1 Introduction

Maize (Zea mays L.) is a major staple food crop worldwide (Andorf et al. 2019). The kernels contain large amounts of starch, consisting of essentially linear amylose and highly branched amylopectin (Myers et al. 2000). The concentration and proportion of amylose and amylopectin determine the quality and usage of the maize (Glaring et al. 2006). Native starch granules in cereal endosperm usually comprise 20–30% amylose and 70–80% amylopectin (Goren et al. 2018). In waxy maize, which is grown as a food crop, especially in East and Southeast Asia, the proportion of amylopectin can be as high as almost 100% (Liu et al. 2007; He et al. 2020), whereas high-amylose varieties, with amylose proportions of more than 50%, are not only particularly valuable for the starch industry but are also used to control dietary carbohydrate intake (Cai et al. 2014; Lin et al. 2016; Wang et al. 2017). These two advantages have led to increased interest in the use of high amylose starches. High-amylose starch is widely used in numerous industrial applications, such as starch-based gels, films, textiles, gum candies and cosmetics (Zhang et al. 2020). It also has great potential for use in the food and nutraceutical fields because of its high levels of resistant starch (RS), which is refractory to gelatinization and hydrolysis, and cannot be readily digested in the upper gastrointestinal tract (Koch et al. 2010; Raigond et al. 2015; Lin et al. 2016; Zhang et al. 2016; Song et al. 2019). Resistant starch has been reported to provide many health benefits in humans, including decreasing the risk of developing diabetes and obesity (Cai et al. 2014). Therefore, maize improvement to achieve higher percentages of amylose has attracted considerable attention in recent years (Bear et al. 1958; Jeon et al. 2010; Jiang et al. 2013).

The proportion of amylose in maize endosperm starch is controlled by a series of starch synthesis genes (Liu et al. 2007; Goren et al. 2018; Li et al. 2018; Song et al. 2019; He et al. 2020). One of the most critical genes is the SBEIIb gene that encodes the starch-branching enzyme SBEIIb (one of the SBE isozymes: SBEI, SBEIIa, and SBEIIb). The full-length sequence of SBEIIb is 23,449 bp long and it contains 22 exons (Bear et al. 1958; Guan and Preiss 1993; Liu et al. 2007; Goren et al. 2018; Li et al. 2018; Song et al. 2019; He et al. 2020; Zhong et al. 2020). The amylose-extender (ae/sbe2b) locus is the structural gene for SBEIIb and the amylose proportion in its recessive mutant ae is greater than 60% (Kim et al. 1998; Fuwa et al. 1999; Li et al. 2008; Jiang et al. 2010; Liu et al. 2015). For example, the total activities of SBEs in an ae mutant decreased to about 71% of those in its wild-type (WT) control because of the deletion of the ninth exon of the SBEIIb gene, containing 84 bases (Kim et al. 1998; Fuwa et al. 1999; Li et al. 2008; Jiang et al. 2010; Liu et al. 2015). When SBEIIb is mutated, the enzyme complex, consisting of starch synthesis enzymes, may become looser or may even not form (Liu et al. 2007; Lin et al. 2016; Wang et al. 2017), leading to a decrease in amylopectin percentage and a consequent increase in amylose percentage. Thus, the amylose proportion of this line was as high as 70% (Liu et al. 2007; Lin et al. 2016; Wang et al. 2017; He et al. 2020). Furthermore, the inactivation of SBE activity and the increased amylose concentration may induce important effects on the physicochemical properties of starch. The inhibition of SBE expression tends to alter starch morphology and structure, thereby resulting in heteromorphous starch granules. For example, the starches from the ae mutant consist of approximately 7% elongated granules (Jiang et al. 2010). Likewise, starch with high-amylose concentration has been demonstrated to show enhanced resistance to gelatinization and hydrolysis (Wei et al. 2011).

