MeSAUR1, Encoded by a Small Auxin-Up RNA Gene, Acts as a Transcription Regulator to Positively Regulate ADP-Glucose Pyrophosphorylase Small Subunit1a Gene in Cassava

Cassava, being one of the top three tuberous crops, features highly efficient starch accumulation in the storage root to adapt the tropical resources and environments. The molecular mechanism for the process, however, is still unclear. ADP-glucose pyrophosphorylase, the first and rate-limited enzyme in starch biosynthesis pathway, is a heterotetramer comprised of two small/catalytic and two large/modulatory subunits. To understand the regulation of MeAGPase, the promoter of a highly expressed small subunit, MeAGPs1a, was used as bait for a yeast one-hybrid assay to screen storage root cDNA library. One cDNA, coding for a small auxin-up RNA protein, named MeSAUR1, was isolated from cassava. MeSAUR1 could bind to the promoter of MeAGPS1a in yeast one-hybrid test and in vitro, and was located in cell nucleus. MeSAUR1 displayed a higher transcript level in cassava root cortex, and its expression was induced by indole-3-acetic acid, gibberellin and ethylene, but repressed by abscisic acid. A dual-luciferase interaction test further convinced that MeSAUR1 could bind to the promoter of MeAGPS1a, and positively regulate the transcription of MeAGPS1a in cassava.


INTRODUCTION
Cassava (Manihot esculenta Crantz), originated from Amazon tropical rainforest in South America, is now one of the top three tuberous crops in the global world. Cassava storage root/starch is not only the main staple for African and South American people, but also an important source for food processing and bio-energy industry. Starch is an insoluble polymer of glucose residues produced by the majority of higher plant species, and mainly stored in seeds and storage organs (Tetlow et al., 2004). ADP-glucose pyrophosphorylase (AGPase) is a rate-limited enzyme by utilizing ADP-glucose as the glucosyl donor for the elongation of α-1,4-glucosidic chains in starch biosynthesis pathway, and it consists of two small/catalytic and two large/modulatory subunits as a heterotetramer in higher plants (Ballicora et al., 2004;Saripalli and Gupta, 2015). The AGPase gene family has been characterized in several higher plants, e.g., two small subunits (SS) and several large subunits (LS) in rice and sweet potato, respectively (Akihiro et al., 2005;Zhou et al., 2016). Over-expressing SS gene or/and LS gene could enhance AGPase activity, increase seed weight and starch content (Li et al., 2011).
Auxin is an important plant hormone which can regulate cell elongation and division, root development and the formation of vascular bundle, flower and other organs. In plant, the Auxin responsive genes are mainly divided into three groups: auxin/indole-3-acetic acid (Aux/IAA), gretchenhagen-3 (GH3) and small auxin-up RNA (SAUR). The SAUR genes are the largest auxin responsive gene family, normally with the characteristics of no intron, and one or several auxin responsive elements in their promoter regions (Hagen and Guilfoyle, 2002). SAUR proteins are generally located in plasma membrane, cytoplasm and nucleus, and over-expressing SAUR genes could cause the phenotypic changes in leaf size, hypocotyl and root length, and distortion of inflorescence (Chae et al., 2012;Hou et al., 2013;Stamm and Kumar, 2013;Spartz et al., 2017). Moreover, SAUR proteins, e.g., OsSAUR39 and OsSAUR45, also acted as a negative regulator of auxin synthesis and transport in rice, possibly through the repression of OsYUCCA and OsPIN gene expression (Kant et al., 2009;Xu et al., 2017).
In this study, an Auxin responsive factor gene from cassava, named MeSAUR1, was identified through a yeast one-hybrid screening by using the promoter of MeAGPS1a as bait and the cDNA expression library of storage root as prey. The MeSAUR1 protein could bind to the bait sequence in yeast and their interaction was also confirmed in vitro. Furthermore, a dualluciferase reporter assay demonstrated that MeSAUR1 could up-regulate the promoter activity of MeAGPS1a, suggesting that MeSAUR1 protein might be a positive regulator of MeAGPS1a and involved in cassava starch biosynthesis.

