Cite this Article
Chao Shufen, Cai Yicong, Feng Baobing, Jiao Guiai, Sheng Zhonghua, Luo Ju, Tang Shaoqing, Wang Jianlong, Hu Peisong, Wei Xiangjin. 2019, Editing of Rice Isoamylase Gene ISA1 Provides Insights into Its Function in Starch Formation . Rice Science 水稻科学(英文版), 26(2): 77-87.  
Doi:10.1016/j.rsci.2018.07.001
Permissions
Copyright©2019, Editorial Office of Rice Science
China National Rice Research Institute
Received:2018-06-01 Revised:2018-07-27
Editing of Rice Isoamylase Gene ISA1 Provides Insights into Its Function in Starch Formation
Chao Shufen1,,#, Cai Yicong1,,#, Feng Baobing1, Jiao Guiai1, Sheng Zhonghua1, Luo Ju1, Tang Shaoqing1,2, Wang Jianlong2, Hu Peisong1, Wei Xiangjin1
1State Key Laboratory of Rice Biology, China National Rice Research Institute, Hangzhou 310006, China
2Southern Regional Collaborative Innovation Center for Grain and Oil Crops, College of Agronomy, Hunan Agricultural University, Changsha 410128, China
#These authors contributed equally to this work
Corresponding author: WEI Xiangjin (weixiangjin@caas.cn); HU Peisong (peisonghu@126.com; hupeisong@caas.cn)
Abstract

Isoamylase 1 (ISA1) is an isoamylase-type debranching enzyme which plays a predominant role in amylopectin synthesis. In this study, the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated endonuclease 9 (CRISPR/Cas9) system was used to edit ISA1 gene in rice via Agrobacterium-mediated transformation. We identified 36 genetic edited lines from 55 T0 transgenic events, and classified the mutation forms into 7 types. Of those, two homozygous mutants, cr-isa1-1(type 1, with an adenine insertion) and cr-isa1-2(type 3, with a cytosine deletion) were selected for further analysis. Seed sizes of both cr-isa1-1and cr-isa1-2were affected, and the two mutants also displayed a shrunken endosperm with significantly lower grain weight. Electron microscopy analysis showed that abnormal starch granules and amyloplasts were found in cr-isa1-1and cr-isa1-2 endosperm cells. The contents of total starch, amylose and amylopectin in the endosperm of the cr-isa1mutants were significantly reduced, whereas sugar content and starch gel consistency were observably increased compared to the wild-type. The gelatinization temperature and starch chain length distributions of the cr-isa1mutants were also altered. Moreover, transcript levels of most starch synthesis-related genes were significantly lower in cr-isa1mutants. In conclusion, the results indicated that gene edition of ISA1affected starch synthesis and endosperm development, and brought potential implications for rice quality breeding.

Key words: ISA1; rice; starch biosynthesis; starch granule; physicochemical property; CRISPR; Cas9

Rice is a main food crop which provides essential carbohydrates and energy to nearly half of the world’ s population. Starch, as the main component (up to 90%) of a milled rice kernel, plays a crucial role in rice quality. With the progress of peopleʼ s living standards, the demand for improved rice quality has increased in recent years, and therefore, cultivating high-quality cultivars has received more attention (Rao et al, 2014).

Starch is composed of two different glucose polymers: amylose, formed by linear α -1, 4-glucosidic chains, and amylopectin, the major component (approximately 65%-85%) forming cluster structures that consist of α -1, 4-linked glucose residues highly branched through α -1, 6-linkages (Thompson, 2000). Studies have shown that there are various enzymes involved in amylose and amylopectin biosynthesis, including ADP-glucose pyrophosphorylase (AGPase), granule bound starch synthase (GBSS), soluble starch synthase (SS), starch branching enzyme (BE), starch debranching enzyme (DBE) and plastidial starch phosphorylase (Pho) (Jeon et al, 2010; Pfister and Zeeman, 2016). AGPase catalyzes an adenylyl transfer reaction with adenosine 5'-triphosphate and glucose-1-phosphate to produce ADP-Glc and pyrophosphate, which is regarded as the first rate-limiting step in starch synthesis (Haugen et al, 1976). Previous studies have shown that mutations in specific AGPase genes lead to serious decline in starch synthesis, resulting in a shrunken endosperm phenotype (Lee et al, 2007). GBSS consists of two isoforms, GBSSI which functions primarily in stored tissues (such as the seed endosperm), and GBSSII which plays a role in non-storage plant tissues (e.g. leaves) to accumulate starch temporarily (Vrinten and Nakamura, 2000). Mutation of the Waxy(Wx) gene, which encodes GBSSI, causes significant reduction in amylose content (Han et al, 2004). There are at least three SS classes in plants (SSI, SSII and SSIII), which are responsible for the elongation of the glucan chains by creating α -1, 4-linked glucose bonds (Smith et al, 1997). A mutant of SSIIIa displays a white-core floury endosperm, alters starch granules morphology, and causes changes in the proportion of the degree of polymerization (DP) values of amylopectin chains (Ryoo et al, 2007). BE has at least three isoforms in rice endosperm (BEI, BEII and BEIII), which forms branch points by introducing α -1, 6-linkages with glucose residues into α -1, 4-glucosidic chains (Mizuno et al, 1992; Nakamura et al, 1992). Phosphorylase and disproportionating enzymes (DPE) are present in plastidic (Pho1, DPE1) and cytoplasmic (Pho2, DPE2) forms. Both Pho1 and Pho2 catalyze the transfer of glucosyl units to non-reducing ends of α -glucan chains. OsPho1 and OsDpe1 assemble together to form a stable protein complex that participate in the synthesis of oligo-maltose, and play an important role in starch synthesis and degradation (Akdogan et al, 2011; Hwang et al, 2016). Genetic and biochemical analyses have been performed in DBE in Arabidopsis (Delatte et al, 2005; Wattebled et al, 2005), rice (Kubo et al, 1999), maize (James et al, 1995), barley (Burton et al, 2002), wheat (Sestili et al, 2016) and potato (Ferreira et al, 2017). Two types of DBE have been identified, isoamylase (ISA1, ISA2 and ISA3) and pullulanase (PUL) (Nakamura et al, 1996; Kubo et al, 2005). Mutations in the ISA1 gene show a dramatic change in the structure of amylopectin (James et al, 1995; Kubo et al, 1999; Burton et al, 2002; Sestili et al, 2016; Ferreira et al, 2017). Knocking down ISA1 viaRNAi in durum wheat alters the starch composition in endosperm, resulting in decreasing in starch content, increasing in contents of phytoglycogen and phytoglucan, and altering the fine structure of amylopectin (Sestili et al, 2016). Chimeric RNAi of three isoamylase genes in potato displays tissue-specific impairment in starch metabolism, which is translated into significant decrease of starch content and reduction of starch granule size in tuber but without affecting the starch in leaves (Ferreira et al, 2017). ISA1 can exist as a homo-oligomer and can also form hetero-oligomer with ISA2, which is also important in the synthesis of starch (Kawagoe et al, 2005; Utsumi and Nakamura, 2006). Kawagoe et al (2005) used an amyloplast- targeted green fluorescent protein to track amyloplasts and starch granules formation in ISA1 mutant and found no granules in the sugary endosperm of the isa1mutant, indicating that ISA1 is crucial in the early stages of starch granule formation. In addition, ISA1 is found to be directly interact with FLO6 (FLOURY ENDOSPERM 6, a CBM-domain protein that binds starch) to affect starch synthesis in developing rice seeds (Peng et al, 2014).