Overall, we hypothesized that SBEIIb plays an important role in increasing amylose concentrations and altering physicochemical properties of starch. To test this hypothesis, we genotyped the high-amylose germplasm line GEMS-0067, released as a parent for high-amylose maize breeding by Truman State University (Campbell et al. 2007), and its wild-type (WT) control for the ae/sbe2b mutation, and performed phenotypic analysis of major agronomic traits. We additionally assessed the starch granule morphology, starch concentration and amylose proportion, gelatinization and thermal properties, and degree of starch chain polymerization through physiological and biochemical techniques, including scanning electron microscopy (SEM), light microscopy, rapid viscosity analysis (RVA), differential scanning calorimetry (DSC) and high-performance anion-exchange chromatography (HPAEC). We  provide information about the effects of mutations in key genes involved in starch synthesis, starch concentration and physicochemical properties of maize. Moreover, the results improve the understanding of SBEIIb and the physicochemical properties of high-amylose starch. Our findings may be useful for applications of high-amylose starches in industrial and maize breeding settings.

2 Material and methods

2.1 Plant material

The high-amylose GEMS-0067 line (ae mutant) and its WT control were used in this study. GEMS-0067 (Reg. no. GP-550, PI 643,420) is a partially inbred germplasm line and is currently maintained by sib-mating within the F5-derived line from the pedigree GUAT209:S13 × (H99ae × OH43ae). GEMS-0067 was released by Truman State University (Kirksville, MO, USA) for use in the development of genetically diverse, elite, Amylomaize Class VII (starch amylose > 70%) parental lines. GUAT209:S13 is a 50% tropical, exotic breeding cross, which was derived from crossing Guatemala 209 (PI 498583) to an inbred from a private cooperator designated as company 13. Only lines from GUAT209:S13 × (H99ae × OH43ae) were found to survive inbreeding and therefore used in subsequent breeding studies. Kernels of the original GUAT209:S13 × (H99ae × OH43ae) F1 seed was planted. Resulting plants were self-pollinated and ears with segregating F2 kernels (3 normal-type:1 ae-type). The WT was from segregating F2 kernels (Ae/Ae homozygous genotype from normal-type) and GEMS-0067 was from segregating F2 kernels (ae/ae homozygous genotype from ae-type) (Campbell et al. 2007).

2.2 Genotyping of the ae/sbe2b mutation

Genomic DNA was extracted separately from leaf samples of the germplasm line GEMS-0067 and the WT, using CTAB buffer (Maroof et al. 1985). The quality and quantity of DNA samples were determined on a NanoPhotometer™P-Class (Implen, Schatzbogen, Germany) and visualized by electrophoresis on 1% (w/v) agarose gel. The reference sequence of the WT SBEIIb gene (GenBank: AF072725.1) was downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/gene/?term=AF072725.1). Referring to the method of Kim et al. (1998), the genomic sequence from the 8th to 10th exon of the SBEIIb gene was amplified with a pair of specific primers (5′-CCAGCCTGGATCAAGTACTC-3′/5′-CTTGGATACAATGCAGTGCAA-3′), with the products separated by 1.5% agarose gel electrophoresis, subcloned into pMD-19-T vector, and sequenced at Sengon Biotech (Chengdu, China). The possible mutation of the ae/sbe2b gene was genotyped by alignment of the sequences between the line GEMS-0067 and the corresponding WT by using the SnapGene software (GSL Biotech LLC, San Diego, CA, USA).

2.3 Scanning electron microscopy

As described by Lin et al. (2016) and He et al. (2020), starch was isolated from mature endosperm, dried completely at 40 °C, mounted on an aluminum specimen holder using double-sided adhesive tape, sputter-coated with gold in a vacuum evaporator, and viewed under a scanning electron microscope (TM4000 Plus, Hitachi, Tokyo, Japan). The images of starch granules were obtained at an accelerating voltage of 10 kV and a magnification of ×4000.