Plant Material
Manihot esculenta Crantz cv.KU50 was used for construction of full-length cDNA library and gene expression analysis. The plants were cultivated in the experimental station of the Institute of Tropical Biosciences and Biotechnology, Chinese Academy of Tropical Agricultural Sciences (Chengmai County, Hainan Province, China). Eight tissues, including tip and mature leaves, flowers, petiole, stems rind, root stele, root cortex, tuber roots and primal roots were collected at 180 days after planting. Tissue cultured plantlets of KU50 were grown in MS medium, and the culture condition was 25 • C and 16/8 h in light/dark photoperiod. One-month-old plantlets were treated with 100 µM abscisic acid (ABA), or 10 mM Ethephon (2chloroethyl phosphoric acid, ET), or 100 µM Gibberellin A3 (GA3) or 10 µM 3-Indole acetic acid (IAA, Sigma, United States). Leaves were harvested at 1, 3, 6, and 12 h after hormone treatments, respectively. As control, plantlets were collected from 0 h and each sample included nine plantlets. All materials were immediately stored at −80 • C after liquid nitrogen freeze.

Yeast One-Hybrid
Yeast One-Hybrid assay was conducted according to the method described by Li et al. (2016). The MeAGPS1a promoter (1025 bp) was amplified by PCR with V1 primer pair (Table 1), and then ligated into the pAbAi vector (Clontech, United States) via the Hind III and Sal I restriction sites, constructed yeast one-hybrid bait vector pAbAi-MeAGPS1a. Cassava storage root total RNA, samples from root early development, enlargement and mature stages, was extracted using plant RNeasy extraction kit (Tiangen, CHN). The mRNA was purified using the NucleoTrap mRNA Mini Kit (Macherey Nage, GER) and the single-strand cDNA was synthesized using the RevertAid First-Strand cDNA Synthesis Kit (Fermentas, GER) following the steps provided by manufacturer. The double-stranded cDNA was amplified by long-distance PCR and size selected (>200 bp) using Chroma Spin TM TE-400 columns (Clontech, United States). Cassava double-stranded cDNAs, the prey vector pGADT7-Rec with Sma I-linearized and the bait vector pAbAi-MeAGPS1a were transformed into the yeast strain Y1HGold (Clontech, United States), then cultivated on SD/-Leu medium with 300 ng/ml Aureobasidin A (AbA) at 30 • C for 3 days. The drug-resistant colonies were identified by plasmid PCR and sequence analysis.

Bioinformatic Analyses of MeSAUR1
The amino acid sequences of SAUR gene families in cassava and Arabidopsis were downloaded from Phytozome v12 1 and NCBI 2 , respectively. A multiple alignment analysis was performed using DNAMAN software. The phylogenetic tree was constructed by MEGA 5.0 using neighbor-joining method (Tamura et al., 2011). Gene or protein accession numbers of cassava and Arabidopsis SAUR genes used in this study were listed in Supplementary  Table S1.
Expression Profiles Analysis of MeSAUR1 by Quantitative PCR Total RNA was extracted by using plant RNeasy extraction kit (Tiangen, CHN) and the first-strand cDNA was reversesynthesized using the RevertAidTM First-Strand cDNA Synthesis

Subcellular Localization
To confirm the subcelluar location of MeSAUR1, the coding sequence of MeSAUR1 was amplified with V3 primer pair which contained Bam HI and Sal I sites ( Table 1). The coding sequence of MeSAUR1was introduced into the pCAMBIA1302 vector to generate CaMV35S::MeSAUR1-GFP. Then CaMV35S::MeSAUR1-GFP and pCAMBIA1302 were transformed into Agrobacterium tumefaciens and infected onion epidermal cells. The transformed onion epidermal was cultured on MS medium in darkness at 25 • C for 24-36 h, and then observation was performed by using confocal microscopy.

MeSAUR1 Protein Expression and Extraction
The full-length coding sequence of MeSAUR1 was amplified with V2 primer pair ( Table 1). The PCR product was ligated into the pCold Pros2 vector (TaKaRa, JPN) at the site of Nde I/Sal I, generating pCold Pros2-MeSAUR1. pCold Pros2-MeSAUR1 was introduced into Escherichia coli strain BL21 (DE3) for protein expression. E. coli cells containing pCold Pros2-MeSAUR1 were cultured in LB medium supplied with 100 mg/L Ampicillin at 37 • C. When the OD 600 of the culture reaches 0.4-0.8, quickly cool the culture to 15 • C in ice water, and then 1.0 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) was added and the cultures were incubated at 15 • C, 120 rpm for 24 h. The cells were collected and resuspended in the BugBuster R Protein Extraction Reagent (Novagen, GER) and incubated at 30 • C for 1 h. The supernatant was collected and purified with Ni-Charged MagBeads (GenScript, United States). pCold Pros2 was expressed and purified as a control in accordance with the above methods.
The purified protein was verified by SDS-PAGE and Western Blotting (Sambrook and Russell, 2001). Tag Anti-ProS2 (Takara, JPN) and Goat Anti-Mouse IgG/HRP (Boster, United States) were the primary and secondary antibodies used in Western Blot, respectively.

