.02 0.46 ± 0.04 0.98 ± 0.005 36.5 ± 7.87 25.1 ± 0.7
Frequency sweep (0.01–10 Hz) was carried out at the end of the network formation (30 min) at 5% strain and 20 ◦ C. To measure the apparent viscosity of the • OH oxidized AX solutions, aliquots were collected from the incubated AX solutions with micropipettes (110 l) at different incubation times (see 2.4) and were placed on the plate of the rheometer. The temperature of the measurements was 20 ◦ C. The flow curves were obtained in rotational mode over a shear rate range of 10–1500 1/s. Comparison of the results was done at 40 1/s. Due to the high deviation of the absolute values of viscosity, remaining viscosity was calculated according to Eq. (1). remainingviscosityat401/s (%) =
apparentviscosityat40 1s atgivenoftheoxidation apparentviscosityat40 1s atthebeginningoftheoxidation
∗ 100 (1)
2.6. Spin trapping and electron spin resonance measurement The formation of • OH was indirectly monitored with EPR measurement using POBN-EtOH spin trapping with initial addition of the spin trap according to Faure, Andersen, and Nyström (2012). In this method, the spin trap POBN was added together with EtOH to the sample of interest right after the beginning of the oxidation (see 2.4) to reach a final concentration of 2% (v/v) and 50 mM respectively. Afterwards, the mixture was vortexed and incubated. Sample aliquots were collected (50 l micropipettes (Brand GmbH, Wertheim, Germany)) and EPR signal was measured at different incubation times (see 2.4). The EPR spectra were recorded with a High Sensitive Benchtop EPR Spectrometer MiniScope MS300 (Magnettech, Berlin, Germany) at room temperature using the following settings: B0-field: 3350 G; sweep width: 100 G; sweep time: 30 s; steps: 4096; number of passes: 3; modulation frequency: 1000 mG; microwave attenuation: 10 dB. Relative EPR signal corresponding to the content of spin adducts was obtained by calculating the ratio between the peak-to-peak- amplitude of the first doublet in the EPR signal of POBN-hydroxyethyl radical adduct and the peak-to-peak-amplitude of the first singlet in the EPR signal of a TEMPO ((2,2,6,6-tetramethylpiperidin-1-yl)oxidanyl) reference solution (2 M in H2 O). The TEMPO solution was measured in triplicate on each day. 3. Results and discussion 3.1. Characterisation of the AX preparation 3.1.1. Composition The AX preparation produced had outstandingly high arabinoxylan content (Table 2). This purity (92.3 g/100 g in dry matter) is comparable to that of commercially available arabinoxylan preparations. Yet, it must be mentioned that the high purity was achieved
266
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
Fig. 1. Gel development of 0.02 g/ml AX solution upon oxidation with peroxidase/H2 O2 at 20 ◦ C, 0.25 Hz, and 5% strain (A) and the mechanical spectrum of AX (0.02 g/ml) gel at 5% strain (B). Phase angle (␦ (G /G ) ), and loss modulus (G ) are presented.
