Resistant starch formation and changes in physicochemical properties of waxy and non-waxy rice starches by autoclaving-cooling treatment

ABSTRACT

Resistant starch (RS) has drawn great attention from both starch research and industrial communities due to its potential health benefits. Physical starch modification approach, such as autoclaving and cooling cycle treatment, offers high potential method of enhancing resistant starch fraction due to its simplicity and cost-effectiveness. Here, the non-waxy KDML105 and waxy RD6 rice starches were physically modified by one, three, and five autoclaving and cooling cycles. The comparative impacts of the modification on the RS content and starch physicochemical properties were examined. Their native starch granules were transformed into a larger continuous matrix with suppressed thermal and pasting properties. However, modified KDML105 starch gels with higher amylose content and the presence of amylose lipid complexes were able to retrograde and maintain viscosity better than the modified RD6 starch. The RS content was also increased with the addition of treatment cycles in KDML105 starch. As a result, the RS content of native KDML105 starch was enhanced from 0.13% to 2.59% after five cycles of the treatment. The dissimilarities observed between the two modified starches could be due to the different degrees of increased proportions of ordered (R1047/1022) and double helical (R995/1022) structures, and the disruption of A- and V-type crystals. Therefore, the RS content was attributed to an interplay of alterations in ordered structure, amylose, and amylose-lipid complex during autoclaving-cooling cycles.

KEYWORDS:

• Rice
• starch
• resistant starch
• autoclaving-cooling cycles
• physicochemical properties

Introduction

Rice starch is an important agricultural commodity, which provides a crucial ingredient for various starch-rich food and other biomaterial industries (e.g., noodles, confectionery, bakery, beverages, pharmaceutical and supplement, cosmetics, and animal feeds).^[1] In Thailand, KDML105 and RD6 are elite rice cultivars that are vastly produced for both domestic consumption and industrial applications.^[2] The non-waxy KDML105 and waxy RD6 cultivars are grown on a large scale, spanning 4.47 and 2.58 million hectares, respectively. They yield long grains with soft texture and pleasant aroma upon cooking.^[2] However, their native starches typically possess low degree of crystallinity and amylose-to-amylopectin ratio, and are hence susceptible to gelatinization,^[2] shear thinning,^[3] slow retrogradation,^[2] and digestive enzyme hydrolysis.^[4] Therefore, their native starches may lack the properties that are required for certain kinds of industrial processing.^[5,6] Particularly, they release reducing sugars quickly upon digestion, which is undesirable in consumers with diabetics and obesity conditions.^[3,4] To date, rice starch has been modified using multiple methods to acquire functionalities that are suitable for different manufacturing applications. The modifications not only broaden its range of applications, but also satisfy customer desires for health benefits.^[5,6]

Four different types of starch modifications are generally applied including genetic, enzymatic, chemical, and physical methods.^[5] Physical modification approaches, such as autoclaving and cooling cycle treatment, have drawn great attention due to its simplicity, cost-effectiveness, and safety.^[7] Under the autoclaving-cooling cycle treatment, the native starch is subjected to cycles of high temperature and pressure (i.e., autoclaving) followed by cooling temperature.^[7] This could transform the small granule of native starch into a large, densely compact microstructure of insoluble three-dimensional gels containing highly ordered, double-helical structures that are less susceptible to digestive enzymes.^[8,9] The resistant starch fraction generated by this autoclaving-cooling cycle was classified as resistant starch type 3 (RS3), which has potential health benefits.^[8] RS3 displayed prebiotic and hypoglycemic properties.^[8] Since it can bind to bile acids, the presence of RS3 in foods helps minimizing the reabsorption of bile acids in the large bowel, thereby lowering blood cholesterol levels and reducing the risk of cardiovascular diseases.^[10,11]

In non-waxy rice starches with varying amylose content, the dual autoclaving-retrogradation treatment can enhance the proportion of starch ordered crystalline and double-helical structures by destroying the amorphous regions.^[3] The formation of the more tightly packed ordered helical structures in the retrograded starch pastes also improved their thermal stability, pasting properties, and enzymatic resistance in the small intestine.^[3,11] There have been several studies that recently demonstrated the influences of autoclaving-cooling cycle treatment on non-waxy rice,^[3] high-amylose oat,^[11] corn,^[9] potato,^[6] pea,^[7] and tacca^[12] starches. Nonetheless, to the best of our knowledge, the impact of multiple autoclaving-cooling cycles on the RS formation, molecular composition and structure, and physicochemical properties of waxy and non-waxy rice starches has never been thoroughly compared. Therefore, the objective of this study is to (i) evaluate the effects of autoclaving-cooling cycle treatment on RS content and morphological, structural, thermal, and pasting properties of non-waxy KDML105 and waxy RD6 rice starches and (ii) to compare the altered properties of two rice starches that were modified by varied cycle numbers of autoclaving-cooling treatment. Our overarching hypotheses were that (i) the difference in amylose content and molecular structure of the waxy and non-waxy rice starches would distinctively contribute to the modified starch properties, and (ii) the increasing treatment cycles would further alter the modified starch properties. Furthermore, the obtained results may explain structural alteration of the starches with different amylose content as a result of autoclaving-cooling cycle treatment, and how it could lead to RS3 formation.

Materials and methods

Materials and reagents

Seeds of Thai non-waxy and waxy rice cultivars, namely KDML105 and RD6, were obtained from Department of Agronomy, Faculty of Agriculture, Khon Kaen University (Khon Kaen, Thailand). The seeds were harvested during the harvesting season in November 2021. The resistant starch assay kit (K-RSTAR) and amylose/amylopectin assay kit (K-AMYL) were purchased from Megazyme (Wicklow, Ireland) and used to determine the RS content and amylose content according to the manufacturer’s recommendations, respectively. All other reagents were of analytical grade.