Although there is established information about the effect of up- and down-regulation of ISA/DBE genes on starch formation in rice and other species (James et al, 1995; Kubo et al, 1999; Burton et al, 2002; Delatte et al, 2005; Wattebled et al, 2005; Sestili et al, 2016; Ferreira et al, 2017), detailed effects on amylopectin characteristics controlled by ISA1 in rice are still incipient. To further understand this, we generated ISA1-deficient mutants (isa1) using the CRISPR/Cas9 genomic editing system. Such mutants displayed shrunken endosperm, significantly reduced amylose content and increased content of total soluble sugar, altered expression of genes associated to starch synthesis, reduced grain weight and altered starch granule morphology.

MATERIALS AND METHODS
Rice materials and growth conditions

T0 and T1 generations of cr-isa1transgenic lines generated using the clustered regularly interspaced short palindromic repeats/CRISPR-associated endo- nuclease 9 (CRISPR/Cas9) system and wild-type (WT) (Oryza sativa subsp. japonicacv. Zhonghua 11, ZH11) were grown under natural conditions in the fields of China National Rice Research Institute, Fuyang, Hangzhou, China in 2016 (T0 and WT) and 2017 (T1 and WT), respectively. ZH11 was selected as the genetic background due to its efficient transformation rate, moderate growth period and high seed-setting rate. Initial sowing was carried out on 20 May, and seedlings were transplanted into the paddy field on 15 June, with an interplant separation of 13 cm and an inter-row separation of 27 cm. Fertilizer and water managements were used as standard field production. At least five mature plants from wild-type and mutants were used to record the agronomic traits, and the seeds were used to examine the grain shape and weight. All assays consisted of three biological replicates.

Knock out ISA1 gene by CRISPR/Cas9 system

The target site (TGGACGGCGTGAGCACGATC) in the 18th exon of ISA1 (LOC_Os08g40930) was designed via CRISPR direct website (http://crispr.dbcls.jp) (Naito et al, 2015). The gRNA was controlled by the rice U6 promoter, whereas mpcas9 by the Ubiqutin promoter and the hygromycin resistance gene (Hyg) by the CaMV 35S promoter, all in the VK005 vector (Primers Target-F and Target-R, Beijing Viewsolid Biotech Co., Ltd., http://www.v-solid.com/Catalog.No.VK005-01). The Sq-primer was used to sequence the vector. The construct was introduced into wild- type rice using the Agrobacterium tumiefaciens strain EHA105 (Hiei et al, 1997). Positive transgenic lines were identified by PCR amplification of a Hyg fragment (Primers Hyg-F and Hyg-R), and the analysis of the genotypes was performed by sequencing an amplified PCR product of a specific transgenic fragment (500 bp) near the protospacer adjacent motif (PAM) sequence (with the primers Cas9-F and Cas9-R). Primer pairs for PCR amplification and sequencing data are shown in Supplemental Table 1. All the mutation forms were identified by the DSDecode website (http://dsdecode.scgene.com/home/) (Liu et al, 2015; Ma et al, 2015). Prediction of the three-dimensional protein structures was conducted using the website Phyre2 (http://www.sbg.bio.ic.ac.uk/phyre2), and the comparison of the three-dimensional protein structures between the wild-type and each mutation types was performed using the tool Swiss-PdbViewer 4.1.0 (https://spdbv.vital-it.ch/).

Microscopy analysis

Brown rice grains of wild-type and the cr-isa1mutants were cut transversely with a sharp blade, and gold was coated on the surface of the ruptured seed to prepare samples according to Kang et al (2005). The photographs were examined with a scanning electron microscope (S-3400N, Hitachi, Tokyo, Japan). Transverse sections of developing endosperms at 10 d after flowering (DAF) were performed as following: the endosperms were cut into approximately 1 mm thick sections, and then placed overnight in 2.5% glutaraldehyde fixation buffer [0.1 mol/L phosphate buffer (pH 7.2), 2% glutaraldehyde and 2% paraformaldehyde]. After dehydration in a series of progressively concentrated ethanol solution, samples were embeded into LR white resin (London Resin, Berkshire, UK, http://www.2spi.com/), sectioned with an ultramicrotome (Leica UC7; http://www.leicamicro systems.com) and then imaged with a transmission electron microscope (H-7650, Hitachi, Japan).