), storage modulus (G
at the expense of the yield. The yield by raw material dry basis was 1 g AX/100 g aleurone-rich flour. Other substances present in the AX preparation were mainly proteins and moisture. The low amounts of glucose and galactose demonstrate that negligible amount of other carbohydrates were present in the sample. The ferulic acid content of the AX preparation was 1.11 mg/ g AX. The feruloylation level of extracted AX samples reported in the literature is quite variable. For instance, Carvajal-Millan, Guigliarelli, Belle, Rouau, and Micard (2005) reported a ferulic acid content of 2.3 mg/g AX, extracted from a wheat milling fraction (endosperm of wheat kernel), while Martínez-López et al. (2011) reported 0.25 mg/g AX, produced from nixtamalized maize bran. The feruloylation level of AX preparations is highly dependent on the applied extraction method, in particular on the conditions of the alkali treatment (Kale et al., 2013). 3.1.2. Gel-forming capacity Oxidative gelation of AX with a treatment of peroxidase/H2 O2 is well documented, (Carvajal-Millan et al., 2005; Dervilly-Pinel, Rimsten, Saulnier, Andersson, & Åman, 2001; Vansteenkiste, Babot, Rouau, & Micard, 2004), and also, the AX sample used in this study exhibited a weak gel behaviour typical to AX gels after such treatment (Fig. 1). At low frequencies, G was obviously higher than G indicating pronounced elastic behaviour (Fig. 1B). At higher frequency range (10 Hz), these values equalized. While G was highly frequency-dependent indicating AX chain mobility in the gel, G increased only slightly with frequency (from 19.9 to 25.6 Pa). The development of AX gel took place rapidly; a visco-elastic material had already been formed when the time-sweep measurement started, in spite of the low temperature applied (Fig. 1A). At the beginning of the measurement the values of ␦ (G /G ), G , and G were 14.4◦ , 16.6 and 6.1 Pa respectively. After the introduction of the peroxidase and H2 O2 , gel development took place in the first 12 min of the measurement, followed by a plateau region that
Fig. 2. EPR signal of POBN-hydroxyethyl radical spin adduct generated by ascorbatedriven Fenton-reaction in AX solution at 20 ◦ C, 60 min after the beginning of the oxidative reaction, at three different oxidation levels (non-oxidized: NOX, 0.02 g/ml AX; oxidation level 1: OX1, 0.02 g/ml AX, 50 M Fe 2+ , 500 M AH2 and H2 O2 ; oxidation level 2: OX2, 0.02 g/ml AX, 100 M Fe 2+ , 500 M AH2 and H2 O2 ).
stayed stable until the end of the measurement. The values of ␦ (G /G ), G , and G at the plateau region were 4.6◦ , 21 and 1.7 Pa, respectively. The ferulic acid content of AX has a great impact on its gelforming capability since AX gelling occurs as a result of ferulic acid dimerization (Kale et al., 2013). The presented gelling behaviour along with the measured ferulic acid content (1.11 mg/g AX d.m.) is in accordance with the findings of Carvajal-Millan et al. (2005); who reported similar gel-forming behaviour for a ferulic acid content of 1.40 mg/g AX. Overall these results demonstrate that the examined AX preparation contains a sufficient amount of ferulic acid for oxidative gelation. 3.2. Formation of • OH in AX solutions The formation of hydroxyl radicals generated in the Fenton reaction was monitored using EPR spin trapping. The EPR signal increased in the course of time, which shows that stable spin adducts were formed and accumulated to detectable levels. The recorded EPR spectra (Fig. 2) showed a characteristic signal of the POBN-hydroxyethyl radical spin adduct (hyperfine coupling constants: aN = 15.5 G and aH = 2.6 G) (Buettner, 1987), which reflect the amount of hydroxyl radicals trapped by the POBN-ethanol system, and thereby the radicals present in the reaction mixture. The oxidation levels used in this work were chosen based on preliminary studies at room temperature in order to reach different levels of • OH development (data not shown). Three different conditions with varying levels of • OH formation were selected (Table 1).
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
Fig. 3. Flow curve of 0.02 g/ml arabinoxylan (AX) solutions at different oxidation levels, 180 min after the beginning of the oxidation. Non-oxidized sample (NOX, , 0.02 g/ml AX, 50 M Fe2+ , 500 M AH2 and H2 O2 ) at 20 ◦ C, and oxidation level 1 at 80 ◦ C ( ) are presented.. AX) and oxidation level 1 (OX1
267
, 0.02 g/ml
Fig. 4. Relationship between the relative EPR signal and the remaining viscosity (%) in hydroxyl radical (• OH) oxidized feruloylated arabinoxylan (AX) solution (0.02 g/ml). Squared Pearson’s correlation coefficient is presented.
