Keywords

9.1 Introduction

The gibberellins (GAs) form a family of tetracylic diterpenoid plant hormones which impinge on various aspects of plant growth and development, from germination through stem, hypocotyl and root growth to the switch from vegetative to reproductive growth (Sun and Gubler 2004; Jiang and Fu 2007; Gao et al. 2008, 2011). The analysis of GA-insensitive mutants in both Arabidopsis thaliana and rice has identified a number of GA signaling components, in particular, the GA receptor GID1 (GIBBERELLIN INSENSITIVE DWARF1), the DELLA proteins (the major repressors of GA signaling), and F-box-containing proteins such as SLY1 [SLEEPY1] in A. thaliana and GID2 [GIBBERELLIN INSENSITIVE DWARF2] in rice (Dill et al. 2004; Fu et al. 2004; Ueguchi-Tanaka et al. 2007; Itoh et al. 2008; Shimada et al. 2008). On this basis, a derepression model for GA signaling has been elaborated, according to which once bioactive GA has been perceived and bound by GID1, the resulting complex can drive the formation of a GA-GID1-DELLA protein complex capable of efficiently binding the E3 ubiquitin ligase SCFSLY1/GID2 complex (Griffiths et al. 2006). This binding results in the polyubiquitination of the DELLA proteins, thereby directing their degradation via the 26S proteasome, and so relieving the growth suppression exerted by them (Fig. 9.1) (Gao et al. 2011; Daviere and Achard 2013).

Fig. 9.1
figure 1

The GA signaling pathway involving DELLA proteins. In the absence of GA, the DELLA proteins repress plant growth. The binding of GA to its receptor GID1 permits the interaction between GID1 and DELLA. The formation of the GID1-GA-DELLA complex enhances the interaction between DELLA and the SCFSLY1/GID2 F-box component, leading to the polyubiquitination of DELLA and their targeting for degradation via the 26S proteasome pathway. M, post-translationally modified DELLA; U, ubiquitinated DELLA

Full size image

The gas ethylene also regulates a wide range of growth and developmental processes. It is perceived by the ETR1 (ETHYLENE RESPONSE1) family of ethylene receptors (Hua et al. 1995, 1998; Hua and Meyerowitz 1998; Sakai et al. 1998; Guo and Ecker 2004). In its absence, ETR1 activates CTR1 (CONSTITUTIVE TRIPLE RESPONSE1), a Raf-like Ser/Thr protein kinase which suppresses ethylene signaling (Kieber et al. 1993; Huang et al. 2003; Mayerhofer et al. 2012). EIN2 (ETHYLENE INSENSITIVE2) acts downstream of CTR1 and represents a critical component of the ethylene signaling pathway (Alonso et al. 1999). In the presence of ethylene, its receptors become inactivated, in turn switching off CTR1 and permitting EIN2 to function. The basis for the EIN2-mediated transduction of the ethylene signal from the ethylene receptors associated with the endoplasmic reticulum to its downstream transcription factors EIN3 and EIL1 (EIN3-LIKE1) has been described independently by three research groups (Ju et al. 2012; Qiao et al. 2012; Wen et al. 2012). EIN3 and EIL1 mediate a wide array of the plant responses to ethylene (Chao et al. 1997; Solano et al. 1998; Stepanova and Alonso 2009). The levels of EIN3 and EIL1 in the nucleus are finely tuned by the Skp1-Cullin-F-box protein (SCF) E3 ubiquitin ligase isoforms EBF1 and EBF2 (Guo and Ecker 2003; Potuschak et al. 2003; Gagne et al. 2004).

The existence of crosstalk between ethylene and GA signaling has been recognized for a long time, but its molecular basis remains to be determined. Both ethylene and GA affect one another at the level of synthesis, signaling and gene expression (De Grauwe et al. 2007, 2008; Dugardeyn et al. 2008). In this chapter, we focus on the interaction between ethylene and GA signaling in the context of the regulation of germination, root growth, hypocotyl elongation, apical hook development, and floral induction (Fig. 9.2). In addition, we also describe their role in the plant response to abiotic stress (Fig. 9.3).

Fig. 9.2
figure 2

The modulation of A. thaliana growth and development, as controlled by ethylene and GA (modified from Weiss and Ori (2007)). Ethylene and GA in concert promote germination and induce the elongation of the hypocotyl in the presence of light. Ethylene and GA also together regulate apical hook development in the absence of light by up-regulating HLS1. Ethylene stabilizes EIN3/EIL1, while GA relieves the DELLA-imposed repression on EIN3/EIL1. EIN3/EIL1 links the ethylene and GA pathways to activate HLS1 transcription. Ethylene inhibits root growth by controlling the abundance/stability of DELLA. Activation of ethylene signaling reduces the level of bioactive GA, thereby promoting the accumulation of DELLA, which in turn delays flowering via the regulation of LFY and SOC1 transcription. CTR1 CONSTITUTIVE TRIPLE RESPONSE1; EIN2 and EIN3, ETHYLENE INSENSITIVE2 and 3; EIL1, EIN3-LIKE1; HLS1 HOOKLESS1; LFY LEAFY; SOC1 SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1

