Core clock, SUB1, and ABAR genes mediate flooding …...7130 | Syed et al. breeding and development programmes from around the world have identified drought-tolerant soybean lines

Journal of Experimental Botany, Vol. 66, No. 22 pp. 7129–7149, 2015doi:10.1093/jxb/erv407 Advance Access publication 27 August 2015

RESEARCH PAPER

Core clock, SUB1, and ABAR genes mediate flooding and drought responses via alternative splicing in soybean

Naeem H. Syed1,*,†, Silvas J. Prince2,†, Raymond N. Mutava2,†, Gunvant Patil2,†, Song Li2, Wei Chen2,‡, Valliyodan Babu2, Trupti Joshi4, Saad Khan3 and Henry T. Nguyen2,*1 School of Human and Life Sciences, Canterbury Christ Church University, Canterbury CT1 1QU, UK2 National Center for Soybean Biotechnology and Division of Plant Sciences, University of Missouri, Columbia, MO 65211, USA3 Christopher S. Bond Life Sciences Center, University of Missouri, Columbia, MO 65211, USA

† These authors contributed equally to this work.‡ Present address: Beijing Tongzhou International Seed Science and Technology Co. Ltd, China* To whom correspondence should be addressed. E-mail: [email protected], [email protected], or [email protected]

Received 8 June 2015; Revised 3 August 2015; Accepted 6 August 2015

Editor: Ramanjulu Sunkar

Abstract

Circadian clocks are a great evolutionary innovation and provide competitive advantage during the day/night cycle and under changing environmental conditions. The circadian clock mediates expression of a large proportion of genes in plants, achieving a harmonious relationship between energy metabolism, photosynthesis, and biotic and abiotic stress responses. Here it is shown that multiple paralogues of clock genes are present in soybean (Glycine max) and mediate flooding and drought responses. Differential expression of many clock and SUB1 genes was found under flooding and drought condi-tions. Furthermore, natural variation in the amplitude and phase shifts in PRR7 and TOC1 genes was also discovered under drought and flooding conditions, respectively. PRR3 exhibited flooding- and drought-specific splicing patterns and may work in concert with PRR7 and TOC1 to achieve energy homeostasis under flooding and drought conditions. Higher expres-sion of TOC1 also coincides with elevated levels of abscisic acid (ABA) and variation in glucose levels in the morning and afternoon, indicating that this response to abiotic stress is mediated by ABA, endogenous sugar levels, and the circadian clock to fine-tune photosynthesis and energy utilization under stress conditions. It is proposed that the presence of multiple clock gene paralogues with variation in DNA sequence, phase, and period could be used to screen exotic germplasm to find sources for drought and flooding tolerance. Furthermore, fine tuning of multiple clock gene paralogues (via a genetic engineering approach) should also facilitate the development of flooding- and drought-tolerant soybean varieties.

Key words: ABA, ABAR, alternative splicing, circadian clock, drought, flooding, jasmonic acid, SUB1.

Introduction

Changing weather patterns around the world cause huge var-iation in average rainfall and storms that result in drought and flooding conditions. Drought and flooding have a major effect on yield reduction in crop plants (Valliyodan and

Nguyen, 2006; Bailey-Serres and Voesenek, 2008; Manavalan et al., 2009; VanToai et al., 2012). Increasing food supply is critical to meet the growing demand, especially under chang-ing environments. Different soybean [Glycine max (L) Merr.]

© The Author 2015. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For permissions, please email: [email protected]

mailto:[email protected]?subject=




7130 | Syed et al.

breeding and development programmes from around the world have identified drought-tolerant soybean lines with good yield potential under drought stress. Soybean has abun-dant genetic diversity in exotic germplasm for yield under drought conditions (Manavalan et al., 2009); however, US soybean cultivars have a very narrow genetic base (Carter et al., 2006). Thankfully, genetic variation exists for flooding tolerance in the soybean germplasm, and various mapping populations have now been developed to map and under-stand the underlying genetic variation for flooding tolerance (VanToai et al., 1994, 2001; Sayama et al., 2009; Nguyen et al., 2012). Quantitative trait loci (QTLs) for flooding toler-ance have recently been mapped using recombinant inbred lines derived from a cross between the flooding-susceptible and tolerant parents S99-2281×PI 408105A, respectively, under greenhouse and field conditions (Sayama et al., 2009; Nguyen et al., 2012). However, no detailed knowledge exists for the genetic and molecular mechanisms involved in flood-ing and drought tolerance in soybean.

Since flooding or submergence tolerance has been well stud-ied in rice (Oryza sativa) and the genes underlying response to shallow or deep flood have been characterized, these gene sequences were used as queries to find their orthologues in soybean and study their potential role in flooding tolerance in this species. In rice, two opposite adaptive responses involv-ing escape [elongation of internodes mediated by gibberellic acid (GA)] or quiescence (slow growth and starch metabo-lism to conserve energy and survival) are mediated by the multigenic SNORKEL-1, SNORKEL2 (SK1 and SK2) and SUBMERGENCE-1A, B, and C (SUB1A, SUB1B, and SUB1C), respectively (Xu et al., 2006; Hattori et al., 2009). All O. sativa accessions contain SUB1B and SUB1C genes, but SUB1A is only present in limited indica and aus cultivars. Increased levels of ethylene under submerged conditions trigger expression of both SUB1 and SK genes. In submer-gence-tolerant rice varieties, expression of SUB1A ultimately limits ethylene production and reduces responsiveness to GA through up-regulation of two negative regulators (SLR1 and SLRL1) of GA signalling (Fukao et al., 2011). This mecha-nism results in reduced consumption of carbohydrates and growth (quiescence), allowing plants to survive complete sub-mergence for up to 2 weeks. On the other hand, SK1 and SK2 trigger accumulation of bioactive GA in the submerged inter-nodes and initiate their elongation and rapid gain in height to escape flood waters. SUB1A also confers drought toler-ance by reducing leaf water loss and lipid peroxidation, and by increased expression of genes associated with acclimation to drought, namely DREB1 genes and AP59 (Fukao et al., 2011). SUB1A increases sensitivity to abscisic acid (ABA) and reduces accumulation of reactive oxygen species (ROS) dur-ing drought as well as de-submergence after a recent flood has subsided. This is supported by the SUB1A-mediated increase in ROS scavenging enzymes increasing oxidative stress toler-ance (Xu et al., 2006; Perata and Voesenek, 2007; Fukao and Bailey-Serres, 2008; Fukao et al., 2011).

One third of Arabidopsis thaliana genes are under the con-trol of circadian clock genes as assessed by enhancer trap studies (Michael and McClung, 2003). The ability of the

circadian clock to predict environmental changes and fine-tune physiology and metabolism presumably buffers against environmental variations and provides fitness advantages (Green et al., 2002; Dodd et al., 2005). Furthermore, there is a significant overlap between clock-regulated genes and those affected by different abiotic stresses (Covington and Harmer, 2007; Matsui et al., 2008; Mizuno and Yamashino, 2008). Studies have shown that the circadian clock also con-trols many genes that are directly or indirectly involved in ABA biosynthesis, which is a key hormone involved in abi-otic stress responses (Covington et al., 2008; Matsui et al., 2008; Mizuno and Yamashino, 2008). The circadian clock regulates the level of ABA throughout the day/night cycle, and many arrhythmic clock mutants and transgenics are compromised in their response to abiotic stresses (Covington and Harmer, 2007; Sanchez et al., 2011). Furthermore, the circadian clock controls the rate of degradation of starch to ensure an optimal carbon supply and continued growth dur-ing the night (Graf et al., 2010; Graf and Smith, 2011). Since slowing down of starch metabolism is a general mechanism in rice under flooding conditions during the quiescence strat-egy, it is conceivable that starch metabolism during flooding and drought response may also be controlled/gated by the circadian clock in soybean. Drought responses in Arabidopsis are mediated by the cross-talk between the circadian clock, levels of ABA, and an ABA receptor-like protein (ABAR/CHLH/GUN5) (Legnaioli et al., 2009). An evening-phased core clock component TIMING OF CAB EXPRESSION 1 (TOC1) binds to the promoter of ABAR/CHLH/GUN5 and controls its circadian expression. TOC1 is acutely induced by ABA, and this induction advances the phase of TOC1 binding and modulates ABAR circadian expression. This gated induction of TOC1 by ABA is disrupted in ABAR and TOC1 RNAi/overexpression lines, indicating that reciprocal regulation between TOC1 and ABAR is important for sensi-tized ABA activity in plants undergoing drought conditions (Legnaioli et al., 2009).

Flooding tolerance in rice is mediated via a few master controllers, SUB1A, and SK1 and SK2 genes for quiescence and escape strategies, respectively. SUB1A also play a role under drought conditions (Fukao et al., 2011); however, this mechanism is totally unique in rice as, after a flood-ing event and subsequent re-oxygenation, vegetative tissues undergo dehydration although the soil contains sufficient water (Fukao and Bailey-Serres, 2008; Bailey-Serres et al., 2012; Fukao and Xiong, 2013). This is fascinating as flood-ing and dehydration tolerance is achieved by SUB1A and it is possible that species which encounter similar patterns of drought and flooding during their lifetime and possess ortho-logues of SUB1-like genes may use similar strategies to coun-ter flooding and/or water deficit conditions. Furthermore, with recent changing weather patterns, US soybean-growing areas experience periods of flooding and drought in the same season, which significantly affects crop quality and yields. Therefore, understanding the mechanism of flooding and drought tolerance and breeding cultivars with both traits is extremely important. Furthermore, single master control-lers have not been found for drought and flooding stresses

Clock-mediated abiotic stress responses in soybean | 7131

because plant responses mediated by various phytohormones elicit a cascade of complex transcriptional, biochemical, and physiological responses. This is why progress to develop drought-tolerant soybean varieties has been slow despite huge investments and decades of hormone, physiology, and genomics research (Deshmukh et al., 2014). It was therefore reasoned that if soybean has orthologues of SUB1, SK1, and SK2 genes, they may also play a role in flooding and drought tolerance in soybean. However, the roles of SUB1 and SK genes may have been changed and/or diversified after poly-plodization due to the fact that soybean plants are seldom completely submerged. Several orthologues of SUB1 and SK genes along with Arabidopsis orthologues of the circadian core clock and their output genes have been found. Three orthologues of Arabidopsis TOC1 and ABAR genes, which have been shown to mediate drought responses in Arabidopsis, were also found. The plant circadian clock is based on gene expression negative feedback loops (Harmer, 2009). The first loop, proposed in 2001, involves two dawn-expressed, closely related and partially redundant Myb transcription factors, CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY), and the evening-expressed pseudo-response regulator TOC1. CCA1 and LHY play important roles in growth and vigour, and also impact starch accumulation and utilization (Ni et al., 2009; Graf et al., 2010; Graf and Smith, 2011) which are important traits under stressful conditions. Herein, gene expression data of important drought- and flooding-related genes along with core clock genes and their output genes, and the ABAR and TOC1 gene expression pattern throughout the day/night cycle (biologically relevant conditions, de Montaigu et al., 2015) are presented for soybean plants experiencing mild or severe drought/flooding stress and during the recovery phase when plants were re-watered or water was drained to stop the flood-ing stress. Endogenous levels of ABA and jasmonic acid (JA) were also measured to see their effects on TOC1 and ABAR regulation and plant responses during flooding and drought treatments. Since plants slow down their metabolism and sugar/starch consumption under stress, glucose quantities were also measured before dusk and soon after dawn to see whether their endogenous levels also play any role and medi-ate drought and/or flooding response mediated by ABAR, SUB1, and clock genes.

