Identification of a novel submergence response gene regulated by the Sub1A gene

Submergence is one of the major constraints to rice production in many rice growing areas in the world. The Sub1A gene has been demonstrated to dramatically improve submergence tolerance in rice. Here, we report the identification of a novel submergence response (RS1) gene that is specifically induced in the Sub1A-mediated submergence tolerance response. Under submergence, RS1 was upregulated in M202 (Sub1A) but downregulated in M202 in RNA-seq and microarray assays. Expression analyses of various tissues and developmental stages show that RS1 mRNA levels are high in leaves and sheaths, but low in roots, stems, and panicles. Our results also show that RS1 is highly expressed under submergence, drought, and NaCl stresses, but not under cold or dehydration stress. Hormone ABA treatment induces, whereas GA treatment decreases, RS1 expression. The RS1 and Sub1A genes are co-regulated under submergence. Overexpression of RS1 in transgenic Kitaake (without Sub1A) and M202(Sub1A)×Kitaake do not result in enhanced submergence tolerance. Conversely, down-regulation of RS1 in M202(Sub1A)×Kitaake lead to weaken submergence tolerance. We hypothesize that RS1 may play a role in the Sub1A-mediated submergence tolerance pathway.


INTRODUCTION
Submergence is one of main environmental stresses to rice growth and productivity in large rice-growing areas, especially in the flood-prone rainfed lowlands in South and Southeast Asia, where it regularly affects about 15 to 20 million hectares of rice land. Although rice is well adapted to aquatic environments, submergence causes annual losses of over US $1 billion in Asia (Xu et al., 2006). Fortunately, some indica cultivars, such as FR13A, can survive 10 to 14 days of complete submergence and renew growth after the floods recede. Sub1A was identified as the major gene conferring submergence tolerance in FR13A (Xu et al., 2006). Sub1A, an AP2/ERF transcription factor, was mapped on chromosome 9 and subsequently isolated from FR13A. Two alleles of Sub1A, *Corresponding author. E-mail: duhewei666@163.com. Tel: 86-716-8066182.
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License Sub1A-1 and Sub1A-2, exist that differ only by a SNP variation at position 556. Allele Sub1A-1 confers submergence tolerance while null allele Sub1A-2 loses the function (Xu et al., 2006). Sub1A limits ethylenepromoted GA responsiveness during submergence by augmenting accumulation of the GA signaling repressors SLR1 and SLRL1 (Fukao and Bailey-Serres, 2008). Sub1A also enhances the expression levels of PDC and ADH1, thus providing the ATP needed for rice survival under submergence (Perata and Voesenek, 2007).
Studies on Sub1A have resulted in significant advancement in understanding submergence tolerance in rice. However, submergence tolerance is a complex trait orchestrated by many genes (Xu et al., 2006;Fukao and Bailey-Serres, 2008). The differences in submergence tolerance between different rice cultivars are only partly resolved since the identification of the Sub1A gene (Perata and Voesenek, 2007). For example, FR13A and Goda Heenati are two indica cultivars that carry the Sub1A-1 gene. FR13A is a submergence tolerant cultivar, which can survive approximately two weeks under submergence whereas Goda Heenati does not (Xiong et al., 2012). The difference of submergence tolerance between FR13A and Goda Heenati indicate that other genes may interact with Sub1A-1 that are necessary for Sub1A-1 to confer submergence tolerance. Towards this goal, genes differentially expressed between FR13A and Goda Heenati under submergence, including two novel submergence responsive genes (Os09g0269900 and Os12g0202700), were identified using suppression subtractive hybridization and microarray approaches (Xiong et al., 2012). In addition, Seo et al. (2011) identified 24 proteins that interacted with Sub1A-1 through yeast two-hybrid (Y2H) screens. However, the Sub1A-1 gene is not sufficient for rice to tolerate more than two weeks under submergence (Septiningsih et al., 2012). Therefore, it is necessary to identify novel genes involved in Sub1A-mediated tolerance in order to investigate the molecular mechanism and further improve submergence tolerance.
RNA-seq technology has provided an effective way to obtain transcript sequences and abundance at the genomic scale, facilitating the identification of genes differentially expressed between different varieties (Grabherr et al., 2011;Trapnell et al., 2010). M202 (Sub1A) is an inbred rice line carrying the Sub1A gene, which leads to enhanced tolerance to submergence. Here, we performed RNA-seq to identify differentially expressed genes in M202(Sub1A) under submergence. We selected 38 increased genes and further tested their differential expression patterns under submergence both in M202 (a rice variety without Sub1A gene, and intolerance to submergence) and M202(Sub1A) by qRT-PCR. We identified a novel gene responsive to submergence, called RS1. The expression patterns of RS1 in various tissues were also determined. The responses of RS1 to various abiotic stresses, such as submergence, drought, cold, dehydration, and high NaCl, and to hormones, including ACC, GA, ABA, were assessed. However, overexpression of RS1 did not result in significant improvement of submergence tolerance in transgenic Kitaake (a rice inbred line without Sub1A-1 gene). In M202(Sub1A)×Kitaake background, the seedling height was significant elongated with the downregulation of RS1, indicating that the down-regulation of RS1 resulted in weakened waterlogging tolerance. These results will help to explore the molecular mechanism of submergence tolerance in rice and facilitate breeding of submergence tolerant varieties.