Isolation and purification of rice starch

The milled rice samples were thoroughly homogenized to a fine powder using a manual coffee grinder (Chestnut C, Timemore, Shanghai, China), and sifted through a 35 mesh (0.50 mm) sieve (Brass Frame Stainless Mesh, Humboldt, Raleigh, NC, USA). The starch was then isolated from the rice flour by alkaline extraction as previously described^[13] with slight modifications. Briefly, 25 mL of 0.2% (w/v) sodium hydroxide were mixed with 10 g of rice flour. The suspension was stirred at 25°C for 3 hour and filtered through four layers of cheesecloth. After centrifugation at 4490 × g for 20 minute, the supernatant was removed. The starch pellet was washed twice with double-distilled water and subsequently resuspended in 50 mL double-distilled water. The suspension was adjusted to neutral pH with 1 M hydrochloric acid and centrifuged at 4490 × g for 20 minute. The lower starch layer was then washed twice with double-distilled water. The starch layer was suspended in 10 mL toluene and then washed twice with double-distilled water to completely remove toluene residues. Lastly, the white starch pellet was dried in a hot air oven (RF115, Binder, Tuttlingen, Germany) at 40°C for 48 hour, ground with mortar and pestle, and then passed through a 35 mesh (0.50 mm) sieve (Brass Frame Stainless Mesh, Humboldt, Raleigh, NC, USA). The isolated starch was stored at room temperature with silica gel desiccant until analysis.

Preparation of autoclaved-cooled starch

The autoclaved-cooled starch samples were prepared following the previously described method^[7] with slight modifications. A suspension containing 25 g of native rice starch and 75 mL of double-distilled water (starch: water, 1:3) was prepared in a 250 mL Duran® glass bottle. The suspension was pre-heated on a hot plate stirrer at 70°C for 15 minute to completely disperse starch granules. For a single cycle of autoclaving and cooling treatment, the starch paste was autoclaved at 121°C for 20 minute, cooled to 25°C for 20 minute, and stored at 4°C for 24 hour. This autoclaving-cooling cycle was repeated for one, three, and five times before the modified starch was finally dried at 50°C for 24 hour, crushed to powder, and passed through a 0.50 mm sieve (Brass Frame Stainless Mesh, Humboldt, Raleigh, NC, USA). The modified samples were designated AC1 (one cycle), AC3 (three cycles), and AC5 (five cycles), respectively.

Determination of starch chemical composition

Moisture content

The moisture content was measured according to AACC Air Oven Method 44–19^[14] in a hot air oven (RF115, Binder, Tuttlingen, Germany) at 135°C for 2 hour.

Amylose content

Approximately 20 mg of the starch samples were used to analyze for amylose content by Amylose/Amylopectin Assay Kit (K-AMYL, Megazyme, Wicklow, Ireland) according to the manufacturer’s recommendations.

Scanning electron microscopy (SEM) analysis

The KDML105 and RD6 native, AC1, AC3, and AC5 starch particle morphology was analyzed according to our previous method.^[15] The starch powders were dispersed on the aluminum stub covered with adhesive carbon discs. The samples were then coated with gold particles using a gold sputter coater (108A sputter coater, Cressington Scientific Instruments, Hertfordshire, UK). The electron micrographs were taken using a scanning electron microscope (LEO 1450VP, Carl Zeiss, Baden-Württemberg, Germany) at an accelerating voltage of 22 kV with the magnification of 2000 ×.

Particle size distribution analysis

The starch particle size distribution analysis followed the procedure of Sangwongchai et al.^[13] with slight modifications. The native, AC1, AC3, and AC5 starch samples (30 mg) were dispersed in 1 mL of double-distilled water. The starch suspension was used to determine the percentage volume of particle size distribution and to measure the granule median diameter (d(0.5)), which is the granule size at which 50% of all granules by volume are smaller, using a laser diffraction particle size analyzer (Mastersizer 2000, Malvern Panalytical, Worcestershire, UK).

X-ray diffraction pattern analysis

The starch samples were packed into a sample holder at room temperature. The crystal structure of each starch sample was scanned on an X-ray diffractometer (SmartLab 9 kW, Rigaku, Tokyo, Japan) with the following operational settings: voltage, 40 kV; current, 30 mA; scanning range, 5–40° diffraction angles (2θ); scanning rate, 2.4°/minute.^[16]

Fourier-transform infrared spectral analysis

To determine changes in the inter- and intra-molecular structures of autoclaved-cooled starch samples, 1 mg of starch samples was homogeneously dispersed in 10 mg of potassium bromide (KBr) and then pressed to form thin slice using a KBr Quick Press Accessory (PerkinElmer, Waltham, MA, USA). The KBr-starch thin slice was scanned with a spectrum range of 4000–400 cm⁻¹ at 25°C on the FTIR spectrometer (Tensor 27, Bruker, Leipzig, Germany).^[7]

Determination of resistant starch content

Approximately 100 mg of starch samples were used to determine the RS. content using the AACC Method 32–40.01^[14] following the recommendations of the resistant starch assay kit (K-RSTAR, Megazyme, Wicklow, Ireland).

Thermal properties

The thermal stability of starch and retrograded starch samples was characterized using a differential scanning calorimeter (DSC) (STARe System DSC 1, METTLER-TOLEDO, Zürich, Switzerland) as previously described.^[13,17] Briefly, starch sample (3 mg) was thoroughly mixed with deionized water (9 μL) in an aluminum pan, and the sample pans containing starch slurry with a starch-to-water ratio of 1:3 were hermetically sealed and equilibrated at room temperature overnight. The sample pan and an empty reference pan were heated from 30°C to 95°C at a rate of 10 ºC/minute in the DSC chamber. The resulting DSC thermograms were used to extract the transition temperatures (i.e., the onset (To(g)), peak (Tp(g)), and conclusion (Tc(g)) temperatures) of gelatinization and calculate the temperature range (ΔT(g)) and enthalpy change (ΔH(g)) of gelatinization. The sample pans containing gelatinized starches were then stored at 4°C for 14 days. The pans containing retrograded starch were reheated using the same temperature program as the gelatinization analysis. The transition temperatures (i.e., the onset (To(r)), peak (Tp(r)), and conclusion (Tc(r)) temperatures) of the retrograded starch melting and the temperature range (ΔT(r)) and enthalpy change (ΔH(r)) of the retrograded starch melting were acquired.