2.4 Determination of starch concentration and percentage amylose

According to Yang et al. (2013), three replicates of kernel samples of each line (100 kernels per replicate of each line) were wetted with a calcium chloride: acetic acid solution (500 g calcium chloride were dissolved in 600 mL distilled water, with the pH value adjusted to 2.3 with acetic acid), hydrolyzed in a glycerin bath for 30 min at 119 ± 1 °C. After precipitating the protein by addition of zinc sulfate and potassium ferrocyanide solutions (30 g 100 mL−1 [w/v] zinc sulfate, 15 g 100 mL−1 [w/v] potassium ferrocyanide), the samples were adjusted to 100 mL volume with distilled water, filtered, and the rotation angle of the filtrate read in a polarimeter (MCP150, Anton Paar, Shanghai, China). Starch concentration was determined as:

$${\text{starch}}\,{\text{concentration}}~\left( {{\text{g}}/100{\text{g}}} \right) = \frac{{\alpha ~ \times ~10^{6} }}{{{\text{L}}~ \times ~{\text{W}}~\left( {100 - {\text{H}}} \right) \times ~203}}$$

where α, L, W, H, and 203 stand for the rotation angle, length (dm) of the polarizing tube, mass of sample, moisture concentration of sample, and specific rotation of starch, respectively.

The percentage amylose of the three replicates of each maize line was determined by using an amylose/amylopectin assay kit (Megazyme, Bray, Ireland), according to the manufacturer’s instructions. After eliminating the possible lipids, proteins, and amylopectin by using ethanol and concanavalin A, the remaining amylose was enzymatically hydrolyzed to D-glucose, oxidized with glucose oxidase/peroxidase (GOPOD) reagent (Gibson et al. 1997), and the absorbance at 510 nm read in a spectrophotometer (UV5Nano, Mettler Toledo, Zurich, Switzerland).

2.5 Determination of starch physicochemical properties

As described by Wang et al. (2018a), the starch samples were homogenized and passed through a 200-mesh sieve, moisture concentration was determined in a rapid moisture meter (HR83-P, Mettler Toledo, Zurich, Switzerland), a 5-g (error < 0.05 g) sample was weighed out, transferred onto aluminum foil, mixed well with 25 mL sterile water, and the viscosity determined in a rapid viscosity analyzer (RVA Super 4, Newport Scientific, Jessup, MD, USA).

As described by Wang et al. (2018a), a 10-mg homogenized and sieved starch sample was added to 30 µl sterile water, incubated at room temperature for 24 h, heated from 30 to 95 °C at a heating rate of 10 °C/min and the transition temperatures (To, Tp, and Tc) were determined in a differential scanning calorimeter (DSC) (TA Instruments, New Castle, DE, USA). The gelatinization enthalpy (ΔH) was calculated using the software of the instrument.

As described by Hanashiro et al. (1996), the starch samples were dissolved in 50 mM sodium acetate and debranched with isoamylase (Sigma, Darmstadt, Germany). The degree of polymerization (chain length distribution) was determined and analyzed by high-performance anion-exchange chromatography (HPAEC) (Thermo Fisher Scientific, Waltham, MA, USA) via pulsed amperometric detection. A Dionex™ CarboPac™ PA10 (250 × 4.0 mm I.D., 10 μm particle size) liquid chromatographic column was used, with an injection volume of 20 µl and a mobile phase A (200 mM NaOH) and mobile phase B (200 mM NaOH/200 mM NaAc). The isocratic elution conditions of the mobile phase are as follows: 0–14.0 min 90% phase A/10% B, 14.1–17.0 min 40% A/60% B, 17.1–22.0 min 90% A/10% B, the column temperature was 30 °C, the flow rate was 0.3 mL min−1 and the monosaccharide components were analyzed and detected by an electrochemical detector.

2.6 Phenotyping of agronomic traits

The field experiments comparing the two lines were conducted in the summer and autumn of May and November 2018–2019, respectively, and in the summer of May 2020. The field experiments were planted at Shunyi, Beijing, China Experimental Station (40° 23 × 74.3″ N, 116° 57 × 03.5″ E) in the summer and were planted at Huangliu, Hainan, China Experimental Station (18° 52 × 65.1″ N, 108° 83 × 81.8″ E) in the autumn. Both stations are affiliated to the Chinese Academy of Agricultural Sciences.