DNA-Protein-Interaction Enzyme-Linked Immunosorbent Assay (DPI-ELISA)
DNA-protein-interaction enzyme-linked immunosorbent assay was according to the method described by Brand et al. (2010). The MeAGPS1a promoter biotinylated probe was obtained by PCR with Biotin-MeAGPS1a primer pair (Table 1). Tag Anti-ProS2 (TaKaRa, JPN) and Goat Anti-Mouse IgG/HRP (Boster, United States) were the primary and secondary antibodies used in DPI-ELISA.

Dual-Luciferase (Dual-LUC) Assay
The assay was performed according to the method of Hellens et al. (2005). LUC and REN were derived from pGreen II 0800-LUC in this study. replicates and retaining the majority of 8-10 repetitions for analysis.

Identification of MeSAUR1 in Cassava
AGPase is the rate-limiting enzyme in plant starch biosynthesis, and it is a heterotetramer, containing two small subunits and two large subunits. The amino acids sequences of Arabidopsis AGPase family genes were used to BLAST in cassava genome database.
Three genes encoded small subunits and five genes encoded large subunits were found in cassava (Supplementary Table S2).
The transcript level of MeAGPS1a was found to be the highest one in AGPase gene family among all tested tissues, and only the MeAGPS1a protein could interact with the large subunits of cassava AGPase by yeast two-hybrid assay (Supplementary Figure S1).  In order to screen certain transcription factors (TFs) which regulate MeAGPS1a, and understand more about the regulation mechanism of starch biosynthesis in cassava storage root, a yeast one-hybrid assay was performed with the MeAGPS1a promoter as bait and the storage root cDNA library as prey. Finally, over 300 positive colonies with inserted fragments longer than 750 bp were sequenced after re-checking the primary positive colonies on the same selective medium. Eight of them were annotated as candidate TFs by BLAST analysis. One of them was designated as MeSAUR1 (Manes.05G149400), coding for an Auxin responsive SAUR protein. The binding specificity between MeSAUR1 and MeAGPS1a promoter was further assayed by one to one interaction, only the yeast clones harboring MeSAUR1 and pAGPS1a-AbAi or positive control could grow on the SD/-Leu selective medium containing 300 ng/ml AbA (Figure 1). The results illustrated that MeSAUR1 was able to bind to the MeAGPS1a promoter specifically.

Molecular Characterization of MeSAUR1
The full-length of MeSAUR1 cDNA is 930 bp long, including 318 bp CDS, 125 bp 5 UTR and 488 bp 3 UTR, without any intron. It encodes a protein of 105 amino acid residues with a predicted molecular mass of 12.04 kD. The deduced protein contains a conserved SAUR-specific domain (SSD, Figure 2). The MeSAUR protein family has 132 members in cassava genome, and can be divided into three clades based on a phylogenetic tree. MeSAUR1 is located in the clade I (Figure 3). In addition, there are eight auxin responsive elements (ARE: TGTCNN, Mironova et al., 2014) in the promoter of MeSAUR1, especially two TGTCNN elements (−37, −128) locate nearby its transcription start site (Supplementary Table S3).

Expression Analysis of MeSAUR1
The transcript levels of MeSAUR1in different cassava tissues were measured by qPCR assay, and the results indicated that  The y-axis is the scale of the relative transcript expression. Error bars represent the SD of four technical replicates, the significant difference is assessed by ANOVA at P < 0.05. a-e means the significant difference at p < 0.05 level.
MeSAUR1 was highly expressed in root cortex, root stele and petiole, but relatively less expressed in other organs ( Figure 4A). Furthermore, in order to examine the response to plant hormone for MeSAUR1, the cassava tissue cultured seedlings were treated with ABA, IAA, GA, and ET, respectively. The expression level of MeSAUR1 was reduced by ABA, but induced by IAA, GA, and ET, and reached to its peak all at 3 h after treatment, then remarkably decreased 6 h afterward ( Figure 4B).

Sub-cellular Location of MeSAUR1
To determine the sub-location of MeSAUR1 protein, the green fluorescent protein (GFP) reporter gene was fused in frame to the N-terminus of MeSAUR1 and transiently expressed into onion epidermal cells by particle bombardment and observed under a fluorescent microscope. The MeSAUR1-GFP fusion protein was located in the nucleus of onion epidermal cell, while the control GFP was dispersed in the onion epidermal cell (Figure 5).
The result clearly indicated that MeSAUR1 is a nuclear-localized protein.