Plant growth and submergence treatment
Dehusked seeds were treated with 70% ethanol for 1 min, soaked in 1% sodium hypochlorite for 15 min, and rinsed with sterile water 4 to 5 times. Sterilized seeds were transferred to MS medium and grown in a chamber at 28°C for one week. Six one-week-old seedlings were planted at a silica sand pot (18 cm×16 cm), and were grown in the greenhouse at 25 to 28°C in a 16 h/8 h light-dark cycle. Three-week-old M202(Sub1A) seedlings were subjected to submergence for 0 or 3 days. Approximately 0.5 g leaves from six seedlings were detached. The leaf samples were ground to a fine power in the presence of liquid nitrogen. Total RNA was isolated using the Trizol reagent (Invitrogen, Carlsbad, CA, USA). All the experiments were performed with three biological replications. In total, 6 RNA samples were used for RNA-seq at BGI.
In order to accurately measure the increased seedling height under submergence, six sterilized seeds were planted into the tube (30 cm×2 cm). Six tubes each genotype were selected, and grown in the growth chamber at 25 to 28°C in a 16 h/ 8 h light-dark cycles. All experiments were performed with three biological replications. Seedlings of approximately 3 to 4 cm height were subjected to submergence treatment. Seedling heights were recorded every 24 h during submergence.

Identification of RS1
Under submergence, the identification of differentially expressed genes in M202 (Sub1A) was conducted according to Reiner et al. (2002). In previous research, our groups conducted microarray assays to identify the differentially expressed genes between M202 and M202 (Sub1A) under submergence (Jung et al., 2010). We combined the RNA-seq and microarray results and selected 38 differentially expressed genes that were upregulated in M202 (Sub1A) compared to M202 upon submergence treatment. The expression levels of these 38 genes under submergence were determined by qRT-PCR for both M202 and M202(Sub1A). Under submergence, the expression level of RS1 was increased in M202 (Sub1A) but decreased in M202. Therefore, RS1 was chosen as a candidate gene responsive to submergence.

Various environmental stresses and hormone treatment
All submergence, drought and hormone treatments were conducted with at least 3 independent biological replications. For gibberellin (GA), abscisic acid (ABA) and 1-aminocyclopropane-1-carboxylic acid (ACC) treatments, M202 and M202 (Sub1A) seedlings of 14 days old grown in germination paper were transferred into solutions containing mock (0.1% DMSO), GA (10 µM), ABA (10 µM), and ACC (10 µM), respectively, and incubated for 6 h. Submergence treatment was carried out as described by Fukao and Bailey-Serres (2008). M202 and M202 (Sub1A) of 35 days old were subject to 0 or 5 days of drought stress in greenhouse. Seedlings of 14 days old were treated with stresses including dehydration (the intact seedlings were exposed in the air without water supply), salinity stress (using 150 mM NaCl solution), and cold exposure at 4°C (Xiong et al., 2014).

RNA isolation and expression analysis
All experiments were performed with three biological replications. Approximately 0.2 g tissues selected from six seedlings were used for RNA isolation. Leaves were selected from various treatments, including submergence, drought, cold, dehydration, NaCl, GA, ACC, and ABA. In addition, various tissues, including leaves, roots, shoots, stems, and panicles at different growth stages were also collected. First strand cDNA was synthesized using SuperScript-II reverse transcriptase according to the manufacturer's instructions (Invitrogen). The actin 1 gene (Os03g50890.1) was used as an endogenous control to normalize expression data (Supplement Table 1). The qRT-PCR primers specific for the RS1 and Sub1A genes are listed on Supplementary Table 1. Real-time PCR was conducted using the SYBR real-time PCR kit (Takara Japan) with IQTM SYBR® Green Supermixture according to the manufacturer's instructions (Bio-Rad USA). The reaction conditions were as follows: 94°C for 1 min; followed by 40 cycles of 95°C for 10 s, and then 55°C for 10 s.

