Waxy and Soluble Starch SynthaseII-3 Alleles Regulating Rice Resistant Starch Contents from Different Processing Status

Resistant Starch (RS) is a healthy dietary fiber that has functions of regulating diabetes, hypertension and obesity. Previous studies mainly focused on investigating RS in raw rice or cooked rice separately, which may receive different results. In this study, ninety-nine lines from a recombinant inbred line (RIL) were selected to investigate the effects of starch synthesis-related genes on the RS content in different process status. RS content in rice will change by different processing ways. Waxy ( Wx ) played an important role in controlling RS content and Wx a could elevate RS content, and soluble starch synthaseII -3 ( SSII-3 ) had an impact on RS2. Additionally, interaction of Wx and SSII-3 was responsible for variations of RS content in three sample types and RS2. Wx could affect RS in cooked rice and retrograded rice under the same SSII-3 allele. Moreover, the correlation analysis results indicated that RS was closely relative with many indexes of physicochemical properties. at 0.01 level (Fig.3c and d). The result indicated that SSII-3 had an influence on RS2. It revealed that SSII-3 had important effect on rice RS2.

glucose pyrophosphorylase (AGPase), granule-bound starch synthase (GBSSI), soluble starch synthase (SS), starch branching enzymes (SBEs), isoamylase (ISA) and pullulanase (PUL) (Tian et al. 2009). Among them, GBSSI is encoded by the Waxy ( Wx), which plays an important role in determining the synthesis of amylose. SS, SBEs, ISA and PUL can control the synthesis of amylopectin. RS content can be affected by the ratio of amylose to amylopectin and the fine structure of starch (Eggum et al. 1993;Hu et al. 2004). Thus, these starch synthesis-related genes (SSRGs) naturally have great influence on the synthesis of RS. A high-RS mutant has been developed by influencing the expression of starch branching enzyme 3 (SBE3) (Yang et al. 2012). Zhou et al. (2016) believed that soluble starch synthaseIII-2 ( SSIII-2) can regulate RS content and further revealed that RS yield is dependent on the high expression of the Wx a allele, which is prevalent in indica varieties. In other cereals, the down-regulation of soluble starch synthaseII-3 (SSII-3) and SBE could increase RS content in barley (Topping et al. 2003; Bird et al. 2004) and wheat (Hazard et al. 2014; Ahmed et al. 2015). Due to the molecular basis underlying the RS synthesis is largely unclear, identifying the RS genes is crucial for clarifying the RS synthesis and for the breeding of high-RS varieties.
It is generally considered that eating and cooking qualities (ECQs) are the most crucial rice qualities, which are mainly comprised of three physicochemical properties: apparent amylose content (AAC), gel consistency (GC), and gelatinization temperature (GT) (Tian et al. 2009). In recent years, because of its advantages of simple and rapid, the rapid visco analyzer (RVA) has been widely employed to evaluate the rice ECQs. The AAC shows a significantly positive correlation with RS content (Kong et al. 2015). Most of rice varieties with eminent RS content tend to have high AAC and poor taste, which is not acceptable to most consumers. Therefore, it is an important goal to study the relationships between RS and physicochemical properties in rice, which can make a foundation for obtaining rice varieties with elevated RS content and good taste. In this study, the recombinant inbred lines (RILs) were employed to study the effects of Wx and SSII-3 on RS content, and the relationships between RS and physicochemical properties were analyzed, aiming to provide useful information for molecular breeding of rice RS.

Plant Materials
In present study, CG133R (Oryza sativa ssp. indica, a restorer line) and Javanica22 (a natural mutant from javanica rice variety Xiangdali), whose SSRGs differed in ADPglucose pyrophorylase small unit ( AGPsma), ADPglucose pyrophorylase large unit ( AGPlar), Wx, SSII-3, soluble starch synthaseIII-1(SSIII-1) and soluble starch synthaseIV-2 ( SSIV-2) ( Fig. 1), were used to construct RIL. At seventh generation, a total of 99 lines with homozygous genetic background of SSRGs were identified from 142 lines and then a single plant was harvested from each line and taken as experimental materials when they were mature. All the materials were grown in an experimental agricultural field of the Southwest University of Science and Technology (Sichuan, China) and maintained normal management.