MeSAUR1 Can Bind to the MeAGPS1a Promoter In Vitro
In order to determine whether MeSAUR1 interacts with the MeAGPS1a promoter in vitro, the purified recombinant MeSAUR1 protein was isolated from E. coli and its molecular mass was identical to the theoretical value when a 23 kD tagged protein ProS2 was added on (Figures 6A,B). Then, we performed DPI-ELISA assay using the recombinant protein and the doublestrand biotinylated MeAGPS1a promoter DNA probes. The results suggested that MeSAUR1 bound to the MeAGPS1a promoter, as the absorbance of MeSAUR1+MeAGPS1a promoter was over fourfold higher than those of the two controls ( Figure 6C). It demonstrated that there was an interaction between MeSAUR1 and MeAGPs1a promoter in vitro.  Although we demonstrated that MeSAUR1 was able to interact with the MeAGPS1a promoter, its exact binding region was unclear. Therefore, the MeAGPS1a promoter was digested into five segments, with each segment about 200 bp, named as Seg1 to Seg5 (Figure 7A). Their PCR products were obtained with primer pairs seg1 to seg5 (Table 1). These segments were linked to pAbAi vector and transformed into Y1HGold yeast, and the SD/-Ura medium with 50 ng/ml AbA could effectively inhibit the growth of positive transformed Y1HGold yeasts. Then, the vector of pGADT7-AD-MeSUAR1 was transformed into pSeg1-AbAi to pSeg5-AbAi yeast cells, respectively, and pGADT7-AD-MeSUAR1+pSeg2-AbAi yeast cell could grow well in the Yeast cells carried pGADT7-MeSAUR1 + p(Seg1 to Seg5)-AbAi (1-5), pGADT7-Rec53 + p53-AbAi (6) and pGADT7-MeSAUR1 + pAbAi (7) were grown in SD/-Leu selective medium with 50 ng/ml AbA for 3 days at 30 • C; the three dots in each line mean three replications. (C) MeSAUR1 interacts with the Seg2 by DPI-ELISA assay. Seg2: double-strand biotinylated Seg2DNA probe + purified MeSAUR1 fusion protein; CK1: double-strand biotinylated Seg2 DNA probe + tagged protein ProS2; CK2: purified water + purified MeSAUR1 fusion protein. Error bars represent the SD of four technical replicates, the significant difference is assessed by ANOVA at P < 0.01. A,B means the significant difference at p < 0.01 level.
SD/-Leu selective medium with 50 ng/ml AbA, but other four co-transformed yeast cells could not survive in the same medium ( Figure 7B). Furthermore, we performed DPI-ELISA assay using the MeSAUR1 recombinant protein and the biotinylated Seg2 DNA probe again (Biotin-seg2, Table 1), the results indicated that MeSAUR1 could bind to the region of −400 to −201 in MeAGPS1a promoter accurately ( Figure 7C).

Activation of the MeAGPS1a Promoter by MeSAUR1
To provide evidence that MeSAUR1 protein regulates the transcription of MeAGPS1a, a dual-luciferase reporter assay was introduced. An integrated vector was created which contained 35S::MeSAUR1 and pMeAGPS1a::Luc (Figure 8A), then transformed into Agrobacterium line LBA4404 for transient expression in tobacco leaves. The relative luciferase activity of pCaMV35S::MeSAUR1-MeAGPS1a pro::LUC was higher than that of pMeAGPS1a pro::LUC (Figure 8B), indicating that MeSAUR1 could upregulate the transcript level of MeAGPS1a in cassava.

