Novel mechanisms of ethylene-gibberellin crosstalk revealed by the gai eto2-1 double mutant

Abstract

DELLA proteins have been shown to act as integrators of the signaling network controlling plant growth. In the January issue of New Phytologist (2008), we analyzed the gai eto2-1 double mutant and corresponding single mutants, with defects in the ethylene-biosynthesis and/or in the gibberellin (GA)-signaling cascade. This research revealed yet unknown modes of cross-talk between the ethylene and GA pathways. Two hypotheses have been put forward. Both essentially suggest the existence of reciprocal posttranslational control of ethylene-GA crosstalk.

Although crosstalk between ethylene and GA has been demonstrated in the past,^1–3 the interaction between these two hormones had not been investigated throughout the life-cycle of the plant. In an attempt to make a detailed investigation of the influence of ethylene-GA crosstalk on plant development, the gai eto2-1 (gibberellin insensitive; ethylene overproducing) double mutant was created. A dominant gain-of-function mutation in the DELLA gene GAI⁴ negatively affects GA-signaling, while the dominant eto2-1 mutation leads to ethylene overproduction due to enhanced stability of the ACC synthase 5 (ACS5) protein by disruption of its C-terminal 12 amino acids.^5,6 Our data revealed synergistic responses in several phases of plant development, leading to two hypotheses, both inferring the existence of reciprocal post-translational control of ethylene-GA crosstalk.

First, since the gai eto2-1 double mutant does not overproduce ethylene, we suggest that the stability of the ACS5 protein is dependent on GA.⁷ ACS is responsible for the conversion of SAM (S-adenosyl-L-methionine) to ACC, which is the rate limiting step in the ethylene-biosynthesis.^8,9 In Arabidopsis, the enzyme is encoded by a multigene family of 12 members, 10 of which are functionally active and differentially regulated at the transcriptional level.^8,9 The gene family can be divided into 2 subclasses, according to the absence (e.g., ACS5 and ACS9) or presence (e.g., ACS2 and ACS6) of a C-terminal extension of about 17 amino acids, carrying 3 serine residues. These are subject to stress-induced mitogen-activated protein-kinase (MPK6) phosphorylation. Ten to twelve amino acids upstream of the C-terminal extension typical for ACS2 and ACS6, appears a conserved serine residue, target for a calcium-dependent protein kinase activity (CDPK). Phosphorylation enhances the stability of ACS enzymes. The non-phosphorylated C-terminal domain acts as a tag for substrate-specific interaction with the ETO1/CUL3 E3 ligase, labeling ACS for subsequent degradation by the 26S proteasome.⁵ Analysis of the eto2 (mutation in ACS5) and eto3 (mutation in ACS9) mutants, suggests that cytokinin (CK) could induce ethylene by modifying the ACS5/ACS9 C-terminal domain such that its interaction with ETO1 is blocked; hence stabilizing the ACS protein. However, since CK still affects the half-life of the stabilized Eto⁻ versions of ACS5 and ACS9, there must exist an ETO1-independent CK-controlled degradation pathway.⁵ Besides the 2 above-mentioned routes, our findings support the existence of an additional mode of ACS degradation; which is not mediated by ETO1, and GA controlled (Fig. 1).⁷ Measurements of ethylene emanation showed that ethylene biosynthesis was not increased in gai eto2-1, as opposed to eto2-1. This observation implies that a functional GA response pathway is required for the increased ethylene biosynthesis in eto2-1. Importantly, this degradation pathway probably targets domains different from the C-terminal domain of the ACS protein. Whether or not this pathway is the same as the ETO1-independent cytokinin-inhibited pathway mentioned above,⁵ is subject of our current investigation.

Figure 1.