Full size image
Fig. 9.3
figure 3

The rice ethylene and GA regulatory network during an episode of abiotic stress (modified from Bailey-Serres and Voesenek (2010)). In deepwater rice, ethylene up-regulates SK1 and SK2 and elevates the level of GA, thereby inducing a rapid elongation of the stem internode, allowing the canopy to rise above the water’s surface. In submergence-tolerant rice, ethylene activates Sub1A, which promotes DELLA accumulation, enhancing survival by limiting stem elongation and restricting assimilate consumption. Sub1A also promotes the tolerance of mild osmotic stress by preventing the build-up of damaging levels of reactive oxygen species. Mild osmotic and salinity stress activates ethylene signaling and stabilizes the DELLA proteins, which act to limit plant growth. ERF1 ETHYLENE-RESPONSIVE FACTOR; ROS reactive oxygen species; SK1 and 2 SNORKEL1 and 2; Sub1A Submergence 1A

Full size image

9.2 Ethylene and GA Coordinately Promote Seed Germination

Seed of the A. thaliana GA-deficient mutant ga1-3 (which synthesizes little or no ent-kaurene because it lacks a functional copy of GA1) is unable to germinate without the provision of exogenous GA. Ethylene influences germination in A. thaliana (Kepczynski and Kepczynska 1997), since the ethylene synthesis precursor 1-aminocyclopropane-1-carboxylic acid (ACC) can compensate for the absence of GA in the ga1-3 seed, allowing it to germinate without GA supplementation in the light; its promotive effect is less marked in the dark (Karssen et al. 1989; Vriezen et al. 2004). Several ethylene response genes, including ACO (encoding 1-AMINOCYCLOPROPANE-1-CARBOXYLIC ACID OXIDASE), are up-regulated in imbibed ga1-3 seed following the exogenous supply of GA4 (Ogawa et al. 2003). The transcription of both the ethylene-inducible gene HOOKLESS1 (HLS1) (Lehman et al. 1996) and the ethylene receptor gene family’s ERS1 (ETHYLENE RESPONSE SENSOR1) (Hua et al. 1998) is similarly increased by the exogenous supply of GA4 (Ogawa et al. 2003). The implication is that GA may activate ethylene synthesis and/or production, thereby promoting the ethylene responses in the imbibed A. thaliana seed.

The dominant A. thaliana etr1-1 mutant is ethylene insensitive. Freshly harvested etr1-1 seed germinates less readily than does wild-type (WT) seed (Beaudoin et al. 2000). Only bioactive GA4 is present in mature WT seed at a concentration of 25 ng/g dry weight (DW), while the concentrations of GA1, GA4 and GA7 are all higher in the etr1-2 mutant seed than in WT seed. WT seed accumulates GA1 over the first 18 h of germination, but thereafter its concentration declines, with a concomitant modest rise in the concentration of GA4. Germinating and germinated WT seed contains 7–12 ng/g DW GAs, whereas etr1-2 seed contains more GA1 and GA4 (Chiwocha et al. 2005). Thus, it appears that while GA may promote ethylene synthesis in imbibed ga1-3 seed, the enhanced level of GA present in etr1 seed indicates that ethylene may suppress GA synthesis. The indication is that a feedback mechanism regulates GA and ethylene synthesis, and that these two phytohormones synergistically influence germination (Fig. 9.2).

9.3 Ethylene Inhibits A. thaliana Root Growth by Interacting with DELLA

The DELLA proteins are members of the plant-specific GRAS family [named from the first three members of the family to be identified, namely GAI (GIBBERELLIN INSENSITIVE), RGA (REPRESSOR OF GA1-3), and SCARECROW] and are thought to represent transcriptional regulators. The A. thaliana DELLA proteins include GAI, RGA, RGL1 (RGA-LIKE1), RGL2, and RGL3 (Jiang and Fu 2007), all of which act to repress GA signaling (Dill et al. 2001; Silverstone et al. 2001; Fu and Harberd 2003). The root is a key organ because it is responsible for the acquisition of water and nutrients. The ga1-3 mutant develops a shortened primary root, which can be made to grow to a WT length either by providing the plant with exogenous GA or by knocking out DELLA. These observations show that the GA-DELLA regulatory system operates in the context of seedling root growth (Fu and Harberd 2003). Ethylene also inhibits root growth. When A. thaliana seedlings are exposed to either ethylene or ACC, primary root growth is inhibited (Abeles et al. 1992), and this inhibition involves the DELLA proteins GAI and RGA (Achard et al. 2003) (Fig. 9.2). When ACC is not supplied to seedlings of DELLA protein-lacking mutants (gai-t6, rga-24, and gai-t6 rga-24), the root length is identical to that of the WT. Conversely, in the presence of ACC treatment, the WT roots are shorter than those of rga-24 seedlings, and the double mutant gai-t6 rga-24 roots grow even longer than those of seedlings of the two single mutants. These data suggest that GAI and RGA together mediate ethylene-induced root growth inhibition. In addition, GA treatment can overcome the ACC-induced inhibition of seedling root growth (Fig. 9.2).

GFP (green fluorescent protein)-RGA fusion protein is detectable in root cell nuclei, but rapidly disappears in response to GA treatment (Silverstone et al. 2001). Ethylene appears to inhibit root growth in A. thaliana by delaying the GA-mediated degradation of GFP-RGA (Achard et al. 2003) (Fig. 9.2). When pRGA::GFP-RGA containing transgenic plants are treated with GA, the GFP signal is markedly attenuated within 90 min at both the root tip and the root elongation zone and disappears completely by 3 h. However, when the same seedlings are grown in an atmosphere containing ethylene gas, the fusion protein becomes more stable, remaining readily detectable even 3 h after the GA treatment. Ethylene promotes many responses which are antagonized by CTR1, and the GFP-RGA fusion product is less readily degraded as a result of GA treatment in the mutant than in WT plants. The implication is that ethylene delays the GA-mediated degradation of GFP-RGA via a CTR1-dependent signaling pathway (Achard et al. 2003) (Fig. 9.2).

