Eleocharis dulcis ((N. L. Burman) Trinius ex Henschel), commonly known as Chinese water chestnut, belongs to Cyperaceae, a perennial herb in shallow water. It is native to southern China and India, with its underground corms as edible organs. It has been cultivated for more than 2000 years in China (Li et al., 2006; Changjiang Vegetables, (8): 39-43). The corm of water chestnut is crisp and juicy, containing 65~85 g water per 100 g fresh corm, 21.8 g carbohydrate, 1.5 g protein, 0.1 g fat and 0.6 g crude fiber (Kong, 2004, Hubei Science and Technology Press, pp.250-269). Carbohydrate is the main nutrient of the corm, while starch is its main storage form, which varies greatly among different varieties (Jiang et al., 2009). At least 4 enzymes are involved in plant starch biosynthesis, including ADP-glucose pyrophosphorylase (AGPase), starch synthase (SS), starch branching enzyme (BE) and debranching enzyme (DBE) (Abe et al., 2014). Among them, the activity of AGPase affects the synthesis rate of starch. It was significantly positively correlated with starch content (Min et al., 2010).

AGPase in higher plants consists of two large subunits (AGPL) and two small subunits (AGPS) to form a heterotetramer (α2β2). It is widely recognized that the small subunit plays a catalytic role while the large subunit focuses on the conformation regulation of the enzyme activity (Cheng et al., 2015). However, with the further study, the large subunit was also shown to have catalytic functions (Ventriglia et al., 2008). Due to the importance of AGPase in starch biosynthesis, many studies have focused on improving the catalytic activity of AGPase to improve starch content and yield (Tuncel and Okita, 2013). For example, overexpression of AGPase-related genes in rice, maize, wheat, potato and other crops can improve starch content in seeds or tubers (Smidansky et al., 2003; Song et al., 2005; Li et al., 2011; Kang et al., 2013).

Currently, there are few studies on starch biosynthesis in water chestnut, and only a few annotated information of genes involved in starch biosynthesis was obtained by high-throughput sequencing of leaves, corms and other tissues (Liu et al., 2015; Cheng et al., 2016; Song et al., 2019). However, there were no reports on the cloning of water chestnut AGPase gene, and the molecular biological characteristics such as sequence structure and expression pattern of water chestnut AGPase gene are still unclear. Therefore, the cDNA sequence of water chestnut AGPase gene (EdAGPL1) was cloned and bioinformatics analysis was conducted. The expression of EdAGPL1 gene in different corm development stages and tissues was detected by real-time quantitative PCR, which provided theoretical basis for studying the starch synthesis mechanism of water chestnut.

1 Results and Analysis

1.1 The gene squence encoding the large subunit gene AGPase was cloned

The cDNA obtained by reverse transcription of total RNA of water chestnut corm was used as the template for PCR amplification and the target gene fragment with length of 1 744 bp was obtained (Figure 1). Positive clones were obtained by ligating, transforming and colony PCR identification. The nucleotide sequence encoding the large subunit of water chestnut AGPase was obtained by sequencing, and it was named EdAGPL1 (GenBank Accession: MT151684).

Figure 1 PCR amplification of EdAGPL1 gene of Eleocharis dulcis Note: M: Trans5K DNA Marker; 1: The fragment of PCR

