Flg22 regulates the release of an ethylene response factor substrate from MAP kinase 6 in Arabidopsis thaliana via ethylene signaling

Abstract

Mitogen-activated protein kinase (MAPK)-mediated responses are in part regulated by the repertoire of MAPK substrates, which is still poorly elucidated in plants. Here, the in vivo enzyme-substrate interaction of the Arabidopsis thaliana MAP kinase, MPK6, with an ethylene response factor (ERF104) is shown by fluorescence resonance energy transfer. The interaction was rapidly lost in response to flagellin-derived flg22 peptide. This complex disruption requires not only MPK6 activity, which also affects ERF104 stability via phosphorylation, but also ethylene signaling. The latter points to a novel role of ethylene in substrate release, presumably allowing the liberated ERF104 to access target genes. Microarray data show enrichment of GCC motifs in the promoters of ERF104-up-regulated genes, many of which are stress related. ERF104 is a vital regulator of basal immunity, as altered expression in both erf104 and overexpressors led to more growth inhibition by flg22 and enhanced susceptibility to a non-adapted bacterial pathogen.

Fig. 1.

In vivo protein-protein interaction based on FRET analysis. (A) Localization of the indicated tagged proteins in transfected Arabidopsis protoplasts. Scale bar, 5 μm. (B) FRET analysis of the indicated components. A CFP-YFP fusion (positive control) and CFP alone (negative control) served as the reference. Note that data with no interaction typically show negative values because of donor bleaching during imaging. The statistical significance is indicated (Mann-Whitney test: samples with the same letters are not significantly different). (C) Co-treatment with inactive flg15d5 peptide (300 μM) blocks the ability of flg22 to disrupt the MPK6/ERF104 complex. (D) Kinase-inactive MPK6s interacted with ERF104 independently of flg22 treatment (mutations in ATP-binding pocket, MPK6KR [K₉₂R], or activation loop, MPK6AEF [S₂₂₉A, S₂₃₂A]). In (B-D), analysis was performed before (white bars) or after flg22 treatment (5–15 minutes; black bars).

Fig. 2.

Role of ET and MPK6 activity in the ERF104/MPK6 interaction. (A) To study the role of ET signaling in ET-insensitive mutants required the use of plant-derived protoplast. FRET was first tested in protoplasts of wild-type Arabidopsis to show that leaf- and cell-culture-derived protoplasts show similar results in FRET (cf. Fig. 1B). Besides flg22 (black bars), ACC treatment (shaded bars) also disrupts the MPK6/ERF104 complex. Pretreatment with AVG or AgNO₃ abrogated the flg22-mediated disruption. Letters above each bar mark the statistically distinct groupings as described in Fig. 1. (B) Genetic evidence, based on strong ET-insensitive mutants (ein2–1 or the ein3–1/eil1–1 double mutant), showed that ET signaling is required for loss of FRET after flg22 or ACC addition. (C) Lack of ERF104/MPK6KR complex disruption by flg22 treatment is not caused by dominant-negative interference of the inactive MPK6 on ET biosynthesis, as ACC addition would otherwise replace this function. (D) MPK6 activity in various genetic background was measured by immune-complex MBP kinase assays. Minor changes in the activity kinetics are experimental variations and unlikely to be real differences (A representative of 3 experiments is shown). UT = untreated.

Fig. 3.