9.4 Ethylene and GA Synergistically Induce the Elongation of the A. thaliana Hypocotyl

In A. thaliana, hypocotyl elongation is achieved predominantly via cell elongation and is tightly regulated by various phytohormones. GA promotes the process, whereas ethylene inhibits it (Cowling and Harberd 1999; Collett et al. 2000). Ethylene has been shown to stimulate hypocotyl elongation when seedlings are grown in a low nutrient medium (LNM) (Smalle et al. 1997). In this situation, most of the elongation occurs over the first three days following imbibition. Treatment with ACC extends the period of rapid growth for an additional day. In contrast, treatment with GA3 has no effect on the duration of rapid growth, but does increase the growth rate between days two and three (Saibo et al. 2003). The effect of a combined application of ACC and GA3 is at the very least additive, and in most cases synergistic, thereby resulting in the highest increase in hypocotyl length (Saibo et al. 2003; vandenBussche et al. 2007) (Fig. 9.2). Furthermore, the hypocotyl elongation in the gai etr1-3 double mutant plant is completely insensitive to either phytohormone (De Grauwe et al. 2007).

The ethylene-mediated regulation is now known to depending on blue light and cryptochrome signaling, and that GA is required for the ethylene-stimulated hypocotyl elongation to occur (vandenBussche et al. 2007). Treatment with ACC in the presence of blue light up-regulates the transcription of the GA synthesis genes GA20ox1 (GA 20-oxidase) and GA3ox1 (GA 3-oxidase) and down-regulates that of the GA metabolism genes GA2ox1 (GA 2-oxidase) and GA2ox7, as well as that of the GA synthesis gene GA3ox2 (Vriezen et al. 2004; Vandenbussche et al. 2007). There is no substantial effect on the abundance of transcript of genes encoding either GA receptors or the DELLA proteins, although the accumulation of GFP-RGA fusion protein is enhanced in ACC-treated pRGA::GFP-RGA plants (Vandenbussche et al. 2007).

GA and ethylene have an effect on endoreduplication frequency (Traas et al. 1998; Gendreau et al. 1999). Ethylene induces endoreduplication in both light- and dark-grown hypocotyl, and the process involves the ethylene signaling pathway genes CTR1 and EIN2 (Gendreau et al. 1999). Hypocotyl cells in LNM-grown seedlings contain nuclei with three ploidy levels: 2C, 4C, and 8C (C is the DNA content of the haploid genome in the G1 phase). ACC treatment increases the 8C:4C ratio and a small fraction of cells undergo an additional round of endoreduplication to deliver about 2 % 16C nuclei. Similarly, GA3 treatment increases the 8C:4C ratio, but no 16C nuclei are produced. A combined ACC and GA3 treatment has the greatest enhancing effect on the 8C:4C ratio and on the representation of 4 % 16C nuclei (Saibo et al. 2003).

Although cell elongation is mainly responsible for hypocotyl growth, cell division also should be considered. Treatment with either ACC or GA3 on their own increases cortical cell number by, respectively, 1.4 and 1.3, while in combination, they enhance cortical cell number by 2.5 (Saibo et al. 2003). However, the measurements of GUS activity generated by CycB1;1::GUS transgenics (the CycB1;1 product is involved in cell division) imply that cell division occurs solely in the hypocotyl epidermal layers, which contributes to the development of stomata. Treatment with either ACC or GA increases stomata number in the hypercotyl by, respectively, 33 % and 21 %, while the combined treatment increases it by 55 % (Saibo et al. 2003).

9.5 Ethylene-Induced Apical Hook Development Is Dependent on GA

The apical hook is a transient curvature of the hypocotyl tip assumed to protect the cotyledons and the shoot apical meristem from mechanical damage as the seedling grows through the soil. The curvature is generated by asymmetric growth, specifically via differential cell division and elongation of the inner and outer sides of the hypocotyl (Lehman et al. 1996; Raz and Ecker 1999; Raz and Koornneef 2001). The formation of the apical hook involves three developmental phases: formation, maintenance, and opening (Raz and Ecker 1999; Abbas et al. 2013). Ethylene is known to be an important regulator of apical hook development. Dark-grown seedlings treated with exogenous ethylene or ACC produce a shortened root, a shortened and radially expanded hypocotyl, and an exaggerated apical hook (Abeles et al. 1992; Ecker 1995). GA is also involved in apical hook development. The ga1-3 mutant does not form an apical hook in three-day-old etiolated seedlings, whereas in the absence of either GAI and/or RGA, a hook is formed by this mutant. These data indicate that GA opposes DELLA proteins repression and can promote the apical hook formation (Achard et al. 2003) (Fig. 9.2). In addition, the GA-DELLA regulatory system also affects apical hook maintenance, involving both ethylene and auxin signaling.