Pasting properties

The viscosity change of 8% (w/v) starch paste was measured using a rheometer (Physica MCR 51, Anton Paar, Graz, Austria) simulating a RVA temperature profile.^[18] The temperature program used was as follows: hold at 50°C for 1 minute, heat to 95°C at a rate of 10ºC/min, hold at 95°C for 3 minute, cool to 50°C at a rate of 10ºC/minute, and final hold at 50°C for 1 minute. A constant rotating speed of 160 rpm was applied. The pasting parameters consisting of pasting temperature (P_Temp), peak time (P_Time), peak viscosity (PV), trough viscosity (TV), breakdown (BD), final viscosity (FV), and setback (SB.) were obtained from the viscoamylograph. The viscosity was expressed in millipascal-second (mPa.s).

Statistical analysis

All experimental data were displayed as means ± standard error of the mean (SEM) of four replicates. Data were analyzed by a one-way analysis of variance (ANOVA) followed by a Tukey’s post hoc test or by an independent-samples student’s t-test. These statistical analyses were conducted for the comparison of the significant difference (p < .05) between the mean values using the SPSS® statistics v.23 software (IBM, Armonk, NY, USA).^[15]

Results and discussion

Effects of autoclaving-cooling cycle treatment on amylose content of non-waxy and waxy rice starches

Starch from non-waxy KDML105 rice cultivar was previously shown to possess a higher amylose content than starch from waxy RD6 rice cultivar.^[2,13] Consistently, amylose content was in the range of 6.95 ± 0.43–9.27 ± 0.63% and 1.21 ± 0.34–2.71 ± 0.36% for native and modified KDML105 and RD6 starches, respectively, in this study. Notably, both native and modified (i.e., AC1, AC3, and AC5) KDML105 starches displayed greater amylose content than the native and all modified RD6 starches (Table 1). Interestingly, the amylose content of KDML105 AC1 (6.95 ± 0.43%) showed a significant decline from that of native KDML105 starch (9.27 ± 0.63%) (Table 1). Similar reduction in amylose content was previously seen in oat starches that underwent dual autoclaving-cooling cycle treatment.^[11] However, the amylose content of KDML105 AC3 and AC5 (8.18 ± 0.58% and 9.03 ± 0.35%, respectively) did not vary from the amylose content of native KDML105 starch (Table 1). Therefore, the amylose reduction found in KDML105 AC1 starch could be attributed to either the result of AC treatment or the underestimation of amylose content.^[19] The underestimation of amylose can be attributed to the presence of the amylose-lipid complex,^[19] the method used to isolate starch, or the method used to estimate it.^[11] Interestingly, no significant difference between the amylose content of native and modified RD6 starches was observed in this study (Table 1), which is consistent with previous studies reporting that the dual AC and three and five AC cycles did not affect the amylose content of rice and cowpea starches, respectively.^[3,7] This could be because autoclaving, a physical process, causes slight chemical changes to starch.^[11]

Table 1. Starch composition, granule size distribution, median particle size, relative crystallinity, R1047/1022 and R995/1022 values, and resistant starch content of native and modified non-waxy KDML105 and waxy rice RD6 starches.

Starch samples	Treatments	Amylose content (%)	Moisture content (%)	Starch granule size distribution (%volume)			d(0.5) ($\mathit{\mu }$m)	R1047/1022	R995/1022	Resistant starch content (%)
				C-type < 5 ($\mathit{\mu }$m)	B-type 5–15 ($\mathit{\mu }$m)	A-type >15 ($\mathit{\mu }$m)
Non-waxy KDML105 starch	Native	9.27 ± 0.63^a	9.09 ± 0.17^a	40.45 ± 0.72^a	50.10 ± 1.09^b	9.45 ± 1.38^b	5.21 ± 0.08^b	0.85 ± 0.01^b	0.73 ± 0.02^c	0.13 ± 0.01^de
	AC1	6.95 ± 0.43^b	7.54 ± 0.23^a	0.62 ± 0.02^c	3.23 ± 0.10^c	95.90 ± 0.15^a	248.90 ± 9.04^a	0.93 ± 0.00^a	0.88 ± 0.01^a	0.65 ± 0.02^c
	AC3	8.18 ± 0.58^ab	7.22 ± 0.34^a	0.62 ± 0.04^c	3.01 ± 0.09^c	96.12 ± 0.18^a	270.51 ± 7.27^a	0.94 ± 0.00^a	0.90 ± 0.01^a	1.27 ± 0.06^b
	AC5	9.03 ± 0.35^a	8.23 ± 0.46^a	0.80 ± 0.04^c	3.50 ± 0.28^c	95.69 ± 0.31^a	271.21 ± 8.84^a	0.93 ± 0.00^a	0.89 ± 0.01^a	2.59 ± 0.02^a
Waxy RD6 starch	Native	2.71 ± 0.36^c	7.75 ± 0.65^a	37.06 ± 0.92^b	56.39 ± 1.10^a	6.55 ± 0.52^b	5.46 ± 0.06^b	0.88 ± 0.02^b	0.79 ± 0.02^b	0.17 ± 0.01^d
	AC1	1.21 ± 0.34^c	8.22 ± 0.85^a	1.07 ± 0.09^c	4.68 ± 0.24^c	94.27 ± 0.33^a	214.31 ± 7.37^a	0.94 ± 0.00^a	0.91 ± 0.00^a	0.10 ± 0.01^def
	AC3	1.48 ± 0.19^c	8.58 ± 0.68^a	0.99 ± 0.26^c	4.08 ± 0.66^c	94.93 ± 0.92^a	210.73 ± 15.13^a	0.94 ± 0.00^a	0.92 ± 0.00^a	0.06 ± 0.00^ef
	AC5	1.83 ± 0.12^c	7.64 ± 0.81^a	1.53 ± 0.26^c	5.45 ± 0.59^c	93.03 ± 0.85^a	210.00 ± 32.21^a	0.93 ± 0.01^a	0.89 ± 0.01^a	0.02 ± 0.00^f