MPK6 phosphorylates ERF104. (A) MBP kinase assays with immunoprecipitated MAPKs showing that transfection with a “constitutive-active, CA,” but not an inactive “loss-of-function, LF,” MKK5 led to the activation of MPK3/MPK6 but not MPK4. (B) Western blot showing that in vivo MPK3/MPK6 activation by CA-MKK5 (but not the LF form) led to a second band of HA-tagged ERF104 (triangle). This band is lost when a phosphatase (At2g40180) is included (CA+P) and is not seen in mpk6-derived protoplasts (bottom). The right panel shows enhanced mobility shift of the upper band in gels with polymerized Phos-tag, which retards mobility of phosphorylated proteins in SDS-PAGE. (C) Recombinant GST-ERF104 protein is phosphorylated in vitro by active MPK6, but not MPK3 and MPK4, immunoprecipitated from flg22-elicited Arabidopsis. The activities of the three kinases are shown by using the general substrate, MBP (lower panel). No MPK6 activity was precipitated from mpk6 plants (Fig. S2C), showing the specificity of the assay. (D) Scheme of the AP2-domain in ERF104 and sequence of peptides used for kinase assays. Pep1 (amino acid 226–237) and pep2 (amino acid 202–212) contain two or one predicted MAPK phosphorylation sites, respectively. (E) The peptides (in D) were used as substrates for in vitro assays with immunoprecipitated MPK6. Pep1 served as substrate, whereas pep2 is only weakly phosphorylated. Pep1 derivatives, m1 and m2, are still phosphorylated but with different strength, while m3 is no longer used as substrate (lower panel). (F) Full-length GST-ERF104m (mutated to correspond to pep1m3) is no longer phosphorylated by MPK6.

Fig. 4.

ERF104 is a transcriptional activator, and “phospho-site” mutation reduces its stability. (A) ERF104m is less stable than ERF104. Transgenic plants of Strep-tagged ERF104 variants were treated with cycloheximide (to block protein synthesis) and the protein stability monitored by Western blot. Reduced stability of ERF104m is obvious mainly after flg22 treatment. (B) Plasmids with GUS driven by the indicated promoter elements were transiently transfected into protoplasts together with (+) or without (-) 35S::ERF104. The promoter activity is shown as a ratio of the GUS activity normalized to that of constitutively expressed luciferase (LUC). The core binding sequences of the promoter elements are indicated. (C) Stable 35S::ERF104 transgenics were crossed into GUS reporter plants. GUS activity (blue staining) can be seen in crosses with the lines with GCC promoter elements but not with mutated GCC (GCCmut) or WRKY elements (4xW2). A weaker staining of plants with jasmonate-responsive elements (JERE) is likely an indirect effect as ERF104 does not bind to JERE elements in EMSA.

Fig. 5.

Effects of modulating ERF104 expression. (A) Chromatin immunoprecipitation (ChIP) shows higher levels of the PDF1.2 promoter in immunoprecipitates of Strep-tagged ERF104 from ERF104^OE compared with Col-0 plants. (B) Botrytis cinerea disease progression, as monitored by symptom development in leaves (left) or biomass determination (right). Each biomass data point is an independent experiment, depicts fungal biomass in 18 leaf disks (with technical duplicates) and is shown as fold changes relative to Col-0 (horizontal dash indicates mean). The fungal biomass in the hypersusceptible mutant, pad3, is ≈2.3 fold higher. Note that enhanced susceptibility of the overexpressors seems to correlate with the transgene expression level (except for ERF104^OE line 2). (C) Both erf104 and ERF104^OE show enhanced susceptibility to the bean pathogen, P. syringae pv. phaseolicola (infiltrated at 2 × 10⁸ cfu/ml). t test P values: *<0.05, **<0.01, ***<0.001. (D) Root growth inhibition by flg22 (10 μM) is enhanced in plants with altered ERF104 expression. Statistically significant differences (marked as in (C) in root length, compared with Col-0 roots, are indicated.

Fig. 6.

Model of flg22 effect on MPK6/ERF104 interaction. The ERF104/MPK6 complex disruption requires flg22 stimulation of MPK6 activity (1) that also positively affects ERF104 stability, as well as ET signaling (2). The rapid ET signal may be upstream (A) or downstream (B) of MPK6 activation; but the inhibitor/mutant analyses support the latter. Yoo et al. reported bifurcate pathways downstream of ETR (ET receptors) and CTR1, one of which includes MKK9, to antagonistically control EIN3 stability (10). Question marks denote the report that active MKK9 can raise ET levels to feedback positively into ET signaling (25); but this conflicts with results reported by Yoo et al. (10). The ERF104/MPK6 complex disruption by flg22 is independent of the MKK9 pathway but dependent on EIN2 and EIN3/EIL (EIN3-like) proteins.