Analysis of starch physicochemical properties in mature grains

Total starch content of the brown rice flour was determined using a starch assay kit (Megazyme, Wicklow, Ireland; http://www.megazyme.com) according to the manufacturerʼ s instructions. First, 50 mg mature grain flour samples were washed in 5 mL of 80% ethanol to remove sugar. Amylose content was determined using the method described by Liu et al (2009). Total soluble sugar content was quantified using the phenol-sulfuric acid method (Masuko et al, 2005). Briefly, 30 mL ddH2O was added into screw-capped tubes containing about 50 mg brown rice flour. Samples were incubated in a boiling water bath for 20 min, and filtrated, and then the volume was adjusted to 100 mL. One mL each sample and a series of glucose standards were mixed with 5 mL phenol-sulfuric acid, and incubated in boiling water bath for 10 min. Total soluble sugar content was calculated with the absorbance values recorded at 620 nm using a spectrophotometer (DU800, Beckman Coulter, USA). Gel consistency and alkali values were determined and analyzed according to the China Agriculture Industry Standard NY/T147-88 (Agricultural Industry Standard of China, 2002). Thermal characteristics were measured with a Modulated Differential Scanning Calorimeter (MDSC, DSC1 STARe system, METTLER- TOLEDO) as described by Kweon et al (2000) with minor modifications. About 5 mg dry grain flour samples were placed in a sample tray, and 10 μ L ddH2O was added with gently mix. The tray was then sealed and subjected to heat treatment for 5.5 min from 35 º C to 90 º C, increasing 10 º C per minute. An empty tray was used as reference. Differences in thermal characteristics between the wild-type, cr-isa1-1and cr-isa1-2 were shown using MDSC curves. All assays were done with three biological replicates. Determination of the chain length distribution of amylopectin was performed according to Li et al (2017).

RNA extraction and quantitative real-time PCR (qRT-PCR) analysis

To investigate the expression of genes associated to starch synthesis, total RNA was extracted from 10 DAF seeds (due to most of these genes have a relatively stable and high expression levels at this stage) by using the Trizol reagent (Invitrogen, https://www.thermofisher.com). First-strand cDNA was synthesized with 2 μ g total RNA using the ReverTra Ace qPCR RT Kit (Toyobo, http://www.bio-toyobo.cn) according to the manufacturerʼ s protocol. qRT-PCR was performed using a Light Cycler 480 device (Roche, http://www.roche-applied-science.com) with the SYBR Green Real-time PCR Master Mix (Toyobo, http://www.bio-toyobo.cn) in 20 µ L reaction volume. The qRT-PCR conditions were as follows: 95 º C for 30 s, 40 cycles of 95 º C for 5 s, 60 º C for 35 s and 95 º C for 15 s. Assays included three biological replications. The gene-specific primers related to starch synthesis were those described by She et al (2010). The Actin gene (Os03g0718150, Supplemental Table 1) was used as an internal control, and the relative expression level was calculated by the 2-Δ Δ CT method (Schmittgen and Livak, 2008).

Supplemental Table 1 Primers used in this study
RESULTS
Knock out of ISA1 by CRISPR/Cas9 system

To deepen our knowledge on the function ofISA1(LOC_Os08g40930), a CRISPR/Cas9 vector containing a gRNA driven by the rice U6 promoter (Fig. 1-A) and carrying one target site for the 18th exon of ISA1 was constructed (Fig. 1-B). The plasmid was then inserted into wild-type (ZH11) calli via Agrobacterium- mediated transformation. A total of 55 T0 transgenic plants were ultimately obtained. Sequencing analysis of the ISA1genomic locus in each T0 transgenic plant was performed to determine whether the targeted mutation had occurred. Results showed that 36 independent plants presented an edited sequence near the protospacer adjacent motif (PAM) region. Six types of homozygous and one heterozygous mutations at the target site were found: an adenine (A) insertion, a cytosine (C) deletion, and other types of heterozygous mutations (Fig. 1-C). Alignment of putative amino acid sequences also showed different edited forms of ISA1 (Supplemental Fig. 1). Comparison of the predicted three-dimensional protein structure between wild-type and each type of edited sequencesvia Swiss-PdbViewer 4.1.0 indicated that there were differences in the protein between wild-type and mutants (Supplemental Fig. 2). Two transgenic lines, cr-isa1-1 (type 1 mutation form, with an adenine (A) insertion) and cr-isa1-2 (type 3, with a cytosine (C) deletion) with enough grains were selected for further phenotypic analysis (Fig. 1-C and Supplemental Fig. 2).

Fig. 1. CRISPR/Cas9 mediated editing of ISA1.
A, The structure of the T-DNA region of the Cas9/guide RNV (gRNA) vector. Marker gene Hygromycin (Hyg) was driven by the CaMV35S (35S) promoter whereas the gRNA was driven by the rice U6 promoter and the mpCas9 was driven by the Ubiquitin (Ubi) promoter. LB, Left border; RB, Right border. B, The structure of ISA1 gene. The primer pairs (Cas9-F and Cas9-R) were used to amplify the region that was sequenced in the genotypes. C, Identification of generated mutation forms in ISA1by sequencing of the target site (protospacer adjacent motif region) in T0 transgenic events. PAM, Protospacer adjacent motif. The mutation forms in all 36 T0 transgenic events can be divided into 7 categories. Homozygous mutations included type 1 to type 6, while a heterozygous mutation was classified as type 7. Type 1 and type 3 mutations were selected for further analysis.

Supplemental Fig. 1. Amino acid sequence alignment of all mutation forms.
Different background colors indicate different identities of multiple sequence alignments by DNAMAN. Residues in white with blue background means 100% identity; and those in black color with red background indicates 80% identity and those in black color with emerald blue background indicate 60% identity.

cr-isa1 mutants have shrunken endosperms and defects in seed development

The seed appearance and yield traits of wild-type plants and cr-isa1 mutants were compared. Results showed that the lengths of grain and brown rice in both cr-isa1-1 andcr-isa1-2 were increased (Fig. 2-A, -B and -H, Supplemental Fig. 3). Grain width of brown rice rather than the widths of whole seeds of those mutants were significantly reduced when compared to wild-type (Fig. 2-C and -D). The 1000-grain weight and seed thickness were also clearly decreased in the two mutants (Fig. 2-J and -K, Supplemental Fig. 3). In addition, both cr-isa1-1 andcr-isa1-2 presented a shrunken endosperm (Fig. 2-E and -F, Supplemental Fig. 3). Cross-sections of brown rice from wild-type and the two cr-isa1mutants were examined by iodine staining (Fig. 2-F, Supplemental Fig. 3). The wild-type stained uniformly with dark blue color, suggesting that the endosperm was filled with starch, whereas the endosperm fractions of cr-isa1-1 and cr-isa1-2were barely stained and only traces of blue staining were visible near the aleurone layer in the mutant endosperms. It suggested that the endosperm of two cr-isa1mutants were mainly filled with the phytoglycogen. Besides, the grain yield per plant, plant height and seed-setting rate of cr-isa1-1were also significantly decreased, but there was no obvious differences in total grain number per panicle compared with the wild-type (Fig. 2-G and -L to -O).

