Salicylic Acid Engages Central Metabolic Regulators SnRK1 and TOR to Govern Immunity by Differential Phosphorylation of NPR1

SUMMARY

Immunity requires a delicate balance between combating infection and preserving metabolic functions. However, the mechanisms by which immune responses are coordinated with cellular metabolism remain largely unknown. Here, we show that NONEXPRESSER OF PR GENES 1 (NPR1), the central plant immune regulator of salicylic acid (SA)-mediated defense responses, is controlled by a cascade of posttranslational modifications (PTMs) involving two master nutrient-sensing kinases. In the absence of pathogen challenge, TARGET OF RAPAMYCIN (TOR) inhibits NPR1 through phosphorylation at Ser-55/59. During defense responses, elevated SA reduces sugar phosphate levels and enhances SNF1-RELATED KINASE 1 (SnRK1) activity, which in turn inhibits TOR signaling and phosphorylates NPR1 at Ser-557, a modification required for subsequent PTMs and NPR1 activation. Together, our findings identify SA as a coordinator of growth and immunity in response to pathogen challenge by modulating the central metabolic regulators SnRK1 and TOR, which antagonistically control NPR1 immune activity through differential phosphorylation.

Graphical Abstract

In brief

The immune signal salicylic acid triggers metabolic changes to activate SnRK1 while inhibiting TOR. These central metabolic kinases antagonistically phosphorylate the key immune regulator NPR1, thereby coupling cellular metabolic state to systemic acquired resistance.

INTRODUCTION

Plants have evolved a sophisticated innate immune system mediated by plasma membrane-associated pattern recognition receptors upon detection of microbe-associated molecular patterns^1,2 and by intracellular nucleotide-binding leucine-rich repeat immune receptors which respond to specific pathogen effectors³. The resulting pattern-triggered immunity (PTI) and effector-triggered immunity (ETI) restrict local infection⁴ and can induce systemic acquired resistance (SAR), a long-lasting, broad-spectrum defense response mediated by the immune hormone salicylic acid (SA)^5–7. Elevation in SA levels induces massive transcriptional reprogramming, leading to the production and secretion of antimicrobial proteins, such as PATHOGENESIS-RELATED PROTEIN 1 (PR1), and establishment of SAR^8–12. This SA-mediated immune response requires the function of NONEXPRESSER OF PR GENES 1 (NPR1), which was first identified in Arabidopsis^8,9,13–15 and has orthologs in most angiosperms¹⁶. NPR1 is required for initiating the SA-mediated transcriptional cascade as a cofactor for the TGA group of transcription factors (TFs)^17–21.

Though the NPR1 protein is constitutively expressed, its activity and stability are tightly controlled through post-translational modifications (PTMs)²². In the uninduced state, NPR1 is phosphorylated at Ser-55/59²³ and sequestered in the cytoplasm as an oligomer via formation of intermolecular disulfide bonds²⁴. Upon SA induction, NPR1 is reduced by thioredoxins to antagonize the effect of S-nitrosylation²⁵ and dephosphorylated at Ser-55/59²³. This enables NPR1 translocation into the nucleus, where it is SUMOylated²³ and phosphorylated at Ser-11/15²⁶ to promote its interaction with the ubiquitin/proteasome system, leading to its full activation and turnover^26,27. Disrupting these PTMs causes either SAR deficiency or autoimmunity^23,26,27. However, few enzymes have been implicated in regulating NPR1. While PKS5/SnRK3.22 phosphorylates NPR1 at the C-terminal region²⁸ and SnRK2.8 phosphorylates NPR1 at Thr-373 and Ser-589²⁹, the full identity and regulation of enzymes which modify NPR1 in response to SA remain obscure.

Although SA has been studied primarily as a defense hormone, it has also been shown to affect plant metabolism. SA induces alternative respiration via alternative oxidase (AOX)^30,31, which regulates redox and metabolic homeostasis³², and binds to CATALASE 2 to inhibit H₂O₂ scavenging^33,34. SA-induced redox changes also facilitate NPR1 nuclear translocation²⁴. However, how metabolism influences SA-mediated immunity remains unclear.

Similar to other eukaryotes, plant metabolism is regulated by central nutrient-sensing kinases TARGET OF RAPAMYCIN (TOR)^35–37 and SNF1-RELATED PROTEIN KINASE 1 (SnRK1)^38–40, the latter being orthologous to yeast SNF1 and mammalian AMPK. Nutrient-activated TOR promotes anabolic processes and growth by activating translation^37,41,42. In contrast, SnRK1 activity is inversely correlated with glucose-6-phosphate (Glc6P) and trehalose-6-phosphate (Tre6P) levels^43–45. Abiotic stress activates SnRK1^39,46,47, which suppresses TOR and shifts growth toward stress tolerance⁴⁸. Although SnRK1 and TOR are also implicated in immunity^48–56, their mechanistic links to plant defense have yet to be elucidated.