Possible mechanisms of ethylene-GA crosstalk revealed by the analysis of the gai eto2-1 double mutant. Arrows represent the influences of ethylene on the GA-biosynthesis and signaling and the effect of GA on its own signaling and on the ethylene-biosynthesis (by influencing ACS stability). Dotted lines represent hypothetic modes of regulation, yet to be proven. Ub indicates addition of ubiquitin units to label ACS5 for degradation. Degradation of ACS5 can either be regulated through the ETO1-CUL3 complex; or by yet unknown pathways (indicated by ?), which might be affected by CK and/or GA. Besides the suggested GA-specific pathway, GA may also influence the ETO1/CUL3 pathway in a CK-independent way or counteract the inhibitory action of CK on the ETO1/CUL3 dependent and independent pathways. Ethylene itself might be involved in the degradation/stabilization of the DELLA-proteins, either through a direct effect on the DELLA-proteins or through an effect on the SCF^SLY1/2 F-box proteins.

Reciprocally, ethylene affects the GA pathway. It has been proven that proteins from the DELLA family, which comprises 5 members in Arabidopsis (GAI, repressor of ga1-3 (RGA), RGA-like (RGL1/2/3)), are rapidly destabilized after GA treatment through degradation by the 26S proteasome.^10–12 The effect of ethylene on the prevalence of these proteins was dual. First, it has been shown that ethylene delays the disappearance of a green fluorescent protein (GFP)-RGA fusion from root cell nuclei via a constitutive triple response1 (CTR1)-dependent signaling pathway.¹ Likewise, ACC treatment stabilized a GFP-RGA fusion protein in hypocotyl cell nuclei.³ More recently; it was demonstrated that active ethylene signaling results in decreased GA content; thus stabilizing DELLA proteins.^13,14 In addition, it was proven that ethylene affects DELLA stability primarily via changes in GA-concentration, supposedly by post-transcriptional control of GA20ox/GA3ox/GA2ox genes.^13,14 Furthermore, analysis of microarray data using Genevestigator has shown that in 28d-old plants none of the genes encoding DELLA-proteins, neither the F-box protein SLEEPY (SLY) are transcriptionally affected by ethylene. The only gene from the GA-signalling cascade which appears to be regulated by ethylene is GIBBERELLIN INSENSITIVE DWARF1 (GID1c) (more than 2-fold induction after ethylene-treatment).^15,16 This induction-albeit more modest- was confirmed by analysis of microarray data from ACC treated seedlings (Toronto database).¹⁷

The gai eto2-1 double mutant showed an enhanced GA-response as compared to the gai mutant, which is either indicative of an ethylene-driven gai-degradation or of the degradation of (one of) the other DELLA-proteins. Alternatively, this could be explained through the existence of a DELLA-independent control of the GA response; as suggested by Cao et al., (2006).¹⁸ Our hypothesis is that ethylene affects the strength of the DELLA-SLY1/2 interaction by controlling the post-translational modification of either the DELLA or the SLY partner, affecting SLY1/2 accumulation (Fig. 1). The latter possibility is likely, since it has been proven that the SCF-E3 Ubiquitine ligase activities are usually regulated by the level of the F-box factor.¹⁹ Phosphorylation or another post-translational modification may be involved. Future research will reveal which type of regulation is predominant in the reciprocal control of GA and ethylene response.

Abbreviations

ACS

ACC-synthase

CDPK

calcium-dependent protein kinase

CK

cytokinin

CTR

constitutive triple response

CUL

cullin

eto2-1

ethylene overproducing

GA

gibberellin

GAI

gibberellin insensitive

GID

gibberellin insensitive dwarf

GFP

green fluorescent protein

MPK6

mitogen-activated protein-kinase

RGA

repressor of ga1-3

RGL

RGA-like

SAM

S-adenosyl-L-methionine

SCF

SKP1/cullin/F-box

SLY

sleepy

Addendum to: De Grauwe L, Chaerle L, Dugardeyn J, Decat J, Rieu I, Vriezen WH, Baghour M, Moritz T, Beemster GT, Phillips AL, Harberd NP, Hedden P, Van Der Straeten D. Reduced gibberellin response affects ethylene biosynthesis and responsiveness in the Arabidopsis gai eto2-1 double mutant. New Phytol. 2008;177:128–141. doi: 10.1111/j.1469-8137.2007.02263.x.