The effect of high hydrostatic pressure treatment on the molecular structure of starches with different amylose content

Food Chemistry 240 (2018) 51–58

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

The effect of high hydrostatic pressure treatment on the molecular structure of starches with different amylose content _ Artur Szwengiel a, Grazyna Lewandowicz b, Adrian R. Górecki c, Wioletta Błaszczak c,⇑ a

´ University of Life Sciences, Poznan ´ , Poland Institute of Food Technology of Plant Origin, Poznan ´ University of Life Sciences, Poznan ´ , Poland Department of Biotechnology and Food Microbiology, Poznan c Institute of Animal Reproduction and Food Research, Polish Academy of Sciences, Tuwima 10, 10-748 Olsztyn, Poland b

a r t i c l e

i n f o

Article history: Received 12 April 2017 Received in revised form 3 July 2017 Accepted 17 July 2017 Available online 18 July 2017 Keywords: Starch Amylopectin High hydrostatic pressure Molecular structure

a b s t r a c t The effect of high hydrostatic pressure processing (650 MPa/9 min) on molecular mass distribution, and hydrodynamic and structural parameters of amylose (maize, sorghum, Hylon VII) and amylopectin (waxy maize, amaranth) starches was studied. The starches were characterized by high-performance sizeexclusion chromatography (HPSEC) equipped with static light scattering and refractive index detectors and by Fourier Transform Infrared (FTIR) spectroscopy. Significant changes were observed in molecular mass distribution of pressurized waxy maize starch. Changes in branches/branch frequency, intrinsic viscosity, and radius of gyration were observed for all treated starches. The combination of SEC and FTIR data showed that a-1,6-glycosidic bonds are more frequently split in pressurized amaranth, Hylon VII, and waxy maize starch, while in sorghum and maize starches, the a-1,4 bonds are most commonly split. Our results show that the structural changes found for pressurized starches were more strongly determined by the starch origin than by the processing applied. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Starch is a biodegradable polymer with simple and well-defined chemical properties. There are two major starch constituents: amylose, which is primarily linear with a few long-chain branches, and amylopectin, a highly branched molecule. Both the molecular mass of amylose/amylopectin and the amylose/amylopectin ratio can be considered the most important structural factors directly affecting the technological properties of a starch (Kurzawska et al., 2014; Yoo & Jane, 2002). It is generally known that waxy starches with trace amount of amylose require higher energy to achieve gelatinization, as a result of their higher crystallinity, than do starches with regular amylose content (Singh, Inouchi, & Nishinari, 2006; Van Hung, Maeda, & Morita, 2006). The differences in the physicochemical properties of starches such as swelling, solubility, and thermodynamic properties (melting temperature and melting enthalpy) are directly assignable to the differences in amylopectin molecular structure (molecular mass, M and its distribution) (Guo et al., 2015). ⇑ Corresponding author at: Institute of Animal Reproduction and Food Research, Polish Academy of Sciences, ul. Tuwima 10, 10-748 Olsztyn, Poland. E-mail addresses: [email protected] (A. Szwengiel), [email protected] (G. Lewandowicz), [email protected] (A.R. Górecki), [email protected] (W. Błaszczak). http://dx.doi.org/10.1016/j.foodchem.2017.07.082 0308-8146/Ó 2017 Elsevier Ltd. All rights reserved.

Various techniques have been used to study the structures of starch macromolecules with an emphasis on the branch structure of amylopectin—for example, high-performance size-exclusion chromatography (Li, Prakash, Nicholson, Fitzgerald, & Gilbert, 2016; Yoo & Jane, 2002) and high-performance anion-exchange chromatography (Kong, Corke, & Bertoft, 2009; Raghunathan, Hoover, Waduge, Liu, & Warkentin, 2017). High-performance size-exclusion chromatography (HPSEC), equipped with multiangle laser-light scattering (MALLS) and refractive index (RI) detectors, is a powerful technique that allows the absolute molecular mass of starch macromolecules to be characterized (Yoo & Jane, 2002). However, determining amylopectin molecular mass is challenging, due to the fact that it is an extremely high molecular mass polymer. It is thus difficult using any available technique to obtain a solution with a molecular dispersion of such starch without it undergoing degradation. On the other hand, physical treatment preceding the dissolution of starch granules could affect the analysis of molecular mass distribution and hydrodynamic parameters. For this reason, the results of molecular structure studies of starch may be considered and compared only if the same methods of granule dissolution and separation (of amylopectin from amylose) were used (Kurzawska et al., 2014; Yoo & Jane, 2002). As mentioned, knowledge of the molecular architecture of starch is crucial, since as well as permitting its technological use-