In the presence of either ethylene or ACC, the size of the apical hook is exaggerated in WT seedlings, but this is not the case in the ga1-3 mutant (Achard et al. 2003). The ctr1-1 mutant does not form a hook in the presence of inhibitor of GA synthesis paclobutrazol (PAC), consistent with the notion that GA is involved in apical hook formation, even in a constitutive ethylene signaling status (Vriezen et al. 2004). The development of the apical hook requires differential cell division and elongation at the hypocotyl apex (Raz and Koornneef 2001). Although GA and ethylene have little effect on cortical cell division in the hypocotyl, the influence of both ACC and GA on cell division during apical hook development is clear. PAC almost completely inhibits cell division in two- or three-day-old dark-grown seedlings, and CycB1;1::GUS transgenic seedlings harbor almost no GUS-stained cells. Thus, GA3 counteracts the PAC effect and stimulates cell division. The combined application of ACC and GA3 has the strongest stimulatory effect on cell division; but when compared with the effect of GA3 on its own, the extent of the enhancement is not significant. The conclusion is that ethylene acts to enhance the effect of GA on cell division (Vriezen et al. 2004).

Similar to the effect on GFP-RGA level in the root and hypocotyl, ethylene also induces RGA accumulation in the nucleus within the apical hook. In addition, ACC treatment clearly up-regulates the transcription in the apical region of the GA-responsive gene GASA1 (Herzog et al. 1995). GA homeostasis is regulated by feedback mechanisms, triggered by changes in GA-signaled GA synthesis and metabolism (Hedden and Phillips 2000). For example, the transcription of GA1 is strongly up-regulated in PAC-treated seedlings. ACC treatment also markedly up-regulates GA1 in the apical hook endodermis, which suggests that ACC, like PAC, attenuates the GA response and modulates the level of feedback over GA synthesis (Vriezen et al. 2004).

Gallego-Bartolome et al. (2011) have described a kinematic analysis of GA action during apical hook development. A comparison between the impact of PAC, ACC, and GA on WT (A. thaliana ecotype Landsberg erecta), the gai-1 (GA signaling pathway gain-of-function mutant), della (a complex mutant which does not produce any of the five DELLA proteins), and the ein2-1 (ethylene-insensitive mutant) (Guzman and Ecker 1990; Peng et al. 1997; Feng et al. 2008) demonstrates that GA plays a prominent role during the formation and opening the apical hook. PAC-treated WT seedlings fail to form a hook and gradually enter the opening phase, whereas the della mutant develops an exaggerated apical hook which is not induced to open by ACC treatment. Seedlings of ein2-1 form no hook, although this failure can be partially reversed by GA treatment. The conclusion is that GA and ethylene probably act independently of one another during hook formation, whereas the two phytohormones cooperate to prevent hook opening. In further experiments, the same authors have established that GA activity in the endodermis is essential for normal hook development, at least during the late formation phase, whereas epidermal GA activity is non-essential.

The gai-1 mutant displays a reduced level of transcription of the ethylene synthesis genes ACC SYNTHASE8 (ACS8) and ACS5/ETO2 (Vogel et al. 1998; Yamagami et al. 2003), as well as of the ethylene-inducible gene HLS1 (Lehman et al. 1996), which all contribute to ethylene-induced hook development. Given that GAI and PIF5 (PHYTOCHROME INTERACTING FACTOR5) interact with one another in vivo, and that PIF5 binds the ACS8 promoter in a GA-dependent manner, a possibility is that the DELLA proteins repress ACS8 expression by inhibiting PIF5 activity. Moreover, GA-regulated ACS expression is associated with ethylene production in etiolated seedlings, and the della mutant produces more ethylene than does the WT plant (Gallego-Bartolome et al. 2011). GA also up-regulates HLS1 expression (Gallego-Bartolome et al. 2011; Abbas et al. 2013). Analysis has shown that there is a temporal coincidence in the requirement of GA and HLS1 activity during hook development (Gallego-Bartolome et al. 2011).

EIN3/EIL1 are the primary transcription factors in the ethylene signal transduction pathway (Alonso et al. 2003), and overexpression of either EIN3 or EIL1 results in exaggerated hook curvature (Chao et al. 1997; An et al. 2010). An and her colleagues find that GA3 enhances, whereas PAC represses, ethylene- and EIN3-overexpression-induced hook curvature in six-day-old etiolated seedlings, and della mutant exhibits exaggerated hook curvature, which requires an intact ethylene signaling pathway (An et al. 2012). HLS1 encodes a protein with sequence similarity to N-acetyltransferase, and mutation in HLS1 has no effect on the growth of the hypocotyl apex in the presence of exogenous ethylene (Lehman et al. 1996). HLS1 is also required for GA to have any effect on hook development (Gallego-Bartolome et al. 2011). The hls1 mutation overrides the exaggerated hook curvature phenotype shown by della, and the hls1 della sextuple mutant forms no hook, which has been taken to imply that GA- and DELLA-regulated hook development is dependent on HLS1 (An et al. 2012) (Fig. 9.2). Furthermore, ethylene and GA induce HLS1 transcription in an EIN3/EIL1-dependent manner and neither can induce HLS1 transcription in the ein3 eil1 mutant. Other experiments have shown that EIN3 binds directly with the HLS1 promoter to induce its expression (An et al. 2012). The DELLA proteins RGA and GAI may interact with the DNA-binding domains of EIN3/EIL1 in vivo and repress EIN3/EIL1-regulated HLS1 expression (An et al. 2012) (Fig. 9.2).