The GEMS-0067 and the WT were transplanted to the field in a randomized complete block design with three replicates. Both lines were isolated from genetically modified and other maize materials to avoid cross-pollination and material mixing. Seedlings of uniform size and development status were selected for transplanting to ensure the survival rate and decrease environmental variation invariables. Each line was cultivated in three rows (630 cm long and 30 cm apart) per plot. Twenty-one plants were retained in each row. The study field is a sandy soil. The N-P-K compound fertilizer (at an N:P:K ratio of 15:15:15) as a base fertilizer was applied once at 50 kg per 0.067 hectare before sowing; additional fertilizer consisting of 20 kg urea per 0.067 hectare was applied in fertilization management. Manual and chemical methods were used for weed control, and conventional local management protocols, including irrigation, for maize cultivation, were followed. Growth status and disease resistance were observed over the entire growth period. To avoid edge-row effects, only the plants in the middle row of each plot were measured with respect to anthesis stage, silking stage, growth period, plant height, and ear height. At maturity stage, the flowers were bagged to avoid any cross pollination before self-pollination, and the ears were harvested from the fifteen middle plants in the middle row of each plot, air dried, photographed, and row number per ear, kernel number per row, 100-kernel mass, and kernel mass per plant were determined for each replicate of each line. Subsequently, sample kernels were cut transversely and longitudinally, stained with I2/KI solution (0.2 g I2 and 2 g KI dissolved in 100 mL of distilled water) for 5 min as described by Hunt et al. (2013), and photographed under an optical microscope (Mshot, Guangzhou, China) at ×2 magnification.

2.7 Data analysis

The data were analyzed by Microsoft Excel 2013 (https://www.microsoft.com/) and Origin 8.0 software (https://www.originlab.com/) and presented as mean ± standard deviation (SD). The significance of the difference between the two lines was tested using Student’s t-test, using SPSS software 11.2 (IBM, Armonk, NY, USA), as each of the variables approximated to a normal distribution, as analyzed by the Shapiro-Wilk (n ≤ 2000) and Kolmogorov-Smirnov (n > 2000) tests, using SPSS software 11.2. The non-significance of these two tests indicated that the data satisfied the normal distribution for each variable, with a p-value of > 0.05.

3 Results

3.1 The ae/sbe2b gene

Genotype-specific 750-bp (GEMS-0067) and 1576-bp (WT) bands were separated from the amplified products of GEMS-0067 and its WT control, respectively (Fig. 1a). The full-length genomic sequence of the SBEIIb gene was 23,449 bp in the WT, and the structural organization of the gene in the two genotypes is shown in Fig. 1b. The results of sequencing and alignment studies showed that the entire 84 bases of the ninth exon in SBEIIb were deleted in GEMS-0067, as were 742 bases flanking regions (101-bp deletion in the intron between exons 8 and 9 and a 641-bp deletion in the intron between exons 9 and 10) (Fig. 1c).

Fig. 1
figure 1

The ae/sbe2b gene in GEMS-0067. a Specific bands separated from the amplified products of GEMS-0067 (750-bp) and WT (1576-bp). b The structural organization of the genomic sequences of the ae/SBEIIb gene in GEMS-0067 and WT. c The results of sequencing and alignment of the ae/SBEIIb gene between GEMS-0067 and WT by SnapGene software. The homologous sequences, deleted sequences, and the alignment results are marked by green, red, and gray boxes, respectively. The reference sequence of the WT SBEIIb gene (GenBank: AF072725.1) was downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/gene/?term=AF072725.1). The entire 84-bp sequence of the ninth exon in SBEIIb was deleted in GEMS-0067, together with the 742-bp flanking regions (101-bp deletion in the intron between exons 8 and 9 and a 641-bp deletion in the intron between exons 9 and 10)

