Research on the Relationship between the Amylopectin Structure and the Physicochemical Properties of Starch Extracted from Glutinous Rice

Glutinous rice has very low amylose content and is a good material for determining the structure and physicochemical properties of amylopectin. We selected 29 glutinous rice varieties and determined the amylopectin structure by high-performance anion exchange chromatography with the pulsed amperometric detection method. We also determined the correlation between amylopectin structure and the physicochemical properties of starch extracted from these varieties. The results showed that the amylopectin chain ratio Σdegree of polymerization (DP) ≤ 11/ΣDP ≤ 24 of 29 glutinous rice varieties was greater than 0.26, signifying that these varieties contained type II amylopectin. The results of the correlation analysis with gelatinization temperature showed that ΣDP 6–11 was significantly negatively correlated with the onset gelatinization temperature (GT) (TO), peak GT (TP), and conclusion GT (TC). Among the thermodynamic properties, ΣDP 12–24 was significantly positively correlated with To, Tp, and Tc, ΣDP 25–36 was significantly negatively correlated with To, Tp, and Tc, and ΣDP ≥ 37 had no correlation with the thermodynamic properties. The results of correlation analysis with RVA spectrum characteristic values showed that ΣDP 6–11 was significantly negatively correlated with hot paste viscosity (HPV), cool paste viscosity (CPV), consistency viscosity (CSV), peak time (PeT), and pasting temperature (PaT) among the Rapid Visco Analyzer (RVA) profile characteristics, ΣDP 12–24 was significantly positively correlated with HPV, CPV, CSV, PeT, and PaT, and ΣDP ≥ 25 had no correlation with the viscosity characteristics. Therefore, we concluded that the amylopectin structure had a greater effect on the TO, TP, TC, ΔH and peak viscosity, HPV, CPV, CSV, PeT, and PaT. The glutinous rice varieties with a higher distribution of short chains and a lower distribution of medium and long chains in the amylopectin structure resulted in lower GT and RVA spectrum characteristic values.


Introduction
The structure of amylopectin has an important effect on the physicochemical properties of starch [1,2], which is the main reason for the quality difference between rice varieties with similar amylose contents (AC) [3]. Previous studies have shown that the amylopectin structure has a significant effect on rice gelatinization temperature (GT) [4]. Nishi et al. [5] found that the decrease in the short chain of amylopectin leads to an increase in GT. Vandeputte et al. [2] and He [6] found that the short chain and the medium-long chain in amylopectin were significantly negatively and significantly positively correlated with rice gelatinization heat, respectively. The relationship between the amylopectin structure and Rapid Visco Analyzer (RVA) spectrum characteristic values has been different in different studies. By performing gel-filtration chromatography, Cai et al. [7] found that FrIII (DP 10-17 glucose unite) in the amylopectin structure was significantly positively correlated with peak viscosity (PKV) and breakdown viscosity (BDV) in the RVA spectrum characteristic values, whereas FrI (DP > 100 glucose unite) and FrII (DP 44-47 glucose unite) showed an opposite trend. Han et al. [8] used the same method to conclude that FrI was significantly negatively correlated with BDV, whereas FrIII was significantly positively correlated with BDV. He [9] used the improved fluorophore-assisted carbohydrate electrophoresis (FACE) method based on a DNA sequencer analysis and found that the proportion of branch chains in different chain length ranges was mainly related to the pasting temperature and the relative crystallinity of starch but was not closely related to the gel consistency and RVA spectrum characteristic values of starch. According to previous studies, we found that AC is not the only determinant of rice quality and the physicochemical properties of starch. Significant differences exist in the physicochemical properties of starch among AC-similar varieties, and the amylopectin structure also has an important effect on rice quality and the physicochemical properties of starch [9]. As the amylopectin content of glutinous rice is very high, accounting for 95-100% of the total starch content, the amylopectin structure of glutinous rice is more complex, which has a great effect on its quality. At the same time, the AC of most glutinous rice is very low (<2%). By using glutinous rice as an experimental material, the effect of amylose can be considerably reduced, and the relationship between the fine structure of amylopectin and the physicochemical properties of starch can be studied more effectively. Although glutinous rice is not suitable for staple food, it is widely used to make traditional food, drinks and condiments, such as tangyuan, rice wine, vinegar, etc. It can be used as raw material of biological composite materials. It is also used in the medical industry (capsule), skin care products, printing, and other fields [10,11]. Therefore, the research on quality breeding of glutinous rice has more potential. In this study, the distribution of amylopectin chain length and the physicochemical properties of amylopectin obtained from 29 varieties of glutinous rice were determined, and the correlation analysis was performed to determine the effect of the crystalline structure of amylopectin from glutinous rice on the physicochemical properties of starch. We aimed to reveal the genetic basis of the formation of amylopectin structure in glutinous rice and provide a theoretical basis for the breeding and improvement of glutinous rice starch quality.