1.2 Bioinformatics analysis of EdAGPL1 gene

According to NCBI online analysis, the cDNA sequence of EdAGPL1 gene contained a 1 599 bp open reading frame encoding 532 amino acids (Figure 2A). The predicted molecular weight of EdAGPL1 protein was 58.84 kD, and the theoretical isoelectric point was 7.54. The total number of negatively charged residues (Asp+Glu) and positively charged residues (Arg+Lys) were 60 and 61, respectively. The total atomic number is 8 244, the molecular formula is C₂₅₉₅H₄₁₁₂N7₂₀O₇₉₄S₂₃; The extinction coefficient is 0.669 (assuming cysteine forms cystine) or 0.660 (assuming cysteine is reduced). The estimated half-life in vitro is 30 h, >20 h in yeast and >10 h in Escherichia coli. The instability index of the protein was 40.53, which was classified as unstable protein. The lipid index of the protein was 81.67, and the total mean hydrophilicity was -0.227. According to hydrophilic prediction, EdAGPL1 showed hydrophilicity, with the lowest negative value appearing at the 67th position of amino acid sequence (-2.922) and the highest hydrophilicity. The highest positive value occurred at the 107^th position of amino acid sequence (2.156), and the hydrophobicity was the strongest (Figure 2B). The physicochemical properties predicted above have reference significance for studying the physiology and biochemistry of the protein.

Figure 2 Bioinformatics analysis of EdAGPL1 gene Note A: EdAGPL1 gene cDNA ribotide sequence and its deduced amino acid sequence; B: Analysis of hydrophobicity / hydrophilicity of amino acid sequence of EdAGPL1; C: Prediction of EdAGPL1 signal peptide; D: Prediction of EdAGPL1 amino acid conserved domain; E: Analysis of EdAGPL1 protein domain; F: Secondary structure and solvent accessibility prediction for EdAGPL1 protein (The circle represents the DNA binding region and the rhombus represents the protein binding region; The blue represents Strand / Exposed, and the kermesinus represents helix and the yellow represents buried); G: Prediction of EdAGPL1 protein tertiary structure

The results of signal peptide prediction showed that EdAGPL1 had no signal peptide (Figure 2C). Subcellular localization indicated that the protein might be localized in cytoplasm (0.450), microbody (0.368), mitochondrial stromal space (0.360) and chloroplast thylakoid membrane (0.100).

The conserved domain of EdAGPL1 was predicted in NCBI, and it was found that the amino acid sequence of EdAGPL1 was highly conserved and similar to that of Glucose-1-phosphate adenylyltransferase, indicating that EdAGPL1 had similar functions. It belongs to the PLN02241 superfamily (Figure 2D). According to SMART online database analysis, EdAGPL1 contains one NTP_transferase domain (Ala¹⁰⁴~Glu³⁷⁹) and two PbH1 domains (Ser⁴¹⁵~Asp⁴⁴⁴, Gly⁴⁷³~Ser⁵¹⁵) (Figure 2E). The NTP_transferase domain could transfer nucleotides to sugar phosphate. PbH1 is a parallel β-helix repeat sequence, which contributes to the overall stability of proteins. Proteins with this domain are usually enzymes based on polysaccharides.

The secondary structure and solvent accessibility of EdAGPL1 were predicted (Figure 2F), and the results showed that the secondary structure was composed of helix (13.72%), strand (25.19%) and loop (61.09%). Solvent accessibility results showed that exposed accounted for 34.77%, buried accounted for 59.77%, intermediate accounted for 5.45%. It has one DNA and six protein binding regions. Phyre2 was used to model 423 residues of EdAGPL1 protein (covering 80% of the sequence) with 100.0% confidence and potato tuber AGPase (c1yp3C_) as template. The results showed that tertiary structure of the protein consists of 17 α-helices, 34 β-folds, and 51 β-turns (curls) (Figure 2G).

1.3 Homology alignment and phylogenetic analysis of EdAGPL1 amino acid sequences

The amino acid sequence of EdAGPL1 was submitted to NCBI for comparative analysis of BlastP homology. The results showed that the amino acid sequence of EdAGPL1 had low homology with AGPL protein of other plants, and the highest consistency was the amino acid sequence of AGPL of Jujube (Phoenix dactylifera L.), which was only 68.94%. DNANAN was used to perform multiple comparison analysis on the amino acid sequences of AGPL proteins in 11 plants. The results showed (Figure 3) that the homology of the amino acid sequences of the AGPL proteins in water chestnut EdAGPL1 and other plants was low, ranging from 57.54%~61.64%. The results indicated that EdAGPL was relatively conserved in the conserved domain such as NTP_transferase and PbH1, but had a large variation range in other regions of the sequence.