Materials and methods

Plant material and growth conditionsTen flooding- and drought-tolerant/susceptible lines were selected and evaluated under drought and flooding conditions explained in detail by Mutava et al. (2015). Four contrasting soybean genotypes for flooding and drought experiments were selected for this study as described by Mutava et al. (2015): PI 567690, drought tolerant; PANA (PI 597387), drought susceptible; PI 408105A, flooding toler-ant; and S99-2281 (PI 654356), flooding susceptible. These lines were grown under 14 h light and 10 h dark cycles and day/night tempera-tures of 28/18 °C as described by Mutava et al. (2015). Briefly, at the V5 stage, drought stress was imposed by withdrawing water, and 5 d later another set of plants were flooded by placing the smaller pots into larger pots tightly wrapped with a bin liner. Larger pots

were then filled with water to flood the smaller pots with plants in them in such a way that all root tissues and part of the stem were submerged. After 21 d of drought stress, plants were re-watered and allowed to recover. Flooding was done for 15 d and then the bin liners were punctured and subsequently removed on day 16 to drain all the water and allow the plants to recover. This was a ran-domized complete block design experiment with three replications. Leaf samples were taken every 4 h for 24 h during mild and severe stress and the recovery phase when water was drained from flooded plants and plants under drought were re-watered. Mild flooding and drought stress means that plants were flooded for 2 d and irrigation was stopped for 11 d, respectively. On the other hand, severe flood-ing and drought stress is when plants were under flooding conditions for 15 d and water was withdrawn for 21 d, respectively. The recov-ery phase is when samples were taken 3 d after draining of water and re-watering the plants from flooded and plants under drought, respectively.

Soybean clock, SUB1, SK1, and ABAR gene discoveryArabidopsis core clock and ABAR, and rice SUB1, SK1, and SUK2 gene sequences were used as queries for blast searches in the soy-bean database (http://soykb.org/blast.php). Since soybean and Arabidopsis are both dicots, orthologous genes were easily found; however, several hits for each gene were obtained owing to the paleo-ploid genome of soybean. Protein sequences of all genes were used to perform ClustalW alignments and construct Neighbor–Joining trees. Interestingly, soybean ABAR orthologues were highly similar to the Arabidopsis ABAR gene; however core clock, SUB1, SK1, and SK2 genes were only conserved in their DNA-binding and ethylene response factor (ERF) domains, respectively. To clarify further and differentiate the relationship and similarity between soybean LHY/CCA1 and different PRR genes, ClustalW align-ments were performed using Arabidopsis and soybean CCT/pseudo-receiver (PR) and MYB domains for PRR and CCA1/LHY genes, respectively.

RNA extraction, cDNA synthesis, RT–PCR, and qPCRTotal RNA was extracted from 100 mg of leaf tissue, flash-frozen, and stored at –80 °C using a Qiagen RNeasy kit (cat# 74104) fol-lowing the manufacturer’s instructions. RNA preparations were on-column DNase treated to remove DNA using a Qiagen RNase-Free DNase kit (cat# 79254). The reverse transcription–PCRs (RT-PCRs) were performed as described (Simpson et al., 2008; James et al., 2012a). Briefly, 4 μg of total RNA was used in first-strand cDNA synthesis by reverse transcription with oligo(dT)18 and random hexamer primers using the Sprint RT Complete-Double PrePrimed kit (Clontech) in a final volume of 20 μl. For each gene, a number of overlapping primers were designed to cover the complete gene sequences. These gene-specific primer pairs (with one 6-carboxyfluorescein-labelled primer) were designed to amplify between ~400 bp and 1200 bp, and to measure gene expression and capture different splicing events. cDNA reactions were diluted by adding 150 μl of water to each reaction, and 1 μl of each cDNA was aliquoted into 96-well plates containing the gene-specific primers. RT–PCRs were performed for 24 cycles as described by Simpson et al. (2008). Quantitative real-time PCR (qRT-PCR) was performed using the cDNA product corresponding to 25 ng of total RNA in a 25 μl reaction volume using the Maxima SYBR Green/ROX qPCR Master Mix (2×) (Cat# K0223,Thermo, USA) on an ABI7900HT detection system machine for the qRT-PCR (ABI). In the qRT-PCR experiments, three biological replicates and two technical repli-cates were used for analysis. The PCR conditions were as follow: 50 °C for 2 min, 95 °C for 10 min, then 40 cycles of 95 °C for 15 s and 60 °C for 1 min. To normalize the gene expression, the ACT11 (Glyma18g52780) gene was used as an internal control. All novel primers were designed using the Primer3 web-interface (http://frodo.wi.mit.edu/primer3/input.htm; Rozen and Skaletsky, 1998).

http://soykb.org/blast.php

http://frodo.wi.mit.edu/primer3/input.htm;

http://frodo.wi.mit.edu/primer3/input.htm;

7132 | Syed et al.

Circadian clock HR RT–PCR panelTotal RNA was extracted and cDNA was prepared from three bio-logical replicates of 18 time points for four lines over 3 d during the mild and severe stress and recovery phases of flooding- and drought-treated plants (432 samples). To measure the expression pattern and identify alternative splicing (AS) events for each gene, a number of overlapping primer pairs (each with one 6-FAM fluorescently labelled primer) were designed to cover the whole gene sequence and were used on samples derived from plants undergoing drought or flooding stress throughout the diurnal cycle. RT–PCRs were carried out for 24 cycles as described (Simpson et al., 2008; James et al., 2012a), and the products were detected on an ABI3730 automatic DNA sequencer along with a GeneScan™ 1200 LIZ size standard. PCR products were accurately sized, and mean peak areas calcu-lated using GeneMapper software. To measure changes in AS of particular isoforms, the peak areas were normalized relative to the peak area values for two reference genes, ACT11 (Glyma18g52780) and FBOX (Glyma12g05510). Statistical analysis to test the signifi-cance of different alternative and fully spliced transcripts was per-formed as described by Simpson et al. (2008).

High resolution RT–PCR (HR-RT–PCR) is a powerful and very sensitive technique which can detect various canonically and alterna-tively spliced transcripts with a single base resolution. Furthermore, quantification of different transcripts is as reliable as a qPCR, and with the additional advantage of detecting and reliably measuring the expression of multiple transcripts from a single reaction (Simpson et al., 2008; James et al. 2012a; Kalyna et al., 2012; Marquez et al., 2012). HR-RT–PCR was performed on 432 samples (216 each for flooding and drought) from different time points across the diurnal cycle during mild flooding/drought, severe stress, and when plants were going through the recovery phase after irrigation or drainage of excess water from drought- and flooding-treated plants.

Analysis of gene expression by RNA-seqThe same RNA samples used for expression and splicing analysis were used for the RNA sequencing (RNA-seq) experiment. The RNA (200 ng μl–1) samples from the 24:00 h time point from control plants and plants experiencing severe drought and flooding stress were used for RNA-seq. Three replicates for each of the drought-sensitive, PANA (control and drought samples); drought tolerant, PI 567690 (control and drought samples); flood-sensitive, S99-2281 (control and flood samples); and flood-tolerant, PI 408105A (con-trol and flood samples) lines were multiplexed and sequenced in an Illumina Hi-Seq 2000 platform.

Hormone and glucose analysisLeaf tissues were collected at midday from these genotypes after 20 d of drought stress and 15 d of flooding. The samples were imme-diately frozen in liquid nitrogen and stored at –80 °C until the time of processing. Hormone and glucose analysis was performed as described previously (Mutava et al., 2015).

In brief, a liquid chromatography–tandem mass spectrometry (LC-MS/MS) method was used for assay of acidic hormones [salicylic acid (SA), ABA, JA, jasmonoyl-isoleucine (JA-Ile), and 12-oxo-phytodien-oic acid (OPDA)]. For hormone extraction, a mixture of deuterium-labelled standards (D5SA, D6ABA, D2JA, and D5IAA) at 2.5 mM each was added to frozen leaf tissue sample and then 900 ml of ice cold MeOH/CAN (1:1 v/v) was added to each sample. Samples were then homogenized with a TissueLyserII (Qiagen, CA, USA) for 2 min at a frequency of 15 Hz s–1 and centrifuged at 16 000 g for 5 min at 4 °C. The supernatants were then transferred to new 2 ml tubes and the pel-lets were re-extracted. The second supernatant was combined with the first one and dried down. The pellets were dissolved in 200 ml of 30% methanol and centrifuged again to remove undissolved material, and the supernatant was transferred to vials for LC-MS/MS analysis.

For sugar analysis, standards of fructose, glucose, sucrose, raffi-nose, and stachyose were dissolved in water at a concentration of

10 mg ml–1. A mixture of 1 mg ml–1 of the five sugar solutions was then prepared and further diluted to 50, 100, 200, 300, 400, and 500 mg ml–1. Leaf samples weighing 0.5 g were lyophilized for 2 d in a Labconco Freeze Dry System (FreeZone 6 litre console freeze dry system with a stoppering tray dryer; Labconco Corporation, Kansas City, MO, USA) and then ground using a Geno/Grinder2010 grinder (SPEX SamplePrep, Metuchen, NJ, USA). Sugar extraction and analysis were performed using a modified HPLC-ELSD (evapora-tive light scattering detector) method developed for soybeans at the Nguyen laboratory, University of Missouri (Valliyodan et al., 2015).

Results

High conservation of core clock and ABAR genes in soybean

Protein sequence alignments are presented for the products of the LHY/CCA1, ABAR, TOC1, and SUB1 genes in Supplementary Figs S1–S5 available at JXB online. It is difficult to differentiate soybean LHY from CCA1 as all four orthologues have ~48% similarity to Arabidopsis CCA1 and LHY genes, and amino acids which differentiate CCA1 from LHY are randomly shared between all the soybean orthologues. However, MYB domains in the products of soybean CCA1/LHY-like genes are highly con-served (>90%) compared with Arabidopsis CCA1 and LHY genes (Fig. 1A). Although all four CCA1/LHY genes are >90% similar to each other, they are still distinguishable, and sequence differ-ences may affect their function in fine tuning the clock. TOC1-1 (Glyma04g33110), TOC1-2 (Glyma05g00880), and TOC1-3 (Glyma06g21120) have ~45–50% overall similarity to Arabidopsis TOC1; however, this sequence conservation increases to ~85 and ~95% in the PR and CCT domain, respectively (Fig. 1B, C). Interestingly, the product of TOC1-2 (Glyma05g00880) lacks 43 upstream amino acids compared with Arabidopsis and two other orthologues of TOC1, and uses an alternative methionine downstream, starting with the MKLL sequence (Fig. 1B). It was assumed that this gene may be wrongly annotated and thus potential upstream ATG start sites were searched for; however, none was found which would preserve the open reading frame of this gene (data not shown). PRR5 and PRR9 genes were more similar to Arabidopsis orthologues, whereas PRR7 and PRR3 were more conserved in the CCT domain (Fig. 1B). Furthermore, the phylogenetic tree shown in Fig. 1D illustrates that Arabidopsis clock genes are quite similar to soybean orthologues and cluster together with the same genes from these two species. Sequence similarity between Arabidopsis ABAR and three soybean ortho-logues across entire coding sequence is ~86%; however, among three soybean genes, the sequence similarity reaches 93–98% (Supplementary Fig. S2). This might indicate diversification after polyploidization and may lead to neofunctionalization as evident from their differential expression patterns during flooding and drought stress (see below).

