- Top of page
- Materials and Methods
- Supporting Information
Flooding drastically reduces rice yields worldwide. Although rice is adapted to standing water, most rice varieties are susceptible to complete submergence resulting from transient flooding in lowland rice growing regions. During submergence, the diffusion rate of carbon dioxide and oxygen between the plant and its environment is significantly reduced (Armstrong, 1980; Colmer, 2003). The resulting hypoxic condition limits photosynthesis, necessitating a metabolic switch to glycolysis and anaerobic respiration (Setter et al., 1989, 1997; Voesenek et al., 2006). Flood intolerant rice varieties typically exhibit rapid elongation upon prolonged submergence, consuming their carbohydrate reserves during the growth phase, and plants that breach the water surface are usually spindly and prone to lodging once the flood waters recede (Ismail et al., 2009). Most intolerant genotypes do not survive several days of submergence. However, a lowland flood-tolerant landrace, Flood Resistant 13A (FR13A) can survive up to 2 wk of submergence and still recover to resume growth (Mackill et al., 1993). FR13A adopts a quiescent strategy that restricts growth, conserving energy until flood waters recede (Xu et al., 2006; Fukao & Bailey-Serres, 2008).
The quantitative trait locus (QTL) primarily responsible for the tolerance of FR13A was mapped to a small region called Sub1 (Xu & Mackill, 1996; Xu et al., 2006; Hattori et al., 2009). Flood intolerant rice varieties become tolerant after introgression of the Sub1 locus, demonstrating that the Sub1 QTL is sufficient for flood tolerance (Xu & Mackill, 1996). The Sub1 locus from FR13A and bioengineered tolerant genotypes with the introgressed Sub1 locus has three Ethylene Response Factor (ERF) genes: SUB1A, SUB1B and SUB1C (Xu et al., 2006). All three genes belong to a large transcription factor family referred to as ERFs and they group within the abiotic stress responding ERF subfamily VII (Nakano et al., 2006; Jung et al., 2010). The intolerant parent variety M202 used in previous submergence studies and in the experiments discussed here has the SUB1B and SUB1C genes but lacks the SUB1A gene. The tolerant near isogenic line used in our experiments is the M202-Sub1 that has the Sub1 locus from FR13A containing the SUB1A gene that confers submergence tolerance (Fukao et al., 2006; Xu et al., 2006). Upon complete submergence, SUB1A is rapidly induced within 1 d. Expression of SUB1C is induced more strongly in the intolerant M202 when compared to the tolerant M202-Sub1. SUB1B transcript abundance was reported to be induced by submergence in both genotypes but its transcript abundance was slightly higher in M202-Sub1 compared to M202 (Fukao et al., 2006). Expression of the ubiquitin promoter driven SUB1A transgene (LG(SUB1A)) in an intolerant variety, Liaogeng (LG) was sufficient to introduce the submergence tolerance response, providing conclusive evidence that SUB1A was the tolerance conferring gene in the FR13A Sub1 haplotype (Xu et al., 2006).
In contrast to the quiescent strategy exhibited by FR13A, deep water rice survives season-long flooding by rapidly elongating internodes to maintain contact with air. This strategy is dependent upon a quantitative trait locus encoding two major genes, SNORKEL1 and SNORKEL2 (SK1 and SK2) (Hattori et al., 2009). Similar to Sub1 locus, SK1 and SK2 are two ERFs and their expression is induced by submergence stress and the gaseous plant hormone ethylene. Submergence decreases the rate of diffusion of ethylene, which accumulates inside the plants and leads to induction of several members of the ERF family including SK1, SK2, SUB1A and SUB1C among others. However, the downstream mechanisms directly affecting both SUB1A and SNORKEL-mediated tolerance is reported to be dependent upon the plant hormone gibberellin (GA), which is critical for many aspects of plant growth and development as well as stress adaptation (Gao et al., 2011). Deep water rice has elevated GA levels that contribute to the rapid growth of internodes and leaves. Although GA content has not been reported for SUB1A-expressing plants, it was suggested that the GA-signaling mechanism is altered due to increased protein concentrations of the GA signal-repressing GRAS family transcription factors SLR1 (SLENDER RICE1) and SLRL1 (SLR1-Like1) during submergence (Fukao & Bailey-Serres, 2008). SLR1 is a DELLA domain protein that is destabilized by GA (Itoh et al., 2002). This occurs when the GA receptor, GID1 (Gibberellin Insensitive Dwarf), binds GA and promotes interaction of GID1-GA with SLR1. Next, the DELLA domain of SLR1 is recognized by the SCFSLY1 E3 ubiquitin ligase complex, which results in the poly-ubiquitination and subsequent degradation of SLR1 through the 26S proteasome. GA-mediated derepression of growth can also occur independently of SLR1 degradation (Ariizumi et al., 2008). This may occur in part through DELLA protein modification, or through protein–protein interactions with factors that may block DELLA function (Sun, 2011).