The NOX (non-oxidized) sample at 20 ◦ C did not show EPR signal, which shows that there were no detectable radical formation, while OX2 (oxidation level 2) sample showed higher relative EPR signal than OX1 (oxidation level 1) (see discussion in 3.5). Since the applied spin trapping system has been shown to be selective to • OH (Pou et al., 1994), it can be concluded that the applied oxidizing levels produced • OH at different speeds, which was successfully detected.
3.3. Apparent viscosity of AX solution with and without oxidation Within the examined range of shear rate the apparent viscosity of NOX solution showed a pseudoplastic behaviour. At low shear rate the viscosity showed moderate shear rate dependency, while at higher shear rate substantial shear thinning was observed (Fig. 3). This behaviour is characteristic of linear polymer solutions and has been presented with AX solutions (Izydorczyk & Biliaderis, 1992; Shelat et al., 2010). The flow curve of the oxidized solutions changed compared to NOX sample (see discussion in 3.4). The apparent viscosity (40 1/s) of the freshly dissolved AX preparation (0.02 g/ml) varied between 0.08 and 0.11 Pa s among batches prepared for different measurements. Due to this high deviation we calculated the remaining viscosity in percentage and compared the values during oxidation to the initial viscosities to present the effect of oxidation. The measured values are in accordance with ear-
lier published results (Izydorczyk & Biliaderis, 1992; Shelat et al., 2010).
3.4. Relationship between relative EPR signal and the viscosity change of AX solution A strong negative correlation was found between relative EPR signal of oxidized AX solutions and remaining viscosity (Fig. 4). Since the 0 min viscosity values were arbitrarily set to 100%, they were excluded from the analysis. This result suggests that • OH formation is almost exclusively responsible for the viscosity loss of AX solution. Not only the remaining viscosity, but also the shape of the flow curve changed with increasing EPR signal: the pseudoplastic behaviour decreased in samples with more hydroxyl radicals and therefore higher EPR signal (Fig. 3). This observation along with the decreasing viscosity suggests a lower molecular weight in the solution and therefore indicates that • OH oxidation causes the depolymerisation of AX (Izydorczyk & Biliaderis, 1992). Based on these findings it can be stated that • OH oxidation did not induce AX gel formation, but resulted in polymer degradation, although at lower extent of viscosity decrease other attributes (i.e. conformational changes) might also have played a role (see 3.5). The lack of gel formation does not necessarily mean that ferulic acid dimerization did not happen; it is also possible that ferulic
268
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
Fig. 5. The effect of hydroxyl radical (• OH) oxidation on the remaining viscosity (%) and the relative EPR signal of feruloylated arabinoxylan (AX) solution at different temperatures and at different oxidation levels. I.A: viscosity change at 20 ◦ C; I.B: relative EPR signal at 20 ◦ C; II.A: viscosity change at 50 ◦ C; II.B: relative EPR signal at 50 ◦ C; , 0.02 g/ml AX), oxidation level 1 (OX1, , 0.02 g/ml AX, 50 M Fe2+ , 500 M AH2 III.A: viscosity change at 80 ◦ C; III.B: relative EPR signal at 80 ◦ C. Non-oxidized (NOX, and H2 O2 ), and oxidation level 2 (OX2, , 0.02 g/ml AX, 100 M Fe2+ , 500 M AH2 and H2 O2 ) samples are presented. Means of triplicates and 95% confidence intervals are presented.