9.6 Ethylene-Induced Flowering May Result in Part by a Modulation of DELLA Activity

A number of mechanisms have evolved to take account of endogenous signals and environmental cues to time the plant’s transition from vegetative to reproductive growth, a critical moment in the life cycle of the flowering plant (Achard et al. 2007). That ethylene modulates the vegetative growth of A. thaliana in response to changes in the environment has been amply demonstrated (Abeles et al. 1992; Wang et al. 2002; Achard et al. 2006). However, its participation in the regulation of the switch from vegetative to reproductive growth is less clear (Thomas and Vince Prue 1997; Achard et al. 2007). On the other hand, GA is clearly important for controlling this transition in A. thaliana (Mutasa-Gottgens and Hedden 2009; Srikanth and Schmid 2011). The ga1-3 mutant is unable to flower under short-day (SD) conditions without exogenously supplied GA, and the flowering of the gai mutant is substantially delayed under SD conditions, a phenotype which cannot be rescued by GA3 treatment (Wilson et al. 1992). GA promotes flowering in A. thaliana by activating floral meristem identity genes such as LEAFY (LFY) and SUPPRESSOR OF OVEREXPRESSION CONSTANS1 (SOC1). The DELLA proteins act to delay flowering under SD conditions by repressing LFY and SOC1 transcription (Moon et al. 2003; Achard et al. 2004, 2007; Mutasa-Gottgens and Hedden 2009).

Ethylene delays flowering in a DELLA-dependent manner (Fig. 9.2). In the presence of ACC or in an ethylene-rich atmosphere, flowering is delayed in WT A. thaliana, although the effect is less marked in both the gai-t6 rga-24 double mutant and the quadruple DELLA mutant gai-t6 rga-t2 rgl1-1 rgl2-1 (Achard et al. 2006). Floral transition is also delayed in the gai eto2-1 double mutant, which overproduces ethylene, and even more strongly than in WT plants exposed to ethylene (De Grauwe et al. 2008). The loss-of-function ctr1-1 mutant is late flowering under both long-day (LD) and SD conditions. Compared with WT plants, the level of GA1 and GA4 is significantly reduced in LD-grown ctr1-1 mutant plants, while the content of intermediate GA species (such as GA24 and GA53), which act as substrates for GA 20-oxidase in the synthesis of bioactive GA (Hedden and Phillips 2000), is significantly enhanced. Both AtGA3ox1 and AtGA20ox1 transcript abundance is higher in the ctr1-1 mutant than in WT plants, but in the triple mutant ctr1-1 gai-t6 rga-24, the levels of transcript revert to those seen in the WT, thereby implicating DELLA function in the up-regulation of the two GA oxidase genes in ctr1-1 (Achard et al. 2007).

The abundance of both LFY and SOC1 transcript is lower in the ctr1-1 mutant than in WT plants, but GA-treated ctr1-1 plants or ctr1-1 plants lacking both GAI and RGA exhibit a relatively normal LFY and SOC1 transcript level. The inference is that the ethylene-mediated inhibition of CTR1 activity reduces the level of bioactive GA, resulting in a down-regulation of LFY and SOC1, and hence a delay in the switch to reproductive growth (Achard et al. 2007) (Fig. 9.2). The gai-t6 rga-24 double mutant, however, rescues the late-flowering phenotype expressed by ctr1-1 plants exposed to SD conditions. Similarly, the delayed flowering of ctr1-1 grown under LD conditions can be negated either by GA treatment or by deleting GAI and RGA. SPINDLY (SPY), an O-linked N-acetylglucosamine transferase, acts as a negative regulator of GA signaling, and this has been demonstrated by showing that the loss-of-function spy mutation partially suppresses the late-flowering phenotype of the GA-deficient ga1-2 mutant exposed to SD (Jacobsen and Olszewski 1993). However, the ctr1-1 spy-5 double mutant flowers much earlier than the ctr1-1 single mutant. Thus, the greater sensitivity to GA brought about by the lack of SPY at least partially suppresses the delay to floral transition conferred by the ctr1-1 mutation (Achard et al. 2007). Ethylene activates ethylene responses by inhibiting SCFEBF1/EBF2 activity, which in turn increases the stability of EIN3 and EIN3-LIKE proteins (Guo and Ecker 2003; Potuschak et al. 2003; Gagne et al. 2004). The constitutive ethylene response phenotype shown by the loss-of-function ebf1-1 ebf2-1 mutant is stronger than that of ctr1-1, and the flowering delay can be overridden by exogenously supplied GA. The level of EIN3 present is unaffected both by GA treatment of the ebf1-1 ebf2-1 mutant, and by the absence of GAI and RGA (as in the ctr1-1 gai-t6 rga-24 mutant), indicating that the DELLA proteins act downstream of CTR1/EIN3 in the ethylene-dependent regulation of flowering (Achard et al. 2007) (Fig. 9.2).

9.7 The Interaction of Ethylene and GA Signaling in Plant Response to Abiotic Stress

Environmental stresses, such as drought, soil salinity, and flooding, can all depress plant growth and seed production, and a suite of protective mechanisms, many of which involve phytohormones, have been evolved to cope with these stresses. By modifying the level, distribution, and/or signal transduction activity of these phytohormones, a plant is able to adjust its physiology and biochemistry quite rapidly, a critical requirement for its survival (Colebrook et al. 2014). Abscisic acid (ABA) and ethylene are the two most frequently encountered phytohormones associated with the abiotic stress response. In A. thaliana, the response to salinity is triggered (at least in part) by an accumulation of ABA, which activates a variety of ABA signaling pathways (Zhu 2002; Shinozaki et al. 2003). Salinity inhibits growth (at least in part) via the ABI1-dependent, ABA-mediated enhancement of DELLA (Achard et al. 2006). In deepwater rice, submergence stimulates the degradation of ABA, thereby enhancing the plant’s responsiveness to GA and promoting the elongation of the internode (Hoffmann-Benning and Kende 1992). A brief account follows describing what is known regarding the interaction between ethylene and GA signaling during the abiotic stress response.