Orthologues of SUB1, SK1, and SK2 genes are highly conserved in the ERF domain and possibly are of SUB1B and SUB1C type

The soybean database was searched for orthologues of rice SUB1, SK1, and SK2 genes, and 18 orthologues were found

http://jxb.oxfordjournals.org/lookup/suppl/doi:10.1093/jxb/erv407/-/DC1




Fig. 1. ClustalW alignment of soybean and Arabidopsis clock genes. (A) MYB domain of soybean and Arabidopsis LHY and CCA1 genes. (B) Pseudo-receiver domain of TOC1 and PRR genes. (C) CCT domain; the PRR7 and PRR3 CCT domain is not shown as it is not conserved. (D) Phylogenetic relationship between soybean and Arabidopsis TOC1 and PRR genes. The numbers at the nodes represent bootstrap percentage values. (This figure is available in colour at JXB online.)

7134 | Syed et al.

(15 SNORKEL and 3 SUB1; Supplementary Fig. S5 at JXB online). Protein sequences encoded by these genes were aligned with those encoded by rice SUB1 and SK1 genes. Even though sequence conservation was very low across whole gene sequences, the ERF domain was highly conserved (Supplementary Fig. S4). To understand the relationship between these genes, a maximum likelihood tree was con-structed according to Dereeper et al. (2008) and is presented in Supplementary Fig. S5B. Rice SK1 and SUB1 genes form a distinct group at the bottom of the tree along with three orthologues from soybean (Glyma03g42450, Glyma16g01500, and Glyma07g04950). To obtain better insight, these three genes were realigned with rice SUB1, SK1, and SK2 genes; however, sequence gaps (data not shown) in SK1 and SK2 showed that these genes are more closely related to SUB1 genes (Supplementary Fig. S4). Sequence conservation was ~30% for complete sequences, but reached 85– 90% in the ERF domain. To decipher whether three orthologues of SUB1 genes belong to the A, B, or C type, the signature amino acids for SUB1B and SUB1C, namely alanine and aspartic acid at position 13 (SUB1A has a serine at this posi-tion) and 18, respectively, were investigated. Interestingly, all three orthologues in soybean have alanine and aspartic acid at position 13 and 18, respectively, indicating that they belong to SUB1B or SUB1C as they lack serine at position 13 of the ERF domain. This probably makes sense as soybean plants are grown in dry regions and are seldom completely sub-merged in water.

SUB1 genes mediate flooding and drought responses in soybean via alternative splicing

The HR-RT–PCR system (James et al., 2012a) was used to quantify gene expression as well as AS in the SUB1 genes. The names SUB1-1–SUB1-3 were randomly assigned to Glyma07g04950, Glyma03g42450, and Glyma16g01500, respectively. Data for Glyma07g04950 (SUB1-1) and Glyma03g42450 (SUB1-2) are presented in Fig. 2; however expression was very low for Glyma16g01500 (SUB1-3) and is therefore not shown. Similarly, very low to no expression was detected for all SK genes and is not shown. The expression of the fully spliced transcript of SUB1-1 remains low during mild flooding stress but increases dramatically towards the end of the night during severe flooding stress (15 d under water) for the flooding-tolerant variety (PI 408105A). A splice variant of SUB1-1-AS1 (see Supplementary Table S1 at JXB online for a description) also shows the same pattern and increases dur-ing severe flooding and then remains high even in the recov-ery phase. The data of qPCR for SUB1 are also presented in Fig. 1 (inserts); qPCR data for all other genes are not shown as they were identical to the HR-RT–PCR results. It is not known how changes in the expression of these transcripts (SUB1-AS1, Glyma07g04950) take place during the 3 days when plants were recovering from flooding stress. Expression of SUB1-1 (Glyma07g04950) shows a completely different pattern and remains low for the drought-susceptible variety under control and drought conditions (during mild stress); however, it significantly increases in the drought-tolerant line

(PI 567690). Expression of SUB1-1 remained high for line PI 567690 during severe drought stress (21 d) compared with PI 567690 under control conditions (normal irrigation), indicat-ing its role during mild and severe drought stress. Expression of SUB1-1 drops after re-watering during the recovery phase and starts resembling that of plants grown under control con-dition, further supporting its role during drought transition and acclimation. However, SUB1-1-AS1 only seems to be playing a role during drought acclimation as its expression increases during the stress period and then is almost abol-ished during severe drought, with even much lower expression under recovery. Notably, SUB1-1 clearly displays a circadian expression pattern (albeit under diurnal conditions) and may be gated by the circadian clock. SUB1-2 (Glyma03g42450) expression seems to be quite similar for flooding-susceptible and tolerant lines grown under normal or flooding condi-tions. In contrast, SUB1-1 expression increases during mild drought conditions (acclimation) and then significantly drops during severe drought and the recovery phase, and also seems to be gated by the clock.

The role TOC1 and ABAR in modulation of flooding and drought response in soybean

In Arabidopsis, TOC1 binds to the promoter of the ABA-related gene (ABAR/CHLH/GUN5) and controls its circa-dian expression. TOC1 is acutely induced by ABA and this induction advances the phase of TOC1 binding and modu-lates ABAR circadian expression. This gated induction of TOC1 by ABA is disrupted in ABAR and TOC1 RNAi and overexpression lines, indicating that reciprocal regulation between TOC1 and ABAR is important for sensitized ABA activity in plants subjected to drought conditions (Legnaioli et al., 2009). Since soybean has two and three orthologues of Arabidopsis TOC1 and ABAR, respectively, the diur-nal expression pattern of these genes was examined among flooding- and drought-tolerant lines undergoing flooding or drought stress and growing in control conditions. The TOC1-3 gene Glyma06g21120 (EU076435), previously reported by Liu et al. (2009), showed robust circadian expression under free running conditions; however, no significant expression of this gene was detected among all lines tested here. Expression of two other TOC1 genes (Glyma05g00880-TOC1-1 and Glyma05g00880-TOC1-2) across the day–night cycle showed strong oscillations and peaked towards the end of the day (Fig. 3). No expression of TOC1-1 in the flooding-susceptible or tolerant lines was detected; however, it was readily detecta-ble and strongly expressed in drought-tolerant and susceptible lines. This shows that some duplicated genes have signifi-cantly diverged in different lines after polyploidization, lead-ing to varied responses to abiotic stresses and possibly other functions as well. Expression of TOC1-2 in flooding-tolerant and susceptible lines (PI 408105A and S99-221) increases slightly during mild flooding conditions; however, it drops significantly during severe stress compared with lines growing under control conditions (Fig. 3). The expression of TOC1-2 in the flooding-tolerant line (PI 408105A) increases signifi-cantly during the recovery phase, whereas its expression in the







susceptible line during recovery remains somewhat lower. The expression levels of TOC1-1 and TOC1-2 increase during mild drought stress, but become much lower during severe drought stress, and this drop in expression is much more pronounced for TOC1-1 than for TOC1-2 (Fig. 3). Expression levels of TOC1-1 and TOC1-2 increase again during the recovery phase when plants are re-watered and probably recover to the mild stress level. The expression of both TOC1 genes during severe stress drops significantly and may modulate drought response mediated via the soybean ABAR gene. The expression pat-tern of three soybean ABAR genes which are almost identical to Arabidopsis orthologous genes is presented in Fig. 4. The expression patterns of ABAR1–ABAR3 show that they peak close to dawn. There are no significant differences among the flooding-susceptible and tolerant lines; however, the expres-sion of ABAR2 and ABAR3 is significantly higher in the lines undergoing mild flooding stress whereas they are very similar for the drought-susceptible lines growing under drought or control conditions. Expression of ABAR2 and ABAR3 drop

significantly during severe drought stress, but this is more pro-nounced for ABAR3. Expression during the recovery phase drops even further for both genes, and a similar trend is also apparent for the drought-susceptible line grown under nor-mal and drought conditions; however, fold changes are not as marked as for drought-tolerant lines. ABAR and TOC1 genes in soybean are dawn and dusk phased, respectively which is reminiscent of Arabidopsis where TOC1 binds to the ABAR promoter and co-ordinates the drought response. ChIP data are not available; however, it can be speculated that TOC1 genes may also bind individual and/or multiple ABAR genes in soybean and co-ordinate stress responses.

Role of ABA and JA in flooding and drought response

In Arabidopsis, elevated quantities of ABA (exogenous) activate TOC1 expression and this activation is gated by the circadian clock and determines the timing of TOC1 binding to the ABAR promoter and changes in ABAR expression.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0

0.1

0.2

0.3

0.4

0.5

0.0

0.1

0.1

0.2

0.2

0.3

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

2

4

Mild Stress Severe StressRecovery

SUB1-1-Glyma07g04950-FS

SUB1-Glyma03g42450

Flooding Drought

S99-C

S99-FLD

105A-C

105A-FLD

Mild Stress Severe StressRecovery

0246

0.0

0.0

0.0

0.1

0.1

0.1

0.1

0.1

0.2

0.0

0.1

0.1

0.2

0.2

0.3

0.3

0.4

0.4Rela

tive

Exp

ress

ion

ACT

11/F

BOX

SUB1-1-Glyma07g04950-AS1

SUB1-2-Glyma03g42450

PANA-CPANA-DR690-C690-DR

S99-CS99-FLD105A-C105A-FLD

S99-C

S99-FLD

105A-C

105A-FLD

Fig. 2. Expression profiling of SUB1 genes (Glyma07g04950 and Glyma03g42450) across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress measured using HR-RT–PCR. Primers are listed in Supplementary Table S2 at JXB online. Data at peak expression points show the mean and SD (n=3). Small insert panels show qPCR results for flooding-tolerant and susceptible lines at 08:00 h during the recovery phase. Similarly, qPCR results for drought-tolerant and susceptible lines are presented on the right side for the 20:00 h samples during the mild drought stress condition. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.


7136 | Syed et al.