52

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

fulness to be predicted, it also allows control of granules behavior under physical, chemical, and enzymatic treatment. Although the high hydrostatic pressure (HHP) process has been described in the literature as a ‘‘mild technology”, it can result in significant changes in the physicochemical properties of starch granules (Hu, Zhang, Jin, Xu, & Chen, 2017; Pei-Ling, Qing, Qun, Xiao-Song, & Ji-Hong, 2012). Intensive HHP treatment is expected to affect not only the morphology/properties of granules, but also of starch molecules (amylose and/ or amylopectin). It was found that the HPLC profile of molecular mass distribution of waxy maize starch (composed mainly of amylopectin) pressurized at 650 MPa for 9 min differed significantly from that of native starch (Błaszczak, Fornal, Valverde, & Garrido, 2005). A HPSEC-MALLS-RI analysis performed by Guo et al. (2015) revealed a distinct decrease in the values of weight molar mass, number molar mass, and dispersity index of lotus starch (40% of amylose) subjected to HHP treatment (100–600 MPa/30 min). The effects of HHP treatment on the structural properties of starch have been widely documented in the literature, (Oh, Pinder, Hemar, Anema, & Wong, 2008; Yang et al., 2016), but very little is known about their molecular structure, in particular the structure of starches with varied amylose content. Therefore, the aim of this study was to determine the effect of HHP processing (650 MPa/9 min, 30 ± 2 °C) on molecular mass distribution, hydrodynamic and structural parameters (intrinsic viscosity (IV), gyration radius (Rg), hydrodynamic radius (Rh) and number of branches) of starches with varied amylose content (maize, sorghum, Hylon VII) and pure amylopectin starches (waxy maize, amaranth starch). The HHP-treated starches were characterized by HPSEC equipped with static light scattering (SLS) and RI detectors. Fourier Transform Infrared (FTIR) spectroscopy was additionally performed to identify the relationship between the hydrodynamic and FTIR parameters of the starches.

sodium nitrate (reagent plus grade), were obtained from Sigma– Aldrich (USA). Other chemicals were reagent grade and used without further treatment.

2. Materials and methods

2.3. Molecular mass distribution and hydrodynamic parameters of native and HHP-treated starches, as determined by SEC

2.2. High hydrostatic pressure treatment of starch granules Our previous study demonstrated that starch processing at 650 MPa for 9 min in excess water (3 g d.m./10 mL) may considerably influence starch crystallinity and/or the molecular mass distribution profile of the analyzed material already at ambient temperature (Błaszczak et al., 2005). In view of the above, starch was pressure-processed under the same conditions in this study. The starch–water suspensions (3 g d.m./10 mL) were thoroughly mixed and homogenized with a Polytron Ultraturrax homogenizer IKA-T18 (IKA works, Wilmington, USA) for 1 min at 12,000 rpm (revolutions per minute). The homogenized samples were closed into 50 mL Teflon tubes, thoroughly mixed, deaerated, sealed tightly, and subjected to HHP treatment using a high pressure Unipress U-303 device (Warsaw, Poland). The Teflon tubes were placed into a high pressure chamber (capacity approximately 100 mL) filled with pressuretransmitting medium, which also minimized adiabatic heating. The samples were then pressure-treated at 650 MPa for 9 min. The time taken to reach the working pressure was 2 min. The temperature inside the pressure chamber averaged 30 ± 2 °C. The pressure treatment was performed in two repetitions for each combination. After pressure treatment, the starch material was frozen in liquid nitrogen, freeze-dried, ground in a laboratory grinder (Sadkiewicz Instruments, Poland), and sieved through a screen with a mesh of 170, so as to unify the diameters. This homogeneously granulated material was sealed in tubes and stored until analysis in a dark, cold, dry place.

2.1. Materials The experimental materials were commercial maize starch (20.5% of amylose; donated by the Department of Food Concentrates, Institute of Agricultural and Food Biotechnology, Poznan´, Poland), Hylon VII starch (68% amylose; donated by the National Starch & Chemical Co.), and waxy maize starch (with trace amounts of amylose; Sigma, S-9679). The other materials, sorghum starches (19.2% amylose) and amaranth starches (pure amylopectin starch), were isolated in our laboratory from sorghum grains and amaranth seeds, respectively. The grains of Sorghum bicolor (v. Rona 1) were purchased from the Kutno Centre for Sugar Beet Breeding in Straszkow, Poland, and the seeds of Amarantus cruentus L. were donated by the Szarłat _ Poland). Metro Industrial Centre (Łomza, The isolation procedure described by Olayinka, Adebowale, and Olu-Owolabi (2008) was used to obtain starch from sorghum grains; the starch from the amaranth seeds was isolated and purified according to the method developed by Walkowski, Fornal, Lewandowicz, and Sadowska (1997). For clarity, the following preparation codes have been proposed for the starches: maize starch (MS), sorghum starch (SS), Hylon VII (Hylon), waxy maize starch (WMS), and amaranth starch (AS). The chemical composition of these starches has been fully characterized in our previous work (Błaszczak, Misharina, Fessas, Signorelli, & Górecki, 2013); the results are not presented here. Pullulan standards were purchased from Showa Denko K.K. (Tokyo, Japan). Dimethyl sulfoxide (DMSO, HPLC grade) and

Following Jackson (1991), an aqueous DMSO solution (90:10, v/ v) was used to obtain the starch dispersion, as the most effective procedure for achieving maximum dispersibility of starch granules. Starch samples of 30 mg were dissolved in 5 ml of DMSO/H2O mixture at 100 °C with gentle stirring at 125 rpm in a Reacti-Therm heating system and stirring modules (Thermo Fisher Scientific, USA). The samples were further diluted with DMSO (the final concentration of the samples being 2.4 mg/mL) and filtered through 5 mm filters prior to analysis (Han & Lim, 2004a, 2004b). SEC equipment (Malvern, TX, USA) with triple detection (Viscotek 305 TDA, Triple Detector Array) was used for starch separation. Conventional dual cell refractometer, viscometer (VIS), and light scattering (lowangle light scattering, LALS, and right-angle light scattering, RALS) detectors were employed to act in concert. The aqueous SEC analysis was performed using three aqueous SEC columns (Shodex OHpak SB-800HQ series) with a guard SB-G type column (Showa Denko, Japan). The chromatography parameters were described by Kurzawska et al. (2014). The starch was analyzed in 0.1 M aqueous sodium nitrate with a 0.3 mL/min flow rate. The RI of the solvent was 1.3340, while that of the sample was 0.160 (In, Ibanez, & Shoemarker, 2007). Intrinsic viscosities (IV) were calculated based on the viscosity signals obtained from the viscometric detector. The calculations of molar mass average (Mn – number-average molar mass; Mw – weight-average molar mass; Mz – Z average molar mass), polydispersity index (Mw/Mn), Rh and the Mark–Houwink a value were performed using OmiSEC 4.7 software (Malvern, TX, USA). The amylopectin fraction was coded as AP and the amylose fraction as A.

