Arabidopsis GCN2 kinase contributes to ABA homeostasis and stomatal immunity

General Control Non-derepressible 2 (GCN2) is an evolutionarily conserved serine/threonine kinase that modulates amino acid homeostasis in response to nutrient deprivation in yeast, human and other eukaryotes. However, the GCN2 signaling pathway in plants remains largely unknown. Here, we demonstrate that in Arabidopsis, bacterial infection activates AtGCN2-mediated phosphorylation of eIF2α and promotes TBF1 translational derepression. Consequently, TBF1 regulates a subset of abscisic acid signaling components to modulate pre-invasive immunity. We show that GCN2 fine-tunes abscisic acid accumulation and signaling during both pre-invasive and post-invasive stages of an infection event. Finally, we also demonstrate that AtGCN2 participates in signaling triggered by phytotoxin coronatine secreted by P. syringae. During the preinvasive phase, AtGCN2 regulates stomatal immunity by affecting pathogen-triggered stomatal closure and coronatine-mediated stomatal reopening. Our conclusions support a conserved role of GCN2 in various forms of immune responses across kingdoms, highlighting GCN2’s importance in studies on both plant and mammalian immunology.

AtGCN2 is essential for pathogen-triggered eIF2α phosphorylation without affecting TBF1 transcript accumulation. a. Transcript accumulation of AtGCN2 was measured in four-week-old plants by real-time RT-PCR. The box plots extend from the 25 th to 75 th percentiles and the whiskers extend from the minimum to the maximum level. Median values are plotted in the box with data generated from three independent biological replicates, each of them representing means of two technical replicates. Statistical analysis was performed with Student's ttest, p-value is listed. b. Detection of phosphorylated form of eIF2α in the samples prepared from two-week-old plants treated with Pst DC3000 or Pst DC3118 (OD 600nm = 0.02) at the indicated time points in hours (h) (full blots for Fig. 1a). Phosphorylation state-specific (S51) anti-human eIF2α antibody was used. Ponceau S staining shows loading amounts. c. Detection of phosphorylated form of eIF2α in the samples prepared from two-week-old plants treated with Pst DC3000 (OD 600nm = 0.02) or control H 2 O at the indicated time points in hours (h). Phosphorylation state-specific (S51) anti-human eIF2α antibody was used. Ponceau S staining was used to determine loading. d.  fig. 1h. j. Transcript accumulation of eIF2α was measured in four-weekold plants by real-time RT-PCR. The box plots extends from the 25 th to 75 th percentiles and the whiskers extend from the minimum to the maximum level. Median values are plotted in the box with data generated from three independent biological replicates, each of them represented as means of two technical replicates. Statistical analysis was performed with Student's t-test and no significant difference was detected. k. Time course total eIF2α protein accumulation in two-week-old plants upon Pst DC3000 or Pst hrcC (OD 600nm = 0.02) challenge. Time points are shown in hours (h) (full blots for Fig. 1c). Ponceau S staining shows loading amounts. l. Transcript accumulation of iudA (encoding the GUS reporter) was measured by real-time RT-PCR in four-week-old transgenic T 3 plants uORF1-uORF2-GUS (Ler), uorf1-uorf2-GUS (Ler) and uORF1-uORF2-GUS (atgcn2) treated with Pst DC3000 (OD 600nm = 0.02), sampled at indicated time points. Data represent the mean and standard error of three independent biological replicates. m. Transcript accumulation of iudA (encoding the GUS reporter) was measured by realtime RT-PCR in four-week-old transgenic T 3 plants uORF1-uORF2-GUS (Ler), uorf1-uorf2-GUS (Ler) and uORF1-uORF2-GUS (atgcn2) treated with Psm ES4326/avrRpm1 (OD 600nm = 0.02), sampled at indicated time points. Data represent the mean and standard error of three independent biological replicates. Median values are plotted in the box with data generated from three independent biological replicates. Statistical analysis was performed by two-way ANOVA followed by Tukey's test, letters above the bars signify statistically significant differences among groups (p≤0.05). c. Stomatal aperture width was measured in epidermal peels of fourweek-old plants that were treated with Pst DC3118 (OD 600nm = 0.2) for three hours. The box plots extend from the 25 th to 75 th percentiles and the whiskers extend from the minimum to the maximum level. Median values are plotted in the box with data generated from 180 stomata derived from three independent biological replicates. Two-way ANOVA with Tukey's test was performed, letters above the bars signify statistically significant differences among groups (p≤0.05). d. Transcript induction of SLAC1 was determined in four-week-old plants upon treatments with Pst DC3118 (OD 600nm = 0.2) spray inoculation using real-time RT-PCR. The box plots extend from the 25 th to 75 th percentiles and the whiskers extend from the minimum to the maximum level. Median values are plotted in the box with data generated from three independent biological replicates. Two-way ANOVA with Tukey's test was performed, asterisks indicate significant differences compared to wild-type Ler (***p≤0.001). jasmonate-zim-domain protein 6 (JAZ6).

AT5G02240
Protein is tyrosine-phosphorylated and its phosphorylation state is modulated in response to ABA in Arabidopsis thaliana seeds. 2 2

AT5G24030
The SLAH3 protein has similarity to the SLAC1 protein involved in ion homeostasis in guard cells. 2 0

AT5G63980
Encodes a bifunctional protein that is involved in the response to cold, drought (negative regulator of drought tolerance), and ABA.  Table 1. Summary of genes associated with ABA response whose transcription is TBF1-dependent upon elf18 challenge. Numbers of exact and degenerate sequence variants corresponding to TL1 (Translocon 1; TBF1 binding site) are shown to the right.

Supplementary Tables
Supplementary Table 2. List of primers used in this study. From the host perspective, the ability to suppress the action of coronatine on stomata without impairing COI1-JAZ-dependent plant defense responses would be highly desirable in an event of an impending attack of herbivores and necrotrophs. In response to this host defense mechanism, it is plausible that P. syringae might have evolved an additional, COR-independent method of inducing JA signaling. Normally, this mechanism could be masked by the action of GCN2 while it would become de-repressed and active in the atgcn2 mutant. If such a mechanism were to exist, it would be likely short-lived, and occur during very early stages of infection, as evidenced by the trends of our expression data (Figs. 3d-f). The enhanced expression of MYC2 in the Pst DC3118-treated atgcn2 wears off over time, reaching the levels observed in Pst DC3000-treated Ler at 4 hr time point (Fig. 3d). Similarly, the ANAC019 and ANAC055 expression levels decline in the atgcn2 plants after a transient induction at the 2 hr time point (Figs. 3e and 3f). However, these transcript levels remain consistently higher than expression observed in the Pst DC3118-treated Ler.
This transient early mis-regulation of MYC2 and NACs could be, in part, responsible for the fact that the atgcn2 plants spray-inoculated with Pst DC3118 no longer show enhanced disease resistance, and have bacterial loads comparable to those of Ler ( Fig. 3b and Supplementary Fig. 2d).
In line with our findings, another node of convergence between ABA and JA signaling was recently reported 6 . HAI1 PP2CA was shown to be induced by ABA through the AREB/ABF transcription factors as well as by Pst DC3000 through COR-mediated activation of MYC2. In turn, HAI1 dephosphorylates MPK3 and MPK6 and is necessary for COR-mediated suppression of MPK3/MPK6 activation and immunity. Taken together and consistent with the previous reports, we propose that AtGCN2 is a positive regulator of ABA biosynthesis and signaling, contributes to MYC2-mediated JA signaling, and virulent pathogens potentially target AtGCN2 to establish disease susceptibility.