Values are means ± SEM. Means within each column with different superscripts are significantly different (p < .05) by Tukey’s test (n = 4). Abbreviations: AC1 = 1 cycle of AC; AC3 = 3 cycles of AC; AC5 = 5 cycles of AC; d(0.5) = the granule size at which 50% of all the granules by volume are smaller; R^1047/1022 = spectral ratio representing the order crystalline to amorphous structures; and R^995/1022 = spectral ratio representing the double helix to amorphous structures.

Effects of autoclaving-cooling cycle treatment on structural features of non-waxy and waxy rice starches

Granule morphology and particle size distribution

The impact of autoclaving-cooling cycle treatment on particle size and size distribution of the KDML105 and RD6 starch granules was observed. The native starch granules can be classified into three distinct types depending on their granular diameter including C- (<5 μm), B- (5–15 μm), and A- (>15 μm) type granules (Table 1).^[2,13] The changes in starch granule shape and size were observed upon the modification with AC1, AC3, and AC5 in comparison to the native granules. Scanning electron micrographs revealed that the native KDML105 starch granules were small-sized, disintegrated, polyhedral, spherical, and irregular shapes with a rugged granular surface (Figure 1). The granules were partially melted upon heating under high pressure conditions of the AC1, AC3, and AC5 treatments (Figure 1). These partially damaged starch granules in the gelatinized paste had reaggregated, recrystallized, and transitioned into a larger continuous matrix of starch particles with cluster-like structures and rough surfaces upon autoclaving and cooling for 1, 3, and 5 cycles (Figure 1). Similarly, the native RD6 starch granule showed disintegrated small-sized, polyhedral, spherical, and irregular shapes with a craggy granular surface, and they also appeared to be incompletely melted upon AC1 and AC3 treatments (Figure 1). Previous studies noted that the autoclaving-cooling treatment over numerous cycles required a large amount of water.^[7,12] Hence, insufficient water may cause the starch granules in the slurry to be incompletely disturbed in the following cycles.^[12] Moreover, the disrupted starch granules would recompose into the expanded matrix of starch granules with the formation of irregular granular aggregations and broken granular surfaces during heating and cooling processes of AC1 and AC3 (Figure 1). However, the original granular structure of native RD6 starch completely disappeared in AC5, which was instead replaced by a densely larger continuous structure with streaked and multi-layered strips on their homogeneous surfaces (Figure 1). The complete collapse of native RD6 starch granules in AC5 may be because the native RD6 starch granules contained a smaller amount of amylose with an absence of amylose-lipid complexes. They could absorb water, swell, and disrupt upon heating better than the native KDML105 starch granules, which possessed higher amylose content.^[13,17] Accordingly, the median particle size (d(0.5)) values of native KDML105 and RD6 starches (5.21 ± 0.08 μm and 5.46 ± 0.06 μm, respectively) greatly increased to 248.90 ± 9.04 μm, 270.51 ± 7.27 μm, and 271.21 ± 8.84 μm and 214.31 ± 7.37 μm, 210.73 ± 15.13 μm, and 210.00 ± 32.21 μm after the AC1, AC3, and AC5 treatments, respectively (Table 1). The particle size distribution also revealed obvious reductions in their volume distributions (%) of small C- and B-granules and a great increase in the volume distributions of large A-granules due to the autoclaving-cooling cycles (Table 1 and Figure S1A and S1B). The increment in particle size may be attributed to the reaggregation and recrystallization of amylose linear and amylopectin branch chains, which were dispersed in hot gelatinized starch during cooling. This in turn induced the reassembly of the mixed crystal by accelerating the formation of crystal nuclei and the growth of continuous gel network microstructures to produce larger particles.^[7,11,20]

Figure 1. Scanning electron micrographs of native and modified non-waxy KDML105 and waxy RD6 rice starches.

The SEM images were obtained at a magnification of 2000×. Scale bar = 8 μm. Abbreviations: AC1 = 1 cycle of AC; AC3 = 3 cycles of AC; AC5 = 5 cycles of AC.