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

2.4. Fourier Transform Infrared (FTIR) spectroscopy The FTIR measurements were performed in the solid state with a FT-IR Bruker IFS 113v/s spectrometer using Opus 5.5 software, under the following conditions: KBr pellet (1.5 mg/200 mg), resolution 2 cm 1. All spectra were standardized on the 2930 cm 1 stretching vibration of the CAH bond. 2.5. Statistical analyses One-way ANOVA and the Tukey test was used to test the significance of differences at alfa = 0.05. Hierarchical cluster analysis was performed to examine the samples based on SEC results using the Ward amalgamation rule with the Euclidean distance (d) measure. The tree plots were scaled to the standardized scale (dlink/ dmax*100). Principal component analysis (PCA) was used to reduce the dimensionality of data and to present the samples in a new coordinate system. Statistical analysis was performed using Statistica software (Version 12, Stat-Soft Inc., OK, US). 3. Results and discussion 3.1. Molecular structure of native and HHP-treated starches The molecular mass distribution of amylose starches (MS, SS, Hylon) and amylopectin starches (WMS, AS), defined based on the signals of the LS, VIS and RI detectors, are shown in Fig. 1A, C, E, G, I. The SEC profiles obtained for MS, SS, and Hylon starch (Fig. 1A, C, E) displayed two peaks at a retention volume (RV) of 20–22 mL and 24 mL, respectively. The first peak, recorded at an RV of 20–22 mL, was assigned to the high molecular fraction (amylopectin), whereas the second one (RV = 24 mL) was assigned to amylose. On the other hand, the SEC profiles of WMS and AS (Fig. 1G, I) demonstrated only a single peak at an RV of 20.5 mL. The character of the chromatograms noted for the analyzed starches was similar to that found by Lee, Han, & Lim, 2009 for waxy, normal, and high-amylose maize (70% amylose) starch. Average molar mass was calculated based on SEC data to describe the distribution of chain lengths in samples. The values of Mw strongly responded to the presence of low-molecularweight molecules in samples. The properties linked with large deformations, such as melt and solution viscosity, are determined by Mw which is influenced by high-molecular-weight polymers. Typical viscoelastic properties are determined by Mz which describes the part of the sample with very high-molecularweight polymers. The Mw/Mn ratio corresponds to the breadth of molecular mass distribution (Van Krevelen & Te Nijenhuis, 2009). In the group of the analyzed native starches, the lowest values of Mw (determined by the light scattering) were determined in both fractions (amylopectin and amylose) of Hylon starch (Table 1). The molecular mass of amylopectin of Hylon starch was almost half that of the molecular mass of the amylopectin of all other starches. The weight average molecular mass of the amylose of Hylon starch was as much as one level of magnitude lower than the Mw of the amylose of other normal starches. It has previously been reported that high-amylose starches with type B crystal structure are characterized by lower values of Mw for amylopectin than normal and waxy corn starches of type A crystal structure (Yoo & Jane, 2002). Lee et al. (2009) explain the differences in the molecular masses of amylopectin fractions by the differences in branch chain length. The mean value of chain length of type A crystal structure starches is usually lower than that of type B crystal structure starches. Type A crystal structure starches contain twice as many short-length chains (with degree of polymerization