Full size image

3.2 Morphology of starch granules and amylose proportion

We observed the morphology of starch granules in mature maize endosperm and determined the starch concentrations and amylose proportions of the two genotypes. The scanning electron microscopy (SEM) images showed that the starch granules isolated from WT endosperm were spherical or elliptical, were large and had smooth surfaces, whereas the starch granules of GEMS-0067 contained individual or aggregated, elongated granules with irregular depressions, as well as an increased frequency of small sub-granules, and were smaller than the normal individual granules (Fig. 2). The starch concentration and amylose proportion of GEMS-0067 on a dry mass basis were 72.63 (g 100−1 g) and 72.95%, respectively, both of which were significantly different from the 74.21 (g 100−1 g) and 26.26% values of WT, respectively (Table 1).

Fig. 2
figure 2

SEM images of starch granules of GEMS-0067 and WT. The starch granules of WT were spherical or elliptical with smooth surfaces. The starch granules of GEMS-0067 were elongated with irregular depressions, as well as an increased frequency of small sub-granules. Scale bars: 10 μm, magnification: ×4000, and accelerating voltage: 10 kV. 4000× -1, 4000× -2 and 4000× -3 represent three different fields of view at ×4000 magnification

Full size image
Table 1 Starch concentration and amylose proportion in the starch of GEMS-0067 and WT
Full size table

3.3 Gelatinization properties

Rapid viscosity analysis (RVA) is a useful tool to measure the pasting properties of maize starch. The results of RVA presented obvious differences in the viscosity curves between GEMS-0067 and WT starch (Fig. 3). In response to increasing temperature, the viscosity of WT starch gradually increased as a result of gelatinization, and presented a typical double-peak curve, whereas GEMS-0067 starch exhibited almost a horizontal straight line with only a small peak at the beginning (Fig. 3). The parameters of peak viscosity, through viscosity, break viscosity, final viscosity, and setback viscosity of GEMS-0067 starch were significantly lower than those of WT starch, whereas the peak time and gelatinization temperature values of GEMS-0067 starch were significantly higher than those of WT starch (Table 2).

Fig. 3
figure 3

Viscosity curves of GEMS-0067 and WT starch. The WT starch presented a typical double-peak curve, whereas GEMS-0067 starch demonstrated almost a horizontal straight line with a small peak at the beginning

Full size image
Table 2 Gelatinization properties of starch of GEMS-0067 and WT
Full size table

3.4 Thermal properties

Differential scanning calorimetry indicated that the phase transition temperatures, including onset temperature (To), peak temperature (Tp) and conclusion temperature (Tc), of GEMS-0067 starch were significantly higher than those of WT starch, and that the gelatinization enthalpy of GEMS-0067 starch was significantly lower than that of WT starch (Table 3). In response to increasing temperature, the differential scanning calorimetry thermogram of WT starch presented an obvious absorption peak at 70 °C, whereas the absorption peak of the GEMS-0067 starch thermogram was almost a straight line (Fig. 4) because its gelatinization enthalpy was one-sixth that of WT (Table 3).

Table 3 Thermal properties of starch of GEMS-0067 and WT
Full size table
Fig. 4
figure 4

Differential scanning calorimetry thermogram of GEMS-0067 and WT starch. The thermogram of WT starch presented an obvious absorption peak at 70 °C, whereas the GEMS-0067 starch thermogram was almost a straight line in response to increasing temperature. This result indicates that the starch of GEMS-0067 was more thermally stable than the WT starch

Full size image

3.5 Degree of polymerization

The distribution of chain fractions in debranched amylopectin was determined by HPAEC. The result of HPAEC showed that the degree of polymerization (DP) depended heavily on the proportion of amylose and amylopectin in the starch. The percentage of short chains (6 ≤ DP ≤ 24) was lower in GEMS-0067 starch than in WT starch, whereas the percentages of medium chains (25 ≤ DP ≤ 36) and long chains (DP ≥ 37) were higher in GEMS-0067 starch than in WT starch (Fig. 5).