Materials and Instruments
In this study, 29 glutinous rice varieties of different qualities were used as study materials, including 17 varieties of indica rice and 12 varieties of japonica rice (Table 1). Test varieties were planted in the Hainan South Breeding Base of Jiangxi Agricultural University on 5 December 2019. The field was managed routinely, wherein sowing was performed in stages, and the grains were harvested at the same time. After the rice matured, every single plant was harvested and stored for three months. After the quality of the harvest was assessed, the brown rice was obtained after processing the harvest using a three-vertical husker (SY88-TH, Korea Shuanglong Machinery Manufacturing Co., Ltd., Taegu, Republic of Korea), which was then polished using a rice polisher machine (SY2001 + NSART100, Korea Shuanglong Machinery Manufacturing Co., Ltd., Taegu, Republic of Korea) to obtain the polished rice. The polished glutinous rice was ground into rice flour using a cyclone mill (CT293, FOSS, Hilleroed, Denmark) and passed through a 100-mesh sieve. The rice flour was dried in an oven set at 37 • C for 48 h, placed at room temperature (25 • C) for 24 h to maintain its water content at 13 ± 1%, and then sealed for subsequent experiments. All indexes were repeated three times for each sample during determination. The following experimental instruments were used in this study: TU-1810D ultravioletvisible spectrophotometer: Beijing Puxi Company, Beijing, China; Thermo ICS5000+ ion chromatography system: Thermo Fisher Scientific, Waltham, MA, USA; Matersizer 3000 Laser Particle Size Analyzer: Malvern Instruments Ltd., Worcestershire, UK; X'Pert Pro X-ray diffractometer: PANalytical, Almelo, The Netherlands; RVA-TecMaster viscometer: Perten, Stockholm, Sweden; and DSC 4000 differential scanning calorimeter (DSC): Perkin-Elmer, Waltham, MA, USA.

Wx Genotyping
DNA was extracted according to the cetrimonium bromide method [12], and sequencetagged site (STS) markers were developed by Sun et al. [13] for the different sequence in the second exon of the rice waxy (Wx) gene to detect glutinous rice varieties. The polymerase chain reaction (PCR) was performed under the following conditions: 5 min at 94 • C, followed by 30 s at 94 • C, 30 s at 55 • C, and 30 s at 72 • C for 35 cycles, and 10 min at 72 • C for a final extension. Only waxy varieties have 492 bp PCR products, while non waxy varieties have no PCR products [13]. PCR amplification primers used are as follows: Glu-F: 5 -GGGTGCAACGGCCAGGAT-3 Glu-R: 5 -TGGAACCCGTGGGCTTGA-3

Determination of AC
AC was determined according to the Chinese national standard GB/T15683-2008 [14].

Extraction and Purification of Starch
Starch was extracted and purified according to the method of Wei et al. [15].

Determination of the Chain-Length Distribution of Starch
According to the method of Zhang et al. [16], the amylopectin chain-length distribution was determined by high-performance anion exchange chromatography with pulsed amperometric detection (HAPED-PAD) using the Thermo ICS5000+ ion chromatography system equipped with pulsed Abe detection (Thermo Fisher Scientific, Waltham, MA, USA). Dionex™ CarboPac™ PA10 (250 × 4.0 mm, 10 µm) liquid chromatographic column is adopted for chromatographic system, with the injection volume of 20 µL. Moving phase A: 200 mM NaOH; Phase B: 200 mM NaOH/200 mM NaAC, the column temperature is 30 • C, and the monosaccharide components are analyzed and detected by electrochemical detector.