Fig. 2. Phenotypes of the wild-type (WT) and cr-isa1-1 mutant (T1 generation).
A, Appearance of mature seeds for WT (above) and cr-isa1-1(below). B, Appearance of brown rice for WT (above) and cr-isa1-1(below). C, Grain widths of WT (above) and cr-isa1-1(below). D, Brown rice widths of WT (above) and the cr-isa1-1transgenic line (below). E, Cross sections of WT (left) and cr-isa1-1transgenic line (right) brown rice. F, Iodine staining of brown rice in WT (left) and cr-isa1-1transgenic line (right). G, Representative plants of WT (left) and cr-isa1-1(right) after heading. H, Comparison of grain length. I, Comparison of grain width. J, Comparison of grain thickness. K, Comparison of 1000-grain weight. L, Comparison of grain yield per plant. M, Comparison of plant height. N, Comparison of total grain number per panicle. O, Comparison of seed-setting rate. Scale bars are 2 mm in A to D, 1 mm in E and F, and 10 cm in G. Values in H to O are Mean ± SD from three biological replicates, with no less than 50, 50, 50 and 200 seeds in each replication for H, I, J and K, and no less than 5 plants for L to O, respectively. Asterisks indicate statistical significance by the Student’ s t-test (* , P < 0.05; * * , P < 0.01).

Supplemental Fig. 2 The predicted three-dimensional protein structure of all mutation forms which was compared by using the tool Swiss-PdbViewer 4.1.0.
The structure in yellow color represented the wild-type and structures in other colors all stand for the lines ofcr-isa1. The red arrows indicated the differences of protein structure between wild-type and each cr-isa1lines.

cr-isa1-1 and cr-isa1-2 endosperm cells present abnormal starch granule

Scanning electron microscopy (SEM) of the cross- sections of mature endosperms showed that starch granules in the central region of wild-type endosperm were closely packed with an irregular polyhedron shape (Fig. 3-A and -B, Supplemental Fig. 4). In contrast, no obvious normal starch granules were found in the endosperm of cr-isa1-1 andcr-isa1-2 (Fig. 3-D and -E, Supplemental Fig. 4). Transmission electron microscopy (TEM) of developing endosperm cells at 10 DAF showed that amyloplasts in the wild-type were filled with polyhedral starch granules to form a typical complex structure (Fig. 3-C and Supplemental Fig. 4). However, no complete compound starch granules or even no individual starch granule were found in the centre region of endosperm cells of cr-isa1-1 and cr-isa1-2 (Fig. 3-F and Supplemental Fig. 4). These results suggested that ISA1 has essential function in the formation of starch grains and amyloplasts in the endosperm cells.

Fig. 3. Electron microscopy images of wild-type and cr-isa1-1 (T1generation).
A, Scanning electron microscopy (SEM) analysis of mature endosperm of wild-type. B, Central region of mature endosperm in the wild-type. C, Amyloplast in endosperm cells of wild-type at 10 d after flowering visualized by transmission electron microscopy analysis. D, SEM analysis of mature endosperm of the cr-isa1-1transgenic plant. E, Central region of mature endosperm in cr-isa1-1. F, Amyloplast in endosperm cells of cr-isa1-1 at 10 d after flowering visualized by transmission electron microscopy analysis. Scale bars are 0.2 mm in A and D, 10 μ m in B and E, and 1 μ m in C and F.

Supplemental Fig. 3 Phenotype of the cr-isa1-2 mutant (T1 generation).
A, Appearance of the wild-type (WT). B, Appearance of cr-isa1-2. C, Cross section of WT brown rice. D, Cross section of cr-isa1-2 brown rice. E, Iodine stained phenotype of brown rice of WT. F, Iodine stained phenotype of brown rice of cr-isa1-2. G, Representative plants of WT and cr-isa1-2 after heading. H, Grain length, I, Grain width, J, Grain thickness. K, 1000-grain weight. Scale bar: 2 mm in A-B, 1 mm in C-F and 10 cm in G. Values are Mean ± SD with three biological replicates, at least 50, 10 and 200 seeds in each replication in H, I, J and K, respectively. Asterisks in H-K indicate statistical significance compared with the WT, as determined by Student’s t-test (*P < 0.05, **P < 0.01).

Fig. 4. Starch physicochemical characteristics analysis in T1 generation of cr-isa1 mutants.
A, Total starch content. B, Amylose content. C, Amylopectin content. D, Total soluble sugar. E, Gel consistency. F, Starch solubility in KOH solution (Scale bar is 5 mm). G, Thermal characteristics presented by modulated differential scanning calorimetry (MDSC) curves. To, Onset temperature, Tp, Peak temperature; Tc, Conclusion temperature. H, Chain length distributions of amylopectin in wild-type (WT) and cr-isa1. The picture in the upper right corner shows the enlarged region with degree of polymerization ranging from 37 to 60. Data are shown as Mean ± SD (n = 3). Asterisks indicate the statistical significance by the Student’ s t-test at the 0.01 level.

Supplemental Fig. 4 Electron microscopy images of cr-isa1-2 (T1) endosperm.
A, SEM image of the central region of mature endosperm of the wild-type. B, TEM image of an amyloplast in 10 DAF endosperm cells of the wild-type. C, SEM image of the central region of mature endosperm of cr-isa1-2. D, TEM image of an amyloplast in 10 DAF endosperm cells of cr-isa1-2. Scale bar: 5m. SEM, Scanning electron microscopy; DAF, days after fertilization; TEM, Transmission electron microscopy.

