Reentry of the Ethylene MPK6 Module

All higher plants produce the gaseous hormone ethylene, which serves as an important regulator of growth, development, and disease resistance (Bleecker and Kende, 2000). Beginning with the identification of etr1 (for ethylene response 1), an ethylene insensitive mutant (Bleecker et al., 1988), and continuing until today, the reference plant Arabidopsis has been used to identify numerous mutants and the corresponding genes encoding components in the ethylene biosynthesis and signal transduction pathways (reviewed in Guo and Ecker, 2004).

3, 2, 1, WE HAVE LIFTOFF!

The first ethylene signal transduction pathway gene to be cloned was called CONSTITUTIVE TRIPLE RESPONSE1 (CTR1) (Kieber et al., 1993). Genetic studies showed that CTR1 was a negative regulator of ethylene responses in Arabidopsis. Its predicted protein sequence showed similarity to mammalian Raf kinase, a mitogen-activated protein kinase kinase kinase or MAPKKK for short. This observation immediately suggested that the plant ethylene-signaling pathway might contain a MAP kinase cascade or module (Widmann et al., 1999). This module of three protein kinases is equivalent to a three stage rocket where MAPKKK phosphorylation of MAPKK activates its kinase activity, and, in turn, MAPKK phosphorylation of MAPK activates its kinase activity.

The decade-long “moon shot,” of sorts, that followed the identification of CTR1 saw tremendous advancements in our understanding of the ethylene-signaling pathway. Notable “missions” included the identification and characterization of each of the five membrane bound ethylene receptors right down to the key transcription factors that control ethylene-responsive gene expression in the nucleus (reviewed in Guo and Ecker, 2004). However, no further biochemical or genetic evidence emerged to support the existence of a MAPK module in the ethylene response pathway.

THE EAGLE HAS LANDED?

Transport to the year 2003, when a report by Ouaked et al. (2003) appeared that provided new and exciting evidence that, indeed, a MAP kinase module (containing MPK6 and a related protein MPK13) had entered orbit above the ethylene-signaling pathway. This news was heralded by the community of ethylene biologists with the scientific equivalent of a tickertape parade (Chang, 2003). After a decade in flight, it seemed that this previously “lost in space” MAP kinase might finally touch down. In fact, it seemed that the landing was smack on target between CTR1 and EIN2, an essential but still enigmatic integral membrane signaling component in the central part of the ethylene pathway (Alonso et al., 1999; Binder et al., 2004; Guo and Ecker, 2004).

One very interesting and unexpected aspect of the Ouaked et al. (2003) MPK6 study was that all of the biochemical and pharmacological evidence indicated that ethylene was acting as a positive regulator of the kinase, causing increased MPK6 activity in the presence of the hormone. Given that unequivocal evidence from genetics studies previously suggested that the stage 1 MAPKKK component of the module (CTR1) functions as a negative regulator of ethylene responses (Kieber et al., 1993), their finding that an as yet unidentified stage 2 MAPKK component and stage 3 MPK6 must be inhibited by CTR1 would be the first of its kind for any MAPK module (Widmann et al., 1999); in the standard MAP kinase cascade, each stage of the module acts positively to promote the activity of the next stage, equivalent to booster rockets.

HOUSTON, WE HAVE A PROBLEM!

Closer inspection of the pharmacological data supporting the landing of MPK6 has led to the identification of a few “loose tiles” in the “heat shield” (i.e., some flaws in experimental design). In particular, the compound aminooxyacetate was used to inhibit the ethylene biosynthetic enzyme ACC oxidase. Although a demonstrated effect on MPK6 activity was observed, aminooxyacetate is, in fact, an inhibitor of ACC synthase (ACS), the penultimate enzymatic step in ethylene biosynthesis, and other pyridoxal phosphate-requiring enzymes (John et al., 1978). Moreover, another ACS inhibitor, aminoethoxyvinylglycine, was used at a toxic concentration: 1000-fold greater than the effective dose required for reversion of the ethylene-induced triple response phenotype in Arabidopsis (Guzman and Ecker, 1990). Thus, these inexplicable drug studies raise serious questions about the strength of evidence supporting the ethylene-dependent activation of MPK6.

Additional biochemical studies by Ouaked et al. (2003) using various ethylene-insensitive and constitutive ethylene signaling mutants demonstrated that the ACC-evoked MPK6 activity was dependent upon CTR1 but not EIN2. These seemingly irrefutable studies suggested that MPK6 functioned downstream of CTR1 but upstream of EIN2. Absent supporting genetic evidence from loss-of-function MPK6 mutant studies, the reported sighting of the MPK6 module in the ethylene-signaling pathway was viewed by some as a possible UFO, or at least as tentative rather than solid evidence (Guo and Ecker, 2004). In fact, recent studies of MPK6 inhibition by RNA interference failed to show a significant effect on plant ethylene responses (Menke et al., 2004) (Figure 1A). In addition, the identification of multiple T-DNA insertion alleles (Alonso et al., 2003) in both MPK6 and MPK13 genes, along with analysis of single and double mpk6 mpk13 mutants, failed to uncover any significant degree of ethylene insensitivity as would be expected for a loss-of-function mutation in a positive ethylene pathway regulator (Figures 1B and 1C; H. Guo and J.R. Ecker, unpublished data).