Fig. 5
figure 5

Chain length (degree of polymerization, DP) of GEMS-0067 and WT starch. The percentage of short chains (6 ≤ DP ≤ 24) was lower, whereas the percentages of medium chains (25 ≤ DP ≤ 36) and long chains (DP ≥ 37) were higher in GEMS-0067 starch than in WT starch

Full size image

3.6 Agronomic traits

Major agronomic traits were investigated to evaluate whether the agronomic performance changed with the SBEIIb gene mutation. Over the growth period, the two lines grew normally and were not infected by any pathogen. Analysis using Student’s t-test showed that there were no significant differences between GEMS-0067 and WT in terms of the dates of anthesis or silking, duration of growth period, plant height, ear height, and row number per ear, thus indicating that the developmental traits and phenology were similar between GEMS-0067 and WT (Table 4). On the other hand, the GEMS-0067 kernels were slightly shrunken and duller, and had high amylose concentrations: the endosperm stained much darker and appeared predominantly blue after treatment with I2/KI solution for 5 min, whereas the WT kernels were predominantly red (Fig. 6). GEMS-0067 showed a significantly lower 100-kernel mass and total kernel mass per ear, compared with WT, although its kernel number per row was significantly greater than that of WT (Table 4). The single grain mass of GEMS-0067 was 0.25 g, a value 20% less than that of WT (0.3 g); similarly, the 100-kernel mass of GEMS-0067 was 85% that of the WT, whereas the kernel number per row of GEMS-0067 (28.06) was 1.98 more than that of the WT (26.08) on average.

Table 4 Comparison of agronomic traits of GEMS-0067 and WT
Full size table
Fig. 6
figure 6

Shrunken kernels and darker-stained endosperm of GEMS-0067. The kernel color of GEMS-0067 in Fig. 6a (external) and 6c (internal) was darker than that of the WT. The red color in sectioned WT kernels in Fig. 6c reflects the high amylopectin concentration of the WT starch whereas the blue color in sectioned GEMS-0067 kernels reflects the high amylose concentration in this line. a and b The kernels of GEMS-0067 were a little more shrunken and duller than WT. c and d A single maize kernel was sectioned and stained with I2/KI solution for 5 min (0.2 g I2 and 2 g KI dissolved in 100 mL of distilled water, diluted 100-fold with sterile water before use). The endosperm of GEMS-0067 was stained much darker and predominantly blue than that of WT (predominantly red). Scale bars: 3 cm (a) and 1 cm (b, c, d)

Full size image

4 Discussion

The SBEIIb gene is an endosperm-expressed gene that encodes one of the starch-branching enzyme (SBE) isozymes (SBEIIb) which play a critical role in the synthesis of amylopectin (Bear et al. 1958; Guan and Preiss 1993; Kim et al. 1998; Fuwa et al. 1999; Liu et al. 2007; Li et al. 2008, 2018; Goren et al. 2018; Song et al. 2019; He et al. 2020; Zhong et al. 2020). The enzyme complex, which is composed of starch synthesis enzymes, may become loose or even may not form with the mutation of the SBEIIb gene (Liu et al. 2007; Lin et al. 2016; Wang et al. 2017), resulting in a decrease in percentage amylopectin and a subsequent increase in percentage amylose. The recessive mutant ae/sbe2b decreased the total activities of SBEs to approximately 71% of the WT level and increased the percentage amylose of maize kernel starch up to 70%, resulting in a high-amylose maize (Campbell et al. 2007; Jiang et al. 2010; Liu et al. 2015). All the 84 bases of the ninth exon of the SBEIIb gene in GEMS-0067 were completely deleted (Fig. 1b), thereby resulting in the percentage amylose of GEMS-0067 increasing to 72.95% (Table 1), a value similar to that of the high-amylose line (He et al. 2020); these results suggested that this germplasm line was derived from the same source as the high-amylose line. Thus, the increased amylose concentration of the maize germplasm might have been due to the mutation in the SBEIIb gene.