cr-isa1mutants present alteration of starch physicochemical properties

Due to the abnormal starch granules in the endosperm of both cr-isa1transgenic lines, the physical and chemical properties of starch were also examined. The data showed that the total starch content in the endosperms of cr-isa1-1 andcr-isa1-2was decreased by 34.4% and 40.0%, respectively, compared to that of wild-type (Fig. 4-A). The amylose content in the wild-type was 14.0%, while there was almost no amylose detected in both cr-isa1-1and cr-isa1-2(Fig. 4-B). These results were consistent with those obtained with iodine staining of endosperm cross-sections (Fig. 2-F, Supplemental Fig. 3). The amylopectin content in the endosperm of both cr-isa1-1 andcr-isa1-2 were all significantly lower than that in the wild-type (Fig. 4-C). Otherwise, the total soluble sugar contents in cr-isa1-1 and cr-isa1-2 grains were 4.3- and 4.5-fold higher than that in the wild-type, respectively (Fig. 4-D). The gelatinization properties of the starch grains were also analyzed. First, the gel consistency of starch grains in the two cr-isa1 mutants was significantly higher than that in the wild-type (Fig. 4-E). Second, the starch solubility in KOH solution showed that the endosperm starch of cr-isa1-1 and cr-isa1-2was difficult to gelatinize, whereas the endosperm of wild-type seeds was readily dissolved and dispersed (Fig. 4-F). The thermal gelatinization temperature of the starch from wild-type and the two cr-isa1 transgenic lines were also analyzed by differential scanning calorimetry. Results showed that the onset (To), peak (Tp) and conclusion (Tc) temperatures of gelatinization were undetectable in cr-isa1-1and cr-isa1-2, but those behaved normal in the wild-type (Fig. 4-G). To further analyze differences in the fine structure of the polyglucans in endosperm, the chain length distribution was also compared between the genotypes. The short chains with degree of polymerization (DP) values less than 10 glucose units significantly increased in the mutants, whereas the proportion of chains with DP in the range of 10 to 60 was noticeably decreased. Chains with DP in the range of 50-60 were undetectable in both cr-isa1-1 and cr-isa1-2 whereas some were presented in the wild-type (Fig. 4-H, Supplemental Fig. 5). In contrast, the proportion of single glucose units (DP1) in the cr-isa1 mutants was 1.15- to 1.52-fold higher than that in the wild-type (Supplemental Fig. 5). As we know, phytoglycogen was enriched in the chain of DP less than 10 (Fujita, 2015), and therefore, the content of phytoglycogen was possibly increased in the mutants. Together, our data showed that the physicochemical properties and the fine structure of the starch in the two cr-isa1mutants were significantly altered.

Fig. 5. Relative expression levels of starch metabolism related genes in seeds at 10 d after flowering in wild-type (WT) and cr-isa1-1 (T1).
Total RNA extracted from developing seeds at 10 d after flowering was used for quantitative real-time PCR analysis. Expression level of each gene in the wild-type was set as reference value of 1. Data are Mean ± SD (n = 3). Asterisks indicate the statistical significance between wild-type and the cr-isa1-1 as determined by the Student’ s t-test at the 0.01 level.

Supplemental Fig. 5 Short chain length distributions of amylopectin in cr-isa1.
A, Chain length distributions in the range of DP 1-5 of amylopectin in WT and cr-isa1 mutants. B, Differences in the amylopectin chain length distributions in the range of DP 1-5 between cr-isa1 and the WT.

Expression of starch synthesis-related genes is affected in cr-isa1-1

Because the starch content and its physicochemical properties were dramatically changed in cr-isa1, we examined the expression of genes associated with starch synthesis in developing endosperms at 10 DAF (Fig. 5 and Supplemental Fig. 6). Results showed that the expression level of ISA1was markedly reduced in cr-isa1-1andcr-isa1-2. In addition, transcript levels of two AGP genes (AGPL1 andAGPS1), five starch synthase genes (SSI, SSIIa, SSIIIa, SSIVb andGBSSI), two starch branching enzyme genes (BEI and BEIIb), two starch debranching enzyme genes (ISA2 and PUL) and PHOL were significantly lower in cr-isa1compared to the wild-type. Conversely, the expression levels of AGPL2and ISA3 were significantly higher in cr-isa1 and cr-isa2 (Fig. 5 and Supplemental Fig. 6). Overall, compared to the wild-type, the expression of most genes associated with starch synthesis was significantly lower in cr-isa1 mutants.

Supplemental Fig. 6 Expression level of starch metabolism related genes in wild-type and cr-isa1-2 (T1) 10 DAF seeds.
Total RNA was extracted from 10 DAF developing seeds for qRT-PCR analysis. Expression level of each gene in the wild-type was set as reference value of 1. Data are displayed as means ± SD with three biological replicates. Asterisks indicate the statistical significance between wild-type and the cr-isa1-2 as determined by a Student’s t-test (*P < 0.05; **P < 0.01). DAF, Days after fertilization.

DISCUSSION

Rice quality is a complex trait that refers to the characteristics of rice grain or rice-related products including milling quality, appearance, as well as cooking, eating and nutritional qualities. Starch makes up to 90% of refined rice grains. The structure and physico- chemical properties of starch in rice endosperm are usually influenced when there are some defects in starch biosynthesis (Fitzgerald et al, 2009). In addition, starch biosynthesis and quality of rice grains can also be impacted by environmental conditions such as temperature, water and fertilizer (Thitisaksakul et al, 2012; Sun et al, 2018).

CRISPR/Cas9 system has emerged as a new technique for precise edition of genomic DNA (Jones, 2015). On one hand, mutants of several genes can be rapidly obtained by this system, which is of great significance to study gene function. On the other hand, valuable genes for breeding programs can be edited to speed up the breeding process and improve crop traits. Furthermore, the system allows to easily obtaining ʻ cleanʼ materials without the genetically modified organism (GMO) component in one or two generations. CRISPR/Cas9 system has been applied to edit theGBSSI(waxy) gene, which results in a dramatic reduction in amylose content from 14.6% to 2.6% and a phenotype similar to the natural rice waxy mutants (Ma et al, 2015). High-amylose rice is also created by editing theSBEIIb gene via the CRISPR/Cas9 system, resulting in a 25% increase in amylose content (Sun et al, 2017). Shao et al (2017) succeeded in getting badh2 fragrance mutants without the presence of the CRISPR/Cas9 vector in rice, which provides further genetic information and assistance to fragrance breeding. ISA1 is a type of DBE that plays an essential role in the synthesis of amylopectin (Nakamura et al, 1996; Smith et al, 1997). In this study, the specific target on the 18th exon of the ISA1 gene was edited through the CRISPR/Cas9 system, and a series of ISA1-deficient mutants were obtained with premature termination of the ISA1 protein (Supplemental Fig. 1). Mutation in ISA1 accompanied a decrease in expression level of ISA1mRNA (Fig. 5 and Supplemental Fig. 6), and also resulted in frame shift in the translation of protein. The predicted three-dimensional protein structures of cr-isa1 were different with that of the wild-type (Supplemental Fig. 2). Mutation in ISA1may affect its normal protein function. Compared to the wild-type, cr-isa1 mutants presented shrunken endosperm with increased grain length, but reduced grain thickness and weight (Fig. 2-E, -H, -J and -K, Supplemental Fig. 3). Iodine staining of cross-sections of brown rice from wild-type and mutants was extremely different. The endosperm of wild-type was mostly stained with dark blue, while the endosperm of mutants was hardly to stain, and only the aleurone layer can be observed blue (Fig. 2-F and Supplemental Fig. 3), which are in accordance with Kubo et al (1999).