53

DP < 12) than type B crystal structure starches, whereas in the case of long-length chains (with degree of polymerization DP > 25), this relationship is reversed. Moreover, Jane, Wong, and McPherson (1997), comparing the molecular structure of normal, waxy, and high-amylose corn and tapioca starches, suggested that type A crystal structure starches are characterized by a higher number of branches. The branches are scattered in location, being found in both amorphous and crystalline regions of the starch. In contrast, type B crystal structure starches are characterized by a number of branches localized mainly in the amorphous regions. Furthermore, the molecular mass distribution and Rg of both amylose and amylopectin fractions of normal, waxy, and high-amylose corn starches are pH dependent (in the pH range of 5–10) (Lee et al., 2009). The weight-average molecular mass of normal and waxy corn starch changes with pH from 1.0  108 to 2.5  108 Da, whereas in high-amylose starch it changes in a lower range of values from 0.7  108 to 1.5  108 Da. The average molecular mass of amylose of normal corn starch varies from 1.7  106 to 8.8  106 Da, whereas in high-amylose starch, it ranges from 0.6  106 to 1  106 Da. The precise determination of the molecular mass of amylopectin is difficult due to extremely high dimension of its molecules, which is higher than all known natural and synthetic polymers. The observed differences between measured values of Mw and the literature data could also be due to the lack of calibration standards for HPSEC or starch dissolution methods used (Yoo & Jane, 2002). According to Jackson (1991), the best dispersion of starch molecules can be achieved by applying an aqueous solution of DMSO (90:10, v/v) as a solvent. This procedure was therefore applied in our study. The values of hydrodynamic and structural parameters (Rg, IV, and number of branches) point to significant differences between the examined starches (Table 1), especially within the groups of starches containing amylose (Hylon, MS and SS) and the waxy varieties (WMS and AS). Hylon starch showed visibly lower intrinsic viscosity and radius of gyration than MS and SS. Similarly, the number of branches (290) calculated for Hylon starch in the whole range of molecular masses, was 5–18 times lower than for MS and SS respectively. Although the amylopectin peak was observed in the RI profile of AS, the signal resulting from the VIS detector was too weak to characterize the intrinsic viscosity of the samples; this could be the result of poor solubility of the starch (sample recovery = 5%). However, in case of WMS, the intrinsic viscosity of approximately 1 dL/ g was the highest of the values obtained for any of the high molecular fractions of the studied starches. Moreover the number of branches calculated for WMS was almost three times higher than for AS (Table 1). These observed differences in hydrodynamic parameters could be due to differences in the configuration (primary structure) of the investigated starches unrelated to the molecular mass distribution of starch. It is known that starches, especially waxy varieties, differ significantly in their primary structure in terms of branch chain lengths and the ratio of long (type A) to short (type B) amylopectin branch chains (Kong et al., 2009). Information on the conformation of starch molecules can also be derived from HPSEC analysis. Two parameters are especially useful here: Rg and Rh. Rg is generally used to describe the dimensions of a polymer chain. It refers to the distribution of the mass of a rotating polymer molecule around a fixed axis. Rh is defined as the radius of an equivalent hard sphere diffusing at the same rate as the molecule under observation. Because Rg depends on the mass distribution, while Rh reflects the shape of the molecules, the Rg/Rh ratio is a very useful parameter to distinguish between compact structures, such as hard spheres (0.778), statistic coils of linear molecules (1.78), and rods (2.0) (López-Franco et al., 2004).

54

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

Fig. 1. HPSEC chromatograms of native (A, C, E, G, I) and HHP-treated (B, D, F, H, J) amylose starches: A, B: maize starch (MS); C, D: sorghum starch (SS); E, F: Hylon VII; G, H: waxy maize starch (WMS); I, J: amaranth starch (AS). RI (Refractive Index), LALS (Low Angle Light Scattering), and Vis (Viscometer) signals were standardized to compare particular signals between native and HPP-treated samples.

Table 1 HPSEC results of native starches (AP – amylopectin, A – amylose retention volume range), ANOVA and the Tukey test was used to perform comparisons between the means.*

Mn** – (Da) Mw – (Da) Mz – (Da) Mw/Mn IV – (dl/g) Rh(w) – (nm) Rg(w) – (nm) Rg/Rh Branches/Molecule Branch Freq. Recovery (%) * **

MS (AP)

MS (A)

SS (AP)

SS (A)

Hylon (AP)

Hylon (A)

WMS (AP)

AS (AP)

5.30
107d ± 4.41
106 5.54
107f ± 1.47
106 5.81
107de ± 1.37
106 1.04a ± 0.06 1.79g ± 0.05 104e ± 2 121b ± 6 1.16a,b,c ± 0.03 1.6
103b ± 97 0.41b ± 0.03 40.7d ± 1.7

1.97
107b ± 2.57
106 2.05
107c ± 3.88
105 2.13
107b ± 1.72
106 1.05a ± 0.12 0.68b ± 0.05 66b ± 1 168c ± 10 2.53e ± 0.13

6.04
107d ± 7.39
106 6.18
107g ± 3.72
106 6.32
107e ± 9.43
105 1.02a ± 0.06 1.35f ± 0.16 111f ± 6 117b ± 10 1.06a ± 0.03 5.3
103c ± 332 1.28d ± 0.07 27.6c ± 3.1

5.47
106a ± 4.14
105 1.46
106b ± 1.39
106 2.39
107b ± 2.09
106 2.67c ± 0.05 0.83b,c±0.09 68b,c ± 3 159c ± 12 2.35d ± 0.09

2.25
107b,c ± 2.31
106 2.99
107d ± 1.37
106 4.01
107c ± 1.60
106 1.33b ± 0.08 0.93c,d ± 0.11 73c ± 3 84a ± 8 1.16a,b ± 0.06 2.9
102a ± 35 0.22a ± 0.03 72.9e ± 0.2

2.02
106a ± 3.70
105 2.20
106a ± 1.27
105 2.37
106a ± 3.89
105 1.09a ± 0.14 0.44a ± 0.04 26a ± 1 128b ± 3 4.85f ± 0.03

5.35
107d ± 5.53
106 5.36
107f ± 2.81
105 5.38
107d ± 4.83
106 1.00a ± 0.1 1.07e ± 0.11 102e ± 1 130b ± 0 1.28c ± 0 5.2
103c ± 575 0.58c ± 0.08 16.8b ± 1.9

2.92
107c ± 8.26
105 4.23
107e ± 3.19
106 5.48
107d ± 4.85
106 1.45b ± 0.07 No cal. 93d ± 5 115b ± 8 1.24b,c±0.01 1.8
103b ± 8 0.38b ± 0.05 5.3a ± 0.4

Different letters next to average values indicate significant difference between means in rows for particular parameter (p < 0.05). (Mn – number-average molar mass; Mw – weight-average molar mass; Mz – Z average molar mass; Mw/Mn – polydispersity index; IV – intrinsic viscosity, Rh – hydrodynamic radius; Rg – gyration radius; Rg/Rh – shape factor).