acid dimerization occurred, but the gel development was inhibited by the simultaneous depolymerisation of the xylan chain. In order to clarify this point we examined the ferulic acid content of • OH oxidized AX in comparison with the ferulic acid content of NOX sample. The results showed that the ferulic acid content did not change upon hydroxyl radical oxidation (NOX, 20 ◦ C, 180 min: 0.99 ± 0.01 mg/g; OX1, 20 ◦ C, 180 min: 0.99 ± 0.01 mg/g; OX1 80 ◦ C, 180 min: 0.98 ± 0.03 mg/g), confirming that • OH oxidation did not induce ferulic acid dimerization. Depolymerisation through • OH radicals was also likely the cause of AX degradation when subjected to ␥-irradiation in the study of Micard, Surget, Raffi, and Rouau (2003) since ␥-radiation is known to produce • OHs. Carvajal-Millan et al. (2005) also reported radicalmediated AX degradation working with laccase-induced AX gels. In their study laccase-induced phenoxy radicals caused gel-hardness reduction, which was shown to be caused by the depolymerisation of AX chains.
Viscosity decrease provoked by • OHs was observed also in the case of beta-glucan oxidation (Faure et al., 2012; Fry, 1998). Based on other studies working with beta-glucan, it can be reasonably assumed that apart from polymer degradation, • OH oxidation leads to other chemical changes in AX as well, such as new carbonyl and carboxylic groups on the polymer backbone (Faure et al., 2014; Kivelä et al., 2012). 3.5. The effect of temperature on • OH oxidation of AX The viscosity changes were shown to be dependent on the amount of radicals formed at different temperatures (Fig. 5). In the NOX sample with no added oxidizing agents at 20 ◦ C no EPR signal was detected (Fig. 5I.B) and no viscosity decrease occurred (Fig. 5I.A). At 50 ◦ C there was a small, but detectable radical formation (Fig. 5II.B), however, viscosity decrease was not observed (Fig. 5IIA) in the NOX samples. At 80 ◦ C a remarkable increase in
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
relative EPR signal (Fig. 5III.B) and a viscosity decrease up to 73% (Fig. 5III.A) was observed. Since no oxidative chemicals were added to NOX sample, this result indicates that at higher temperature there were suitable conditions in the AX solution for • OH formation. Similar behaviour was observed earlier in the case of beta glucan (Faure, Knüsel, & Nyström, 2013; Faure, Koppenol, & Nyström, 2015). In these studies, it was assumed that the endogenous iron content and the supposed reducing capacity of the carbohydrate led to Fenton reaction and provoked • OH formation. We suggest that the same processes are responsible for • OH formation in AX solution at 80 ◦ C as well: the endogenous Fe2+ oxidizes to Fe3+ in the presence of oxygen, causing • OH formation, while the reducing carbohydrate regenerates Fe2+ from Fe3+ resulting in a self-maintaining process. The AX preparation contains a considerable amount of metal elements (Fe, Cu) that can take part in Fenton reaction (Table 2). These metals either can be complexed metal residues that could not be removed during dialysis or can originate from the ball-milling process, as contamination. The final concentration of the endogenous Cu and Fe in the experimental AX solution was calculated to 7.8 M and 13.2 M respectively, which concentrations are comparable to the concentration of the Fe2+ (50 M) that was added to the oxidized samples. These endogenous metals are likely to participate in the oxidative reaction, and oxidized metals can be reduced by AX. The metal-reducing effect of AX