Figure 3 Multiple alignment of the amino acid sequence of AGPL protein between E. dulcis and other plants

MEGA 5.1 was used to construct the phylogenetic evolutionary tree of water chestnut EdAGPL1 and other plants AGPL. The results showed that 5 monocotyledons such as oil palm and jujube clustered into one group (Figure 4), 5 dicotyledons such as pomegranate and balmoma clustered into one group, and water chestnut became a separate branch. The results suggested that EdAGP had a distant evolutionary relationship with other plants, suggesting that water chestnut EdAGPL1 gene had a different evolutionary mode from other species.

Figure 4 Phylogenetic tree constructed based on amino acid sequence of EdAGPL1 protein

1.4 Expression level analysis of EdAGPL1

Fluorescence quantitative PCR results showed that EdAGPL1 gene was expressed in the roots, leaf stems, stolons and corms of water chestnut, with the highest expression level in corms, followed by leaf stems and the lowest expression level in roots (Figure 5). During corm development, the expression level of corm was higher at the early stage of corm development (90 d), then decreased and maintained a relatively stable expression level (Figure 6). The expression of EdAGPL was mainly concentrated in corm, which was the main organ of starch accumulation in water chestnut. The high expression of this gene at the early stage of corm development indicates that this gene is more active in the early stage, initiating starch synthesis.

Figure 5 Tissue-specific expression of EdAGPL1 gene in Eleocharis dulcis

Figure 6 Changes of EdAGPL1 gene expression during the corms expansion process of Eleocharis dulcis Note: S1 ~ S4 respectively represent the corm samples of transplanted field 90 d, 100 d, 110 d and 120 d

2 Discussion

In this study, EdAGPL1 cDNA sequence 1 744 bp was cloned from water chestnut, containing a 1 599 bp ORF (532 aa) encoding the large subunit of water chestnut AGPase. The molecular weight of EdAGPL1 protein was 58.84 kD and the theoretical isoelectric point was 7.54. It has typical NTP_transferase and PbH1 domains and is most likely to be located in the cytoplasm. The secondary structure was composed of helix (13.72%), strand (25.19%) and loop (61.09%). The tertiary structure consists of 17 α-helices, 34 β-folds and 51 β-turns (crimp). EdAGPL1 has a homology of 57.54%~61.64% with AGPL protein of other plants, and has a distant evolutionary relationship with other plants. The expression of EdAGPL1 gene was mainly concentrated in corm and was highest in the early stage of corm development.

A large number of studies have shown that AGPase plays a key role in starch synthesis (Abe et al., 2014). EdAGPL1 gene cloned in this study belongs to the PLN02241 superfamily, which has the function of transferring nucleotides to sugar phosphate. Involved in anabolic pathways such as starch and sucrose, amino acids and nucleotides (Jensen and Reeves, 1998). Bioinformatics analysis of EdAGPL1 protein showed that its structural characteristics and physicochemical properties were similar to that of potato sAGP protein, with NTP_transferase and PbH1 domains, indicating that EdAGPL1 had similar functions. Subsequent studies will further verify the function of this gene through transgenic technology.

The EdAGPL1 protein had low homology with AGPL proteins of other plants in GenBank database. Phylogenetic results showed that water chestnut EdAGPL1 became a separate branch and had a distant relationship with other plants. It is speculated that the EdAGPL1 gene of water chestnut has a different evolutionary mode from that of other species. Meanwhile, the large subunit of AGPase is later than the small subunit in evolution and less conservative than the small subunit. It has more variability and has great differences in different plant families and genera (Ballicora et al., 2004). Therefore, more conserved water chestnut AGPase subunit genes should be further explored for systematic analysis to clarify its evolutionary position.