Mutations in ISA1altered the physicochemical properties of the endosperm starch. The total starch and amylopectin contents were greatly reduced in cr-isa1 mutants (Fig. 4-A and -C). Simultaneously, there were almost no integral compound starch granules and any single starch grain present in the central region of endosperm cells of cr-isa1-1 and cr-isa1-2 (Fig. 3-D to -F and Supplemental Fig. 4). These results are similar to previous observations (Kawagoe et al, 2005; Utsumi et al, 2011). In addition, the amylose could not be detected in cr-isa1 mutants, which may due to its extremely low content (Fig. 4-B), and the gelatinization properties of starch were also changed. Particularly, there was a significant difference in the thermal starch gelatinization of cr-isa1 mutants from wild-type, showing no detectable enthalpy curve (Fig. 4-G). It also indicated that there were little starch granules gelatinized in mutants. These observations are new for a mutation of ISA1. ISA1 removes the excessive branch points or misplaced branch points in amylopectin introduced by branching enzymes and efficiently promotes the crystallization of neo- amylopectin molecules during starch biosynthesis (Streb et al, 2008). Erlander (1958) proposed that glycogen can be the direct precursor of amylopectin, and amylose can be generated by further debranching amylopectin. Defect inISA1 leads to starch to be replaced by soluble phytoglycogen (Erlander, 1958). With a high degree of branching, the structure of phytoglycogen is similar to amylopectin, but contains more short chains (DP ≤ 10) and branches at more random positions, which cannot be stained by iodine (Kubo et al, 1999; Streb et al, 2008; Fujita, 2015). Here, the short chains with a DP below 10 were significantly increased and the endosperm could not be iodine stained incr-isa1(Figs. 2-F, 4-H to -I, and Supplemental Fig. 3), which suggested that the starch may be replaced by soluble phytoglycogen in cr-isa1. Due to the high water solubility and low viscosity of phytoglycogen (Li et al, 2010), cr-isa1 can be used as a natural food thickener and stabilizer. Moreover, the total soluble sugar content in cr-isa1 mutants was approximately 4-fold higher than that in the wild-type (Fig. 4-D), which is consistent with the sugary endosperm caused by defects in ISA1 (Kawagoe et al, 2005). In addition, the mutations created in ISA1 led to lower expression level of ISA1, and the transcript levels of main starch synthesis genes were also markedly reduced (Fig. 5 and Supplemental Fig. 6). These transcriptional changes are consistent with the physicochemical properties data obtained from the cr-isa1 endosperm. The reduced expression levels of BE, SS and PHOL may weaken the ability to extend the short branches, and then generate amylopectin with more short branch and less long chains in mutant. ISA1 may have feedback physiological regulation on the expression of ISA3. However, up-regulation expression of ISA3 cannot compensate the effect of ISA1 mutation on grain starch synthesis. The considerable changes in cr-isa1mutants were caused by the mutations which likely inhibited ISA1 activity. This is similar to the antisense inhibition of ISA1described by Fujita et al (2003). Together, our results demonstrated the effect of ISA1 in seed development, especially in starch and sugar metabolism, and the structure of starch granules and grain sizes, all of which can contribute to our better understanding and applications of starch synthesis in rice quality improvement.

ACKNOWLEDGEMENTS

This study was supported by the National Nature Science Foundation of China (Grant Nos. 31471472 and 31521064), the National S& T Major Project of China (Grant No. 2016ZX08001006), and the Central Level, Non-Profit, Scientific Research Institutes Basic R and D Operations Special Fund (Grant Nos. Y2017PT46 and 2017RG002-1).

SUPPLEMENTAL DATA

The following materials are available in the online version of this article at http://www.sciencedirect.com/science/journal/16726308; http://www.ricescience.org.

Supplemental Table 1. Primers used in this study.

Supplemental Fig. 1. Amino acid sequence alignment of all mutations.

Supplemental Fig. 2. The predicted three-dimensional protein structure of all mutation forms by using the tool Swiss-PdbViewer 4.1.0.

Supplemental Fig. 3. Phenotype of the cr-isa1-2 mutant (T1 generation).

Supplemental Fig. 4. Electron microscopy images of cr-isa1-2 (T1) endosperm.

Supplemental Fig. 5. Short chain length distributions of amylopectin in cr-isa1.

Supplemental Fig. 6. Expression level of starch metabolism related genes in seeds of wild-type and cr-isa1-2 (T1) at 10 d after flowering.

The authors have declared that no competing interests exist.