X-ray diffractogram

The crystalline packing patterns of native and modified KDML105 and RD6 starches were determined using XRD to explore the details of their long-range ordered double helical structures.^[19] The native KDML105 and RD6 rice starches displayed the typical A-type crystalline polymorph consisting of the strong reflection peaks at the corresponding diffraction angles (2θ) of two single peaks at 15° and 23° and a doublet peak at 17° and 18° (Figure 2a,b). Both native starches also occupied a resolved weak diffraction peak (2θ) at 20° (Figure 2a,b), which is characterized as a V-type crystal pattern.^[2,19] The characteristic peaks of an A-type pattern are attributed to the amylopectin crystallite, whilst the weak diffraction peak of a V-type pattern is intrinsically linked to the structure of the amylose-lipid complex.^[19,21] Previous studies demonstrated that the heat-moisture^[19] and dual autoclaving-cooling cycle^[11] treatments annihilated the A-type crystalline structure and promoted the formation of V-type crystalline structure in oat rice starches. Similar results were observed in our study where all AC treatments greatly suppressed the diffraction peak intensity of the native KDML105 amylopectin crystallite from the strong two single and one double diffraction peaks to weaker ones (Figure 2a). Meanwhile, the A-type polymorph peak intensity of the native RD6 amylopectin crystallite drastically declined to nearly vanish after the AC1, AC3 to AC5 treatments, respectively (Figure 2b). Notably, the 2θ peak characteristic at 20° of the native RD6 amylose-lipid complex arrangement completely disappeared since the first autoclaving-cooling cycle (Figure 2b). This may be because the waxy rice RD6 starch contained a considerably lower amylose content than the non-waxy rice KDML105 starch (Table 1). The native RD6 starch granule may have less amylose-lipid complex than the non-waxy rice KDML105 starch granule.^[2,17] Following the autoclaving-cooling cycle treatment, the amylose-lipid complex in the RD6 starch granule was thus disappeared due to the complete depolymerization of the small amount of amylose.^[22] In contrast, a weak diffraction peak at 20° of KDML105 starches was still present after all AC treatments (Figure 2a). The presence of the amylose-lipid complex arrangement in the modified KDML105 starches might be attributed to the generation of a newly formed amylose-lipid complex, which instead replaced the partially destroyed original amylose-lipid complex due to the autoclaving-cooling cycle treatment.^[11,19,23] This observation also coincided with the scanning electron micrographs showing the formation of aggregated starch particles with a rough surface in KDML105 AC1, AC3, and AC5, and RD6 AC1 and AC3 starches (Figure 1). Therefore, both XRD and SEM results suggested incomplete disruption of the original granular structures in KDML105 under all AC treatments and in RD6 under AC1 and AC3 modification (Figures 1 and 2). Additionally, the XRD pattern revealed that the A-type crystalline polymorph peaks of RD6 starch completely vanished after the AC5 treatment (Figure 2b). This is consistent with the SEM images showing that the AC5 treatment could completely deform the granular structure of the native RD6 starch into a compact matrix microstructure of the starch gel network as shown in Figure 1.

Figure 2. X-ray diffractograms (a and b) and FTIR spectra (c and d) of native and modified non-waxy KDML105 and waxy RD6 rice starches.

Abbreviations: AC1 = 1 cycle of AC; AC3 = 3 cycles of AC; AC5 = 5 cycles of AC.

FTIR analysis

The alterations in the short-range ordered conformational structure of starch granules can be determined with FTIR analysis.^[24] Here, the FTIR spectra of native and modified non-waxy KDML105 and waxy RD6 rice starches are shown in Figure 2c,d, respectively. All samples possessed the wide single absorption peak centered at ~3400 cm⁻¹ and the single absorption peaks centered at 2929.77 cm⁻¹, ~1640 cm⁻¹, ~1500 cm⁻¹, ~1150 cm⁻¹, ~1080 cm⁻¹, and 620–527 cm⁻¹ (Figure 2c,d). These IR absorption peaks are related to O-H bond stretching, O-H bond and C-H bond stretching, COO^– stretching vibration in a carboxylate group, C-O-C (skeletal vibrational mode of α-1,4 glycosidic linkage), coupling of C-O, C-C and O-H bond stretching, bending and asymmetric stretching of the C-O-C glycosidic linkage, and skeletal vibrational modes of the pyranose ring, respectively.^[8] In addition, all samples displayed a doublet absorption peak at 1415.70 cm⁻¹ and 1367.48 cm⁻¹ (Figure 2c,d), which is ascribed to the bending of C-O-H, C-C-H, and C-O-H groups in starch molecule.^[8] All starch samples also had the highest single peak near 1022 cm⁻¹ (Figure 2c,d), which is representative of amorphous molecular structure (disordered structure) proportion in the granules.^[19,24] The IR absorbance peaks closed to 1047 cm⁻¹ and 995 cm⁻¹ were previously reported to signify the amount of order crystalline and double helix structures in the starch granules, respectively.^[19,24] Therefore, the IR intensities at those absorbance peaks were used to calculate the ratio of the intensities at 1047/1022 cm⁻¹ and 995/1022 cm⁻¹, which express the degrees of order crystalline and double helix structures within the starch molecule, respectively.^[19,25] Faridah et al.^[25] demonstrated that the AC treatment enhanced the ordered crystalline structure of the starch by causing breakdown of the amorphous structure upon autoclaving. Similarly, we found that the ratio of the intensities at 1047/1022 cm⁻¹ and 995/1022 cm⁻¹ of KDML105 and RD6 starches significantly increased after the autoclaving-cooling cycle treatment (Table 1). This could be the result of the declined intensity of absorption peaks at 1022 cm⁻¹ in all modified KDML105 and RD6 starches in comparison to their native counterparts (Figure 2c,d). In addition, the diminished intensities of the aforementioned characteristic peaks, particularly those associated with the O-H bond stretching region in the modified KDML105 and RD6 starches compared to corresponding native rice starches also supported this finding (Figure 2c,d). Certain hydrogen bonds linking the neighboring parallel double helices inside the native starch granules are broken down during gelatinization.^[19] Upon retrogradation, the dispersed starch chains containing a large amount of free -OH groups will retrograde by forming new hydrogen bands to regenerate a continuous gel network with a smaller amount of amorphous structure than native starch granules.^[25] Accordingly, the loss of amorphous regions reflects the reduction of intra- and intermolecular hydrogen bonding in the retrograded starch in comparison to the native starch.^[19]

Effects of autoclaving-cooling cycle treatment on the resistant starch content of non-waxy and waxy rice starches