Many studies have shown that AGPase-related genes have different expression patterns in different plants (Xu et al., 2011; Cheng et al., 2015; Li et al., 2018). Fluorescence quantitative analysis showed that EdAGPL1 expression was relatively concentrated in the corm, and this expression pattern had many similarities with that of bare barley and lotus root, which were all concentrated in the starch accumulation site (Cheng et al., 2015; Li et al., 2018). During corm development, EdAGPL1 was expressed at all growth stages, with the highest expression level at the early stage (90 d), and remained stable at other growth stages, which was different from the expression pattern in lotus root and bare barley. In lotus root, the expression level of AGPL gene was basically stable at the early and late expansion stages.In hull-free barley, APL gene expression peaked 15 days after flowering and then decreased (Cheng et al., 2015; Li et al., 2018). The starch accumulation of water chestnut corm increased steadily during development, while EdAGPL1 gene was the highest at the early stage of the development of corm, and maintained a relatively stable level in other periods. It is speculated that the gene plays a conformational regulation role in starch synthesis,, catalysis is completed by AGPase small subunit. The expression pattern of water chestnut AGPase small subunit should be further studied to clarify the role of water chestnut AGPase small subunit in starch synthesis.

In this study, EdAGPL1 gene was successfully cloned from water chestnut, and its bioinformatics and expression analysis were carried out, which provided a basis for research on the regulation mechanism of starch biosynthesis and breeding of high-starch varieties in the later stage.

3 Materials and Methods

3.1 Experimental materials

The experimental material 'Guilin water chestnut' was planted in the demonstration base of Nian Village, Xiuren Town, Lipu City, Guilin City, Guangxi Province in late July 2019. Samples were taken at 90, 100, 110 and 120 days after transplanting according to the development stage of water chestnut corm. The aboveground part is leaf stem, the underground part is corm, stolon and root. Liquid nitrogen is quick-frozen and stored in a refrigerator at -80℃.

3.2 RNA extraction and cDNA synthesis

Total RNA was extracted from tissue samples using the Polysaccharide polyphenol plant total RNA extraction Kit (DP441) of Tiangen Biotech (Beijing) Co., Ltd. The extraction method referred to kit instruction. The extracted total RNA was reversely transcribed into the first strand of cDNA using the TIANScript II cDNA First Strand Synthesis Kit (KR107) of Tiangen Biotech (Beijing) Co., Ltd. cDNA synthesis reaction was carried out according to the kit instruction.

3.3 Cloning of EdAGPL1 gene

Using cDNA of water chestnut corm as template, the annotation information of EdAGPL1 gene of water chestnut was obtained according to transcriptome. Specific cloning primers EdAGPL1-F (5′-ACCGCAAGATCGCTTCTCTCGTG-3′) and EdAGPL1-R (5′-ATAATGTTTGTCTCTCTGTC-3′) were designed.

PCR amplification reaction system was as follows: 5×Phusion HF Buffer 4.0 μL, dNTPs (10 mmol/L) 0.4 μL, upstream and downstream primer (10 μmol/L) 1.0 μL, Phusion DNA Polymerase (2 U/μL) 0.2 μL, cDNA 1.0 μL, supplemented by double steam water to 20.0 μL. The PCR reaction procedure was as follows: pre-denaturation at 98℃ for 30 s, 30 cycles (denaturation at 98℃ for 10 s, annealing at 55℃ for 20 s, extension at 72℃ for 1 min); Finally, it was extended at 72℃ for 5 min. The amplified products were detected by electrophoresis in 1.0% agarose gel. The pEASY-Blunt Cloning Kit of Beijing TransGen Biotech Co., Ltd. was used to connect the target clip to the pEASY-Blunt carrier, and then transform it into trans-T1 Phage Resistant chemoreceptive cells. Coated on LB (100 mg/L Amp) plate. M13 Forward Primer and M13 Reverse Primer were used for PCR positive identification of the selected monoclonal clones, and the positive clones were sent to Guangzhou IGE Biotechnology Co., Ltd. for sequencing.