Mn** – (Da) Mw – (Da) Mz – (Da) Mw/Mn IV – (dl/g) Rh(w) – (nm) Rg(w) – (nm) Rg/Rh Branches/ Molecule Branch Freq. Recovery (%) * ** ***

MS (AP)

MS (A)

SS (AP)

SS (A)

Hylon (AP)

Hylon (A)

WMS (AP)

WMS (A)***

AS (AP)

5.20
107f ± 4.72
106 5.31
107g ± 1.23
106 5.43
107d ± 5.19
106 1.00a ± 0.07 1.01d ± 0.09 98f ± 2 111bc ± 5 1.14a ± 0.04 3.2
103b ± 1.3
102

1.91
107bc ± 1.48
106 2.04
107c ± 1.43
106 2.14
107b ± 1.93
106 0.78a ± 0.01 No cal. 71c ± 3 159f ± 11 2.23c ± 0.14

5.99
107g ± 2.70
106 6.13
107h ± 1.76
106 6.29
107e ± 8.10
106 1.03a ± 0.02 1.21e ± 0.1 107g ± 4 122cd ± 12 1.14a ± 0.1 1.0
104c ± 9.1
102

6.41
106a ± 4.36
105 1.51
107b ± 1.20
106 2.36
107b ± 1.88
106 2.35d ± 0.03 0.78b ± 0.09 70c ± 2 170f ± 12 2.43c ± 0.13

2.78
107d ± 7.19
105 3.36
107d ± 2.16
106 4.06
107c ± 3.27
106 1.21b ± 0.05 0.94bc ± 0.08 79d ± 4 89a ± 7 1.13a ± 0.07 9.8
102a ± 1.1
101

2.81
106a ± 2.17
105 3.00
106a ± 8.69
104 3.18
106a ± 2.18
105 1.07a ± 0.05 0.44a ± 0.04 30a ± 3 137de ± 2 4.64e ± 0.18

4.63
107e ± 2.70
106 4.82
107f ± 1.84
106 5.00
107d ± 2.85
106 1.04a ± 0.02 0.93bc ± 0.1 91e ± 5 108be ± 7 1.19a ± 0.02 1.1
104c ± 2.9
102

1.75
107b ± 6.33
105 1.78
107bc ± 1.42
106 1.81
107b ± 1.48
106 1.02a ± 0.04 0.55a ± 0.05 57b ± 3 154ef ± 8 2.72d ± 0.02

2.32
107c ± 1.05
106 3.78
107e ± 1.03
106 4.91
107d ± 4.19
106 1.63c ± 0.03 0.58a ± 0.05 72c ± 4 100ab ± 5 1.38b ± 0.01 2.1
104d ± 7.2
102

0.91b ± 0.07 43.9c ± 1.5

1.83c ± 0.19 27.1b ± 2.2

0.28a ± 0.04 70.8d ± 5.1

2.12c ± 0.23 42.1c ± 1

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

Table 2 HPSEC results of pressure-treated (HHP) starches (AP – amylopectin, A – amylose retention volume range), ANOVA and the Tukey test was used to perform comparisons between the means*

4.96d ± 0.34 9.2a ± 0.78

Different letters next to average values indicate significant difference between means in rows for particular parameter (p < 0.05). (Mn – number-average molar mass; Mw – weight-average molar mass; Mz – Z average molar mass; Mw/Mn – polydispersity index; IV – intrinsic viscosity, Rh – hydrodynamic radius; Rg – gyration radius; Rg/Rh – shape factor). Low molecular fraction of WMS after HHP was indexed with A to highlight low retention volume, typical for amylose.