Sample preparation
Harvested rice grains were air-dried and stored at room temperature for 3 months to balance water before analyses. Rice was dehulled by using a Dehuller (TR-200, kett, Japan). The brown rice was polished by using polisher (kett, Japan) and subsequently ground to flour in a Laboratory Mill 300 (Perten, Sweden) with a 100 mesh sieve for measuring RS content and physicochemical properties.

Measurement Of Resistant Starch And Physicochemical Properties
Cooked rice was prepared according to the Zhou's method (Zhou et al. 2016) and taken a part of cooked rice stored at 4℃ for a week to product retrograded rice. In this work, RS content was measured in three sample types, including milled samples (raw rice), cooked rice and retrograded rice by using RS Rapid Assay Kit (Megazyme, International Ireland Ltd.), according to the manufacture's assay procedure. Each starch sample was measured in triplicate.
AAC and GC were measured according to the standard of Chinese Ministry of Agriculture, NY/T 2639 − 2014, and Chinese national standards, GB/T 17891 − 1999, respectively. GT was estimated by differential scanning calorimetry (DSC) according to methods reported previously . In order to measure sample percentage of retrogradation (R %), the gelatinized starch was stored at 4℃ for 7 days. Afterwards, it was equilibrated to room temperature for 1 h and then with the same thermal program as the measurement of GT. R was calculated by the following formula:  Table S1). And then, the molecular markers of differential genes between two parents were selected to detect the genotypes of 99 lines. All

8
Analysis on genetic background of materials and grouping of RIL Polymorphism of eighteen SSRGs had been detected between two parents and they differed in 6 genes (AGPsma, AGPlar, SSIII-1, SSIV-2, Wx and SSII-3). Genetic analysis on SSRGs of the 99 lines came from the RIL showed that there were different allelic genotypes for the 6 genes ( Fig. 1). Among the experimental population, statistically significant differences on RS were only detected in different Wx alleles and SSII-3 alleles (P<0.05) (Additional file 2: Table S2). So we mainly focused on the effects of Wx and SSII-3 alleles for RS content. Diverse groups were obtained according to different alleles among 99 lines, and were designated as follows: to Wx alleles: type C (came from CG133R), type J (came from Javanica 22), respectively; to SSII-3 locus: type I (same as Javanica 22) and type II (same as CG133R), respectively ( Fig. 1).

Variation of RS content of rice RIL and their parents
A t-test was implemented in order to compare the difference of parents on RS content. The result showed that the parents differed significantly from each other on the RS content in milled samples (RSm), the RS content in cooked rice (RSc) and the RS content in retrograded rice (RSr) at 0.05 level (Table 1). RS content in three sample types showed a wide variations among the RIL, especially RSm, with an average of 10.61% and ranging from 0.03% to 29.42% (Table 1). Compared with the RSm, the average of RSc and RSr decreased by 9.88% and 9.55%, respectively. The difference of RS content between milled samples and cooked rice was significant in RIL (P<0.01), but there was no statistical difference on RS content between cooked rice and retrograded rice at 0.05 level (Additional file 1: Figure S1). This fact indicated that RS contents of the different lines were decreased sharply after cooking, but no distinct variation was observed in RS content after retrogradated process. Therefore, the subtype of RS in the RIL was mainly RS2, and less RS3.  Table S4). The combined effects of Wx and SSII-3 alleles on rice RS must be further studied.
There were four combinations among the test rice lines: J-I, J-II, C-I and C-II, which came from the combination of Wx and SSII-3 alleles (Fig. 4). RSc of the combination C-I (0.89%) and C-II (1.04%) was significantly higher than that of J-I (0.061%) and J-II (0.037%).
Besides that, the significant differences on RSc between homozygotes of Wx a and wx were found under the same background of SSII-3 alleles. However, the differences on RSc between homozygotes of different SSII-3 alleles were not significant under the same background of Wx. And the same result also showed in RSr. These results indicated that Wx had an important influence on RSc and RSr under the same background of SSII-3 alleles. Furthermore, RSm of J-I (7.79%) was similar to that of C-II (5.59%), which was significantly higher than that of J-II (0.67%), but distinctly lower than C-I (21.33%) at 0.01 level, and RSa with same effect. These results revealed that different combinations of Wx and SSII-3 alleles played vital role in determining RSm and RSa.