Reference
[1] Akdogan G, Kubota J, Kubo A, Takaha T, Kitamura S. 2011. Expression and characterization of rice disproportionating enzymes. J Appl Glycosci, 58(3): 99-105. [Cited Within:1]
[2] Burton R A, Jenner H, Carrangis L, Fahy B, Fincher G B, Hylton C, Laurie D A, Parker M, Waite D, van Wegen S, Verhoeven T, Denyer K. 2002. Starch granule initiation and growth are altered in barley mutants that lack isoamylase activity. Plant J, 31(1): 97-112. [Cited Within:3]
[3] Delatte T, Trevisan M, Parker M L, Zeeman S C. 2005. Arabidopsis mutants Atisa1 and Atisa2 have identical phenotypes and lack the same multimeric isoamylase, which influences the branch point distribution of amylopectin during starch synthesis. Plant J, 41(6): 815-830. [Cited Within:2]
[4] Erland er S R. 1958. A proposed mechanism for the synthesis of starch from glycogen. Enzymologia, 19(5): 273-283. [Cited Within:1]
[5] Ferreira S J, Senning M, Fischer-Stettler M, Streb S, Ast M, Neuhaus H E, Zeeman S C, Sonnewald S, Sonnewald U. 2017. Simultaneous silencing of isoamylases ISA1, ISA2 and ISA3 by multi-target RNAi in potato tubers leads to decreased starch content and an early sprouting phenotype. PLoS One, 12(7): e0181444. [Cited Within:4]
[6] Fitzgerald M A, Mccouch S R, Hall R D. 2009. Not just a grain of rice: The quest for quality. Trends Plant Sci, 14(3): 133-139. [Cited Within:1]
[7] Fujita N, Kubo A, Suh D S, Wong K S, Jane J L, Ozawa K, Takaiwa F, Inaba Y, Nakamura Y. 2003. Antisense inhibition of isoamylase alters the structure of amylopectin and the physicochemical properties of starch in rice endosperm. Plant Cell Physiol, 44(6): 607-618. [Cited Within:1]
[8] Fujita N. 2015. Manipulation of rice starch properties for application. In: Nakamura Y. Starch Metabolism and Structure. Tokyo: Springer: 342-343. [Cited Within:2]
[9] Han Y P, Xu M L, Liu X Y, Yan C J, Korban S S, Chen X L, Gu M H. 2004. Genes coding for starch branching enzymes are major contributors to starch viscosity characteristics in waxy rice ( Oryza sativa L. ). Plant Sci, 166(2): 357-364. [Cited Within:1]
[10] Haugen T H, Ishaque A, Preiss J. 1976. Biosynthesis of bacterial glycogen: Characterization of the subunit structure of Escherichia coli B glucose-1-phosphate adenylyltransferase (EC 2. 7. 7. 27). J Biol Chem, 251(24): 7880-7885. [Cited Within:1]
[11] Hiei Y, Komari T, Kubo T. 1997. Transformation of rice mediated by Agrobacterium tumefaciens. Plant Mol Biol, 35: 205-218. [Cited Within:1]
[12] Hwang S K, Koper K, Satoh H, Okita T W. 2016. Rice endosperm starch phosphorylase (Pho1) assembles with disproportionating enzyme (Dpe1) to form a protein complex that enhances synthesis of malto-oligosaccharides. J Biol Chem, 291: 19994-20007. [Cited Within:1]
[13] James M G, Robertson D S, Myers A M. 1995. Characterization of the maize gene sugary1, a determinant of starch composition in kernels. Plant Cell, 7(4): 417-429. [Cited Within:3]
[14] Jeon J S, Ryoo N, Hahn T R, Walla H, Nakamura Y. 2010. Starch biosynthesis in cereal endosperm. Plant Physiol Biochem, 48(6): 383-392. [Cited Within:1]
[15] Jones H D. 2015. Regulatory uncertainty over genome editing. Nat Plants, 1(1): 14011. [Cited Within:1]
[16] Kang H G, Park S, Matsuoka M, An G. 2005. White-core endosperm floury endosperm-4 in rice is generated by knockout mutations in the C-type pyruvate orthophosphate dikinase gene ( OsPPDKB). Plant J, 42(6): 901-911. [Cited Within:1]
[17] Kawagoe Y, Kubo A, Satoh H, Takaiwa F, Nakamura Y. 2005. Roles of isoamylase and ADP-glucose pyrophosphorylase in starch granule synthesis in rice endosperm. Plant J, 42(2): 164-174. [Cited Within:3]
[18] Kubo A, Fujita N, Harada K, Matsuda T, Satoh H, Nakamura Y. 1999. The starch-debranching enzymes isoamylase and pullulanase are both involved in amylopectin biosynthesis in rice endosperm. Plant Physiol, 121(2): 399-410. [Cited Within:4]
[19] Kubo A, Rahman S, Utsumi Y, Li Z, Mukai Y, Yamamoto M, Ugaki M, Harada K, Satoh H, Konik-Rose C, Morell M, Nakamura Y. 2005. Complementation of sugary-1 phenotype in rice endosperm with the wheat isoamylase1 gene supports a direct role for isoamylase1 in amylopectin biosynthesis. Plant Physiol, 137(1): 43-56. [Cited Within:1]
[20] Kweon M, Haynes L, Slade L, Levine H. 2000. The effect of heat and moisture treatments on enzyme digestibility of AeWx, Aewx and aeWx corn starch. J Therm Anal Cal, 59: 571-586. [Cited Within:1]
[21] Lee S K, Hwang S K, Han M, Eom J S, Kang H G, Han Y, Choi S B, Cho M H, Bhoo S H, An G, Hahn T R, Okita T W, Jeon J S. 2007. Identification of the ADP-glucose pyrophosphorylase isoforms essential for starch synthesis in the leaf and seed endosperm of rice ( Oryza sativaL. ). Plant Mol Biol, 65(4): 531-546. [Cited Within:1]
[22] Li J, Huang L, Yao Y, Ma Y Q. 2010. The progress in the highly branched carbohydrate polymer phytoglycogen. Acad Period Farm Prod Proc, 4: 35-38. (in Chinese with English abstract) [Cited Within:1]
[23] Li S F, Wei X J, Ren Y L, Qiu J H, Jiao G A, Guo X P, Tang S Q, Wan J M, Hu P S. 2017. OsBT1 encodes an ADP-glucose transporter involved in starch synthesis and compound granule formation in rice endosperm. Sci Rep, 7: 40124. [Cited Within:1]
[24] Liu L L, Ma X D, Liu S J, Zhu C L, Jiang L, Wang Y H, Shen Y, Ren Y L, Dong H, Chen L M, Liu X, Zhao Z G, Zhai H Q, Wan J M. 2009. Identification and characterization of a novel waxy allele from a Yunnan rice land race. Plant Mol Biol, 71(6): 609-626. [Cited Within:1]