Determination of the Relative Crystallinity
X'Pert Pro X-ray diffractometer (PANalytical, Almelo, The Netherlands) was used to analyze the X-ray diffraction patterns of the crystallographic structure of starch. The relative crystallinity of starch was calculated using the MDI Jade software.

Determination of Starch Viscosity
The starch viscosity was determined using the RVA-TecMaster viscometer (Perten, Stockholm, Sweden) and its supporting software, Thermal Cycle for Windows, according to the American Association of Cereal Chemists operating procedures (1995 61-02) [17]. The obtained results were based on various RVA spectrum characteristic values, including PKV, hot paste viscosity (HPV), cool paste viscosity (CPV), setback viscosity (SBV), BDV, consistency viscosity (CSV), peak time (PeT), and pasting temperature (PaT).

Determination of the Thermodynamic Properties
The thermodynamic properties were measured using a DSC (DSC 4000, Perkin-Elmer, Waltham, MA, USA), and the sample curves were analyzed using the supporting software Pyris Manager. The onset GT (T O ), peak GT (T P ), conclusion GT (T C ), and gelatinization enthalpy (∆H) were recorded. The specific methods are: weigh 5.0 mg of rice flour sample with 14.0% water content into an aluminum crucible and add 10 µL deionized water, after mixing, use a matching sample press to seal the crucible, and leave it in a 4 • C refrigerator overnight; before the test, take out the crucible stably, place it at room temperature and balance it for 1 h, and then go on the machine for measurement; take the empty disk as the control, and conduct 10 • C/min heating at 30~110 • C; analyze the sample curve and record T O , T P , T C, and ∆H.

Statistical Analysis
The structural and physicochemical properties of each amylopectin were measured in duplicate. Excel 2011 and IBM Statistical Package for Social Sciences Statistics 22.0 data processing systems were used for analysis of the phenotypic data of the tested varieties and analysis of variance (ANOVA), Duncan's multiple comparisons, and Pearson's correlation. The means of duplicated measurements were used for the analysis. Significant differences in the mean values were determined at p < 0.05. Table 1 shows the AC of the tested varieties. The results showed a little difference in AC among the test varieties, with the highest AC in Yunhenuo (2.91%) and the lowest in Anyinuo (0.46%), with an average AC of 2.06%.

Wx Gene Type and AC of the Glutinous Rice Varieties
The glutinous gene wx is an allelic variation of the Wx gene, which has a recessive mutation due to 23-bp fragment deletion in the second exon of the Wx gene. The appearance of glutinous rice greatly differs from that of non-glutinous rice, but distinguishing this appearance among the heterozygous genotypes is difficult using conventional methods [15]. Therefore, we used the STS dominant molecular markers designed by Sun et al. [13] to detect glutinous rice varieties among the test varieties. The glutinous appearance of the test varieties in this study was due to 23-bp fragment deletion in the second exon of the Wx gene, and the amplified fragment length was 492 bp. According to AC determination and the Wx genotyping results of the tested varieties, AC had little effect on the physicochemical properties of starch in glutinous rice, and the amylopectin structure was the determining factor affecting the physicochemical properties of starch.

Amylopectin Structure
In this study, the amylopectin chain length and its chain-length distribution in the 29 test varieties were determined according to the method of He Xiaopeng [6]. The amylopectin chain lengths were divided into short (DP 6-11), medium-long (DP 12-24), long (DP [25][26][27][28][29][30][31][32][33][34][35][36], and extra-long (DP ≥ 37) chain lengths ( Table 2). Their distribution in the test varieties was different based on their DP [6]. The distribution range of short chains was 18.087-25.332%, among which, Guazixiangdao had the highest and Heixiangnuo had the lowest DP. The short-chain distribution, DP 6-11, in the rice starch branch was the main factor affecting its GT. The short-chain distribution with DP ≤ 11 was higher in most of the glutinous rice varieties than that in the common rice. Therefore, the GTs of the glutinous rice varieties were generally low. The distribution range of ΣDP 12-24 was 52.609-60.592%, among which Zaonuo116 had the highest and Xiangyanuo had the lowest DP. The distribution range of ΣDP 25-36 was 10.86-12.25% and that of DP ≥ 37 was 9.39-11.30%. The average DP was between 19.01 and 20.03, of which Heixiangnuo had the highest and Guazixiangdao had the lowest DP.