The changes in RS content of both KDML105 and RD6 starches as a result of AC modification were observed. As displayed in Table 1, the RS content of KDML105 starch gradually increased from the KDML105 native starch (0.13 ± 0.01%) to the KDML105 AC1 starch (0.65 ± 0.02%), to the KDML105 AC3 starch (1.27 ± 0.06%), and finally to the KDML105 AC5 starch (2.59 ± 0.02%), respectively (Table 1). This could be explained by the substantial increase in the proportions of order (R1047/1022) and double-helix structures (R995/1022) (Table 1), which was accompanied by decreasing intensity of absorption peaks at 1022 cm⁻¹ of disordered structures within the modified KDML105 starches compared to the native one (Figure 2c). The disordered amorphous structures of starch molecules are more accessible to digestive enzymes in the small intestine than the ordered crystalline and double helix structures, which are tightly packed.^[3,7] Therefore, the loss of disordered amorphous structures of KDML105 starch after the AC1, AC3, and AC5 treatments is important for the formation of RS3 in this study.^[3] On the other hand, even though the degree of order R1047/1022 and double helix structure R995/1022 values of RD6 starch were enhanced by the AC1, AC3, and AC5 treatments (Table 1), the RS content of modified RD6 starches significantly decreased from that of the native RD6 starch (Table 1). We suggest that the steady decrease in the RS content in the modified RD6 starches with a lower amylose-to-amylopectin ratio compared to KDML105 starch may be attributed to an interplay of the following factors: (i) Upon cooling, the reaggregation of highly branched amylopectin polymers in the modified RD6 starches into newly ordered crystalline and double helical structures may form weaker intra- and intermolecular coherences due to stronger interchain steric hindrance of the very large amylopectin molecules compared to the modified KDML105 starches.^[6,26,27] (ii) The presence of a smaller amount of linear amylose polymers in the modified RD6 starches (Table 1) suggests the diminished formation of tightly bonded packs of double or single helix structures, which confers weaker resistance to enzymatic attack than the modified KDML105 starches.^[6,24] (iii) The absence of an amylose-lipid complex in the modified RD6 starches (Figure 2b) results in the lack of hydrophobic fatty acid tails, which intercalate inside an amylose double or single helix and aid in further linking starch molecules together. The lack of these fatty acid tails may impair resistance to gastrointestinal enzymatic hydrolysis of the modified RD6 starches.^[11,19]

Effects of autoclaving-cooling cycle treatment on the physicochemical properties of non-waxy and waxy rice starches

Thermal properties

The gelatinization behaviors, including the transition temperatures of gelatinization (i.e., onset To(g), peak Tp(g), and conclusion Tc(g) temperatures), temperature range of gelatinization (ΔT(g)), and enthalpy of gelatinization (ΔH(g)), of native and modified KDML105 and RD6 starches were summarized in Supplementary Table S1. The results revealed that the native KDML105 starch displayed higher To(g) and Tp(g) values than the native RD6 starch (Table S1). This could be due to the higher amounts of amylose and C-type granules in the native KDML105 starch than the RD6 starch (Table 1). Our previous studies demonstrated that KDML105 starch, containing higher proportions of amylose and C-type granules, required a higher temperature to initiate the phase transitions of starch granules (To(g)) and a higher temperature to dissociate the crystal structure of starch granules (Tp(g)) than that required by RD6 starch.^[2,13] In addition, the lower proportion of double helix structures (R995/1022), present within the native KDML105 starch (Table 1), may provide less compact and stable molecular crystalline structures compared to those of the native RD6 starch.^[2,6,13] Accordingly, the native KDML105 starch required a lower Tc(g) (Table S1), which is the point at which the starch birefringence entirely disappears,^[6] and showed a narrower range of gelatinization temperature (lower ΔT(g)) than the native RD6 starch (Table S1).^[2,6,13] To assess the altered gelatinization behaviors of modified non-waxy and waxy rice starches, the native starches were autoclaved at 121°C for 20 minute and stored at room temperature for 24 h for one, three, and five autoclaving-cooling cycles. With these heat treatment and storage conditions, a small amount of water molecules may escape from the starch gel (i.e., syneresis) during starch retrogradation.^[28] This may eventually decrease the amount of strong intra- and intermolecular hydrogen bonds of retrograded starch chains.^[28] The FTIR results also support this observation. The AC1, AC3, and AC5 treatments lessened intensities of the absorption peak associated with the O-H bond stretching region in the starches, especially the wide single peak centered at ~3400 cm⁻¹ (Figure 2c,d). As a result, the modified KDML105 and RD6 starches may be gelatinized completely when subjected to a lower temperature than the temperature range used in our DSC temperature program (30°C to 95°C). Therefore, we cannot detect the DSC endothermic peak of amylopectin melting and gelatinization behaviors of all modified KDML105 and RD6 starches (Table S1). Moreover, the XRD analysis showed that the intensities of A-type crystalline polymorph diffraction peaks from KDML105 AC1, AC3, and AC5 starches and RD6 AC1 and AC3 starches considerably declined (Figure 2a,b), while that of RD6 AC5 completely disappeared (Figure 2b). This further supports the low gelatinization transition temperatures required by these modified starches.^[19] Accordingly, previous studies demonstrated that dual autoclaving-cooling and heat-moisture treatments led to irreversible transitions by inducing water absorption, granular swelling, and the collapse of molecular ordered double helical structures inside the starch granule.^[3,19,29] This would allow the starch granules to completely disintegrate and solubilize the original A-type crystalline structure.^[3,19,29,30] After cooling, the dispersed starch chains realigned and recrystallized to form the new crystalline structures, while the original amylopectin crystallite vanishes.^[19,22] Therefore, the XRD diffraction peaks of the A-type crystal arrangement and the DSC endothermic peak of amylopectin melting of the recrystallized starch were not observed.^[19]