A functional ABAR is required for this interaction between TOC1 and ABA and to co-ordinate ABA-mediated changes in gene expression and responses to drought (Legnaioli et al., 2009). Recently, De Ollas et al. (2013) showed that tran-sient accumulation of endogenous JA is necessary for the ABA increase in citrus roots under drought conditions. To understand the potential role of endogenous JA and ABA in drought/flooding acclimation via interactions with TOC1 and ABAR genes, JA and ABA were quantified in leaf sam-ples harvested at 04:00 h and 12:00 h among plants undergo-ing mild and severe stress and in the recovery phase for both flooding and drought tolerance (Fig. 5A). It is apparent that

levels of ABA are significantly higher among lines which are drought tolerant, especially at midday (12:00 h). The ABA levels at 04:00 h are much lower among lines which are flood-ing tolerant and undergoing flooding stress or growing under normal conditions. ABA quantities are much more similar for both flooding-tolerant/susceptible lines at midday, with the only exception of the recovery phase where both lines show elevated levels. This is interesting as it is known that flooding tolerance genes are triggered by ethylene and it probably neg-atively regulates ABA synthesis and/or sensitivity (Hattori et al., 2009; Fukao et al., 2011; Bailey-Serres et al., 2012). JA shows a completely different picture and has high levels in the flooding-tolerant line (PI 408105A) during all phases of stress and recovery at 07:00 h. However, it shows a different profile at midday and drops significantly during mild stress, is marginally higher during severe stress, but remains high dur-ing the recovery phase (Fig. 5B).

ABA quantities in drought-specific lines at 04:00 h and 12:00 h are almost the exact opposite of those in lines under flooding. At 04:00 h, only the drought-tolerant line under-going recovery after irrigation shows a high level of ABA, whereas at midday, the drought-tolerant line has a high level of ABA during severe stress and the recovery phase, and the level is marginally less than in the susceptible line during mild stress. JA levels are somewhat higher (Fig. 5B) for PI 567690 and increase significantly at midday during the mild and severe drought stress; they start to go down, however, dur-ing the recovery phase. It seems likely that in soybean both JA and ABA co-ordinate the drought responses and JA trig-gers more progressive ABA accumulation and induces plant responses to stress (De Ollas et al., 2013). It is clear, however, that flooding does not induce ABA production and is pos-sibly co-ordinated via JA, as is evident from elevated levels during flooding stress especially a few hours before dawn.

Plants under flooding and drought show opposite phenotypes for glucose levels

The glucose levels during mild flooding stress at 16:00 h (4.5 h before dusk) are slightly higher for S99-2281 compared with the flooding-tolerant PI 408105A (Fig. 6); however, this is reversed 1.5 h after dawn (06:30 h). This clearly shows that photosynthetic activity in the flooding-tolerant line is still robust, and PI 408105A is efficient in energy utilization dur-ing mild stress. The overall glucose content drops in this line (PI 408105) growing under control conditions compared with when it was under severe flooding stress. Plants may have less glucose reserves during this active growing period (vegetative) just before the onset of the reproductive stage when the sugar reserves are used for all types of metabolic activities. PI 408105A has less glucose at 16:00 h and 08:00 h compared with the susceptible S99-2281 line. This may be due to the higher efficiency of PI 408105A to use energy reserves for its growth and metabolism. Glucose content is also lower among lines under severe flooding stress; how-ever, the trend is the opposite, and PI 408105A has slightly more glucose than the PI S99-2281 line. Furthermore, a pro-fuse lateral shoot branching phenotype was observed in PI

0

0.1

0.2

0.3

0.4

0.5

0.6

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

Flooding

Drought

Mild Stress Recovery

TOC1-2-Glyma05g00880-FS

Severe Stress

0

0.05

0.1

0.15

0.2

0.25

0.3

TOC1-2-Glyma05g00880-FS

TOC1-1-Glyma04g33110-FSRela

tive

Exp

ress

ion

ACT

11/F

BOX


S99-C S99-FLD 105A-C 105A-FLD

Fig. 3. Expression of TOC1-1 and TOC1-2 genes measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.



408105A under severe flooding conditions, with many new leaves on them (see Supplementary Fig. S6 at JXB online). Perhaps PI 408105A produced new branches and leaves to counter excessive water levels and possibly to increase pho-tosynthetic activity. Drought-susceptible and tolerant PANA and PI 567690, respectively, showed high and very similar levels of glucose during control and mild drought conditions. However, glucose levels were higher in the drought-tolerant PI 567690 line, showing that either some photosynthetic activity could still take place or the energy reserves are rela-tively slowly exhausted, allowing it to survive and initiate a better response to irrigation when ample supplies of water are present for optimal metabolic activities.

Changes in expression and AS of LHY/CCA1 genes are more significant under drought conditions

Plants usually slow down their starch degradation process under stress conditions and use the starch as a survival strat-egy to extend their energy reserves. Furthermore, starch utili-zation during the day/night cycle is tightly under the control of the circadian clock. The core clock genes CCA1 and LHY play a significant role in this, as a mutant lacking the CCA1 and LHY components exhausts its starch reserves much faster at the perceived dawn via its fast running circadian clock than at the actual dawn a few hours later (Graf et al., 2010). Furthermore, a splice variant in the CCA1 gene in Arabidopsis

0.00

1.00

2.00

3.00

4.00

5.00

6.00

-0.50

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

0.00

0.05

0.10

0.15

0.20

0.25

0.00

0.02

0.04

0.06

0.08

0.1

0.12

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

Mild Stress Severe Stress Recovery

Flooding DroughtS99-C

S99-FLD

105A-C

105A-FLD


ABAR1-Glyma19g32070-FS

Severe Stress

0

0.2

0.4

0.6

0.8

1

1.2

1.4

ABAR1-Glyma19g32070-AS1



Rel

ativ

e Ex

pres

sion

AC

T11/

FBO

X PANA-CPANA-DR690-C690-DR


Fig. 4. Expression of ABAR1–ABAR3 genes measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.



7138 | Syed et al.

Fig. 5. ABA and JA concentrations at 04:00 h and 12:00 h among flooding- and drought-susceptible and tolerant lines under control (C), flooding (F), and drought (D) conditions. M, S, and R stand for mild, severe and recovery phases, respectively. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant).

Fig. 6. Glucose concentration at 08:00 h and 16:00 h among lines under control (C), flooding (F), and drought (D) conditions. M and S stand for mild and severe stress, respectively. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant).


mediates the cold temperature acclimation process by slow-ing down the starch degradation process and prolonging usage of its starch reserves as a survival strategy (Park et al., 2012). It was therefore asked whether expression or AS of LHY/CCA1-like genes causes such differences during flood-ing and drought stress responses. Figure 7 shows that three

LHY/CCA1 genes respond to flooding and drought stresses in a different way. The transcript abundance of these three genes showed no significant differences in the flooding experiment; however, it markedly increases during mild drought stress in the drought-tolerant line PI 567690 and remains low or unde-tectable for PANA (Fig. 7). Expression of Glyma16g01980

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0.0

0.0

0.0

0.0

0.0

0.1

0.1

0.1

0

0.5

1

1.5

2

2.5

3

3.5

Severe Stress

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0


Flooding Drought


LHY/CCA1-Glyma16g01980-FS

LHY/CCA1-Glyma16g01980-AS1



Severe Stress

Rel

ativ

e Ex

pres

sion

AC

T11/

FBO

X



Fig. 7. Expression of transcripts of three soybean CCA1/LHY genes and a splice variant (AS1) measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.


7140 | Syed et al.

and Glyma03g42260 for both PI 567690 and PANA (Fig. 7) begin to resemble each other under control as well as severe drought conditions, whereas Glyma19g45030 remains high for both lines under severe drought stress and the recovery phase. An AS event (LHY/CCA1-AS1) in Glyma16g01980 represents a very small proportion of the fully spliced prod-uct for both flooding-tolerant and susceptible lines; however, this AS event represents almost 50% of the fully spliced (FS) transcript of this gene under drought conditions. The tran-script abundance of LHY/CCA1-AS1 remains very similar for both PANA and PI 567690 under control and drought conditions, but the transcript abundance of LHY/CCA1-FS is almost twice as high during mild drought stress for the drought-resistant line (PI 567690).

Significant splicing in PRR3 genes modulates flooding and drought responses

Arabidopsis sequences were used as bait to find soybean orthologues of PRR9, 7, 5, and 3, and their conserved regions were compared using protein alignments (data not shown). Although sequence homology with Arabidopsis orthologues of PRR9, 7, 5, and 3 is 36, 43, 30, and 58%, respectively, they all group together with their corresponding orthologues from Arabidopsis. The soybean PRR3 showed several gene mod-els in the soybean genome browser (http://phytozome.jgi.doe.gov/jbrowse/). The primary gene model, Glyma11g15580.1, showed a truncated transcript encoding 217 amino acids; however, the alternative transcripts Glyma11g15580.2 was 766 amino acids long. Surprisingly, this gene is mapped to the scaffold region (Glyma.U03450) in the new soybean refer-ence assembly (Wm82.a2.v1) and did not show a truncated

version. The primary transcript was used to search homol-ogy with Arabidopsis, and the highest homology of 58% was found with At5g60100.

PRR5 expression shows a somewhat similar trend between flooding- and drought-tolerant lines (Fig. 8) as it is relatively higher during mild flooding and drought stress and drops dur-ing severe flooding and drought stress. It maintains the same amplitude during drought recovery; however, it increases signifi-cantly for plants going through the recovery phase after severe flooding stress. Expression of a splice variant (AS1) of PRR5 (Supplementary Table S1 at JXB online) was down-regulated during severe flooding in PI 408105A and resembles that of con-trol plants. PRR5-AS1 was also down-regulated during severe drought treatment in the PI 567690 line, but retains the same transcript abundance during the recovery phase. PRR3 pro-duces four splice variants that show considerable variation under flooding and drought conditions (Fig. 9; Supplementary Table S1). FS as well as AS1 and AS2 expression of PRR3 is con-siderably higher in the PI 408105A line during mild stress, but it drops during severe flooding conditions and recovers margin-ally during the recovery phase; on the other hand, it maintains a similar expression pattern during mild and severe flooding stress and the recovery phase in S99-2282 and PI 408105A growing under flooding and control conditions. The expression pattern of PRR3-AS2 for both PANA and PI 567690 is higher during mild drought stress but then drops during severe drought stress for PANA; however, it maintains robust rhythms throughout all phases of growth during mild and severe drought and the recovery phase. The expression of FS and AS1 drops during severe drought stress and recovers during the recovery phase in such a way that it resembles control plants. AS3 and AS4 do not show any significant variation in the flooding experiment, but

0.0

0.5

1.0

1.5

2.0

2.5

3.0


Flooding Drought



Severe Stress

PRR5-Glyma04g40640-FS

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40PRR5-Glyma04g40640-AS1

Rela

tive

Exp

ress

ion

ACT

11/F

BOX



Fig. 8. Expression of transcripts of a soybean PRR5 gene and a splice variant (AS1) measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.

http://phytozome.jgi.doe.gov/jbrowse/

http://phytozome.jgi.doe.gov/jbrowse/






AS3 is much higher among lines undergoing mild drought stress, with PI 567690 showing even higher expression compared with PANA. Furthermore, the expression of AS3 in PANA drops

during severe drought whereas it increased a little more in PI 567690 and started resembling that in the control lines after irri-gation. AS4 also presents a similar picture for both PANA and

0.0

0.1

0.1

0.2

0.2

0.3

0.3

0.0

0.1

0.1

0.2

0.2

0.3

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7


Flooding DroughtPRR3-Glyma11g15580-FS

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.0

0.1

0.1

0.2

0.2

0.3

0.3

0.0

0.1

0.1

0.2

0.2

0.3 PRR3-Glyma11g15580-AS1

PRR3-Glyma11g15580-AS2



0.0

0.0

0.0

0.1

0.1

0.1

0.1

0.0

0.1

0.1

0.2

0.2

0.3

0.3

0.0

0.0

0.0

0.1

0.1

0.1

0.1

0.1

0.2

0.2

0.2

0.00

0.05

0.10

0.15

0.20

0.25

0.30


Rela

tive

Exp

ress

ion

ACT

11/F

BOX



Fig. 9. Expression of transcripts of a soybean PRR3 gene and splice variants (AS1–AS4) measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.