Prolonged flooding results in the development of hypoxic conditions in the soil, which has a dramatic impact on levels of respiration and photosynthesis, on redox homeostasis, and on intracellular pH. Paddy rice, and particularly deepwater rice, has developed various strategies to prevent the build-up of anoxia in their tissue (Xu et al. 2006; Fukao et al. 2006). Deepwater rice plants respond rapidly to a rising water level by elongating their stem internodes to maintain the upper canopy above the water’s surface. This process is strongly regulated by ethylene, the rapid accumulation of which in submerged tissue (via physical entrapment and active synthesis) triggers shoot elongation, adventitious root formation, and alterations in carbohydrate metabolism (Steffens et al. 2006; Xu et al. 2006; Hattori et al. 2009). GA appears to be involved in some of these ethylene-mediated responses (Steffens et al. 2006; Xu et al. 2006). Ethylene also coordinates the balance of GA and ABA content during submergence (Xu et al. 2006). In deepwater rice, adventitious roots are induced to grow from a node, and their development in time replaces or at least supports the main root system which will have become increasingly dysfunctional as a result of prolonged hypoxia (Steffens et al. 2006). The growth of these secondary roots is mediated by ethylene (Suge 1985; Lorbiecke and Sauter 1999), while GA has little or no effect on their development. Nevertheless, a combined treatment of ethylene and GA results in a synergistic promotive effect, while the application of a competitive inhibitor of ethylene abolishes root growth. The inference is that ethylene perception is required for adventitious root growth, and that the resulting ethylene-induced growth of adventitious roots is strongly promoted by GA (Steffens et al. 2006).

GA’s involvement in the regulation of the ethylene-mediated growth response of deepwater rice lies not only in inducing adventitious root growth but also in promoting shoot growth. Hattori et al. (2009) have identified that the genes SNORKEL1 (SK1) and SK2 are responsible for allowing the rapid elongation of the stem internode. These genes are specific to deepwater rice germplasm, and are up-regulated by flooding. The constitutive expression of SK1 and SK2 in non-deepwater rice has a positive effect on stem internode elongation even under non-flooded conditions. SK1 and SK2 both possess a single APETALA2/ethylene response factor (AP2/ERF) domain. Their transcription is inducible by exogenously supplied ethylene, but not by any other phytohormone. The SK1 and SK2 promoter regions harbor core sequences for EIN3-binding site and can bind to the rice EIN3-like protein (OsEIL1b). The suggestion is that SK1 and SK2 serve as ethylene response factors and form part of an ethylene signaling pathway which positively regulates stem internode elongation in deepwater rice (Fig. 9.3).

Physiological experiments show that ethylene, ABA, and GA are all involved in the submergence response (Kende et al. 1998). Ethylene has been identified as an initiation factor for internode elongation (Raskin and Kende 1984; Kende et al. 1998). When plants are submerged, although ethylene accumulates in both deepwater and non-deepwater rice, it only induces stem internode elongation in the former type. The exogenous supply of GA induces stem internode elongation in deepwater rice plants even when they are not submerged, while in flooded plants, the level of GA1 at the nodes is heightened in deepwater rice, but not in non-deepwater rice. Stem internode elongation is repressed in deepwater rice by treatment with the GA synthesis inhibitor uniconazole, and this repression can be lifted by supplying the plant with GA. The indications are that GA positively regulates stem internode elongation in deepwater rice, operating through the action of SK1 and SK2 (Hattori et al. 2009) (Fig. 9.3). The probable scenario therefore is that under submergence, the response of deepwater rice is to rapidly accumulate ethylene, which has the effect of up-regulating SK1 and SK2; additionally, this pair of genes may, directly or indirectly, promote GA synthesis and/or GA signal transduction, in order to promote stem internode elongation, thereby allowing the apex of the plant to remain above the water’s surface where it can access the oxygen necessary to escape hypoxic stress (Hattori et al. 2009, 2011) (Fig. 9.3).

Flash flooding also has the effect of submerging the plant, but in this case, the water subsides after a period of some days. Submergence-tolerant rice cultivars respond to temporary submergence by restricting the growth of their shoot and limiting its respiration, recommencing normal growth once the water has subsided (Singh et al. 2001; Das et al. 2005; Fukao et al. 2006; Xu et al. 2006). A major quantitative trait locus designated Submergence 1 (Sub1) has been identified as harboring three distinct genes (Sub1A, Sub1B, and Sub1C) responsible for this mode of submergence tolerance (Fukao et al. 2006; Xu et al. 2006). Each of the Sub1 genes encodes a protein featuring an ERF-like DNA-binding domain. Genotypic variation at Sub1 locus confers distinctions in submergence tolerance. Sub1B and Sub1C are present in both tolerant and non-tolerant paddy cultivars, whereas Sub1A is only found in tolerant ones. Under submergence, the level of ethylene rises rapidly in each of a pair of near isogenic paddy rice lines, one carrying Sub1B and Sub1C and the other carrying Sub1A, Sub1B, and Sub1C, but it rises to a higher level in the former (non-tolerant) line. The constitutive expression of Sub1A in a non-tolerant cultivar results in a marked improvement in submergence tolerance (Xu et al. 2006). There is an indication that the abundance of Sub1A transcript in node and internode regions in some rice accessions containing Sub1A is positively correlated with the degree of submergence tolerance. Sub1A transcription is inducible by low levels of ethylene (up to 100 ppm) but this treatment has the effect of reducing endogenous ethylene production in submerged plants via negative feedback regulation (Singh et al. 2010). The inference is that the presence of Sub1A is necessary and sufficient to confer submergence tolerance (Xu et al. 2006) (Fig. 9.3).