[25] Liu W Z, Xie X R, Ma X L, Li J, Chen J H, Liu Y G. 2015. DSDecode: A web-based tool for decoding of sequencing chromatograms for genotyping of targeted mutations. Mol Plant, 8(9): 1431-1433. [Cited Within:1]
[26] Ma X L, Chen L T, Zhu Q L, Chen Y L, Liu Y G. 2015. Rapid decoding of sequence-specific nuclease-induced heterozygous and biallelic mutations by direct sequencing of PCR products. Mol Plant, 8(8): 1285-1287. [Cited Within:2]
[27] Masuko T, Minami A, Iwasaki N, Majita T, Nishimura S I, Lee Y C. 2005. Carbohydrate analysis by a phenol-sulfuric acid method in microplate format. Anal Biochem, 339(1): 69-72. [Cited Within:1]
[28] Mizuno K, Kimura K, Arai Y, Kawasaki T, Shimada H, Baba T. 1992. Starch branching enzymes from immature rice seeds. J Biochem, 112(5): 643-651. [Cited Within:1]
[29] Naito Y, Hino K, Bono H, Ui-Tei K. 2015. CRISPRdirect: Software for designing CRISPR/Cas guide RNA with reduced off-target sites. Bioinformatics, 31(7): 1120-1123. [Cited Within:1]
[30] Nakamura Y, Takeichi T, Kawaguchi K, Yamanouchi H. 1992. Purification of two forms of starch branching enzyme (Q-enzyme) from developing rice endosperm. Physiol Plant, 84(3): 329-335. [Cited Within:1]
[31] Nakamura Y, Umemoto T, Ogata N, Kuboki Y, Yano M, Sasaki T. 1996. Starch debranching enzyme (R-enzyme or pullulanase) from developing rice endosperm: Purification, cDNA and chromosomal localisation of the gene. Planta, 199(2): 209-218. [Cited Within:2]
[32] Peng C, Wang Y H, Liu F, Ren Y L, Zhou K N, Lv J, Zheng M, Zhao S L, Zhang L, Wang C M, Jiang L, Zhang X, Guo X P, Bao Y Q, Wan J M. 2014. FLOURY ENDOSPERM 6encodes a CBM48 domain-containing protein involved in compound granule formation and starch synthesis in rice endosperm. Plant J, 77(6): 917-930. [Cited Within:1]
[33] Pfister B, Zeeman S C. 2016. Formation of starch in plant cell. Cell Mol Life Sci, 73(14): 2781-2807. [Cited Within:1]
[34] Rao Y C, Li Y Y, Qian Q. 2014. Recent progress on molecular breeding of rice in China. Plant Cell Rep, 33(4): 551-564. [Cited Within:1]
[35] Ryoo N, Yu C, Park C S, Baik M Y, Park I M, Cho M H, Bhoo S H, Ah G, Hahn T R, Jeon J S. 2007. Knockout of a starch synthase gene OsSSIIIa/Flo5 causes white-core floury endosperm in rice( Oryza sativaL. ). Plant Cell Rep, 26(7): 1083-1095. [Cited Within:1]
[36] Schmittgen T D, Livak K J. 2008. Analyzing real-time PCR data by the comparative CT method. Nat Protoc, 3(6): 1101-1108. [Cited Within:1]
[37] Sestili F, Sparla F, Botticella E, Janni M, DʼOvidio R, Falini G, Marri L, Cuesta-Seijo J A, Moscatello S, Battistelli A, Trost P, Lafiand ra D. 2016. The down-regulation of the genes encoding isoamylase 1 alters the starch composition of the durum wheat grain. Plant Sci, 252: 230-238. [Cited Within:4]
[38] Shao G N, Xie L H, Jiao G A, Wei X J, Sheng Z H, Tang S Q, Hu P S. 2017. CRISPR/CAS9-mediated editing of the fragrant gene Badh2 in rice. Chin J Rice Sci, 31(2): 216-222. (in Chinese with English abstract) [Cited Within:1]
[39] She K C, Kusano H, Koizumi K, Yamakawa H, Hakata M, Imamura T, Fukuda M, Naito N, Tsurumaki Y, Yaeshima M, Tsuge T, Matsumoto K, Kudoh M, Itoh E, Kikuchi S, Kishimoto N, Yazaki J, Ando T, Yano M, Aoyama T, Sasaki T, Satoh H, Shimada H. 2010. A novel factor FLOURY ENDOSPERM2is involved in regulation of rice grain size and starch quality. Plant Cell, 22(10): 3280-3294. [Cited Within:1]
[40] Smith A M, Denyer K, Martin C. 1997. The synthesis of the starch granule. Annu Rev Plant Physiol Plant Mol Biol, 48: 67-87. [Cited Within:2]
[41] Streb S, Delatte T, Umhang M, Eicke S, Schorderet M, Reinhardt D, Zeeman S C. 2008. Starch granule biosynthesis in Arabidopsis is abolished by removal of all debranching enzymes but restored by the subsequent removal of an endoamylase. Plant Cell, 20(12): 3448-3466. [Cited Within:2]
[42] Sun T, Tong L G, Zhao S Y, Wang H W, Han Y F, Zhang Z C, Jin Z X. 2018. Effects of nitrogen fertilizer application on starch quality, activities and gene expression levels of related enzymes in rice endosperm. Chin J Rice Sci, 32(5): 475-484. (in Chinese with English abstract) [Cited Within:1]
[43] Sun Y W, Jiao G A, Liu Z P, Zhang X, Li J Y, Guo X P, Du W M, Du J L, Francis F, Zhao Y D, Xia L Q. 2017. Generation of high-amylose rice through CRISPR/Cas9-mediated targeted mutagenesis of starch branching enzymes. Front Plant Sci, 8: 298. [Cited Within:1]
[44] Thitisaksakul M, Jimenez R C, Arias M C, Beckles D M. 2012. Effects of environmental factors on cereal starch biosynthesis and composition. J Cereal Sci, 56(1): 67-80. [Cited Within:1]
[45] Thompson D B. 2000. On the non-rand om nature of amylopectin branching. Carbohyd Polym, 43(3): 223-239. [Cited Within:1]
[46] Utsumi Y, Nakamura Y. 2006. Structural and enzymatic characterization of the isoamylase1 homo-oligomer and the Isoamylase1-isoamylase2 hetro-oligomer from rice endosperm. Planta, 225(1): 75-87. [Cited Within:1]
[47] Utsumi Y, Utsumi C, Sawada T, Fujita N, Nakamura Y. 2011. Functional diversity of isoamylase oligomers: The ISA1 homo- oligomer is essential for amylopectin biosynthesis in rice endosperm. Plant Physiol, 156(1): 61-77. [Cited Within:1]
[48] Vrinten P L, Nakamura T. 2000. Wheat granule-bound starch synthase I and II are encoded by separate genes that are expressed in different tissues. Plant Physiol, 122(1): 255-264. [Cited Within:1]
[49] Wattebled F, Dong Y, Dumez S, Delvallé D, Planchot V, Berbezy P, Vyas D, Colonna P, Chatterjee M, Ball S, DʼHulst C. 2005. Mutants of Arabidopsis lacking a chloroplastic isoamylase accumulate phytoglycogen and an abnormal form of amylopectin. Plant Physiol, 138(1): 184-195. [Cited Within:2]