The intolerant variety, M202 and the corresponding tolerant line M202-Sub1 have been used in several studies to identify SUB1A-specific gene expression profiles (Fukao et al., 2006; Jung et al., 2010). Expression analyses have shown that genes associated with anaerobic respiration and antioxidant protection were upregulated during submergence in M202-Sub1, consistent with the enhanced ability to adjust to submergence stress. These studies indicated that submergence alters hormone levels as well as transcript abundance of genes associated with hormone homeostasis (both synthesis and catabolic genes). Significant changes were reported for ethylene, ABA and cytokinin pathways. Ethylene levels of both rice lines increase when submerged, but were slightly lower in M202-Sub1 relative to M202 (Fukao et al., 2006). Ethylene can reduce levels of ABA by inducing the expression of an ABA catabolic enzyme (Fukao & Bailey-Serres, 2008). Although M202-Sub1 uniquely exhibits ABA hypersensitivity, ABA content is reduced to comparable concentrations in both genotypes during submergence (Fukao & Bailey-Serres, 2008). The ABA hypersensitivity of M202-Sub1 probably reflects an increase in ABA-responsive genes, which are proposed to prevent dehydration after post-submergence by repressing SUB1A expression and inducing genes that prevent water loss (Fukao et al., 2011).
Another hormone regulating cell expansion and plant height is brassinosteroid (BR), but its role in rice submergence tolerance has not been addressed. The plasma membrane-localized receptor Brassinosteroid insensitive (BRI1), a member of the leucine-rich repeat subfamily of receptor-like kinases (LRR-RLK), is primarily responsible for BR recognition in Arabidopsis thaliana. Downstream BR-signaling is dependent upon the transcription factors Brassinozole resistant 1 (BZR1) and Bri1 EMS suppressor 1 (BES1), which bind BR-response element (BRRE) or E-box cis elements in the promoters of BR-regulated genes (He et al., 2005; Yin et al., 2005). In the absence of BR, a kinase called Brassinosteroid insensitive 2 (BIN2) negatively regulates BR-signaling through the phosphorylation of BZR1 and BES1, targeting them for degradation. In rice, studies have shown that OsBRI1, OsBZR1,and GSK3-like/BIN2 have similar function to their Arabidopsis orthologs (Yamamuro et al., 2000; Bai et al., 2007; Koh et al., 2007; Nakagawa et al., 2011; Tong et al., 2012). Several BR biosynthesis genes, such as DWF4 and CPD, as well as the BRI1 receptor are regulated by BR in a negative feedback manner (Nakagawa et al., 2011). BR biosynthesis genes are upregulated in BR signaling mutants and BR biosynthesis mutants become hypersensitive to exogenous BR due to upregulation of BRI1. Although BR-deficient mutants have reduced size similar to GA biosynthesis mutants, they are uniquely identifiable by a reduction in lamina leaf bending (Nakagawa et al., 2011). Recent reports suggest that GA and BR pathways coordinate to regulate plant growth through DELLA proteins, BZR1, and other transcription factors (Bai et al., 2012; Hao et al., 2012; Oh et al., 2012; Yang et al., 2012). Given that abundance of SLR1, a rice DELLA protein was positively correlated with SUB1A expression, it remains to be determined whether BR could be involved in submergence response in rice (Fukao & Bailey-Serres, 2008).