3.4 Bioinformatics analysis of EdAGPL1 gene

Using NCBI ORF Finder online tools (https://www.ncbi.nlm.nih.gov/orffinder) for analysis of the open reading frame (ORF); DNAMAN 9 was used to compare the homology of amino acid sequences. The ExPASy - ProtParam (http://web.expasy.org/protparam/) and SignalP (http://www.cbs.dtu.dk/services/SignalP/) was used to predict the physical and chemical properties of EdAGPL1 and signal peptide; Using the PSORT Prediction for subcellular localization (http://psort1.hgc.jp/form.html); Using the Conserved Domain Database of NCBI for function structure domain analysis (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi); SMART database (http://smart.embl-heidelberg.de/) was used for protein domain prediction analysis. PredictPro software (https://www.predictprotein.org/home) was used to predict the secondary structure of protein. Phyre2 (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index) was used to predict the tertiary structure of protein; Finally, MEGA 5.1 was used to construct the phylogenetic evolutionary tree.

3.5 Analysis of gene expression

Primer 5.0 was used to design fluorescence quantitative primers EDAGPL2-F (5′-CAAGGGTGTAGCGGCAATCAT-3′) and EdAGPL2-R (5′-GGCGATAACATCCTCCAACTGG-3′) based on the obtained EdAGPL1 gene sequence. The reference primers CWC1-F (5′-GGCGATAACATCCTCCAACTGG-3′) and CWC1-R (5′-TCGCTCCACCAACTAAGAACGG-3′) were designed using water chestnut 18S rRNA (Registration number: MG742686.1) as the template. Fluorescence quantitative RT-PCR was performed using 2×ChamQ Universal SYBR qPCR Master Mix kit (Nanjing Vazyme Biotechnology Co., Ltd.), and the reaction system was 20 μL according to the kit instructions.

Quantitative RT-PCR reaction procedure was as follows: 95℃ for 3 min, 95℃ for 15 s, 59℃ for 15 s, 72℃ for 20 s, 45 cycles, 3 replicates per sample.

Authors’ contributions

HFL was mainly responsible for experimental operation, data analysis and writing the first draft of the paper. QZY and LLL participated in sample collection; DWQ is the project proposer and the person in charge, who is responsible for directing the experiment and the writing and revision of the paper. All authors read and approved the final manuscript.

Acknowledgement

This study was supported by the Guangxi Science and Technology Major Project (Guike AA17204045-8).

Reference

Abe N., Asai H., Yago H., Oitome N.F., Itoh R., Crofts N., Nakamura Y., and Fujita N., 2014, Relationships between starch synthase I and branching enzyme isozymes determined using double mutant rice lines, BMC Plant Biology, 14: 80

https://doi.org/10.1186/1471-2229-14-80
PMid:24670252 PMCid:PMC3976638

Ballicora M.A., Iglesias A.A., and Preiss J., 2004, ADP-glucose pyrophosphorylase:a regulatory enzyme for plant starch synthesis, Photosynth. Res., 79(1): 1-24

https://doi.org/10.1023/B:PRES.0000011916.67519.58
PMid:16228397

Cheng L.B., Li S.Y., Chen S.N., Wang Y., Yu M.Z., Chen X.H., Li L.J., and Yin J.J., 2016, Transcriptome analysis of gene expression during Chinese water chestnut storage organ formation, PLoS One, 11(10): 1-19

https://doi.org/10.1371/journal.pone.0164223
PMid:27716802 PMCid:PMC5055346

Cheng N., Zeng X.F., Zheng X.F., Diao Y., Wang Y.W., Xie K.Q., Zhou M.Q., and Hu Z.L., 2015, Cloning and characterization of the genes encoding the small and large subunit of the ADP-glucose pyrophosphorylase in lotus (Nelumbo nucifera Gaertn), Acta Physiol. Plant., DOI: 10.1007/s11738-014-1734-21734

https://doi.org/10.1007/s11738-014-1734-2

Jensen S.O., and Reeves P.R., 1998, Domain organisation in phosphomannose isomerases (types I and II), Biochimica et Biophysica Acta, 1382(1): 5-7

https://doi.org/10.1016/S0167-4838(97)00122-2

Jiang W., Li Y.R., Yang L.T., Chen L.J., and Meng P., 2009, Study on agronomic characters and nutrition of Chinese water chestnut (Eleocharis tuberosa), Zhongguo Shucai (China Vegetables), (2): 51-54