7142 | Syed et al.

PI 567690 during mild drought stress; however, they look very similar during severe drought and the recovery phase, indicating that this AS4 may only be more important during mild stress, in contrast to AS3 whose expression keeps on rising during mild as well as severe drought stress. It is interesting that AS3 and AS4 display very different expression patterns during flooding and drought conditions, which would indicate that they may be play-ing a fine tuning role during flooding and drought treatments.

PRR7 peaks at different times under drought conditions

Recently, Liu et al. (2013), by employing ChIP-seq, identified putative PRR7 target genes that included morning-expressed

circadian-regulated genes, growth regulators, and those involved in light signalling and stress responses in Arabidopsis. Interestingly, many of these target genes were regulated by ABA, and this response to ABA levels was mediated by changes in the expression of PRR7; however, this response to ABA was only partial as PRR7 mutants and overexpress-ing lines were still able to respond to ABA in a time-depend-ent manner to promote stomatal closing. This makes sense as the ABA level also affects TOC1 expression, and TOC1 in turn binds to the ABAR (GUN5) promoter to mediate drought responses in Arabidopsis (Legnaioli et al., 2009). In Arabidopsis, ABA affects TOC1 expression but not that of PRR7, at least under continuous light conditions; however, the length of the period in the PRR7 mutant is more sensi-tive to exogenous ABA application than that in the wild type (Legnaioli et al., 2009). The expression patterns during the day/night cycles for PRR9- and PRR7-like genes are presented in Fig. 10. PRR9 expression is only detectable in the flooding-susceptible and tolerant lines (S99-2281 and PI 408105A) and was no seen under drought conditions for PANA and PI67690. PRR7 was readily detectable under flooding as well as drought conditions among all lines tested. Intriguingly, there is a 4 h difference between PANA and PI 567690 under mild drought conditions, with PRR7 peaking at 08:00 h (1.5 h after subjective dawn) whereas it peaks at 12:00 h for PANA and PI 567690 under control conditions (normal irrigation). S99-2281 and PI 408105A do not show any remarkable dif-ference between them under control or mild flooding condi-tions, but there is evidence of phase advance for PI 408105A (flooding tolerant) during recovery. Both lines peak at 12:00 h during severe drought; however, changes in the PRR7 expres-sion pattern are again visible during the recovery phase in the drought experiment with PI 567690, again peaking at 12:00 h but showing significantly higher expression.

Response of CAB2, ELF3, and LUX during flooding and drought treatment

Lines growing under normal conditions (both drought-tol-erant and susceptible lines) during the mild drought stress phase showed relatively lower expression of the CAB2 gene compared with plants undergoing initial onset of drought (Fig. 11). The expression of the CAB2 gene was higher in the drought-tolerant line compared with the drought-susceptible line; however, it shows that up-regulation of CAB2 is a gen-eral response to mild drought conditions under the experi-mental conditions used here. A similar but opposite trend was observed under severe drought conditions, as the expression of CAB2 dropped for both lines. The expression of CAB2 in the drought-tolerant line remained low even when the plants were re-watered; however, it quickly recovered in the drought-susceptible line. On the other hand, CAB2 expres-sion remained stable under mild and severe flooding condi-tions, but increased significantly when water was drained from both flooding-susceptible and tolerant lines (Fig. 11). These data suggest that CAB2 expression might be more relevant under water-limiting conditions or when plants are adjust-ing their photosynthetic machinery to variable amounts of water. ELF3 and LUX exhibited genotype-specific expression

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7


Flooding

Drought



PRR7-F3-Glyma13g19870-FSPANA-C

PANA-DR

690-C

690-DR

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8Rela

tive

Exp

ress

ion

ACT

11/F

BOX



Fig. 10. Expression of soybean PRR9 and PRR7 gene transcripts measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.




patterns and were only expressed in flooding- and drought-specific lines, respectively (Fig. 11). ELF3 expression dropped in flooding-tolerant and susceptible lines during mild

flooding stress. ELF3 expression was reduced significantly during severe flooding stress; however, it follows a similar trend and dropped further for both flooding-susceptible and tolerant lines. During the recovery period when water was drained from plants, both flooding-susceptible and tolerant lines showed a phase advance of ~4 h compared with plants growing under normal conditions. LUX expression among flooding-specific lines was undetectable, and in the drought-tolerant line it dropped slightly during mild drought condi-tions and was completely abolished under severe drought conditions. However, it showed complete recovery when plants were re-watered and resembled that of the drought-tolerant line growing under normal conditions (Fig. 11).

RNA-seq analysis of sugar metabolism and LEA genes

As photosynthetic efficiency is compromised under stress-ful conditions, energy conservation (sugar metabolism) and osmotic adjustment become very important (Hong-Bo et al., 2005; Manavalan et al., 2009; Qazi et al., 2014). Similar func-tional genes are involved in drought and sugar metabolism and transfer of sugar molecules to various acceptor mole-cules such as flavonoids (Winkel-Shirley, 2002). Furthermore, late embryogenesies abundant (LEA) proteins were found to be up-regulated in drought and heat stress in rice, Brassica, and legumes (Goyal et al., 2005; Hong-Bo et al., 2005; Hundertmark and Hincha, 2008; Dalal et al., 2009; Battaglia and Covarrubias, 2013). Therefore, LEA proteins and sugar synthesis-related genes were analysed for their expression pattern in drought stress. Pairwise comparison of drought and control conditions in a similar drought-susceptible line (our unpublished results) revealed that the sugar-related gene (Glyma20g30620) associated with fructose-1,6-bisphos-phatase activity was down-regulated. This gene is crucial in generating glucose as an energy source through gluconeogen-esis and the Calvin cycle to maintain photosynthetic activity in drought stress conditions. These sugar-related genes (espe-cially sugar transporter genes) are critical in transporting sugar from source (leaves) to sink tissues (roots and seeds) to ensure optimum productivity in drought environments (Lemoine et al., 2013). Recently, Patil et al. (2015) studied the soybean SWEET gene family (a class of sugar effluxer) using transcriptome profiling and identified their role in source to sink transport. In the present study several SWEET genes differentially regulated under drought and flooding treatments were identified (Supplementary Table S3 at JXB online). Under drought conditions, Glyma06g17540 and Glyma15g16030 showed up-regulation in the drought-toler-ant line; on the other hand Glyma06g13110, Glyma09g04840, and Glyma16g27350 (sucrose transporter genes) were down-regulated in the drought-susceptible line (Supplementary Table S3). This expression pattern suggests that sucrose efflux genes are activated under drought conditions and may be serving sugar transport to provide an energy supply to main-tain cell survival under high respiration conditions; however, a detailed study is needed to support these findings. It has been shown that overexpression of sugar transporter genes main-tains osmotic adjustments by regulating sugar flux within

0.0

0.5

1.0

1.5

2.0

2.5

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0


CAB2-Glyma08g08770-FSFlooding

DroughtCAB2-Glyma08g08770-FS

Rela

tive

Exp

ress

ion

ACT

11/F

BOX

Drought0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0.00

0.01

0.02

0.03

0.04

0.05

0.06

FloodingELF3-Glyma14g10530-FS

LUX-Glyma12g06410-FS



Fig. 11. Expression of soybean CAB2, ELF3 and LUX gene transcripts measured by HR-RT–PCR across a 14:10 h diurnal cycle (black and white bars represent light and dark, respectively) under mild, severe, and recovery phases of flooding and drought stress. Data at peak expression points show the mean and SD (n=3). Primers are listed in Supplementary Table S2 at JXB online. S99, S99-2281 (flooding susceptible); 105A, PI 408105A (flooding tolerant); PANA (drought susceptible); 690, PI 567690 (drought tolerant). Values are normalized to the soybean actin gene ACT11 and the FBOX gene, and expressed as relative quantification.






7144 | Syed et al.

cells (Klemens et al., 2013). The reduction in sink demand and disruption of photosynthate transport are well known mechanisms in plants under flooding stress (Lemoine et al., 2013). Up-regulation of sugar effluxer genes (Glyma06g17540, Glyma08g01300, Glyma09g04840, and Glyma15g27530) in a flooding-tolerant genotype might influence sugar homeosta-sis under flooding stress (Lemoine et al., 2013; Mutava et al., 2015). It is noteworthy that Glyma06g17540 was associated with seed sucrose concentration in diverse soybean germ-plasm as revealed by whole-genome re-sequence analysis (Patil et al., 2015).

Higher expression of this gene (Glyma20g30620) in PI 567690 resulted in high total sugar content and the break-down components (fructose and glucose) of sucrose in drought stress conditions (Mutava et al., 2015; and see above). Comparative transcriptome analysis between drought-sus-ceptible and drought-tolerant lines also showed down-reg-ulation of two genes (Glyma05g04290 and Glyma17g14750) associated with invertases involved in the drought-tolerant line under drought. In the drought-tolerant line, most of the LEA genes were found to be down-regulated. Similar down-regulation of LEA genes by ABA was shown in rice under drought conditions (Wang et al., 2007). None of the sugar and LEA genes was identified to be up-regulated in drought and flooding conditions.

Discussion

SUB1 and SNORKEL genes in soybean

Several rice orthologues of SK (15) and SUB1 (three) genes were found in soybean. Intriguingly, none of the SK genes showed any significant expression under drought or flooding conditions among drought- and flooding-tolerant/susceptible lines evaluated in this study (data not shown). Expression of SK genes was either very low or undetectable among all lines under flooding as well as drought stress; hence, they might be pseudo-genes or could have tissue- or condition specific-expression patterns. It can be speculated that many of these SK genes might have lost their function during the evolution of soybean as it is cultivated in largely dry environments. It is notable that soybean-growing areas are now experiencing more onsets of flooding and drought in the same region; how-ever, soybean plants are rarely completely submerged even under flooding conditions. It is therefore not surprising that SK genes are not highly expressed, but their presence is indic-ative that its ancestors might have experienced more flooding in the past. SK genes might have evolved and acquired neo-functionalization and may show significant expression under conditions which have not yet been tested. Circumstantial support for this hypothesis comes from the absence of the SUB1A gene in soybean as inferred from the signature amino acids at positions 13 and 18 in the ERF domain (Xu et al., 2006). SUB1A confers tolerance to prolonged submergence in rice (Xu et al., 2006; Bailey-Serres and Voesenek, 2008; Bailey-Serres et al., 2012), and its absence in soybean indi-cates that SUB1B and SUB1C genes may be more impor-tant in conferring tolerance to partial flooding conditions in

soybean. However, transgenic soybean with the SUB1A rice gene could be an attractive alternative to develop varieties for areas which are frequently flooded on a year by year basis.