55

56

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

Significant differences in the Rg/Rh ratio of amylopectin between the starches containing amylose and the waxy varieties were found in the starches tested in this study (Table 1). In the case of MS, SS, and Hylon starch, the Rg/Rh ratio of amylopectin was in the range 1.06–1.16, indicating the highly branched, dense structure of that fraction. In the case of the waxy starches, the Rg/Rh ratio of amylopectin was higher (1.24–1.28), indicating a looser structure. As expected, the Rg/ Rh ratio found for the amylose of MS, SS, and Hylon was above 2.0, confirming the linear structure of that fraction. It is worth mentioning that the amylose of Hylon starch had a higher Rg/ Rh ratio than that found for the amylose fraction of MS and SS. The effect of high pressure treatment (650 MPa/9 min) on the molecular mass distribution and hydrodynamic parameters of amylose and amylopectin starches was also considered in the present study (Fig. 1B, D, F, H, J and Table 2). As shown by the SEC profiles, MS, SS, Hylon, and AS treated with HHP showed no changes in their molecular mass distributions (Fig. 1B, D, F, J). On the other hand, significant changes in molecular mass distribution were found for the pressurized WMS. The SEC profile of pressurized WMS (Fig. 1H) has two peaks, one occurring by the same RV as in the case of the amylopectin of native WMS (Fig. 1G), and the second pronounced peak at RV = 24.4 mL—that is, in the elution range typical of amylose. These results show that pressurization causes a change in the molecular mass of part of the starch material only. Moreover, the appearance of the new peak related to the fraction of the molecular mass similar to amylose points to the degradation of amylopectin molecules. Similarly, the HHP treatment distinctly affected the hydrodynamic parameters of all the tested starches (Table 2). The changes in the profiles of IV and of Rg versus Mw as an effect of HHP treatment were observed for all the starches. The most spectacular changes were found for WMS (a sample profile is given in Fig. 1S). These changes were accompanied by an increase in the branching number. In case of MS, WMS, and SS, a twofold increase in the number of branches was found, whereas in case on AS, there was an eleven-fold increase. The pressurized Hylon starch also demonstrated an increase (3.5-fold) in the number of branches. Considering this increase in the number of branches in pressurized starches, three hypotheses were proposed. First, it was assumed that applying high pressure increases the solubility of macromolecules with a large number of branches. This hypothesis is highly probable, if only for WMS and AS, since distinctly higher recoveries were observed for pressurized starches than for native starches. However, this does not explain the increase in the number of branches in amylose-containing starches. The second hypothesis suggests that the high pressure significantly affects the compactness of the molecular structure of the samples. Since intrinsic viscosity is inversely proportional to molecular density, the pressure-induced compression of molecules could result in lower IV values. The branching number was calculated using the Zimm–Stockmayer equation for a polydisperse randomly branched polymer. The ratio of the intrinsic viscosities for the branched over the linear polymer can be calculated using this equation. The lower values of intrinsic viscosities for branched polymers result in a higher number of branches. However, in case of Hylon starch, the intrinsic viscosity did not change, so this hypothesis does not explain the increase in the number of branches for that starch. The third hypothesis is that the amylopectin fraction underwent a partial depolymerization upon HHP treatment. This could explain the lower values of intrinsic viscosity of HHP starches. HPSEC data also suggested that HHP treatment resulted in a split of a-1,4glycosidic bonds and in increase in a-1,6-glycosidic bonds as a consequence. This phenomenon is particularly evident in case of WMS. Two ranges can be seen on the Mark–Houwink plot (Fig. 1S).

For high molecular masses, the relationship was typical of native starches—that is, with an increase in molecular mass, the intrinsic viscosity gradually increased. The intrinsic viscosity was constant for low molecular masses. This means that, with the increase in molecular mass, the density of macromolecules (the number of branches) increased but their hydrodynamic volume did not. This could be explained by the assumption that the degradation of amylopectin molecules occurred mainly through the splitting of a-1,4-glycosidic bonds. Otherwise (with splitting of the a-1,6-glycosidic bonds), linear chains would form and the intrinsic viscosity would increase. This explanation is highly probable for data in the RV range from 18 to 22 mL—the most abundant fractions (Fig. 1H)—because the peak with poor viscosity signal was also detected in the RV range (22–26 mL) typical of small molecules (DP about 25). This means that SEC data can explain the debranching mechanism of HPP process for starch fractions where the signal from the viscometric detector is intense enough to perform reliable calculations (for the most abundant fraction). The foregoing discussion assumes that both the a-1,4- and a-1,6glycosidic bonds were probably split during HHP treatment.

3.2. FTIR analysis of native and HHP-treated starches To determine which mechanism of degradation of amylopectin molecules occurred upon HHP treatment, FTIR spectroscopy analyses were carried out (Table 3). Moreover, FTIR analysis was used to determine which of the bonds was split more frequently. It was found that the HHP treatment of the starches did not cause the creation or vanishing of any new band (spectra not shown). The only change observed for Hylon, WMS, and AS was an increase in the intensity of the band at 1647 cm 1 assigned to the a-1,4 glycoside bond (Table 3). These data are similar to those obtained by Le Thanh-Blicharz, Błaszczak, Szwengiel, Paukszta, and Lewandowicz (2016). These observations support the hypothesis that the a-1,6-glycosidic bonds are more frequently split. However, for MS and SS, a decrease in the intensity of that band was observed, which suggests that, in these cases, the a-1,4-glycosidic bonds are more frequently split. Such a phenomenon was recently reported in the literature for starch modified by high pressure homogenization of its paste (Le Thanh-Blicharz, Szwengiel, & Lewandowicz, 2016). The conclusion postulated above is complementary with the results presented by Guo et al. (2015), who examined Nelumbo nucifera starch (type C crystal structure) and indicated that Mw and Mn decrease as a result of high pressure treatment (100– 600 MPa for 30 min). These changes have been attributed to starch degradation. Moreover, the Mw/Mn ratio decreased from 1.28 for native starch to 1.11 for starch pressurized at 600 MPa. A minor reduction in polydispersity was also observed in our study for pressurized MS, SS, and Hylon starch. However, for amylopectin starches (WMS, AS), the opposite relationship was observed (Table 2).

Table 3 Changes in the intensity of 1647 cm

1

caused by HHP treatment of starches.

Starch

Intensity of 1647 cm 1 bands in the spectra of pressurized starches relative to those of native ones

MS SS Hylon WMS AS

0.95 0.89 1.13 1.14 1.06

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

Fig. 2. Cluster analysis results, variation in parameters (Mn, Mw, Mz, polydispersity index, IV, Rg, Rh, Rg/Rh, number of branches, sample recovery) of native (N) and pressure-treated (HHP) starches; A: parameters were calculated in all ranges of the retention volume of the sample, with two repeats; B: parameters were calculated for the amylopectin (AM) and amylose retention volume range (A) fractions separately (the low molecular fraction of WMS after HHP was indexed with A to highlight the low retention volume typical of amylose). The scale tree was normalized to dlink/dmax*100 (d: distance; l: linkage; max: maximum of linkage Euclidean distance). Amalgamation rule: Ward’s method; distance metric: Euclidean distance.