was shown previously with Au3+ and Ag+ ions (Amin et al., 2013). The fact that this behaviour was observed only at 80 ◦ C is in accordance with earlier observations reporting that polysaccharides have reducing effect only at high temperatures (Amin et al., 2013; Faure et al., 2013). An obvious explanation for the observed viscosity decrease would be that • OH-provoked polymer degradation was responsible for it. However, in earlier studies (Andrewartha, Phillips, & Stone, 1979; Carvajalmillan, Guigliarelli, Belle, Rouau, & Micard, 2005) it was found that although viscosity decrease at high temperature (boiled for 15 min) occurred, molecular weight reduction could not be observed. Andrewartha et al. (1979) suggested that instead of degradation conformational changes in WEAX chains are responsible for the observed viscosity decrease, although the reason for this behaviour could not be explained. Whether indeed conformational changes occurred under the present experimental conditions and if • OHs played a role in these conformational changes need further examinations. In the case of the oxidized samples OX1 and OX2, a pronounced increase in relative EPR signal along with considerable viscosity loss was observed. At 20 ◦ C the radical formation (Fig. 5IB) and the viscosity change (Fig. 5IA) occurred gradually throughout the duration of measurement: OX2 samples showed faster • OH formation and slightly larger viscosity loss than OX1. Since these samples differed only in Fe2+ concentration, this result shows that Fe2+ concentration was rate-determining in • OH formation at 20 ◦ C. Increased temperature accelerated the formation of • OH, resulting in larger and faster viscosity loss. It was reported in the case of beta-glucan, that increasing temperature enhances the polymer degradation effect of • OH oxidation (Faure et al., 2013). At 50 ◦ C the relative EPR signal increased gradually (Fig. 5II.B), while the remaining viscosity reached its minimum within 30 min and stayed stable until the end of the measurement (Fig. 5II.A). At 80 ◦ C, the relative EPR signal reached its maximum within 30 min and remained stable (Fig. 5III.B), in accordance with the remaining viscosity, which decreased rapidly as well (Fig. 5III.A). At 50 ◦ C both samples showed a maximum of 55% viscosity loss (Fig. 5II.A), while at 80 ◦ C the viscosity decreased to 70% of the starting value (Fig. 5III.A). It is remarkable that OX1 and OX2 behaved differently only at 20 ◦ C (Fig. 5I.A and I.B), while at higher temperatures these samples did not show any differences (Fig. 5II.A, II.B, III.A, and III.B). This result shows that at high temperatures (50 and 80 ◦ C) Fe2+
269
concentration is not rate determining at the applied concentration range, unlike at 20 ◦ C. 4. Conclusion • OH
oxidation did not induce ferulic acid dimerization and thus gelation processes in aqueous AX solution, although the examined AX sample was capable of gel formation upon enzymatic oxidation. Instead of gelation, • OH oxidation provoked viscosity decrease, which probably originated from the oxidative degradation of the polymer. It was demonstrated that the degradation can be enhanced with a higher concentration of oxidizing agents and by increased temperature. It was observed that high temperature induces • OH formation in AX solution, also without the application of any oxidizing agents. This probably occurs due to the presence of endogenous transition metals and the reducing capacity of AX at high temperature, which together provoke radical-generating Fenton reaction. Merely