Under normal conditions, the non-tolerant and tolerant isogenic lines develop at a similar rate and their stem elongation response to GA3 treatment is similar. In submerged plants, the tolerant line responds positively to GA3 supply with negative implication for its survival. The exposure of the non-tolerant line to PAC reduces its stem elongation, thereby enhancing its survival. The suggestion is that GA-induced stem elongation has a negative impact on tolerance to prolonged submergence (Setter and Laureles 1996; Das et al. 2005). Rice plants engineered to constitutively express Sub1A display a classic GA-insensitive phenotype: germination is delayed, the plants are semi-dwarfed, flowering is late, maturity is slow, and grain set is reduced. The seedling GA3 dose response curve indicates that the ectopic expression of Sub1A compromises the plant’s response to GA, thereby having a negative effect on GA-dependent processes. SLR1 and SLR1 LIKE1 (SLRL1) are both nuclear-localized GRAS proteins which function as suppressors of GA signaling in rice. The transcription of both SLR1 and SLRL1 is submergence-inducible in both submergence-tolerant and non-tolerant plants, but the level of inducibility is greater and its onset is more rapid in the tolerant ones. Engineered constitutive Sub1A expressors display an elevated basal level of SLR1 and SLRL1 transcription under normal growing conditions, but higher levels when the plants are submerged. These observations confirm that the submergence-induced expression of Sub1A enhances the transcription and translation of SLR1 and SLRL1, thereby restraining GA-mediated underwater elongation, prolonging submergence endurance, and sustaining the plant’s capacity to regrow once the flooding has subsided (Fukao and Bailey-Serres 2008) (Fig. 9.3).

The abundance of neither SLR1 nor SLRL1 transcript is increased by the ethylene treatment of a submergence non-tolerant plant, but not so for a tolerant one. At the protein level, the treatment suppresses the production of SLR1 in the non-tolerant plant but not in the tolerant one, while the abundance of SLRL1 is little changed in the former, but markedly increased in the latter. The indication is therefore that ethylene induces Sub1A expression, which has the effect of up-regulating SLR1 and SLRL1 transcription and translation (Fukao and Bailey-Serres 2008). The ethylene does not affect the growth of non-submerged tissue, whether or not Sub1A is present. However, a combined ethylene plus GA3 treatment promotes the GA-mediated elongation of the aerial part of the non-tolerant plant, but significantly attenuates the response in the tolerant one. The interpretation is that in the absence of Sub1A, ethylene enhances GA responsiveness by lowering the level of both SLR1 and SLRL1. On the other hand, ethylene-induced Sub1A expression stimulates the transcription and translation of SLR1 and SLRL1, which results in the suppression of GA-mediated growth (Fukao and Bailey-Serres 2008) (Fig. 9.3). The overall picture is that submergence tolerance in paddy rice is modulated by Sub1A, the product of which dampens ethylene production and responsiveness, resulting in a restriction in ethylene-promoted GA responsiveness achieved via the accumulation of SLR1 and SLRL1; the ultimate result is to limit stem internode elongation and assimilate consumption during a submergence episode.

It is notable that both the SK genes and Sub1A encode ERFs (ETHYLENE-RESPONSIVE FACTORs) and are associated with GA and ethylene action; however, they confer opposite functions in regulating stem internode elongation in response to flooding: while Sub1A restricts shoot elongation during submergence, the SK genes stimulate it (Fig. 9.3). In addition, a novel rice ERF gene OsEATB (for ERF protein associated with tillering and panicle branching) also mediates crosstalk between ethylene and GA to regulate rice internode elongation. Ethylene treatment sharply down-regulates OsEATB expression while increases OsCPS2 expression (a key GA synthesis gene encodes ent-kaurene synthase A). OsEATB suppresses the internode elongation process through the restriction of GA biosynthesis, specifically down-regulating the expression of OsCPS2 (Qi et al. 2011). OsEATB expression is also negatively regulated by ABA and abiotic stress, but further investigations of the molecular crosstalk of the related signaling pathways are still required.

At the end of a submergence episode, rice plants often experience dehydration stress due to reduced hydraulic conductivity in leaf sheaths (Setter et al. 2010). Individuals which harbor Sub1A recover by developing new leaves once the water has subsided (Fukao and Bailey-Serres 2008), and also show markedly enhanced recovery from drought at the vegetative stage (Fukao et al. 2011). During a drought episode, Sub1A contributes to limiting the extent of water loss, and also promotes the recommencement of leaf growth and development once the stress has been lifted. The abundance of Sub1A transcript is markedly enhanced by drought stress, which increases expression of genes associated with acclimation to dehydration. The presence of Sub1A during a period of drought stress is associated with a reduced accumulation of ROS (reactive oxygen species), which limits the stress-induced damage to a range of cellular components, and so aids in survival (Fukao et al. 2011) (Fig. 9.3). Therefore, in addition to providing robust submergence tolerance, Sub1A improves survival of rapid dehydration following submergence subsidence and water deficit during drought (Fukao et al. 2011) (Fig. 9.3).