3.3. Hierarchical cluster analysis Cluster analysis was performed to organize the data into meaningful structures. This is an exploratory data analysis method that discovers structures in a data set without providing explanatory

57

relations between variables. The profile similarity of native (N) and pressure-treated (HHP) starches is shown in Fig. 2A. Generally, the samples were grouped by their botanical origin. However, after HHP, the WMS sample was most similar to the MS and SS samples. The cluster analysis was repeated and the samples were recognized according to parameter similarity between the amylopectin (AM) and amylose (A) fractions (Fig. 2B). It seems that it is possible to distinguish the two groups by cutting the tree diagram at 50%. The first group contains cases indicated as amylose, while the second group contains cases indicated as amylopectin. The position of particular cases was generally determined again by the botanical origin of starches. However, the recorded fraction for WMS in the retention volume typical of amylose after HHP was more similar to the amylose from MS than that from SS or Hylon. The cases form two subgroups when the parameters of amylopectin are recognized. The first group contains AS and Hylon, while the second group contains WMS, MS, and SS. PCA analysis was performed to describe the direction and strength of the interaction parameters describing the samples analyzed using HPSEC technique (Fig. 3). The positions of the vectors in Fig. 3A point to a positive relationship between Rh and the average molecular mass Mz. Both parameters correspond to molecule size, where Mz describes the left side of the peak and the distribution of large chain lengths in the sample (Van Krevelen & Te Nijenhuis, 2009), and Rh is the radius of an equivalent hard sphere (Striegel, 2016). High values of these parameters indicate that the very high-molecular-weight part of the samples has a less compact structure. This is particularly evident in samples where amylopectin aggregation was observed. The aggregation process was also observed by Blennow, Bay-Smidt, and Bauer (2001). Branches are correlated with Mn; branches were higher for samples treated with HPP, which suggests that sample depolymerization occurred during this process. The Rg/Rh ratio was positively correlated with the polydispersity index Mw/Mn, so depolymerization, as well as the dissolved amylopectin/amylose ratio, affected significantly the hydrodynamic volume and the shape of the macromolecules. Moreover, sample recovery positively correlated with abovementioned indexes, confirming the hypothesis that the increase in solubility was accompanied by depolymerization caused by HHP. The relative position of the Rg and IV vectors points to a lack of correlation of these parameters with Mw. They are negatively correlated with number of branches, which confirms that the analyzed samples differ in both configuration and conformation of

Fig. 3. PCA of loadings plot (A) and score plot (B) for native (N) and pressure-treated (HHP) starches; parameters were calculated in all ranges of the retention volume of the sample, with two repeats; supplementary variables are indicated by a superscript (*); the principal components (PC1 and PC2) were computed using only the active variables.

58

A. Szwengiel et al. / Food Chemistry 240 (2018) 51–58

macromolecules. The distribution of the samples in Fig. 3B is determined essentially only by the botanical origin of starch. The only exception was the HHP-treated WMS starch, which was clearly relocated towards normal corn starch. 4. Conclusion In summary, SEC analysis of the data demonstrates that the range of structural changes caused by pressurization strongly depends on the botanical origin of the starch. Surprisingly, this effect did not correlate with amylose content. The smallest changes were undergone by MS and SS—that is, those with typical amylose contents of about 20%. There were almost no changes to the average molecular masses caused by HHP treatment of either of these starches. At the same time, the average molecular mass of Hylon increased. In contrast, in the case of the amylopectin starches (WMS and AS), this parameter decreased. The structural information elucidated from HPSEC demonstrates that HHP treatment caused an increase in branches and in branch frequency for all the starches. However, this does not consider low molecular weight fractions with low abundance that result from the splitting of a-1,6 glycoside bonds. The combination of HPSEC and FTIR data allows the splitting frequency of a-1,4 and a-1,6 glycoside bonds to be assessed in the whole sample. Nevertheless, these structural changes were less significant in case of MS and SS than for the other starches. From the IR data, it can be concluded that the pressurization process led to an increase in the number of a-1,4 glycoside bonds in Hylon, WMS, and AS. For MS and SS starches, the number of a-1,4 glycoside bonds decreased. Acknowledgements The study was financed by a grant from the Polish Ministry of Science and Higher Education (Grant No. N N312 101938). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foodchem.2017. 07.082. References Błaszczak, W., Fornal, J., Valverde, S., & Garrido, L. (2005). Pressure-induced changes in the structure of corn starches with different amylose content. Carbohydrate Polymers, 61, 132–140. Błaszczak, W., Misharina, T. A., Fessas, D., Signorelli, M., & Górecki, A. R. (2013). Retention of aroma compounds by corn, sorghum and amaranth starches. Food Research International, 54, 338–344. Blennow, A., Bay-Smidt, A. M., & Bauer, R. (2001). Amylopectin aggregation as a function of starch phosphate content studied by size exclusion chromatography and on-line refractive index and light scattering. International Journal of Biological Macromolecules, 28, 409–420. Guo, Z., Zeng, S., Lu, X., Zhou, M., Zheng, M., & Zheng, B. (2015). Structural and physicochemical properties of lotus seed starch treated with ultra-high pressure. Food Chemistry, 186, 223–230. Han, J. A., & Lim, S. T. (2004a). Structural changes of corn starches by heating and stirring in DMSO measured by SEC-MALLS-RI system. Carbohydrate Polymers, 55, 265–272.