Correlation Analysis Between RS And Physicochemical Properties In RIL
In this study, the physicochemical properties of the RIL and two parents were tested (Additional file 2: Table S5). The correlation analyses between physicochemical properties and RSm, RSc and RSr were carried out. The AAC, all DSC thermal parameters, PKV, PaT and CSV showed a positive correlation with the RSm ( Table 2). The AAC and all RVA profile parameters, except PKV, BDV and PaT, were significantly positively correlated with RSc, while all DSC thermal parameters showed significantly negative correlation with RSc (P < 0.05), and the same result also showed in RSr. Besides that, all the correlation coefficients of RSr were lower than that of RSc, except for PKV. These data suggested that RS was closely relative with many indexes of physicochemical properties and the correlation between RS and most physicochemical properties were different in raw or gelatinized rice.  Figure S1).
This result revealed that the RS3 in these materials was low. Briefly, the subtype of RS was mainly RS2 in this RILs, while RS3 was less.
Wx is the major gene controlling the amylose synthesis of rice (Tian et al. 2009). Previous studies indicated that Wx could influence the digestibility of starch and it was the major gene controlling Our study showed that SSII-3 had an influence on rice RS except for RS3 in RIL ( Fig. 3a and b), but only affected RS content in raw starch and RS2 in 18 lines (Fig. 3c and d). In 18 lines, the RSm and RSa of lines with SSII-3 allele from Javanica 22 (with a mean of 21.33% and 20.44%, respectively) were significantly higher than that from CG133R (5.56% and 4.52%, respectively) ( Fig. 3c and d), which may be caused by the structure of amylopectin and need further study. Given that Wx and SSII-3 are closely located on chromosome 6, therefore, the interactions of the two genes on RS were analyzed. Apart from RS3, the combined effects of Wx and SSII-3 greatly affected RSm, RSc, RSr and RS2 (Additional file 2: Table S4). The RSc and RSr of C-I (with a mean of 0.89% and 1.283%, respectively) and C-II (1.04% and 1.49%) were higher than that of J-I (0.06% and 0.15%) and J-II (0.04% and 0.05%). Besides that, the statistical differences on RSc were identified between Wx a and wx alleles under the same background of SSII-3 alleles, but the difference between the different SSII-3 alleles was not significant under the same Wx allele (Fig. 4b and c). This result indicated that Wx had a great impact on RS from cooked rice and retrograded rice under the same background of SSII-3 alleles.
In addition to affecting RS content, the amylose content and amylopectin fine structure also determines the physicochemical properties of rice starch. Therefore, it is necessary to analyze the found that the effect of amylose content on RS in raw rice was not as great as that on RS in cooked rice, which is consistent with our result and the correlation coefficient between AAC and RSm was smaller than that between AAC and RSc (R 2 = 0.43, 0.849, respectively, P < 0.01; Table 2). The  ) found that RS was positively correlation with PKV and BDV. In our study, a negatively correlation between RS and BDV was detected (R 2 =-0.672, P < 0.01;

Conclusions
Previous studies mainly focused on investigating RS in raw rice or cooked rice separately, which may receive different results. In present study, the RS contents of 99 lines were measured for them in different processing status. RS contents of 99 lines were sharply decreased after cooking, but no distinct variation was observed in RS content after cooling (Additional file 1: Figure S1). Thus, we deduced that the subtype of RS in present study was mostly RS2 and less RS3 according to the characteristics of RS2 and RS3. The results of genetic analysis demonstrated that Wx had a great effect on RS content and Wx a could elevate RS content in different processing rice (Fig. 2). The different SSII-3 alleles resulted in the variation of RS in raw rice and RS2 (Fig. 3). Furthermore, the combined effects of Wx and SSII-3 was responsible for the variation of RS content in three sample types and RS2. For cooked and retrograded rice, effects of Wx alleles on RS were significant different under the same background of SSII-3 alleles (Fig. 4). RS was closely relative with many indexes of physicochemical properties ( Table 2). The study shows that the RS content in rice will change by different processing ways, moreover, Wx and SSII-3 play an vital role in controlling rice RS, which may provide useful information for molecular breeding of rice RS.

Ethics Approval and Consent to Participate
Not applicable

Consent for Publication
Not applicable

Availability of Data and Materials
All data generated or analysed during this study are available within in this article or its supplementary files.

Competing Interests
The authors declare that they have no competing interests.

Authors' Contributions
HY and LX investigated genetic studies. HY, LX and OZ carried out the measure of resistant starch and physicochemical properties. XX designed the overall project. HY analyzed the data and wrote the manuscript. All authors read and approved the final manuscript.