Figure 1.

MPK6 Loss-of-Function Mutants Do Not Exhibit an Ethylene Response Phenotype That Differs from the Wild Type.

(A) Reproduced from Figure 7 in Menke et al. (2004). Data is shown for two MPK6-silenced lines of Arabidopsis (MPK6ihpL4 and MPKihpL7) and for wild-type (Col-0) and a control transgenic line (MPK3ihpL4, which failed to exhibit silencing of MPK3). Seeds were germinated on medium containing the indicated amount of ACC (which induces ethylene accumulation), and hypocotyl lengths were measured 4 d after germination.

(B) Wild-type (Col-0), ein3-1 mutant, and three independent mpk6 mutant seedlings grown in the presence of ACC. The ein3-1 mutant shows an ethylene response phenotype for comparison. EIN3 encodes a transcription factor that is a positive regulator of ethylene signal transduction.

(C) Wild-type (Col-0), ctr1 mutant, ctr1 mpk13-1 double mutant, and mpk6-2 mpk13-1 double mutant seedlings grown in the presence of ACC.

CRASH AND BURN ON ENTRY

In this issue of The Plant Cell, Liu and Zhang (pages 3386–3399) present new and compelling evidence that MPK6 is not involved in ethylene signal transmission, in direct contrast with the findings of Ouaked et al. (2003). They provide biochemical evidence using several MPK6-specific assays that ethylene (in the form of ACC) has no effect on in planta MPK6 kinase activity. Moreover, they also clearly show that MPK6 activity is not altered in any of the Arabidopsis ethylene-insensitive or constitutive ethylene signaling pathway mutants, refuting the earlier findings. These conflicting results are difficult to reconcile. Landing of the same MPK6 module in both the biosynthetic and signaling pathways could provide a mechanism of communication to couple hormone synthesis with signaling events. However, the additional biochemical evidence and new genetic studies of Liu and Zhang indicating the lack of involvement of MPK6 in ethylene signal transduction are difficult to refute, making a dual MPK6 landing site scenario highly unlikely. Rather, it seems more likely that the first attempt to land the MAPK module burned on its steep nongenetic approach toward the ethylene signaling runway; recovery and analysis of the black box recorder might be insightful in resolving what went wrong.

THE RIGHT STUFF

If MPK6 is not involved in ethylene signaling, then where does it act? Liu and Zhang provide extensive biochemical, genetic, and molecular evidence that supports a role for MPK6 as a key regulator of ethylene biosynthesis in response to stress (Figure 2). Through a logical and elegant series of experiments that included both gain- and loss-of-function mutant studies in MPK6 and ACS6, the authors convincingly demonstrate that Arabidopsis MPK6 acts, in part, to control the rapid stress-induced synthesis of ethylene by modulating the stability of ACS6. To further understand the precise mechanism of this interaction, Liu and Zhang used phylogeny as a guide to direct the construction of a series of site-specific mutations within the C-terminal tail of the ACS6 protein. This approach allowed the identification of three critical sites of phosphorylation within this ethylene biosynthesis enzyme isoform. Furthermore, as determined by both in vitro enzyme assays and in planta gain-of-function studies, they found that phosphorylation of ACS6 at more than one Ser residue is necessary for stabilization of this short-lived protein and for the concomitant rapid increase in ethylene production. Moreover, they show that rapid posttranslational modification of ACS6 can be evoked by ectopic expression of a heterologous (tobacco) activated MAPKK called NtMEK (the suspected ortholog of Arabidopsis MKK4/MKK5) or by wounding of plant tissues. Importantly, they showed that phosphorylation of ACS6 requires functional MPK6. However, evidence is also provided that at least one stress stimulus (Flg22, a bacterial flagellar peptide) that activates ethylene production does not use the MPK6 module. Thus, the stress signaling networks that employ ethylene as an early output are more numerous than previously suspected.

Figure 2.

A Simplified Model Depicting the Induction of Ethylene Biosynthesis and the Downstream Ethylene Sensing/Signaling Pathway.