Kang G.Z., Liu G.Q., Peng X.Q., Wei L.T., Wang C.Y., Zhu Y.J., Ma Y., Jiang Y.M., and Guo T.C., 2013, Increasing the starch content and grain weight of common wheat by overexpression of the cytosolic AGPase large subunit gene, Plant Physiol. Biochem, 73: 93-98

https://doi.org/10.1016/j.plaphy.2013.09.003
PMid:24080395

Li D.M., Yang Z.M., Liu X.C., Song Z., Feng Z.Y., and He Y., 2018, Cloning and expression analysis of cDNAs encoding ADP-glucose pyrophosphorylase large and small subunits from hulless barley (Hordeum vulgare L. var. nudum), Zeitschrift für Naturforschung C, 73(5-6): 191-197

https://doi.org/10.1515/znc-2017-0154
PMid:29455192

Li N., Zhang S.J., Zhao Y.J., Li B., and Zhang J.R., 2011, Over-expression of AGPase genes enhances seed weight and starch content in transgenic maize, Planta, 233(2): 241–250

https://doi.org/10.1007/s00425-010-1296-5
PMid:20978801

Liu H.B., You Y.N., Zhu Z.X., Zheng X.F., Huang J.B., Hu Z.L., and Diao Y., 2015, Leaf transcriptome analysis and development of SSR markers in water chestnut (Eleocharis dulcis), Genetics and Molecular Research, 14(3): 8314-8325

https://doi.org/10.4238/2015.July.27.20
PMid:26345758

Min Y., Wang J., Yao Y., Hu X.W., and Guo J.C., 2010, The determination of AGPase activity and its isozymes in Cassava, Jiyinzuxue Yu Yingyong Shengwuxue (Genomics and Applied Biology), 29(2): 298-302

Smidansky E.D., Martin J.M., Hannah L.C., Fischer A.M., and Giroux M.J., 2003, Seed yield and plant biomass increases in rice are conferred by deregulation of endosperm ADP-glucose pyrophosphorylase, Planta, 216: 656-664

https://doi.org/10.1007/s00425-002-0897-z
PMid:12569408

Song B.T., Xie C.H., and Liu J., 2005, Expression of potato sAGP gene and its effects on contents of starch and reducing sugar of transgenic potato tubers, (Scientia Agricultura Sinica), 38(7): 1439-1446

Song M.B., Shuai L., Huang S.Q., Wu S.J., Cao X.H., Duan Z.H., Chen Z.L., and Fang F., 2019, RNA-Seq analysis of gene expression during the yellowing developmental process of fresh-cut Chinese water chestnuts, Scientia Horticulturae, 250: 421-431

https://doi.org/10.1016/j.scienta.2019.02.081

Tuncel A., and Okita T.W., 2013, Improving starch yield in cereals by overexpression of ADP glucose pyrophosphorylase: Expectations and unanticipated outcomes, Plant Science, 211: 52-60

https://doi.org/10.1016/j.plantsci.2013.06.009
PMid:23987811

Ventriglia T., Kuhn M.L., Ruiz M.T., Ribeiro-Pedro M., Valverde F., Ballicora M.A., Preiss J., and Romero J.M., 2008, Two Arabidopsis ADP-glucose pyrophosphorylase large subunits (APL1 and APL2) are catalytic, Plant Physiol., 148: 65-76

https://doi.org/10.1104/pp.108.122846
PMid:18614708 PMCid:PMC2528121

Xu W.Y., Yang R.D., and Li M.N., 2011, Transcriptome phase distribution analysis reveals diurnal regulated biological processes and key pathways in rice flag leaves and seedling leaves, PLoS One, 6(3): e17613

https://doi.org/10.1371/journal.pone.0017613
PMid:21407816 PMCid:PMC3047585