by increasing the temperature, viscosity decrease of AX solution was also induced, however, it is not clear if depolymerisation or conformational changes were responsible for this behaviour. The exact changes that occur upon • OH oxidation in the polymer structure are unknown and should be the subject of future studies. Acknowledgements The study was carried out within the framework of a Scientific Exchange Programme NMS-CH project (code: 13.026) and was funded by ETH Zurich, Switzerland. We thank Gyermelyi Zrt., Hungary for providing aleurone-rich flour and Diana Góngora Jurado for her technical assistance. We thank Prof. Raffaele Mezzenga (Laboratory of Food and Soft Materials, ETH Zurich) for kindly allowing us to use the rheometer and Christophe Zeder for carrying out the AAS measurements. References AACC International. (1999). Approved Methods of Analysis (Methods 44-15.02, 46-11.02). Amin, M., Iram, F., Iqbal, M. S., Saeed, M. Z., Raza, M., & Alam, S. (2013). Arabinoxylan-mediated synthesis of gold and silver nanoparticles having exceptional high stability. Carbohydrate Polymers, 92(2), 1896–1900. http://dx. doi.org/10.1016/j.carbpol.2012.11.056 Andrewartha, K. A., Phillips, D. R., & Stone, B. A. (1979). Solution properties of wheat-flour arabinoxylans and enzymically modified arabinoxylans. Carbohydrate Research, 77, 191–204. Bagdi, A., Tóth, B., L"orincz, R., Szendi, S., Gere, A., Kókai, Z., . . . & Tömösközi, S. (2016). Effect of aleurone-rich flour on composition, baking, textural, and sensory properties of bread. LWT—Food Science and Technology, 65, 762–769. http://dx.doi.org/10.1016/j.lwt.2015.08.073 Buchanan, C. M., Buchanan, N. L., Debenham, J. S., Gatenholm, P., Jacobsson, M., Shelton, M. C., . . . & Wood, M. D. (2003). Preparation and characterization of arabinoxylan esters and arabinoxylan ester/cellulose ester polymer blends. Carbohydrate Polymers, 52(4), 345–357. http://dx.doi.org/10.1016/S01448617(02)00290-4 Buettner, G. R. (1987). Spin trapping: ESR parameters of spin adducts. Free Radical Biology and Medicine, 3(4), 259–303. http://dx.doi.org/10.1016/S08915849(87)80033-3 Burkitt, M. J. (2003). Chemical, biological and medical controversies surrounding the Fenton reaction. Progress in Reaction Kinetics and Mechanism, 28(1), 75–103. http://dx.doi.org/10.3184/007967403103165468 Carvajal-Millan, E., Landillon, V., Morel, M.-H., Rouau, X., Doublier, J.-L., & Micard, V. (2005). Arabinoxylan gels: impact of the feruloylation degree on their structure and properties. Biomacromolecules, 6(1), 309–317. http://dx.doi.org/ 10.1021/bm049629a Carvajalmillan, E., Guigliarelli, B., Belle, V., Rouau, X., & Micard, V. (2005). Storage stability of laccase induced arabinoxylan gels. Carbohydrate Polymers, 59(2), 181–188. http://dx.doi.org/10.1016/j.carbpol.2004.09.008 Courtin, C. M., & Delcour, J. A. (2002). Arabinoxylans and endoxylanases in wheat flour bread-making. Journal of Cereal Science, 35(3), 225–243. http://dx.doi.org/ 10.1006/jcrs.2001.0433 Dervilly-Pinel, G., Rimsten, L., Saulnier, L., Andersson, R., & Åman, P. (2001). Water-extractable arabinoxylan from pearled flours of wheat, barley, rye and triticale. Evidence for the presence of ferulic acid dimers and their involvement
270
A. Bagdi et al. / Carbohydrate Polymers 152 (2016) 263–270