In this study, we investigated the molecular mechanisms involved in submergence tolerance response in rice. Specifically, our transcriptome analysis revealed a SUB1A-dependent BR response during submergence. We present evidence that BR biosynthesis genes and endogenous BR levels are differentially regulated in the SUB1A rice line compared to the intolerant line during submergence. Experiments suggest that BR could be negatively affecting GA levels during transient submergence. Based on our experiments, we propose a model for SUB1A action during submergence that highlights the role of BR and GA crosstalk in restricting shoot elongation during submergence.
- Top of page
- Materials and Methods
- Supporting Information
A key feature of the SUB1A-mediated response is restricting rice shoot growth during complete submergence to conserve energy until flood waters recede. Shoot growth is driven by cell elongation and/or cell division and responds to environmental factors such as light, temperature, and hormones such as gibberellins, auxin and brassinosteroids. The SUB1A response was linked to GA signaling resulting in decreased shoot elongation when submerged (Fukao & Bailey-Serres, 2008). Our results suggest that brassinosteroids could be involved in SUB1A-mediated dampening of the GA responses during submergence. BR significantly restricts shoot elongation in M202 plants during submergence, an effect similar to the SUB1A-restricted shoot elongation in M202-Sub1 plants (Fig. 4). Consistent with the pretreatment submergence phenotype and lamina leaf bending, we found that BR levels were higher in M202-Sub1 plants during submergence relative to M202. BR treatment increased the expression of a GA catabolic gene (GA2ox7), which is induced in a SUB1A-dependent manner during submergence (Fig. 1c; Jung et al., 2010). Our observations are consistent with another study where BR treatment strongly induced the transcript abundance of another GA catabolic gene, GA2ox3 (De Vleesschauwer et al., 2012). We postulate that BR could decrease bioactive levels of GA by inducing GA2ox7 as part of an early submergence response. Increased bioactive GA, especially GA4, is known to promote SLR1 protein degradation in rice (Ueguchi-Tanaka et al., 2007). These findings suggest that SUB1A-mediated submergence tolerance likely involves crosstalk between the BR and GA pathways.
In Arabidopsis, GA and BR regulate growth interdependently via direct interaction between the BR activated downstream protein BZR1 and the GA degraded DELLA protein (Bai et al., 2012; Gallego-Bartolomé et al., 2012). BZR1 was required for GA-promoted hypocotyl elongation (Bai et al., 2012). Although existence of these interactions remains to be examined in rice, it is likely that some interhormonal crosstalk will be conserved across species. SLR1 induction by BR treatment was observed in both genotypes and is independent of the SUB1A presence (Fig. 5). Although, the SLR1 induction was not different between M202 and M202-Sub1 plants, it is pertinent to point out that during submergence M202-Sub1 plants have higher BR levels compared to M202. Therefore, it is possible that higher SLR1 levels observed by others are a consequence of increased endogenous BR in M202-Sub1 plants. Based on our data we cannot directly associate SLR1 with the crosstalk between BR and GA in the M202-Sub1 plants. Although the exact mechanism remains to be determined, it is likely that BR repression of GA responses is via SLR1 protein induction and/or stability rather than transcript level induction of SLR1. For instance, in the study on the role of JA in repressing GA regulated growth, JA treatment increased the protein stability and had no effect on SLR1 transcript abundance (Yang et al., 2012). In our array analysis and quantitative PCR assays, we also did not detect a significant difference in SLR1 transcript levels between M202 and M202-Sub1 during submergence. We observed that the SLR1 protein levels decrease after 3 d of BR treatment in our time-series experiment. During submergence, BR levels begin to increase within 1 d in M202-Sub1, whereas SLR1 levels increase at later stages of the 2-wk submergence treatment (Fukao & Bailey-Serres, 2008). It is possible that a specific concentration of endogenous BR needs to accumulate during submergence before SLR1 is induced in M202-Sub1 plants. Although we did not observe a differential transcript regulation for SLRL1 between M202 and M202-Sub1, it could still be an important regulator of SUB1A-mediated submergence tolerance with higher protein induction and/or stabilization in M202-Sub1.