Han, J. A., & Lim, S. T. (2004b). Structural changes in corn starches during alkaline dissolution by vortexing. Carbohydrate Polymers, 55, 193–199. Hu, X.-P., Zhang, B., Jin, Z.-Y., Xu, X.-M., & Chen, H.-Q. (2017). Effect of high hydrostatic pressure and retrogradation treatments on structural and physicochemical properties of waxy wheat starch. Food Chemistry. http://dx. doi.org/10.1016/j.foodchem.2017.04.040. In, M. P., Ibanez, A. M., & Shoemarker, C. F. (2007). Rice starch molecular size and its relationship with amylose content. Starch, 59, 69–77. Jackson, D. S. (1991). Solubility behavior of granular corn starches in methyl sulfoxide (DMSO) as measured by high performance size exclusion chromatography. Starch, 43, 422–427. Jane, J., Wong, K. S., & McPherson, A. E. (1997). Branch-structure difference on starches of A- and B-type X-ray patterns revealed by their Naegeli dextrins. Carbohydrate Research, 300, 219–227. Kong, X., Corke, H., & Bertoft, E. (2009). Fine structure characterization of amylopectins from grain amaranth starch. Carbohydrate Research, 344, 1701–1708. Kurzawska, A., Górecka, D., Błaszczak, W., Szwengiel, A., Paukszta, D., & Lewandowicz, G. (2014). The molecular and supramolecular structure of common cattail (Typha latifolia) starch. Starch, 66, 849–856. Le Thanh-Blicharz, J., Błaszczak, W., Szwengiel, A., Paukszta, D., & Lewandowicz, G. (2016). Molecular and supermolecular structure of commercial pyrodextrins. Journal of Food Sciences, 81, C2135–C2142. Le Thanh-Blicharz, J., Szwengiel, A., & Lewandowicz, G. (2016). Molecular structure of starch modified by high pressure homogenisation of pastes. Proceedings of the 12thInternational Conference on Polysaccharides-Glycoscience, p. 189–192. Prague 19–21.10.2016. Lee, J. H., Han, J.-A., & Lim, S.-T. (2009). Effect of pH on aqueous structure of maize starches analyzed by HPSEC-MALLA-RI system. Food Hydrocolloids, 23, 1935–1939. Li, H., Prakash, S., Nicholson, T. M., Fitzgerald, M. A., & Gilbert, R. G. (2016). The importance of amylose and amylopectin fine structure for textural properties of cooked rice grains. Food Chemistry, 196, 702–711. López-Franco, Y. L., Valdez, M. A., Hernández, J., Calderón de la Barca, A. M., Rinaudo, M., & Goycoolea, F. M. (2004). Macromolecular dimensions and mechanical properties of monolayer films of sonorean mesquite gum. Macromolecular Bioscience, 4, 865–874. Oh, H. E., Pinder, D. N., Hemar, Y., Anema, S. G., & Wong, M. (2008). Effect of highpressure treatment on various starch-in-water suspensions. Food Hydrocolloids, 22, 150–155. Olayinka, O. O., Adebowale, K. O., & Olu-Owolabi, B. I. (2008). Effect of heatmoisture treatment on physicochemical properties of white sorghum starch. Food Hydrocolloids, 22, 225–230. Pei-Ling, L., Qing, Z., Qun, S., Xiao-Song, H., & Ji-Hong, W. (2012). Effect of High hydrostatic pressure on modified noncrystalline granular starch of starches with different granular type and amylose content. LWT-Food Sciences and Technology, 47, 450–458. Raghunathan, R., Hoover, R., Waduge, R., Liu, Q., & Warkentin, T. D. (2017). Impact of molecular structure on the physicochemical properties of starches isolated from different field pea (Pisum sativum L.) cultivars grown in Saskatchewan, Canada. Food Chemistry, 221, 1514–1521. Singh, N., Inouchi, N., & Nishinari, K. (2006). Structural, thermal and viscoelastic characteristics of starches separated from normal, sugary and waxy maize. Food Hydrocolloids, 20, 923–935. Striegel, A. M. (2016). Viscometric detection in size-exclusion chromatography: Principles and select applications. Chromatographia, 79, 945–960. Van Hung, P., Maeda, T., & Morita, N. (2006). Waxy and high-amylose wheat starches and flours-characteristics, functionality and application. Trends in Food Science & Technology, 17, 448–456. Van Krevelen, D. W., & Te Nijenhuis, K. (2009). Properties of polymers: Their correlation with chemical structure; their numerical estimation and prediction from additive group contributions (4th ed.). Oxford: Elsevier (Chapter 2). Walkowski, A., Fornal, J., Lewandowicz, G., & Sadowska, J. (1997). Structure, physicochemical properties and potential uses of amaranth starch. Polish Journal of Food and Nutrition Sciences, 6, 11–22. Yang, Z., Swedlund, P., Hemar, Y., Mo, G., Wei, Y., Li, Z., & Wu, Z. (2016). Effect of high hydrostatic pressure on the supramolecular structure of corn starch with different amylose contents. International Journal of Biological Macromolecules, 85, 604–614. Yoo, S.-H., & Jane, J.-L. (2002). Molecular weights and gyration radii of amylopectins determined by high–performance size–exclusion chromatography equipped with multi–angle laser–light scattering and refractive index detectors. Carbohydrate Polymers, 49, 307–314.