Another important topic to be mentioned is the relationship between defective SBEIIb activity and the morphology of starch granules in the mature endosperm. The dose-effect of SBE is the main cause of the production of heteromorphous starch granules in endosperm. The starches subjected to defective SBEIIb activity from weak to strong during starch development gradually increase the number of long branched chains of amylopectin and decrease the degree of branching of amylopectin, thus resulting in formation of heteromorphous starch granules (Wellner et al. 2011; Liu et al. 2013; Chen et al. 2017). SBEIIb contains four heterogeneous starches: polygonal, aggregate, elongated and hollow starch (Wang et al. 2018a). These traits apparently arise from inhibition of SBE activity. The characteristic features of high-amylose maize starch were observed under optical microscopy and scanning electron microscopy (Figs. 2 , 6b–d; ). The kernels of GEMS-0067 were a little more shrunken, wrinkled, and duller than those of WT (Fig. 6a, b). The individual or aggregated, elongated granules with irregular depressions and small sub-granules (Fig. 2), as well as the darker staining of the endosperm by I2/KI solution (Fig. 6c, d), are phenotypes typical of the recessive mutant ae/sbe2b, findings which are consistent with previous reports (Li et al. 2008; Wellner et al. 2011; Jiang et al. 2013; Cai et al. 2014; Wang et al. 2018b). Moreover, because of the differences in physicochemical properties between amylose and amylopectin, the results of amylose concentration determination were fully consistent with the iodine staining results: the higher the amylose concentration, the darker the staining. Thus, iodine staining of the endosperm serves as an appropriate rapid, inexpensive, and non-destructive screening tool to detect starch composition. The advantages of I2/KI solution staining are that it allows for observation of the starch granules in endosperm cells in situ, and that, because the granules are sectioned, variations in staining intensity and color (i.e., amylose concentration) within granules can be discerned according to the color differences.

The SBEIIb gene mutation also significantly affects the physicochemical properties of starch (Bear et al. 1958). Therefore, further research is required to study the correlation between amylose concentration and the pasting properties, thermal properties and chain length distribution. Starch pasting properties are influenced by a variety of factors including amylose concentration, granular morphology, architecture and integrity, amylopectin structure and the composition of non-starch components (Vamadevan and Bertoft 2015). Breakdown viscosity has been reported to decrease in starch with increasing amylose concentration (Case et al. 1998). Shorter chain lengths tend to destabilize the ordered arrangements of double helices in starch granules, thereby decreasing the thermal stability, whereas the interactions of starch molecules in the granules are reinforced by an increase in amylose concentration, thus increasing the thermal stability (Vamadevan and Bertoft 2015).

The gelatinization temperature of GEMS-0067 starch was recorded as > 95 °C because it was higher than the upper limit of the heating temperature of the differential scanning calorimeter. The viscosity of starch was increased, along with its gelatinization, during gradual heating due to the absorption of water and loss of starch structure (Ji et al. 2004; Singh et al. 2010). Moreover, with the influence of SBEIIb-defective lines, the number of long chains increased whereas the branching degree of amylopectin decreased, resulting in the change of amylose concentration and amylopectin chain length distribution, which could further affect the pasting properties of the starch. In other words, the higher gelatinization temperature of GEMS-0067 starch was also related to the greater chain length (Singh et al. 2010). With reference to Hanashiro et al. (1996), in GEMS-0067 starch, the percentage of short chains was lower, whereas the percentages of medium chains and long chains were higher (Fig. 5). The significantly lower peak viscosity, through viscosity, breakdown viscosity, final viscosity, setback viscosity, and gelatinization enthalpy (Fig. 3; Tables 2 and 3), and significantly higher peak time, gelatinization temperature, and transition temperatures (To, Tp, and Tc; Tables 2 and 3), as well as the inconspicuous absorption peak (Fig. 4) of GEMS-0067 starch, implied that gelatinization initiation of GEMS-0067 starch needed much more energy to heat than the WT starch, because of the former’s high proportion of amylose (Table 1), high percentage of medium and long chains, and closer and more orderly arrangement of starch granules (Fig. 5). Moreover, because the SBEIIb gene mutation eliminated a major form of SBE in the endosperm, and because of the unique physicochemical properties of high-amylose starch, long linear chains form double helices that pack together into crystallites during gelatinization and cooling. These structures are resistant to attack by α-amylases in the upper gut. Thus, starches with high amylose concentration are generally more resistant to digestion (Man et al. 2012; Tuncel et al. 2019).