Characterization of circadian clock genes in soybean

Due to the polyploid nature of soybean, many orthologues of Arabidopsis circadian clock genes were found in soybean. Recently, Marcolino-Gomes et al. (2014) reported the expres-sion pattern of circadian clock genes under drought in soy-bean and they identified one orthologue of each Arabidopsis circadian clock and clock-related gene. However, several orthologues/paralogues of most clock and clock-related genes were discovered in this study. There are four paralogues of LHY/CCA1-like genes and all are highly conserved in their MYB domain compared with Arabidopsis LHY and CCA1 genes. Marcolino-Gomes et al. (2014) only showed expression of CCA1/LHY-like gene (Glyma07g05410) for a drought-sensitive line (BR16) under normal and drought conditions. Interestingly, Glyma07g05410 has an additional 20 amino acids immediately after the MYB domain and did not show any consistent and significant expression among lines and conditions used in the present study. Robust expression and circadian rhythms of the three remaining LHY/CCA1-like genes were detected (Fig. 7). Three LHY/CCA1-like genes showed much higher expression among drought-tolerant and susceptible lines; however, there are differences in expression patterns among flooding- and drought-specific lines. Similar trends were observed in Arabidopsis plants growing under cold stress and at 20 °C for CCA1 which is closely related to and partly redundant with LHY, which showed expres-sion in the opposite direction. Similarly, Arabidopsis PRR9 is partially redundant with PRR7, but the mRNAs of these genes also respond to cooling in opposite directions (James et al., 2012a, b). LHY and CCA1 can form both homo- and heterodimers (Lu et al., 2009; Yakir et al., 2009), and dif-ferent combinations are likely to have different properties. Furthermore, heterodimers have been shown to bind more strongly to their cognate promoter elements than their homodimers in vitro (O’Neill et al., 2011). Loss of LHY and CCA1 function causes the circadian period to be shortened and the lhy/cca1 double mutant is arrhythmic (Green and Tobin, 1999; Mizoguchi et al., 2002). The presence of at least four paralogues of LHY/CCA1-like genes in soybean and their differential expression patterns among drought- and flooding-tolerant and susceptible lines presents a tantalizing scenario whereby knocking down or overexpression of one or more of the LHY/CCA1-like genes could be used to mimic the function of a drought- or flooding tolerant line in a high-yielding soybean cultivar which lacks tolerance to abiotic stresses.

Differential expression of clock paralogue genes in different soybean lines

To understand the roles of specific clock genes throughout the diurnal cycle, the expression profiles and splicing patterns of all major soybean clock genes were compared in different


soybean lines, and considerable variation was observed. The adaptive significance of correctly regulated circadian tim-ing was correlated with the variation in the period, phase, and amplitude of 150 Arabidopsis accessions and the period length was correlated with the day-length at the latitude of origin (Michael et al., 2003). In Arabidopsis under free run-ning conditions, the rve8-1 mutant and wild type had a very different circadian phase, and 1557 clock-regulated genes were differentially expressed between them (Hsu and Harmer, 2012). However, after adjusting for the phase difference, rve8-1 mutant and wild-type plants exhibited a very similar gene expression pattern, implying that the difference in gene expression was largely due to a phase difference between these two genotypes. Not only were genotype-specific expres-sion patterns of clock and clock-related genes found, but, in addition, expression of several clock genes was undetectable under drought and flooding treatments. This is intriguing and presents interesting prospects for using these genotype-specific genes and variation in their sequence and expression as markers to develop and/or screen flooding- and drought-tolerant soybean varieties. Phase shifts were observed for TOC-1, PRR7, and ELF3 genes, and may fine-tune drought and flooding response in concert with glucose concentration, ABA level/sensitivity, and net photosynthesis (see below). Similar results were observed by Marcolino-Gomes et al. (2014) who reported phase advance of expression for the GmTOC1-like, GmLUX-like, and GmPRR7-like genes under drought conditions for a drought-sensitive genotype (BR16). No phase advance was detected in the drought-sensitive gen-otype; however, in the drought-tolerant genotype expression of the LUX gene was completely abolished under drought conditions but it showed complete recovery when plants were re-watered. These data show that repression of the soybean LUX gene might be involved in growth inhibition under drought conditions in the drought-tolerant line because the evening complex clock proteins (ELF3, ELF4, and LUX) in Arabidopsis repress the expression of PIF4 and PIF5 in a sucrose-dependent manner (reviewed in Hsu and Harmer, 2012). However, this would need further investigation to understand how these clock proteins interact with each other to drive circadian rhythms and growth in soybean under nor-mal as well as flooding and drought conditions.

Flooding- and drought-specific alternative splicing

Intron retention is a common phenomenon in plants, and ~60% of genes are alternatively spliced in Arabidopsis (Marquez et al., 2012) and 16% in soybean (Shen et al., 2014). AS has been shown to regulate gene expression in response to different tissue development (Patil and Nicander, 2013) and environmental factors, including stresses (Reddy, 2007). The AS in the circadian clock genes is widespread and mediates plant responses to cooler temperatures (James et al., 2012a). Using a similar strategy of employing overlapping prim-ers and the HR-RT–PCR system, a number of splice vari-ants were identified for the soybean core clock, SUB1, and ABAR genes. In soybean, 16% of soybean genes are alterna-tively spliced (Wang et al., 2014); however, some AS events

exhibited very interesting expression patterns/abundance which is indicative of their physiological relevance under abiotic stress conditions. It is however noteworthy that dis-covery of AS in soybean was inherently difficult compared with Arabidopsis due to the presence of many paralogues and the polyploid nature of soybean. High conservation of some paralogous genes further exacerbated this prob-lem as overlapping primers across the whole gene sequences could not be designed due to almost identical sequences (Supplementary Fig. S2 at JXB online). One extreme exam-ple is the three ABAR paralogues which, although were dis-tinguishable from each other, were highly conserved between and with Arabidopsis ABAR orthologues. In total, six AS events were found for core clock genes, comprising one each for LHY/CCA1 and PRR5, and four for PRR3. One AS event was also found for the SUB1 and ABAR1 gene. It is notable that genes with many paralogues (LHY/CCA1, ABAR, and TOC1) did not show significant AS, whereas AS in other genes with only one orthologous copy in Arabidopsis showed significant splicing. For example, PRR3 showed an interesting expression and splicing pattern which appears to be flooding and drought specific.

Phase shift in PRR7 and TOC1 may affect flooding- and drought-responsive genes

An interesting diurnal expression pattern of TOC1 and PRR7 genes was found among flooding- and drought-specific lines during flooding and drought conditions. Intriguingly, expres-sion of the TOC1-1 gene was undetectable in flooding-specific lines but showed robust expression and circadian rhythm in drought-susceptible and tolerant lines under normal and drought conditions (Fig. 3). A phase advance of ~4 h for TOC1-2 genes in the flooding-tolerant line and an increase in expression for the drought-tolerant line under mild drought stress were also observed. The expression of TOC1-2, how-ever, significantly drops during both severe flooding and drought stress in the flooding- and drought-tolerant lines. Interestingly, the expression of TOC1-2 peaks early in both drought-susceptible and tolerant lines under severe drought stress compared with these lines growing under normal con-ditions. TOC1-1 also shows a similar pattern of high and low expression under mild and severe drought stress, respec-tively, and recovers again when plants are re-watered. As explained above, phase shift of clock genes has a profound effect on the differentially expressed genes between an rve8-1 (REVEILLE8) mutant and a wild-type line. PRR7 was recently found to be directly involved in abiotic responses in Arabidopsis by repressing the master regulators of growth and development, light signalling, cold, and drought stress genes (Liu et al., 2013). Drought- responsive genes were over-repre-sented (80%) among PRR7 targets (ChIP analysis) compared with 45% genome-wide targets. Similarly, PRR1 (TOC1) also regulates the diurnal expression of ABAR by directly bind-ing to its promoter, and ABA treatment affects the phase of TOC1 binding and expression of ABAR. TOC1 and ABAR overexpression and RNAi lines show defective responses to drought, and highlight the significance of TOC1-mediated


7146 | Syed et al.

clock response to water-limiting conditions (Legnaioli et al., 2009). Intriguingly, phase shifts were found in TOC1-2 as well as in PRR7 under flooding and drought conditions, respec-tively. This change in phase of TOC1-2 and PRR7 might modulate the expression level and target genes involved in flooding and drought response, respectively. To understand the effect of PPR7 phase advance on their targets in soybean, Arabidopsis orthologues of PRR7 targets in soybean were searched to see whether there is any influence of PRR7 phase advance on them in the drought-tolerant and susceptible lines. However, down-regulation of PRR7 targets involved in the drought response was not detected from RNA-seq data from 12:00 h samples. This may be due to the fact that most of the PRR7 targets are morning phased in Arabidopsis and detect-ing expression differences at noon might be difficult. Further analysis of morning samples and ChIP analysis should pro-vide clues as to whether similar mechanisms of PRR7 target genes exist between Arabidopsis and soybean.

Interplay between clock genes and hormone and glucose levels under stress

Besides the expression level and phase changes, the cross-talk between hormone and glucose levels and clock genes also enforces the functionality (or fine tuning) and response to stress conditions. Significant variations in the circadian expression and phase of TOC1 and PRR7 genes were dis-covered in soybean. The high expression of TOC1 coincided with high levels of ABA. The circadian clock is a homeostatic mechanism which has evolved to maintain its period rela-tively constant even under varying environmental conditions which fluctuate rapidly and constantly (James et al., 2012a). Considerable natural variation exists among circadian clock genes as revealed by the correlation between adaptation and correctly regulated circadian timing with the variation in the period, phase, and amplitude of 150 Arabidopsis accessions (Michael et al., 2003). Carbon fixation via photosynthesis is central to the survival and growth of plants; however, effi-cient use of starch reserves during the day/night cycle to drive growth and metabolism is crucial (Graf et al., 2010; Graf and Smith, 2011). Efficient energy utilization becomes even more important when plants experience various abiotic stresses, such as low temperature, flooding, and drought, and conserve and/or enhance their energy utilization to endure or escape these environmental stresses (Xu et al., 2006; Seo et al., 2011). Sugars produced by photosynthesis entrain circadian rhythms in Arabidopsis by regulating the gene expression of circadian clock components early in the photoperiod by defining the so-called metabolic dawn. The oscillations in the endogenous levels of sugars provide metabolic feedback to the central oscillator via PRR7 to maintain robust circadian rhythms in Arabidopsis (Haydon et al., 2013). Furthermore, the circadian clock also controls the rate of degradation of starch and ensures continuous carbon supply during the day/night cycle in Arabidopsis (Graf and Smith, 2011). CCA1 and LHY are essential for the proper utilization of starch during the night because a cca1/lhy1 double mutant has a circadian period of 17 h and consequently degrades starch at faster

rates. Interestingly, when grown under 17 h light/dark cycles, starch degradation in the cca1/lhy mutant leaves is almost lin-ear and completely exhausted by the end of the 8.5 h night (correct matching of the 17 h circadian period of the cca1/lhy mutant with the 17 h light/dark cycle).