Ethylene biosynthesis is induced in response to a variety of endogenous and environmental cues. One of the early responses after the sensing of these cues is the induction of ethylene, which is achieved via upregulation of ACS activity, the rate-limiting enzyme that catalyzes the first committed step of ethylene biosynthesis. Members in the two major subgroups of the ACS family, represented by ACS5 and ACS6, are regulated by different kinase pathways. Phosphorylation stabilizes ACS6 (and possibly ACS5), and interaction of ACS5 with ETO1 (and perhaps other unknown components) targets ACS5 for degradation by the proteasome. Fine-tuning of ethylene induction is achieved by tissue-specific expression, gene activation, and the formation of heterodimers between different ACS members. ACC oxidase (ACO) activity can also affect levels of ethylene induction. Ethylene gas is perceived by a family of ER-associated receptors (ETR1, ETR2, ERS1, ERS2, and EIN4). CTR1, a negative regulator of downstream responses, is proposed to be activated by the unoccupied receptors via physical interaction and is inhibited upon binding of ethylene by the receptor. CTR1 may function through unidentified MAPKK(s) and MAPK(s). Downstream components in the ethylene pathway include several positive regulators (EIN2, EIN5, EIN6, and EIN3). In the absence of ethylene, EIN3 protein is targeted for degradation by an SCF complex containing one of the two F-box proteins, EBF1 and EBF2. In the presence of ethylene, EIN3 accumulates in the nucleus and activates gene expression, which eventually leads to ethylene-induced responses. Liu and Zhang present strong evidence that lands the MPK6 MAPK module in the pathway regulating ACS6 activity—upstream of ethylene biosynthesis in response to stress.

In addition, we now know that there are at least eight functional homodimeric isoforms of ACS enzymes in Arabidopsis. These proteins may form functional heterodimers (Tsuchisaka and Theologis, 2004a), and the corresponding genes show both distinct and overlapping patterns of expression in plant tissues (Tsuchisaka and Theologis, 2004b). Recent studies of another ACS isoform called ACS5 have revealed a similar mechanism of regulation whereby, like ACS6, this ACS isoform is also constitutively targeted for turnover in the absence of the stimulus (cytokinin). The protein mediating ACS5 turnover, encoded by a gene called ETO1, was found to be a member of a small family of plant-specific BTB/TPR domain proteins (Wang et al., 2004). Additional studies revealed that ETO1 is the first plant member of a newly emerging family of BTB domain proteins that function as E3 ubiquitin ligases in plants and animals. ETO1 functions by interacting directly with the C-terminal tail of ACS5, which targets this protein to the proteasome via subsequent direct interaction with the CUL3 component of an SCF-type complex (Wang et al., 2004). Interestingly, it seems that both ACS5 and ACS6 proteins are produced continuously in plant cells, and in the absence of the appropriate stimulus, they are continuously turned over. Given that ethylene has major developmental and physiological effects, this new example of an energy-expensive negative control of a plant signaling pathway might represent the norm, rather than the exception, in regulation of signal transduction in plants. This mechanism is akin to firing the rocket engines but not releasing it from the gantry and not allowing it to lift off the launch pad.

While there are similarities in the regulation of ACS5 and ACS6, there are also significant mechanistic differences. For example, ETO1 interacts with ACS5, but was shown not to interact with members of other ACS subfamilies, including ACS6/2 (K. Wang and J.R. Ecker, unpublished data), and unlike ACS6, ACS5 does not contain a target site for MPK6 in its C-terminal domain. Thus, it seems that tight regulation of ethylene biosynthesis is important to the plant, and control of its synthesis involves a variety of control levels (e.g., transcriptional, posttranslational) and a host of mechanisms within each level (MAP kinase modules as well as other kinases still to be discovered).

ONE SMALL STEP…

Thus, it seems that Liu and Zhang have successfully navigated reentry and soft-landing of a MAPK module containing MPK6 (and possibly MKK4/MKK5) into the ethylene biosynthesis pathway. This study is important for a number of reasons: (1) it reveals the identity of an in vivo substrate for any plant MAP kinase, that substrate being ACS6 (and likely ACS2, a related subfamily member); (2) although clues suggesting that phosphorylation can control ACS activity have been known for some time, this study provides significant mechanistic details about the regulation of biosynthesis of this critical plant growth regulator; (3) it provides a shining example that, moving forward to future missions, the “better, faster, cheaper” model (i.e., not employing genetic approaches) may not serve NASA or the plant biology community very well; and (4) given that the Arabidopsis genome contains a minimum of 20 MAP kinase modules (MAPK Group, 2002), it is apparent that novel approaches will be needed to piece together all of the stages of each of these rockets (i.e., which MAPKKKs, MAPKKs, and MAPKs are combined to form a kinase module), to determine each of their landing sites (pathways and targets), and to fully understand each of their missions (what signals they are receiving). Clearly, the complexities encountered even in these elegant and relatively simple studies of stress signaling mediated by a single MAPK suggest that we need better infrastructure at mission control (i.e., tools, technologies, and a large number of databases populated with empirical data) if we are to attempt to go where no one has gone before in the plant systems biology nebula!