Physicochemical Properties of Starch
The results of the physicochemical properties of starch in the test varieties are shown in Tables 3 and 4. The results showed that the relative crystallinity of starch granules in the test varieties ranged from 31.75% to 40.47%, which belonged to the typical A-type crystalline structure [18]. The relative crystallinity of most glutinous rice varieties was higher than that of the common rice [19].   Among the RVA spectrum characteristic values of the test varieties, PKV ranged from 1523 centipoise (cP) to 3425.33 cP, which was generally lower than that of the common rice. HPV ranged from 882 cP to 2111.33 cP, CPV ranged from 1085 cP to 2563 cP, BDV ranged from 641 cP to 1529 cP, SBV ranged from −1222.67 cP to −363.67 cP, CSV ranged from 203 cP to 484 cP, PeT ranged from 3.76 min to 4.78 min, and PaT ranged from 70.95 • C to 83.00 • C. The measurement results of the thermodynamic characteristics showed that the ranges of variation of T O , T P , and T C values were 62.54-76.91 • C, 70.25-80.66 • C, and 77.92-88.18 • C, respectively. In the heating process of glutinous rice starch, due to the complexity of amylopectin structure, the pasting temperature of individual varieties is higher. Compared with ordinary rice, most glutinous rice varieties had lower GTs. The ∆H was between 6.33-12.10 J/g, with a mean value of 9.30 J/g. The large difference in ∆H among the test varieties indicated that during the heating process of the glutinous rice starch, due to the complexity of the amylopectin structure, the enthalpy change values among the varieties were different.

Correlation Analysis between the Thermodynamic and Physicochemical Properties of Starch
Further correlation analysis between the thermodynamic and physicochemical properties of starch in different glutinous rice varieties (Table 5) revealed a close correlation between the thermodynamic properties and most of the RVA spectrum characteristic values. Except that BDV and SBV had no correlation with thermodynamic parameters, and PKV had no correlation with T O , T C , and ∆H, the other characteristic values were significantly correlated with the thermodynamic parameters. Relative crystallinity and ∆H were significantly positively correlated. To and Tc were significantly positively correlated with HPV, CPV, CSV, PeT, and PaT. Tp was significantly positively correlated with PKV, HPV, CPV, CSV, PeT, and PaT. ∆H was significantly positively correlated with PaT, HPV, CPV, CSV, and PeT. Among these, the correlation coefficients between PaT and T O , T P , T C , and ∆H were the highest, which were 0.965, 0.944, 0.689, and 0.591, respectively, indicating a great correlation among each of the physicochemical properties of starch. PaT in the RVA spectrum can differ in GTs of different varieties to some extent.

Correlation Analysis between the Amylopectin Structure and Physicochemical Properties of Starch
The correlation analysis results of the amylopectin chain length distribution and physicochemical properties of starch are shown in Table 6. ΣDP 6-11 in the test varieties was significantly negatively correlated with PKV, HPV, CPV, CSV, PeT, and PaT in the RVA spectrum characteristic values and T O , T P , T C , and ∆H in the thermodynamic properties. ΣDP 12-24 was significantly positively correlated with HPV, CPV, CSV, PeT, and PaT in the RVA spectrum characteristic values and T O , T P , T C , and ∆H in the thermodynamic properties. ΣDP 25-36 was significantly negatively correlated with PaT in the RVA spectrum characteristic values and T O , T P , and T C in the thermodynamic properties. No correlation was present between ΣDP ≥ 37 and the physicochemical and thermodynamic properties of each starch sample. The average DP was only significantly positively correlated with PeT and PaT in the RVA spectrum characteristic values, indicating that the lesser the short-chain distribution of glutinous rice amylopectin (DP 6-11), the more the medium-long-chain distribution (DP [12][13][14][15][16][17][18][19][20][21][22][23][24], and the lesser the long-chain distribution (DP 25-36), the higher the RVA spectrum characteristic values, this is consistent with Zhou's study [20]. The higher the average DP of the glutinous rice, the higher the pasting time and PaT of starch. No significant correlation was present between the chain-length distribution of glutinous rice amylopectin and the relative crystallinity of starch, indicating that the complexity of the amylopectin structure did not affect the crystallinity of the glutinous rice starch granules. Note: * indicates a significant correlation at the 0.05 level, and ** indicates a very significant correlation at the 0.01 level.