TOC1 is a transcriptional repressor and affects functioning of almost all genes in the central oscillator of the Arabidopsis clock (Huang et al., 2012). The overlap between TOC1 and ABA-regulated genes and the gated induction of TOC1 by ABA highlights its importance in physiological responses under drought conditions (Legnaioli et al., 2009; Pokhilko et al., 2013). Furthermore, any perturbations in TOC1 expression will have a pervasive effect on other clock and clock-related genes, in turn affecting many different pathways which confer drought or cold tolerance (Alabadí et al., 2001; Legnaioli et al., 2009; Pruneda-Paz et al., 2009; Gendron et al., 2012; Huang et al., 2012; Pokhilko et al., 2013). In agree-ment with the modelling and experimental data (Legnaioli et al., 2009; Pokhilko et al., 2013), higher concentrations of ABA were observed in the drought-tolerant line in the after-noon, probably triggering higher expression of TOC1. The model proposed by Pokhilko et al. (2013) predicts that the up-regulation of TOC1 by ABA should lengthen the circa-dian period. A broadening of the TOC1 peak under severe stress was observed even under long-day growing conditions (Fig. 3). However, no decipherable period change in CAB2 expression was detected under severe stress accompanied by higher levels of ABA and TOC1 expression. ABA also regu-lates the peak and trough phases of stomatal aperture, and this is also in agreement with previous data from the same experiment. It was previously reported that both flooding- and drought-susceptible lines showed reduced stomatal con-ductance and photosynthesis compared with tolerant lines, with these effects being more pronounced under drought conditions (Mutava et al., 2015). This reduction in photosyn-thesis is mediated by ABA and stomatal conductance under drought stress, while under flooding stress, accumulation of starch granules played a major role. Conservation of energy by increasing starch granules is reminiscent of rice under flooding conditions, where rice plants prolong their survival through the quiescence pathway by conserving energy and slowing down their metabolism (Xu et al., 2006). There was not much difference in glucose content between 08:00 h and 16:00 h in the drought-tolerant line under mild drought stress, whereas under severe drought the glucose content was ~40% higher at 16:00 h compared with 08:00 h. This is indicative of the fine tuning of available glucose concentration under mild and severe drought stress in the early morning and in the afternoon. As reported recently, these endogenous glu-cose levels affect the expression of circadian genes and medi-ate photosynthesis via PRR7 (Haydon et al., 2013). It was observed that expression of PRR7 in both drought-suscep-tible and tolerant lines during mild stress peaked 4 h before that of the lines growing under well-watered conditions. It has already been reported that PRR7 might regulate drought- and cold-responsive genes (Legnaioli et al., 2009) and medi-ate photosynthesis by entraining circadian clock genes using fluctuations in endogenous sugar levels during the day/night


cycle in Arabidopsis (Haydon et al., 2013). GIGANTA (GI) also plays a role in circadian period determination and responses to multiple abiotic stresses (Xie et al., 2015); however, no expression of the soybean version of GI (Glyma09g07240) was detected in all lines and conditions tested (data not shown). Allelic diversity in GI is also responsible for naturally occurring variation in the circadian period in Brassica rapa (Xie et al., 2015). Interestingly, natural variation in GI alleles in Arabidopsis also impacts GI sensitivity to light in the even-ing and its periodicity, albeit partly independent of circadian rhythms, and mediates changes in the expression of PIF4 and growth (de Montaigu et al., 2015). Recent studies pro-pose that circadian clock control represent hubs of important transcriptional networks, and fine tuning of these regulatory networks by mutation of individual circadian clock genes has the potential for systemic effects that produce beneficial agronomic traits/phenotypic variation (Bendix et al., 2015; de Montaigu et al., 2015; Xie et al., 2015). Since soybean has many paralogues of clock genes with interesting interplay with sugar levels and hormones, it provides attractive pros-pects of fine tuning different clock paralogues or exploiting natural variation in the phase and period of these genes to develop more resilient soybean varieties for the future.

Supplementary data

Supplementary data are available at JXB online.Figure S1. ClustalW alignment of Arabidopsis and soybean

LHY/CCA1 genes.Figure S2. ClustalW alignment of Arabidopsis and soybean

ABAR genes.Figure S3. ClustalW alignment of Arabidopsis and soybean

TOC1 genes.Figure S4. ClustalW alignment of rice and soybean

SUB1 genes.Figure S5. ClustalW alignment of rice and soybean SUB1

and SK genes.Figure S6. Branching and leaf formation in flooded soy-

bean plants.Table S1. Alternative splicing in core clock genes detected

by high-resolution RT–PCR.Table S2. Primer sequences.Table S3. Differential expression of SWEET genes in

drought and flooding stress.

AcknowledgementsWe thank the sequencing facility at the University of Missouri, Columbia for running the RNA-seq and HR-RT–PCR samples. Funding support from United Soybean Board, USA is greatly appreciated.

ReferencesAlabadí D, Oyama T, Yanovsky MJ, Harmon FG, Más P, Kay SA. 2001. Reciprocal regulation between TOC1 and LHY/CCA1 within the Arabidopsis circadian clock. Science 293, 880–883.

Bailey-Serres J, Fukao T, Gibbs DJ, Holdsworth MJ, Lee SC, Licausi F, Perata P, Voesenek LACJ, van Dongen JT. 2012. Making sense of low oxygen sensing. Trends in Plant Science 17, 129–138.

Bailey-Serres J, Voesenek LACJ. 2008. Flooding stress: acclimations and genetic diversity. Annual Review of Plant Biology 59, 313–339.

Battaglia M, Covarrubias AA. 2013. Late embryogenesis abundant (LEA) proteins in legumes. Frontiers in Plant Science 4, 190.

Bendix C, Marshall CM, Harmon FG. 2015. Circadian clock genes universally control key agricultural traits. Molecular Plant 8, 1135–1152.

Carter TE, Orf JH, Purcell LC, Specht JE, Boerma HR, Chen P, Sinclair TR, Rufty TW. 2006. Tough times, tough plants—new soybean genes defend against drought and other stresses. Proceedings of the 33rd Soybean Seed Research Conference. CD-ROM, American Seed Trade Association, Alexandria, VA.

Covington MF, Harmer SL. 2007. The circadian clock regulates auxin signaling and responses in Arabidopsis. PLoS Biology 5, 1773–1784.

Covington MF, Maloof JN, Straume M, Kay SA, Harmer SL. 2008. Global transcriptome analysis reveals circadian regulation of key pathways in plant growth and development. Genome Biology 9, R130.

Dalal M, Tayal D, Chinnusamy V, Bansal KC. 2009. Abiotic stress and ABA-inducible Group 4 LEA from Brassica napus plays a key role in salt and drought tolerance. Journal of Biotechnology 139, 137–145.

de Montaigu A, Giakountis A, Rubin M, Tóth R, Cremer F, Sokolova V, Porri A, Reymond M, Weinig C, Coupland G. 2015. Natural diversity in daily rhythms of gene expression contributes to phenotypic variation. Proceedings of the National Academy of Sciences, USA 112, 905–910.

De Ollas C, Hernando B, Arbona V, Gómez-Cadenas A. 2013. Jasmonic acid transient accumulation is needed for abscisic acid increase in citrus roots under drought stress conditions. Physiologia Plantarum 147, 296–306.

Dereeper A, Guignon V, Blanc G, et al. 2008. Phylogeny.fr: robust phylogenetic analysis for the non-specialist. Nucleic Acids Research 36, W465–W469.

Deshmukh R, Sonah H, Patil G, Chen W, Prince S, Mutava R, Vuong T, Valliyodan B, Nguyen HT. 2014. Integrating omic approaches for abiotic stress tolerance in soybean. Frontiers in Plant Science 5, 244.

Dodd AN, Salathia N, Hall A, Kévei E, Tóth R, Nagy F, Hibberd JM, Millar AJ, Webb AAR. 2005. Plant circadian clocks increase photosynthesis, growth, survival, and competitive advantage. Science 309, 630–633.

Fukao T, Bailey-Serres J. 2008. Submergence tolerance conferred by Sub1A is mediated by SLR1 and SLRL1 restriction of gibberellin responses in rice. Proceedings of the National Academy of Sciences, USA 105, 16814–16819.

Fukao T, Xiong L. 2013. Genetic mechanisms conferring adaptation to submergence and drought in rice: simple or complex? Current Opinion in Plant Biology 16, 196–204.

Fukao T, Yeung E, Bailey-Serres J. 2011. The submergence tolerance regulator SUB1A mediates crosstalk between submergence and drought tolerance in rice. The Plant Cell 23, 412–427.

Gendron JM, Pruneda-Paz JL, Doherty CJ, Gross AM, Kang SE, Kay SA. 2012. Arabidopsis circadian clock protein, TOC1, is a DNA-binding transcription factor. Proceedings of the National Academy of Sciences, USA 109, 3167–3172.

Goyal K, Walton LJ, Tunnacliffe A. 2005. LEA proteins prevent protein aggregation due to water stress. Biochemical Journal 388, 151–157.

Graf A, Schlereth A, Stitt M, Smith AM. 2010. Circadian control of carbohydrate availability for growth in Arabidopsis plants at night. Proceedings of the National Academy of Sciences, USA 107, 9458–9463.

Graf A, Smith AM. 2011. Starch and the clock: the dark side of plant productivity. Trends in Plant Science 16, 169–175.

Green RM, Tingay S, Wang Z-Y, Tobin EM. 2002. Circadian rhythms confer a higher level of fitness to Arabidopsis plants. Plant Physiology 129, 576–584.

Green RM, Tobin EM. 1999. Loss of the circadian clock-associated protein 1 in Arabidopsis results in altered clock-regulated gene expression. Proceedings of the National Academy of Sciences, USA 96, 4176–4179.

Harmer SL. 2009. The circadian system in higher plants. Annual Review of Plant Biology 60, 357–377.

Hattori Y, Nagai K, Furukawa S, et al. 2009. The ethylene response factors SNORKEL1 and SNORKEL2 allow rice to adapt to deep water. Nature 460, 1026–1030.










7148 | Syed et al.

Haydon MJ, Mielczarek O, Robertson FC, Hubbard KE, Webb AA. 2013. Photosynthetic entrainment of the Arabidopsis thaliana circadian clock. Nature 502, 689–692.

Hong-Bo S, Zong-Suo L, Ming-An S. 2005. LEA proteins in higher plants: structure, function, gene expression and regulation. Colloids and Surfaces B: Biointerfaces 45, 131–135.