in gel formation. Journal of Cereal Science, 34(2), 207–214. http://dx.doi.org/10. 1006/jcrs.2001.0392 Escarnot, E., Aguedo, M., Agneessens, R., Wathelet, B., & Paquot, M. (2011). Extraction and characterization of water-extractable and water-unextractable arabinoxylans from spelt bran: study of the hydrolysis conditions for monosaccharides analysis. Journal of Cereal Science, 53(1), 45–52. http://dx.doi. org/10.1016/j.jcs.2010.09.002 Faure, A. M., Andersen, M. L., & Nyström, L. (2012). Ascorbic acid induced degradation of beta-glucan: hydroxyl radicals as intermediates studied by spin trapping and electron spin resonance spectroscopy. Carbohydrate Polymers, 87(3), 2160–2168. http://dx.doi.org/10.1016/j.carbpol.2011.10.045 Faure, A. M., Knüsel, R., & Nyström, L. (2013). Effect of the temperature on the degradation of -glucan promoted by iron(II). Bioactive Carbohydrates and Dietary Fibre, 2(2), 99–107. http://dx.doi.org/10.1016/j.bcdf.2013.09.003 Faure, A. M., Sánchez-Ferrer, A., Zabara, A., Andersen, M. L., & Nyström, L. (2014). Modulating the structural properties of -d-glucan degradation products by alternative reaction pathways. Carbohydrate Polymers, 99, 679–686. http://dx. doi.org/10.1016/j.carbpol.2013.08.022 Faure, A. M., Koppenol, W. H., & Nyström, L. (2015). Iron(II) binding by cereal beta-glucan. Carbohydrate Polymers, 115, 739–743. http://dx.doi.org/10.1016/j. carbpol.2014.07.038 Fry, S. C. (1998). Oxidative scission of plant cell wall polysaccharides by ascorbate-induced hydroxyl radicals. Biochemical Journal, 332, 507–515. Houben, R., de Ruijter, C. F., & Brunt, K. (1997). Determination of the pentosan content of wheat products by hydrolysis, glucose oxidase treatment and analysis by HPAEC/PAD. Journal of Cereal Science, 26(1), 37–46. http://dx.doi. org/10.1006/jcrs.1996.0110 Izydorczyk, M. S., & Biliaderis, C. G. (1992). Influence of structure on the physicochemical properties of wheat arabinoxylan. Carbohydrate Polymers, 17(3), 237–247. http://dx.doi.org/10.1016/0144-8617(92)90010-N Izydorczyk, M. S., Biliaderis, C. G., & Bushuk, W. (1990). Oxidative gelation studies of water-soluble pentosans from wheat. Journal of Cereal Science, 11(2), 153–169. http://dx.doi.org/10.1016/S0733-5210(09)80117-3 Kale, M. S., Hamaker, B. R., & Campanella, O. H. (2013). Alkaline extraction conditions determine gelling properties of corn bran arabinoxylans. Food Hydrocolloids, 31(1), 121–126. http://dx.doi.org/10.1016/j.foodhyd.2012.09.011 Kiszonas, A., Fuerst, E., & Morris, C. (2013). Wheat arabinoxylan structure provides insight into function. Cereal Chemistry, 509. Retrieved from: http:// cerealchemistry.aaccnet.org/doi/abs/10.1094/CCHEM-02-13-0025-FI Kivelä, R., Henniges, U., Sontag-Strohm, T., & Potthast, A. (2012). Oxidation of oat -glucan in aqueous solutions during processing. Carbohydrate Polymers, 87(1), 589–597. http://dx.doi.org/10.1016/j.carbpol.2011.08.028 Lam, T. B., & Stone, B. A. (1994). Covalent cross-cinks in the cell wall, 315–320. Lloyd, L. L., Kennedy, J. F., Methacanon, P., Paterson, M., & Knill, C. J. (1998). Carbohydrate polymers as wound management aids. Carbohydrate Polymers, 37(3), 315–322. http://dx.doi.org/10.1016/S0144-8617(98)00077-0 Martínez-López, A. L., Carvajal-Millan, E., Lizardi-Mendoza, J., López-Franco, Y. L., ˜ Rascón-Chu, A., Salas-Munoz, E., . . . & Micard, V. (2011). The peroxidase/H2O2 system as a free radical-generating agent for gelling maize bran arabinoxylans: rheological and structural properties. Molecules (Basel, Switzerland), 16(10), 8410–8418. http://dx.doi.org/10.3390/molecules16108410 Martínez-López, A. L., Carvajal-Millan, E., Miki-Yoshida, M., Alvarez-Contreras, L., Rascón-Chu, A., Lizardi-Mendoza, J., & López-Franco, Y. (2013). Arabinoxylan microspheres: structural and textural characteristics. Molecules (Basel, Switzerland), 18(4), 4640–4650. http://dx.doi.org/10.3390/molecules18044640