Amylopectin Chain Length and Its Distribution in the Glutinous Rice Varieties
Amylopectin is the main component of the rice endosperm. With the development of science and technology, the methods for amylopectin structure determination are also gradually developing. Presently, these methods can be divided into two categories: electrophoresis and chromatography, with each category consisting of various submethods and having its own advantages and disadvantages. The methods can be divided into FACE, enzymatic method, gel chromatography [13], and spectrophotometry [21]. Umemoto et al. [22,23] believed that the distribution of amylopectin short-length chains with DP ≤ 11 and the medium-length chains with DP 12-24 in different rice varieties was relatively different, whereas the number of long-length chains with DP ≥ 25 was the same. Nakamura et al. [24] used the FACE method based on capillary electrophoresis to determine the amylopectin structures of 129 different rice varieties. The amylopectin structures were divided into long-chain type (L-type) and short-chain type (S-type), and the amylopectin chain ratio (ACR) of ΣDP ≤ 10/ΣDP ≤ 24 of the L-type amylopectin was less than 0.20, whereas that of ΣDP ≤ 10/ΣDP ≤ 24 of the S-type amylopectin was greater than 0.24.
He et al. [9] used the FACE method based on a DNA sequencer to determine the amylopectin structure of 50 different indica and japonica rice varieties. Based on the actual differences in the chain length and chain-length distribution in different types of rice varieties, the ACR of ΣDP ≤ 11/ΣDP ≤ 24 was used as the classification basis. All amylopectin varieties were divided into two types: type I and type II. The ACR of type I amylopectin was less than 0.22, corresponding to the L-type amylopectin, and that of type II amylopectin was greater than 0.26, corresponding to the S-type amylopectin. In this study, HAPED-PAD [5], using a Thermo ICS5000+ ion chromatography system equipped with pulsed amperometric detection, was performed to determine amylopectin chain length distribution in the 29 glutinous rice varieties. The HAPED-PAD method is simpler and more reproducible than other methods, such as the enzymatic method, spectrophotometry, and FACE, as it can accurately determine the distribution of amylopectin chain lengths with different DP and can more effectively analyze the correlation between amylopectin chain length distribution and the physicochemical properties of starch. The results showed that except for Heixiangnuo and Zaonuo116, whose ACR were 0.231 and 0.236, respectively, the ACR of other glutinous rice varieties were greater than 0.26 and ranged from 0.301 to 0.324, which was consistent with the results of He et al. [9], thus verifying the accuracy of the HAPED-PAD method.