Hsu PY, Harmer SL. 2012. Circadian phase has profound effects on differential expression analysis. PLoS One 7, e49853.

Huang W, Perez-Garcia P, Pokhilko A, Millar AJ, Antoshechkin I, Riechmann JL, Mas P. 2012. Mapping the core of the Arabidopsis circadian clock defines the network structure of the oscillator. Science 336, 75–79.

Hundertmark M, Hincha DK. 2008. LEA (late embryogenesis abundant) proteins and their encoding genes in Arabidopsis thaliana. BMC Genomics 9, 118.

James AB, Syed NH, Bordage S, Marshall J, Nimmo JA, Jenkins GI, Herzyk P, Brown JW, Nimmo HG. 2012a. Alternative splicing mediates responses of the Arabidopsis circadian clock to temperature changes. The Plant Cell 24, 961–981.

James AB, Syed NH, Brown JWS, Nimmo HG. 2012b. Thermoplasticity in the plant circadian clock: how plants tell the time-perature. Plant Signaling and Behavior 7, 1219–1223.

Kalyna M, Simpson CG, Syed NH, et al. 2012. Alternative splicing and nonsense-mediated decay modulate expression of important regulatory genes in Arabidopsis. Nucleic Acids Research 40, 2454–2469.

Klemens PA, Patzke K, Deitmer J, Spinner L, Le Hir R, Bellini C, Bedu M, Chardon F, Krapp A, Neuhaus HE. 2013. Overexpression of the vacuolar sugar carrier AtSWEET16 modifies germination, growth, and stress tolerance in Arabidopsis. Plant Physiology 163, 1338–52.

Legnaioli T, Cuevas J, Mas P. 2009. TOC1 functions as a molecular switch connecting the circadian clock with plant responses to drought. EMBO Journal 28, 3745–3757.

Lemoine R, La Camera S, Atanassova R, et al. 2013. Source-to-sink transport of sugar and regulation by environmental factors. Frontiers in Plant Science 4, 272.

Liu T, Carlsson J, Takeuchi T, Newton L, Farré EM. 2013. Direct regulation of abiotic responses by the Arabidopsis circadian clock component PRR7. The Plant Journal 76, 101–114.

Liu H, Wang H, Gao P, Xu J, Xu T, Wang J, Wang B, Lin C, Fu YF. 2009. Analysis of clock gene homologs using unifoliolates as target organs in soybean (Glycine max). Journal of Plant Physiology 166, 278–289.

Lu SX, Knowles SM, Andronis C, Ong MS, Tobin EM. 2009. CIRCADIAN CLOCK ASSOCIATED1 and LATE ELONGATED HYPOCOTYL function synergistically in the circadian clock of Arabidopsis. Plant Physiology 150, 834–843.

Manavalan LP, Guttikonda SK, Tran L-S, Nguyen HT. 2009. Physiological and molecular approaches to improve drought resistance in soybean. Plant and Cell Physiology 50, 1260–1276.

Marcolino-Gomes J, Rodrigues FA, Fuganti-Pagliarini R, Bendix C, Nakayama TJ, Celaya B, Molinari HBC, de Oliveira MCN, Harmon FG, Nepomuceno A. 2014. Diurnal oscillations of soybean circadian clock and drought responsive genes. PLoS One 9, e86402.

Marquez Y, Brown JWS, Simpson C, Barta A, Kalyna M. 2012. Transcriptome survey reveals increased complexity of the alternative splicing landscape in Arabidopsis. Genome Research 22, 1184–1195.

Matsui A, Ishida J, Morosawa T, et al. 2008. Arabidopsis transcriptome analysis under drought, cold, high-salinity and ABA treatment conditions using a tiling array. Plant and Cell Physiology 49, 1135–1149.

Michael TP, McClung CR. 2003. Enhancer trapping reveals widespread circadian clock transcriptional control in Arabidopsis. Plant Physiology 132, 629–639.

Michael TP, Salomé PA, Yu HJ, Spencer TR, Sharp EL, McPeek MA, Alonso JM, Ecker JR, McClung CR. 2003. Enhanced fitness conferred by naturally occurring variation in the circadian clock. Science 302, 1049–1053.

Mizoguchi T, Wheatley K, Hanzawa Y, Wright L, Mizoguchi M, Song HR, Carré IA, Coupland G. 2002. LHY and CCA1 are partially redundant genes required to maintain circadian rhythms in Arabidopsis. Developmental Cell 2, 629–641.

Mizuno T, Yamashino T. 2008. Comparative transcriptome of diurnally oscillating genes and hormone-responsive genes in Arabidopsis thaliana:

insight into circadian clock-controlled daily responses to common ambient stresses in plants. Plant and Cell Physiology 49, 481–487.

Mutava RN, Prince SJK, Syed NH, Song L, Valliyodan B, Chen W, Nguyen HT. 2015. Understanding abiotic stress tolerance mechanisms in soybean: a comparative evaluation of soybean response to drought and flooding stress. Plant Physiology and Biochemistry 86, 109–120.

Nguyen VT, Vuong TD, VanToai T, Lee JD, Wu X, Mian MAR, Dorrance AE, Shannon JG, Nguyen HT. 2012. Mapping of quantitative trait loci associated with resistance to Phytophthora sojae and flooding tolerance in soybean. Crop Science 52, 2481–2493.

Ni Z, Kim E-D, Ha M, Lackey E, Liu J, Zhang Y, Sun Q, Chen ZJ. 2009. Altered circadian rhythms regulate growth vigour in hybrids and allopolyploids. Nature 457, 327–331.

O’Neill JS, van Ooijen G, Le Bihan T, Millar AJ. 2011. Circadian clock parameter measurement: characterization of clock transcription factors using surface plasmon resonance. Journal of Biological Rhythms 26, 91–98.

Park M-J, Seo PJ, Park C-M. 2012. CCA1 alternative splicing as a way of linking the circadian clock to temperature response in Arabidopsis. Plant Signaling and Behavior 7, 1194–1196.

Patil G, Nicander B. 2013. Identification of two additional members of the tRNA isopentenyltransferase family in Physcomitrella patens. Plant Molecular Biology 82, 417–426.

Patil G, Valliyodan B, Deshmukh R, et al. 2015. Soybean (Glycine max) SWEET gene family: insights through comparative genomics, transcriptome profiling and whole genome re-sequence analysis. BMC Genomics 16, 520.

Perata P, Voesenek LACJ. 2007. Submergence tolerance in rice requires Sub1A, an ethylene-response-factor-like gene. Trends in Plant Science 12, 43–46.

Pokhilko A, Mas P, Millar AJ. 2013. Modelling the widespread effects of TOC1 signalling on the plant circadian clock and its outputs. BMC Systems Biology 7, 23.

Pruneda-Paz JL, Breton G, Para A, Kay SA. 2009. A functional genomics approach reveals CHE as a component of the Arabidopsis circadian clock. Science 323, 1481–1485.

Qazi HA, Rao PS, Kashikar A, Suprasanna P. 2014. Alterations in stem sugar content and metabolism in sorghum genotypes subjected to drought stress. Functional Plant Biology 41, 954–962.

Reddy ASN. 2007. Alternative splicing of pre-messenger RNAs in plants in the genomic era. Annual Review of Plant Biology 58, 267–294.

Rozen S, Skaletsky HJ. 1998. Primer3. bioinformatics methods and protocols. Methods in Molecular Biology 3, 1–41.

Sanchez A, Shin J, Davis SJ. 2011. Abiotic stress and the plant circadian clock. Plant Signaling and Behavior 6, 223–231.

Sayama T, Nakazaki T, Ishikawa G, et al. 2009. QTL analysis of seed-flooding tolerance in soybean (Glycine max [L.] Merr.). Plant Science 176, 514–521.

Seo PJ, Kim MJ, Ryu J-Y, Jeong E-Y, Park C-M. 2011. Two splice variants of the IDD14 transcription factor competitively form nonfunctional heterodimers which may regulate starch metabolism. Nature Communications 2, 303.

Shen Y, Zhou Z, Wang Z, et al. 2014. Global dissection of alternative splicing in paleopolyploid soybean. The Plant Cell 26, 996–1008.

Simpson CG, Fuller J, Maronova M, Kalyna M, Davidson D, McNicol J, Barta A, Brown JWS. 2008. Monitoring changes in alternative precursor messenger RNA splicing in multiple gene transcripts. The Plant Journal 53, 1035–1048.

Valliyodan B, Nguyen HT. 2006. Understanding regulatory networks and engineering for enhanced drought tolerance in plants. Current Opinion in Plant Biology 9, 189–195.

Valliyodan B, Shi H, Nguyen HT. 2015. A simple analytical method for high-throughput screening of major sugars from soybean by normal-phase HPLC with evaporative light scattering detection. Chromatography Research International 2015, 757649.

VanToai TT, Beuerlein AF, Schmitthenner SK, St. Martin SK. 1994. Genetic variability for flooding tolerance in soybeans. Crop Science 34, 1112.


VanToai TT, Lee JD, Goulart PFP, Shannon JG, Alves JD, Nguyen HT, Yu O, Rahman M, Islam R. 2012. Soybean (Glycine max L. Merr.) seed composition response to soil flooding stress. Journal of Food Agriculture and Environment 10, 795–801.

VanToai TT, St. Martin SK, Chase K, Boru G, Schnipke V, Schmitthenner AF, Lark KG. 2001. Identification of a QTL associated with tolerance of soybean to soil waterlogging. Crop Science 41, 1247.

Wang L, Cao C, Ma Q, et al. 2014. RNA-seq analyses of multiple meristems of soybean: novel and alternative transcripts, evolutionary and functional implications. BMC Plant Biology 14, 169.

Wang X-S, Zhu H-B, Jin G-L, Liu H-L, Wu W-R, Zhu J. 2007. Genome-scale identification and analysis of LEA genes in rice (Oryza sativa L.). Plant Science 172, 414–420.

Winkel-Shirley B. 2002. Biosynthesis of flavonoids and effects of stress. Current Opinion in Plant Biology 5, 218–223.

Xie Q, Lou P, Hermand V, et al. 2015. Allelic polymorphism of GIGANTEA is responsible for naturally occurring variation in circadian period in Brassica rapa. Proceedings of the National Academy of Sciences, USA 112, 3829–3834.

Xu K, Xu X, Fukao T, Canlas P, Maghirang-Rodriguez R, Heuer S, Ismail AM, Bailey-Serres J, Ronald PC, Mackill DJ. 2006. Sub1A is an ethylene-response-factor-like gene that confers submergence tolerance to rice. Nature 442, 705–708.

Yakir E, Hilman D, Kron I, Hassidim M, Melamed-Book N, Green RM. 2009. Posttranslational regulation of CIRCADIAN CLOCK ASSOCIATED1 in the circadian oscillator of Arabidopsis. Plant Physiology 150, 844–857.

Documents

Core clock, SUB1, and ABAR genes mediate flooding …...7130 | Syed et al. breeding and development programmes from around the world have identified drought-tolerant soybean lines