Binary construct and rice transformation
The full coding sequence (CDS) of the RS1 gene (Os02g0134200) encoding 255 amino acids (aa) derived from M202(Sub1A) was isolated by RT-PCR and transferred into the pCAMBIA1300 binary vector under control of the ubiquitin (Ubi-1) promoter to generate the RS1 overexpression construct (RS1ox). The RS1 RNAi construct (part sequence of RS1 cDNA) under control of the CaMV35s promoter was transferred into the pBWA(V)HS vector by Wuhan Biorun Biotechnology company. The RS1ox and RS1 RNAi constructs were transformed into Kitaake (a rice inbred line), respectively. Partial maps of the constructs including the loop sequence and the RS1 sequence in the RNAi construct are shown in Supplementary Figure 1. Rice transformation was also conducted by Wuhan Biorun Biotechnology Company.

Differentially expressed genes specific to Sub1Amediated submergence tolerance were identified
Differentially expressed genes were identified through comparing their expression levels based on RPKM (reads per kilobase of exon model per million mapped reads). The transcriptomic profiles of M202 (Sub1A) were determined by comparing their gene expression levels under normal condition and submergence treatment. Under submergence, 3712 genes were differentially expressed in M202 (Sub1A); 1696 of the 3712 genes were decreased and 2016 genes were increased ( Figure  1). Our previous study using microarray has identified genes differentially expressed between M202 and Du et al. 2745 M202 (Sub1A) under submergence (Jung et al., 2010).
Here, we combined the RNA-seq and microarray results and selected 38 genes that increased in M202(Sub1A). Subsequently, their expression levels both in M202 (Sub1A) and M202 under submergence were determined by qRT-PCR (data no shown). Of particular interest, the expression level of RS1 increased in M202 (Sub1A) but decreased in M202 (Figure 2) upon submergence. Under 1, 3, and 6 days submergence, the expression level of RS1 in M202 (Sub1A) is approximate 2, 6, and 5 folds compared with that in M202 separately ( Figure 2). Therefore, we chose RS1 as a candidate gene for possible involvement in Sub1A-mediated submergence tolerance.

The RS1 gene responds to drought and high salt stresses
We assessed changes of RS1 expression levels by qRT-PCR when subject to various abiotic stresses, including drought, cold, dehydration, and high NaCl. The results show that the RS1 expression levels both in M202 and M202(Sub1A) increased upon drought treatment, peaking at day 7 ( Figure 3a). Under cold and dehydration stresses, the RS1 expression levels remained unchanged (Figure 3b), suggesting that RS1 does not respond to cold or dehydration stress. Under high concentration of NaCl, the RS1 levels were significantly higher than those of untreated controls (Figure 3b). Therefore, the RS1 gene responds to drought and high salt stresses.

RS1 is responsive to ACC and ABA treatments
Hormones play important roles in plant responses to abiotic stresses. Thus, we treated M202 and M202 (Sub1A) seedlings of 14 days old with ABA, GA, and ACC (a precursor of ethylene). The RS1 expression levels were also determined by qRT-PCR. The results show that the RS1 levels were significantly increased both in M202 and M202 (Sub1A) upon ACC treatment compared to untreated controls (Figure 4a). Under GA treatment, the RS1 levels were increased both in M202 and M202 (Sub1A), but not statistically significant (Figure 4b). ABA treatment significantly increased the RS1 levels both in M202 and M202 (Sub1A) (Figure 4c). In summary, the results show that ACC and ABA, but not GA, treatments induce RS1 expression, indicating that the RS1 gene is responsive to ACC and ABA hormones.

RS1 is mainly expressed in leaves and sheaths
The RS1 expression levels in various tissues were determined using qRT-PCR. The transcript levels of RS1 were low in shoots and roots of one-week and two-week old seedlings both in M202 and M202 (Sub1A). In one-    Expression level of RS1 under ABA treatment. Three biological replications were carried out. Approximate 0.2 g leaves collected from six seedlings each genotype was used for RNA extraction and qRT-PCR. The expression level of RS1 was calculated using the 2^ddCt value. Each data point represents the mean±SD of 3 independent biological replicates. The student's t-test was performed. * indicates P<0.05; ** indicates P<0.01. month old plants, the RS1 levels were high in leaves and sheaths, but remained low in roots. At the reproductive stage, the RS1 levels were high in leaf tissues, including the flag leaf, leaves near bolting, and leaves at bolting. The RS1 levels remained low in stems and panicles ( Figure 5). These results show that RS1 is mainly expressed at high levels in leaves and sheaths, while only expressed at low levels in shoot, root, stem, and panicle tissues.