On the other hand, the retrogradation behaviors of the retrograded modified KDML105 starches could be monitored by a DSC (Table S1). For the retrogradation analysis, the gelatinized modified starches were stored at low storage temperatures (4°C) with longer duration (14 days) than the storage temperature and time used for AC treatments (section 2.3). This could promote the syneresis mechanism and give more strong and stable hydrogen bonds between amylose linear chains during retrogradation^[28] of the gelatinized modified KDML105 starches. Nonetheless, similar to their gelatinization behaviors, the retrogradation behaviors of the retrograded modified RD6 starch gels could not be detected by a DSC (Table S1). Since the autoclaved-cooled RD6 starches contained low amylose content with the absence of amylose-lipid complexes (Table 1 and Figure 2b), they tended to have small reassociation of dispersed starch chains in the gelatinized paste and led to the starch gel networks with less structural integrity upon cooling.^[2,31] Consequently, the gel networks within the retrograded modified RD6 starch gels may require lower temperature to dissociate and completely dissolve than the temperature range used for retrogradation analysis in this study. Notably, the retrograded KDML105 AC1, AC3, and AC5 starch gels required melting temperatures between 41.38–57.38°C, 41.71–57.62°C, and 41.68–57.88°C, respectively (Table S1), which fell within the temperature range of the DSC temperature program (30°C to 95°C). Furthermore, the observed recrystallization behaviors of the retrograded modified KDML105 starch gels (Table S1) could be related to their increased RS content (Table 1). The retrogradation of the gelatinized modified KDML105 starch indicated the recrystallization of the linear amylose chains, amylopectin side chains, and amylose-lipid complexes to form densely packed ordered crystalline and double helix structures. These structures are stabilized by hydrogen bonds and van der Waals interactions between the fatty acid methylene groups and the amylose glucose residues during the cooling step of the autoclaving-cooling cycle treatment.^[3,32] These molecular alterations could facilitate the formation of resistant starches in the autoclaved-cooled KDML105 starches relative to their native counterpart by allowing them to resist enzymatic digestion.^[6,19]

Pasting properties

The pasting properties of native and modified non-waxy and waxy rice starches were observed with a Rapid Visco Analyser (RVA) (Table 2 and Figure 3). Post AC treatment at one, three, and five cycles, the P_temp, PV, and BD values of both modified KDML105 and RD6 starches declined from their native values in similar manner (Table 2). This observation indicated that the autoclaved-cooled KDML105 and RD6 starches may possess ordered double helical structures with a lesser degree of perfection than their native counterparts.^[33] The autoclaved-cooled KDML105 and RD6 starches lost their the original A-type crystalline structure (Figure 2a,b) and became more susceptible to heating and shearing force than their native structures (Table 2).^[28,33] The results also revealed that the AC3 and AC5 treatments induced a reduction in the trough viscosity (TV) of the RD6 starch (Table 2). TV measured the susceptibility of starch paste to heat and shear.^[16] Therefore, the RD6 AC3 and AC5 starches with the disrupted original A-type crystal structure (Figure 2b) experienced intensive granular collapse and leaching out of starch polymers when the starch sample was held at a constant high temperature (95°C) with mechanical shear during the holding phase of the test.^[34] The scenario would lead to a marked drop in viscosity, finally reflected in reduced TV value (Table 2). Additionally, the final viscosity (FV) value of the native RD6 starch substantially declined from 2131.50 mPa.s to 1470.75 mPa.s, 1292.50 mPa.s, and 1120.58 mPa.s after the AC1, AC3, and AC5 treatments, respectively (Table 2). The autoclaving-cooling cycle treatment can increase the formation of the amylopectin short-chain region due to the destruction of the amylopectin long-chain region upon autoclaving.^[6] The increased amylopectin short-chain region with unparalleled and irregular short branch chains may impede the molecular interactions between the starch chains dispersed in the gelatinized starch upon cooling.^[35] This scenario could attenuate the formation of a hard gel network through hydrogen bonds, particularly in the RD6 AC1, AC3, and AC5 starch gels containing a very low amylose content with absences of amylose-lipid complexes (Table 1 and Figure 2b).^[19,31] Thus, the decline in FV value was observed in the RD6 AC1, AC3, and AC5 starches (Table 2). Furthermore, this result suggested that the recrystallization of the modified RD6 starch chains during the cooling step may provide lower structural integrity, making them more susceptible to enzyme attack than the native RD6 starch.^[3,6] Therefore, the decline in the RS content of the RD6 starch after addition of autoclaving-cooling cycles was observed (Table 1). Conversely, no significant difference in FV value was observed between the native and modified KDML105 starches (Table 2). Amylose chains play a major role in the starch retrogradation process due to their essentially linear and flexible structure,^[28,33,36] which makes it easier to create hydrogen bonds between molecules and results in the gel viscosity.^[36] Therefore, the native and modified KDML105 starch, containing comparable amounts of amylose, showed no significant variations in their FV values (Tables 1 and 2). Moreover, the mixed starch chains that were dispersed in the modified KDML105 starch gels could readily rebind to form a continuous crystal network with large compact block particles (Figure 1) through the hydrogen-bonded intra- or inter-association during the cooling step of AC treatment.^[3,28] As a result, the presence of the compact continuous gel network in the large KDML105 AC1, AC3, and AC5 starch particles is presumably responsible for their enhanced resistance to enzyme hydrolysis (Table 1).^[7,36]

Figure 3. Viscoamylograph of native and modified non-waxy KDML105 (a) and waxy RD6 (b) rice starches.

Abbreviations: AC1 = 1 cycle of AC; AC3 = 3 cycles of AC; AC5 = 5 cycles of AC.

Table 2. Pasting properties of native and modified non-waxy KDML105 and waxy RD6 rice starches.