DISCUSSION
We used the promoter of MeAGPS1a as bait and the cDNA expression library of storage root as prey to perform a yeast onehybrid screening. Eight of the positive colonies were annotated as candidate TFs. This study is focused on MeSAUR1, one of the candidates, which can bind to the −400 to −201 region in MeAGPS1a promoter in vitro and in vivo, and can up-regulate the promoter activity of MeAGPS1a. SAUR gene family is auxin responsive factor, and has been reported as functional genes involved in tissue development (Park et al., 2007;  2012; Kong et al., 2013). MeSAUR1 protein is located in the cell nucleus. Combining all our results, we could conclude that MeSAUR1 is a TF despite the lack of the typical conserved domains of TFs. In addition, further mutation analysis should be used to demonstrate the specific DNA binding site of the SAUR protein.
MeSAUR1 expression can be induced by the application of exogenous auxin, GA and ET, and repressed by exogenous ABA in cassava. Interestingly, many plant hormone responsive elements are located in the promoter of MeSAUR1, e.g., eight AREs, two ERELEE4 (Rawat et al., 2005), three GAREs (Gubler and Jacobsen, 1992) and two ABREs (AATTATTA, Supplementary Table S3). It is worth mentioning that the ABRE could interact with ATHB6 protein, and the ATHB6 seemed to represent a negative regulator of ABA signal pathway (Himmelbach et al., 2002). Furthermore, other SAUR genes could also be induced or repressed by more than one type of plant hormones, for instance, AtSAUR36 was induced by auxin and repressed by GA; AtSAUR76, 77 and 78 were induced by auxin and ET, and their encoded proteins could interaction with ETR2 and EIN2 as crosstalk points between auxin and ethylene signaling pathway (Stamm and Kumar, 2013;Li et al., 2015). Therefore, MeSAUR1 might be at a crucial crosstalk point in plant hormone signaling pathways, also a mediate for plant hormone to affect starch biosynthesis in cassava.
AGPase is a key enzyme in starch biosynthesis, and overexpressing AGPase genes could increase the amount of its functional proteins, and further improve seed/root starch content and other related traits in cereal and root crops (Sweetlove et al., 1996;Tuncel and Okita, 2013). TF is a protein that can regulate the expression of genes by binding to a specific DNA sequence (Karin, 1990;Latchman, 1997), and many reports have shown that TF can function as transcriptional regulators for starch biosynthesis genes in plant. OsbZIP58 can directly bind to the promoter of six starch biosynthesis genes, OsAGPL3, Wx, OsSSIIa, OsSBE1, OsBEIIb, and OsISA2, and regulate their expression in rice (Wang et al., 2013). Some TFs have shown positive or negative co-expression pattern with starch biosynthesis genes, such as RSR1, a member of AP2/ERF family, could up-regulate starch biosynthesis genes and increase starch content in rice endosperm while it gets knocked out (Fu and Xue, 2010). Over-expression of ZmaNAC36 could up-regulate some starch biosynthesis genes, AGPL2, AGPS2, SSI, GBSSIIb, and SBEI, in maize endosperm, indicating that ZmaNAC36 is involved in the co-expression of many starch biosynthesis genes, and might play a role in regulating starch biosynthesis (Zhang et al., 2016). MeSAUR1 can bind to the MeAGPS1a promoter and activate its promoter activity, suggesting that MeSAUR1 can upregulate the transcript level of MeAGPS1a in cassava. Currently, we are over-expressing or silencing MeSAUR1 in cassava by a transgenic approach to further investigate and analyze its influence to starch biosynthesis in cassava storage root.
Cassava is an important tuberous root crop with enriched starch in its storage root, and cassava starch is a main staple for over 700 million people in the world (Zhang et al., 2010). However, the molecular mechanism of high starch accumulation in cassava is still unclear. Many studies have been focused on genetic modification of AGPase and granule bound starch synthase (GBSS), few reports are involved in regulatory mechanism of starch biosynthesis in cassava (Ihemere et al., 2006;Zhao et al., 2011;Rolland-Sabaté et al., 2013). A TF, MeSAUR1, identified in this study, can up-regulate the transcript level of MeAGPS1a. This is our first step to discover the molecular mechanism of highly efficient starch accumulation in cassava. Subsequently, a series of studies on the TF and cassava starch biosynthesis genes, MeAGPS1a, MeAGPL3, MeSus1, and GBSSI, will be carried out. Eventually, we hope to reveal the regulation system, to find the core TFs which can up-regulate the entire starch biosynthesis pathway and to improve starch yield in cassava.

CONCLUSION
One cDNA, coding for a small auxin-up RNA protein, named MeSAUR1, was isolated from cassava. MeSAUR1 could bind to the promoter in yeast one-hybrid system and in vitro, and was located in cell nucleus. The expression level of MeSAUR1 was high in root cortex, and could be up-regulated under the treatment of IAA, GA, and ET but down-regulated by ABA. MeSAUR1 could activate the promoter of MeAGPS1a in tobacco leaves, indicating that it could positively regulate the transcript level of MeAGPS1a in cassava.

AUTHOR CONTRIBUTIONS
WW and XC designed the research. PM, CLi, YM, ZX, and CLu performed the research. XC and PM wrote the paper. CZ and WW modified the paper. All authors read and approved the final manuscript.

ACKNOWLEDGMENT
Dr. Yong-Mei Bi (Department of Molecular and Cellular Biology, University of Guelph, Canada) has modified the language for the manuscript and given us many constructive suggestions.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.01315/ full#supplementary-material FIGURE S1 | The interaction between small subunit and large subunit by yeast two-hybrid assay.