Overexpression of RS1 in Kitaake did not enhance submergence tolerance
Constructs designed to overexpress (RS1ox) or to knockdown (RNAi) the RS1 gene have been transformed into Kitaake separately. Twenty transgenic RS1ox and 16 transgenic RS1 RNAi independent lines were obtained. The homozygous transgenic lines were chosen in the T 3 generation, and their RS1 expression levels were

RS1
RNAi plants (# 3-5-10, 6-3-7, and 7-2-4) were lower than WT (Figure 7), indicating that the endogenous RS1 was suppressed in the presence of the RS1 RNAi construct. Submergence-tolerant rice plants show stunted growth upon submergence and restart growth after the flood has receded (Xu et al., 2006). Thus, we subjected their seedlings planted in tubes to submergence tests, and measured their increased seedling heights. Under submergence for 10 days, transgenic RS1ox seedlings elongated (12.26 cm for line #3-1-8 and 13.47 cm for #4-1-5) slightly less than the WT (14.17 cm) (Figure 7), but not statistically significant. The similar results were also observed in transgenic RS1 RNAi seedling (Figure 7). We did not observe other differences in submergence tolerance between RS1ox and WT plants, including seedling survival rates (data not shown). We conclude that the seedling elongation in transgenic RS1ox plants was not significantly different than that in WT upon submergence, indicating that RS1ox in rice did not result in significant improvement of submergence tolerance.
To measure the interaction between Sub1A and RS1, M202(Sub1A) was crossed with transgenic RS1ox and RNAi lines (Kitaake background), separately. The seedlings in F 1 generation were subjected to submergence. The increased seedling height was measured during submergence. With overexpression of RS1, the increased seedling height was unchanged compared with that of M202 (Sub1A)×Kitaake.
Conversely, the seedling height was significant elongated with the down-regulation of RS1 compared with the control even with the presence of Sub1A (Figure 8).
Above results indicate that RS1 may be play the role in Sub1A pathway to confer submergence tolerance in rice.

DISCUSSION
Reverse genetics approaches are widely used to identify genes of interest. In our previous and current studies, RNA-seq and microarray have been conducted to identify the candidate genes responsive to submergence in rice. RS1, a gene differentially expressed between M202 and M202(Sub1A) under submergence, was identified by both RNA-seq and microarray assays. M202(Sub1A), carrying the Sub1A gene, is a submergence tolerant line. Previous studies suggest that Sub1A limits ethylene-promoted GA responsiveness during submergence by augmenting accumulation of the GA signaling repressors SLR1 and SLRL1 (Perata and Voesenek. 2007;Fukao and Bailey-Serres, 2008). ADH1 is a submergence tolerance marker gene, whose levels in M202(Sub1A) increased during submergence (Xu et al., 2006). Here we show that RS1, Sub1A, ADH1, SLR1, and SLRL1 genes are co-induced upon submergence, reaching a peak within 24 h ( Figure  9), indicating that RS1 is tightly co-regulated under submergence with these well-characterized submergence-associated genes. The contents of GA and Figure 8. The increased seedling height in M202(Sub1A)×Kitaake background under submergence for 10 days. M202(Sub1A) was crossed with transgenic RS1 or RNAi lines, respectively. In the F1 generation, seedlings of approximately 3 to 4 cm were subjected to submergence. The increases in seedling height were measured every 24 h. Each bar represents the mean±SE of three independent biological replicates. Student's t-test was performed. * indicates p<0.05; ** indicates p<0.01. Figure 9. Co-expression of RS1, Sub1A, ADH1, SLR1, and SLRL1 in M202(Sub1A) during submergence. Fourteen-day old seedlings were subjected to submergence for 0, 1, 3 and 6 days. All experiments were performed with three biological replications. The leaves from six seedlings collected from different time points during submergence were used for RNA isolation. Expression levels of RS1, Sub1A, ADH1, SLR1, and SLRL1 were determined by qRT-PCR, and were calculated using the 2^ddCt value. Each data point represents the mean±SD of 3 independent biological replicates.