Increased amylose concentrations also affect yield. Negative correlations between percentage amylose and starch concentration, as well as kernel-filling characteristics, have been reported in most of the high-amylose germplasm lines, with high-amylose concentration traits exhibiting side effects, and causing a significant reduction in total yield, which was essentially impossible to avoid by conventional breeding (Zhao et al. 2015). In some high-amylose lines with recessive mutations of some other modifier genes (such as sbe1), the kernel-filling characteristics were less affected, but the amylose proportion was not as high as in the ae/sbe2b mutant lines (Zhang et al. 2020). Kernel mass and kernel number were also determined in the current study as measures of maize yield, the most important agronomic trait. In the present study, the shrinkage of GEMS-0067 kernels resulted in the significant decrease in its 100-kernel mass (Fig. 6; Table 4). However, the significant increase in the number of kernels per row in GEMS-0067 (28, versus 26 for WT) offset some of this loss. The kernel mass per ear of GEMS-0067 was only 8.3% lower than its WT (Table 4). A previous study had also confirmed the possibility of creating high-amylose hybrid lines without a serious reduction in total starch (Bear et al. 1958). In our study, the amylose proportion of GEMS-0067 on a dry mass basis was 72.95%, a value 178% higher than that of the WT (26.26%), whereas the single grain mass of GEMS-0067 was 0.25 g, a value 20% less than that of the WT (0.3 g). In the presence of the SBEIIb gene mutation, the grain yield decreased 17%, whereas the amylose proportion per grain biomass increased 131%; the ratio of these two indexes was 0.1298 (Table 4). Although the high-amylose concentration was negatively correlated with yield, the effect of SBEIIb gene mutation on increasing the amylose concentration far exceeded the side effect of decreasing the yield, particularly given the considerable value of amylose in the starch industry and nutraceutical fields, thereby determining the value of GEMS0067 in breeding.

In summary, the ae/sbe2b mutant is defective in the starch-branching enzyme SBEIIb, which plays an important role in maize endosperm starch synthesis. The mutations in this gene present in GEMS-0067 caused a significant increase in the amylose proportion of maize starch and also affected other traits. The present study demonstrated that the phenotypic characteristics and physicochemical properties affected, including agronomic traits, kernel morphology and glossiness, starch concentration, percentage amylose, pasting properties, thermal properties, and chain length distribution, could be largely correlated with the differences in the structure caused by the ae gene mutation. These results are in accordance with our original hypothesis. These experimental results provide information that may guide future studies on the physicochemical properties of high-amylose starch and aid in elucidating the function of the ae gene for the utilization of high-amylose starch and maize breeding. To better understand how the mutant gene affects starch synthesis, and whether enzymes encoded by the gene play important roles in the starch biosynthetic process and starch physicochemical properties, further research by our team is in progress to knock out the SBEIIb gene by CRISPR/Cas9, using GEMS-0067 as the reference material, to create additional maize germplasm with even higher percentage amylose. This future work will focus on the similarities and differences between the gene-edited maize and the GEMS-0067 line (with the ae genetic mutation) in terms of molecular, phenotypic, and physicochemical characteristics. It will be helpful for researchers to analyze the target genotypes, to facilitate the design of gene-modified plants producing high-amylose starches, and to further understand the mechanisms through which the mutant genes affect starch synthesis and physicochemical properties, in order to achieve practical applications.