The GA and ethylene associated brake on growth imposed by abiotic stress is mirrored in A. thaliana (Achard et al. 2006, 2008; Dubois et al. 2013). In response to water deprivation, the growth of the leaf is halted via the rapid up-regulation of the two transcription factors ERF5 and ERF6 (Hruz et al. 2008; Dubois et al. 2013). The growth under osmotic stress of erf5 erf6 mutant leaves is less restricted than that of WT leaves, but the mutant shows no added tolerance to mild salinity stress. The transcription of either ERF5 or ERF6 in WT plants is up-regulated by mild osmotic stress and several other stresses including drought (Hruz et al. 2008; Dubois et al. 2013), but it is not up-regulated by mild salinity stress. ERF6 overexpressors are dwarfed in stature and form dark green leaves and stunted rosettes, and they are highly sensitive to mild osmotic stress. The inference is that ERF6 negatively regulates leaf growth under mild osmotic stress. The GA2ox6 (which encodes an enzyme involved in GA inactivation) activated by mild osmotic stress is rapidly induced in ERF6 overexpressors; meanwhile, in the erf5 erf6 double mutant, the induction of GA2ox6 is delayed (Dubois et al. 2013), which indicates that ERF6 negatively regulates the level of GA by inducing GA2ox6 expression. Overexpressing GA20ox1 (which encodes the rate-limiting GA synthesis enzyme) suppresses the retarded growth phenotype conferred by ERF6 overexpression. In addition, ERF6 stabilizes RGA (the major DELLA protein species expressed in the developing leaf). These data confirm that, under mild osmotic stress, an elevated level of ERF6 reduces the level of endogenous GA through the suppression of GA2ox6. The resulting stabilizing effect on the DELLA proteins eventually helps to negatively regulate leaf growth (Dubois et al. 2013). As mild osmotic stress triggers the accumulation of ACC, it is likely that ERF5 and ERF6 integrate ethylene and GA signaling to regulate plant growth and the mild osmotic stress response (Dubois et al. 2013) (Fig. 9.3).

High levels of soil salinity restrict the ability of the root to take up water; once the salt has entered the root, it compromises many aspects of cell physiology and slows growth (Hasegawa et al. 2000; Munns and Tester 2008). Achard et al. (2006) have shown that the seedlings of the A. thaliana gai-t6 rga-t2 rgl1-1 rgl2-1 quadruple DELLA mutant are less inhibited by salinity (100 mM NaCl) than WT seedlings. WT plants respond to salinity by reducing the endogenous level of GA1 and GA4, and salt treatment promotes the accumulation of GFP-RGA in pRGA::GFP-RGA transgenic plants. Thus, one way in which salinity stress (as well as other forms of stress) can slow growth is by reducing the endogenous level of GA, which consequently favors the accumulation of DELLA proteins. Extreme salt concentrations kill plants, compared with WT and the quadruple DELLA mutant, the ga1-3 and gai mutants show increased tolerance of extreme salt concentration (200 mM NaCl). Thus, DELLA proteins can contribute the survival of salt toxicity. Treatment with ethylene allows the WT plant to withstand a higher level of salinity stress (Wang et al. 2002), while the ctr1-1 mutant shows an enhanced rate of survival when challenged by salinity stress, possibly because of the constitutive activation of ethylene signaling pathways (Guo and Ecker 2003; Potuschak et al. 2003); their increased tolerance is overridden in mutants which also lack GAI and RGA. Thus, salinity slows growth through the activation of ethylene signaling, the effects of which are at least partly integrated with GA signaling at the level of DELLA function (Fig. 9.3). Meanwhile, the accumulation of DELLAs up-regulates a suite of genes encoding ROS detoxification enzymes, thereby limiting ROS-induced damage and enhancing salinity tolerance (Achard et al. 2006). The restraint imposed on growth and respiration allows the plant to ration its carbohydrate reserves during the stress period, so that energy is available when growth can resume (Fukao and Bailey-Serres 2008). A common regulatory mechanism mediating both the restraint on growth and survival of adverse conditions is likely to exist. Phytohormonal signaling pathways (ethylene, GA) may be involved in this mechanism, integrated at the level of DELLA function.

9.8 Conclusions and Perspectives

Ethylene and GA clearly participate in the plant growth, development, and the response to abiotic stress, but many aspects remain to be uncovered. For example, the molecular basis of the interaction between ethylene and GA during germination still needs elucidation. A number of questions are outstanding, such as: do ethylene and GA affect root cell division, cell differentiation, or stem cell fate? Where and how is signaling transduced? Other than GA, are there other hormones involved in ethylene-mediated flowering? How are developmentally separated hormone-mediated responses integrated with DELLA function to bring together several signaling pathways in the response to abiotic stress? Characterization of the many interactions involved and the identification of the roles played by the various phytohormones and gene products will require an integration of cell biological and system biology approaches. The goal is to gain an understanding of how plant growth and development is quantitatively modulated in the face of abiotic stress.