Model for SUB1A action
Our results suggest that BR homeostasis is differentially regulated in the SUB1A plants during submergence compared to M202. A model for SUB1A action that integrates our experimental observations with those by others is shown in Fig. 8 (Fukao & Bailey-Serres, 2008; Jung et al., 2010). Increased ethylene levels in submerged plants trigger SUB1A gene expression in M202-Sub1 (Fukao et al., 2006). Higher SUB1A expression dampens ethylene production (Fukao et al., 2006; Fukao & Bailey-Serres, 2008). Within 1 d of submergence, transcript abundance of two BR biosynthesis genes DWF4 and DWF1 is increased, resulting in higher BR levels in M202-Sub1 plants. In striking contrast, BR biosynthesis gene expression is repressed during submergence in M202. Higher BR levels could further augment SUB1A expression during submergence. We propose that increased BR levels during submergence may modulate the GA responses in a temporal manner during transient submergence events. As part of an early response (within 1 d after submergence), BR promotes the expression of GA2ox7, which catabolizes GA thus reducing the endogenous GA4 levels in M202-Sub1 plants. Upon continued submergence, gradual accumulation of BR induces SLR1 at the protein level. Because GA2ox7 reduces GA levels as an early response to submergence, reduced GA levels diminish the GID1-SLR1 interaction required for SLR1 degradation. Given, that differential elongation is observed fairly early during submergence, it may be that the BR-induced SLR1 induction and stabilization in M202-Sub1 is a downstream effect of altered GA and BR homeostasis occurring as early as 1 d after submergence (Fukao & Bailey-Serres, 2008). Based on previous report, the protein levels of SLRL1, which is resistant to GA-mediated degradation (Itoh et al., 2005), could also be important for SUB1A action (Fukao & Bailey-Serres, 2008). The relationship between BR and SLRL1 induction during submergence needs to be explored in future studies.
Figure 8. Model for SUB1A action. Based on our experimental data, we propose a model for SUB1A action in rice during submergence. Increased ethylene concentration during submergence induces SUB1A (an ethylene response factor) within 1 d of submergence. SUB1A induction under submerged conditions increases the level of brassinosteroid (BR). Increased BR levels can potentially induce and/or stabilize SLR1, the negative regulator of gibberellic acid (GA) responses such as shoot elongation. Increased BR levels in the SUB1A plants also increase the expression of a GA catabolic gene, GA2ox7. Expression of GA2ox7 inversely correlates with the GA level during submergence in the SUB1A plants. In summary, we hypothesize that increased BR levels in SUB1A genotype could restrict shoot elongation via repressing GA signaling responses and by decreasing GA levels.
Download figure to PowerPoint
In our experiments, we focused on shoot elongation during submergence as the key phenotype for submergence tolerance. However, submergence tolerance involves several components critical for submergence recovery such as optimal carbohydrate level regulation, scavenging of reactive oxygen species and minimizing chlorophyll degradation among others. Future studies will focus on hormonal regulation of some of these components. Besides creating a hypoxic condition, complete inundation also alters the light, temperature and carbon status of the rice plants. Based on recent reports from Arabidopsis, it is likely that changes in light status would be a critical component in the crosstalk between BR and GA pathways that restricts growth of M202-Sub1 plants during submergence. In fact, the Arabidopsis ortholog (At4 g21200) of the GA catabolic gene GA2ox7 is induced by low light and dark conditions based on transcriptome data from multiple independent experiments catalogued in Genevestigator (Zimmermann et al., 2008). Future experiments will elucidate the possible role of light in mediating BR and GA crosstalk in SUB1A genotypes during submergence.