Mendis, M., & Simsek, S. (2014). Arabinoxylans and human health. Food Hydrocolloids, 42, 239–243. http://dx.doi.org/10.1016/j.foodhyd.2013.07.022 Micard, V., Surget, A., Raffi, J., & Rouau, X. (2003). Feruloylated arabinoxylan and arabinoxylan-protein solutions do not gel upon ␥-irradiation. Food Hydrocolloids, 17(3), 297–304. http://dx.doi.org/10.1016/S0268005X(02)00080-2 de Moura, F. A., Pereira, J. M., da Silva, D. O., Zavareze, E. D. R., da Silveira Moreira, A., Helbig, E., & Dias, A. R. G. (2011). Effects of oxidative treatment on the physicochemical, rheological and functional properties of oat -glucan. Food Chemistry, 128(4), 982–987. http://dx.doi.org/10.1016/j.foodchem.2011.04.003 ˜ Nino-Medina, G., Carvajal-Millán, E., Rascon-Chu, A., Marquez-Escalante, J. A., ˜ Guerrero, V., & Salas-Munoz, E. (2009). Feruloylated arabinoxylans and arabinoxylan gels: structure, sources and applications. Phytochemistry Reviews, 9(1), 111–120. http://dx.doi.org/10.1007/s11101-009-9147-3 Noort, M. W. J., van Haaster, D., Hemery, Y., Schols, H. A., & Hamer, R. J. (2010). The effect of particle size of wheat bran fractions on bread quality-evidence for fibre-protein interactions. Journal of Cereal Science, 52(1), 59–64. http://dx.doi. org/10.1016/j.jcs.2010.03.003 Pou, S., Ramos, C. L., Gladwell, T., Renks, E., Centra, M., Young, D., . . . & Rosen, G. M. (1994). A kinetic approach to the selection of a sensitive spin trapping system for the detection of hydroxyl radical. Analytical Biochemistry, 217(1), 76–83. http://dx.doi.org/10.1006/abio.1994.1085 Rouau, X., Cheynier, V., Surget, A., Gloux, D., Barron, C., Meudec, E., . . . & Criton, M. (2003). A dehydrotrimer of ferulic acid from maize bran. Phytochemistry, 63(8), 899–903. http://dx.doi.org/10.1016/S0031-9422(03)00297-8 Sárossy, Z., Tenkanen, M., Pitkänen, L., Bjerre, A.-B., & Plackett, D. (2013). Extraction and chemical characterization of rye arabinoxylan and the effect of -glucan on the mechanical and barrier properties of cast arabinoxylan films. Food Hydrocolloids, 30(1), 206–216. http://dx.doi.org/10.1016/j.foodhyd.2012.05.022 Saghir, S., Iqbal, M. S., Hussain, M. A., Koschella, A., & Heinze, T. (2008). Structure characterization and carboxymethylation of arabinoxylan isolated from Ispaghula (Plantago ovata) seed husk. Carbohydrate Polymers, 74(2), 309–317. http://dx.doi.org/10.1016/j.carbpol.2008.02.019 Saghir, S., Iqbal, M. S., Koschella, A., & Heinze, T. (2009). Ethylation of arabinoxylan from Ispaghula (Plantago ovata) seed husk. Carbohydrate Polymers, 77(1), 125–130. http://dx.doi.org/10.1016/j.carbpol.2008.12.009 Shelat, K. J., Vilaplana, F., Nicholson, T. M., Wong, K. H., Gidley, M. J., & Gilbert, R. G. (2010). Diffusion and viscosity in arabinoxylan solutions: implications for nutrition. Carbohydrate Polymers, 82(1), 46–53. http://dx.doi.org/10.1016/j. carbpol.2010.04.019 Stone, B., & Morell, M. K. (2009). Carbohydrates. In K. Khan, & P. R. Shewry (Eds.), Wheat chemistry and technology (4th ed., pp. 322–332). St. Paul, Minnesota: AACC International, Inc. Vansteenkiste, E., Babot, C., Rouau, X., & Micard, V. (2004). Oxidative gelation of feruloylated arabinoxylan as affected by protein. Influence on protein enzymatic hydrolysis. Food Hydrocolloids, 18(4), 557–564. http://dx.doi.org/10. 1016/j.foodhyd.2003.09.004 Wardman, P., & Candeias, L. P. (1996). Fenton chemistry: an introduction. Radiation Research, 145(5), 523. http://dx.doi.org/10.2307/3579270