Correlation between the Amylopectin Structure and Physicochemical Properties of Starch
GT refers to the temperature at which a large number of water-absorbing starch granules undergo irreversible expansion, birefringence, and crystallinity disappearance after being heated in a suspension aqueous solution [25]. Vandeputte et al. [2] reported that ΣDP 6-9 was negatively correlated with GT, whereas ΣDP 12-22 was positively correlated with GT. Qi et al. [26] analyzed the starch chain length distribution in six glutinous rice varieties and found that the higher the ratio of ΣDP 13-24, the higher the GT. Satoh et al. [27] reported a decrease in starch extra-long and long chains in the ae rice mutant and a significant increase in the short chains, resulting in a significant decrease in T O of starch. He et al. [9] found that the short chains with DP 6-11 and medium-length chains with DP 13-24 were significantly negatively and significantly positively correlated with pasting temperature (PT), respectively. The long branch chains with DP 28-34 were significantly negatively correlated with GT, and the extra-long branch chains with DP 39-49 were significantly positively correlated with it in all of the varieties. In this study, T O , T P , and T C of the test varieties were significantly negatively correlated with ΣDP 6-11 and ΣDP 25-36 and significantly positively correlated with ΣDP 12-24. These results were consistent with those of previously reported studies, indicating that the amylopectin structure plays similar and important role in the GT of glutinous rice and non-glutinous rice. The relative number of short chains with DP 6-11 and long chains with DP 25-36 should be increased, whereas the relative number of medium-length chains with DP 12-24 should be reduced to improve the rice amylopectin structure for reducing GT.
The RVA spectrum characteristic values are closely related to the eating quality of rice. The rice varieties with lower HPV, CPV, SBV, CSV, PeT, and PaT and a higher BDV are considered to have better grain quality, softer texture, better viscosity, and cold rice texture [9]. By performing gel chromatography, Cai et al. [7] found that FrIII in the amylopectin structure was significantly positively correlated with PKV and BDV in the RVA spectrum, whereas FrI and FrII were negatively correlated with it. Han et al. [8] used the gel chromatography and determined that FrI was significantly negatively correlated with BDV, whereas FrIII was significantly positively correlated with it. By performing chromatography, Jin et al. [28] found that the amylopectin FrI + FrII content was significantly positively correlated with HPV, CPV, SBV, and CSV and was significantly negatively correlated with BDV. He et al. [9] suggested that the ACR of short chains with DP 6-11 and medium-length chains with DP 13-24 had no significant correlation with the RVA spectrum characteristic values in the low-AC rice varieties, whereas in the high-AC rice varieties, SBV was positively correlated with ΣDP 6-11 and negatively correlated with ΣDP 13-24. ΣDP 28-34 and ΣDP 39-49 also had no significant correlation with the RVA spectrum characteristic values in the general rice varieties. Among the 29 glutinous rice varieties used in this study, the distribution of short chains with DP 6-11 was significantly negatively correlated with PKV, HPV, CPV, CSV, PeT, and PaT. The distribution of mediumlong chains with DP 12-24 was significantly positively correlated with HPV, CPV, CSV, PeT, and PaT. The long chains with DP 25-36 were significantly negatively correlated with PaT. DP was significantly positively correlated with PeT and PaT, which was inconsistent with the results of previous studies using non-glutinous rice varieties. The reason behind this is glutinous rice is considerably different from non-glutinous rice, as it almost does not contain amylose. The difference in the amylopectin structure leads to differences in the physicochemical properties of starch among different varieties. Therefore, in glutinous rice varieties, the distribution and proportion of amylopectin short and medium-long chains are important factors in determining their eating quality.

Conclusions
In this study, the amylopectin chain ratio ΣDP ≤ 11/ΣDP ≤ 24 of 29 glutinous rice varieties was greater than 0.26, signifying that these varieties contained type II amylopectin. The results of the correlation analysis showed that ΣDP 6-11 was significantly negatively correlated with T O , T P , and T C among the thermodynamic properties, ΣDP 12-24 was significantly positively correlated with To, Tp, and Tc, ΣDP 25-36 was significantly negatively correlated with To, Tp, and Tc. ΣDP 6-11 was significantly negatively correlated with HPV, CPV, CSV, PeT, and PaT among the RVA profile characteristics, ΣDP 12-24 was significantly positively correlated with HPV, CPV, CSV, PeT, and PaT. Therefore, we concluded that the amylopectin structure had a greater effect on the T O , T P , T C , ∆H and peak viscosity, HPV, CPV, CSV, PeT, and PaT. The glutinous rice varieties with a higher distribution of short chains and a lower distribution of medium and long chains in the amylopectin structure resulted in lower GT and RVA spectrum characteristic values. These new findings will significantly assist in revealing the genetic basis of the formation of amylopectin structure in glutinous rice and provide a theoretical basis for the breeding and improvement of glutinous rice starch quality.