Samples	Treatments	Pasting properties
		PTemp (°C)	PTime (min.)	PV (mPa.s)	TV (mPa.s)	BD (mPa.s)	FV (mPa.s)	SB (mPa.s)
Non-waxy KDML105 starch	Native	66.27 ± 0.28^a	4.94 ± 0.06^a	4260.75 ± 170.67^a	1229.50 ± 240.08^abc	3031.25 ± 248.01^a	2407.25 ± 96.14^a	1177.75 ± 205.58^a
	AC1	49.77 ± 0.13^b	4.05 ± 0.23^ab	2087.50 ± 179.80^cd	1420.50 ± 192.34^ab	667.00 ± 79.67^c	2572.75 ± 272.94^a	1152.25 ± 163.83^a
	AC3	55.17 ± 5.22^b	4.95 ± 0.39^a	1753.00 ± 103.32^cd	1558.00 ± 78.49^a	195.00 ± 57.91^c	2192.00 ± 106.03^a	634.00 ± 34.90^b
	AC5	49.80 ± 0.09^b	4.60 ± 0.17^a	2202.25 ± 156.72^c	1584.00 ± 20.98^a	618.25 ± 162.23^c	2281.00 ± 55.71^a	697.00 ± 41.84^ab
Waxy RD6 starch	Native	64.86 ± 0.81^a	3.88 ± 0.03^b	3219.75 ± 117.26^b	1399.75 ± 40.90^ab	1820.00 ± 81.65^b	2131.50 ± 74.62^a	731.75 ± 37.54^ab
	AC1	49.35 ± 0.03^b	1.43 ± 0.23^c	1494.00 ± 81.22^de	885.83 ± 34.19^bc	608.18 ± 59.48^c	1470.75 ± 32.27^b	464.64 ± 121.59^b
	AC3	49.42 ± 0.12^b	1.40 ± 0.17^c	1585.50 ± 106.58^de	815.18 ± 21.18^c	770.33 ± 113.47^c	1292.50 ± 64.03^b	477.33 ± 45.21^b
	AC5	49.68 ± 0.26^b	1.28 ± 0.25^c	1090.30 ± 99.57^e	707.15 ± 47.77^c	383.15 ± 61.96^c	1120.58 ± 84.05^b	413.43 ± 47.94^b

Values are means ± SEM. Means within each column with different superscripts are significantly different (p < .05) by Tukey’s test (n = 4). Abbreviations: AC1 = 1 cycle of AC; AC3 = 3 cycles of AC; AC5 = 5 cycles of AC; PTemp = Pasting temperature; PTime = Peak time; PV = Peak viscosity; TV = Trough viscosity; BD = Breakdown; FV = Final viscosity; and SB = Setback.

Conclusion

This study compared changes in the RS content, structural features, and physicochemical properties of non-waxy KDML105 and waxy RD6 rice starches modified by autoclaving-cooling cycles. Overall, our hypothesis was confirmed since the effects of AC treatment on RS content and physicochemical properties of the two starches were significantly different. The AC treatments increased the starch particle size, while minimized the amorphous (1022 cm⁻¹) structures, resulting in enhanced proportions of ordered (R1047/1022) and double helical (R995/1022) structures in both cultivars. Notably, the AC cycles considerably impaired the starch structures of both cultivars so that they cannot be observed for the gelatinization behaviors. However, the modified KDML105 starch gels were able to form retrograded structures that can be observed for their thermal properties. In comparison to their native starches, the pasting profiles of modified KDML105 starch was less affected by AC cycles compared to the modified RD6 starch. This is because the AC cycles incompletely destroyed the A-type crystalline structure found in the native KDML105 starch, while the A-type crystalline structures of the RD6 starch completely disappeared after AC5. This was directly related to a gradual increase and decline in the RS content observed in modified KDML105 and RD6 starches upon the addition of AC cycles, respectively. Therefore, as demonstrated in the modified KDML105 starch, the formation of RS3 during AC cycles was enhanced by amylose and amylose-lipid complexes present in its native counterpart. Our findings suggest that non-waxy or high-amylose rice starches offer a more suitable raw material for starch modification by autoclaving-cooling cycles. Nonetheless, further study is required to optimize the methods for higher RS3 formation in order to produce functional rice starch products with nutritional and health benefits.

Abbreviations

BD	=	Breakdown
Tc(g)	=	Conclusion temperature of gelatinization
Tc(r)	=	Conclusion temperature of the retrograded starch melting
DSC	=	Differential scanning calorimetry
∆H(g)	=	Enthalpy of gelatinization
∆H(r)	=	Enthalpy of the retrograded starch melting
FV	=	Final viscosity
FTIR	=	Fourier transform infrared
TV	=	Trough viscosity
To(g)	=	Onset temperature of gelatinization
To(r)	=	Onset temperature of the retrograded starch melting
Tp(g)	=	Peak temperature of gelatinization
Tp(r)	=	Peak temperature of the retrograded starch melting
PTemp	=	Pasting temperature
PTime	=	Peak time
PV	=	Peak viscosity
RVA	=	Rapid Visco Analysis
RS	=	Resistant starch
RS3	=	Resistant starch type 3
SB	=	Setback
R995/1022	=	Spectral ratio of the double helix to amorphous structures
R1047/1022	=	Spectral ratio of the order crystalline to amorphous structures
∆T(g)	=	Temperature range of gelatinization
∆T(r)	=	Temperature range of the retrograded starch melting
d(0.5)	=	The granule size at which 50% of all the granules by volume are smaller
XRD	=	X-ray Diffractometry

Acknowledgments

The authors thank the Research Instrument Center, Khon Kaen University for instrumental services. We are grateful to Asst. Prof. Jirawat Sanitchon for a gift of rice seeds used in this study. We also thank Mr. Boonsong Kongsook, Mr. Supon Bokum, and Assoc. Prof. Nithima Khaorapapong and Miss Nipaporn Pongkan for their kind guidance and assistance in granule size distribution, SEM, and FTIR spectroscopy analyses, respectively. Miss Jurarat Chantaban, Miss Sirikajorn Pongsiri, and Miss Yanisa Kritsadakanpinyo also aided in starch compositional and structural analyses.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon request.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/10942912.2024.2332357.

Additional information

Funding

This work was supported by the Fundamental Fund of Khon Kaen University, which has received funding support from the National Science Research and Innovation Fund (NRSF). The funder has no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.