Here, we show that SnRK1 and TOR regulate SA-mediated immunity via differential phosphorylation of NPR1. SA decreases sugar phosphate levels and activates SnRK1, promoting Ser-557 phosphorylation, a PTM required for NPR1 function. Furthermore, SA-activated SnRK1 suppresses TOR activity, inhibiting TOR-dependent Ser-55/59 phosphorylation, thereby releasing NPR1 from its quiescent state. Therefore, we demonstrate a direct link between SA-mediated immunity and cellular metabolic pathways through differential phosphorylation of NPR1 mediated by TOR and SnRK1.

RESULTS

Ser-557 phosphorylation is required for SA-mediated activation of NPR1

To detect NPR1 PTMs in an unbiased screen, we employed targeted mass spectrometric analysis to detect NPR1-specific phospho-peptides from mature Arabidopsis plants after SA treatment. The analysis revealed Ser-557 phosphorylation (pSer-557) in NPR1 peptides following proteolytic digestion with either trypsin or GluC (Figure 1A). We then developed a phospho-specific antibody against Ser-557 (α-pS557). Although different antibody aliquots displayed one of two distinct nonspecific banding patterns, the antibody was validated using negative controls (Table S1), such as the phospho-dead 35S:npr1^S557A-GFP (npr1^S557A-GFP) and phospho-mimetic 35S:npr1^S557D-GFP (npr1^S557D-GFP) mutant lines (Figure S1A). We found that pSer-557 was elevated in mature transgenic Arabidopsis (35S:NPR1-GFP/npr1-2) plants following SA treatment in a dose-dependent manner (Figures 1B and S1B). Importantly, the pSer-557 level induced by exogenous application of 1 mM SA was similar to that in systemic tissues after biological SAR induction by the bacterial pathogen Pseudomonas syringae pv. maculicola ES4326 (Psm) carrying the effector gene avrRpt2 (Figure 1B, top). Infection of local tissues by Psm without avrRpt2 produced similar results (Figure S1C). The endogenous NPR1 protein was similarly phosphorylated upon SA treatment (Figure 1B, bottom).

Figure 1. Phosphorylation of NPR1 at Ser-557 is necessary and partially sufficient for SA-induced disease resistance.

(A) Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis of NPR1 phosphorylation. 35S:NPR1-GFP plants were treated with mock or 0.5 mM SA for 24 hours, followed by trypsin (top)/Glu-C (bottom) digestion for LC-MS/MS.

(B) Phosphorylation of NPR1 at Ser-557. Top: 3.5-week-old 35S:NPR1-GFP plants were treated with mock, 0.1 or 1 mM SA 24 hours before collection, or inoculated with Psm avrRpt2 (avrRpt2, OD₆₀₀ = 0.01) in local leaves 2 days before systemic collection. Bottom: 3.5-week-old wild-type (WT) plants were treated with mock or 1 mM SA 8 hours before collection. NPR1-GFP and endogenous NPR1 were immunoprecipitated using GFP-trap and α-NPR1, respectively, and then blotted with an anti-phosphorylated Ser-557 (α-pS557) antibody. Relative phosphorylation levels (top) were calculated by normalizing band intensities to total NPR1-GFP levels.

(C and D) Representative image (C) and NPR1 protein levels (D) of 3.5-week-old WT, npr1-2, 35S:NPR1-GFP, and independent 35S:npr1^S557A-GFP and 35S:npr1^S557D-GFP transgenic lines. Scale bar, 5 cm.

(E) Basal resistance in 35S:NPR1-GFP, 35S:npr1^S557A-GFP, 35S:npr1^S557D-GFP, and npr1-2 plants inoculated with Psm (OD₆₀₀ = 0.0002). Bacterial growth was measured at 3 days post-inoculation (dpi) in all experiments. cfu, colony-forming units. n = 8.

(F) SA-induced resistance in 35S:NPR1-GFP, 35S:npr1^S557A-GFP, 35S:npr1^S557D-GFP, and npr1-2 plants. 3.5-week-old plants were treated with mock or 1 mM SA, followed by inoculation with Psm (OD₆₀₀ = 0.001) 24 hours later. n = 8.

(G) PR1 gene expression. RT-qPCR was performed in 3.5-week-old 35S:NPR1-GFP, 35S:npr1^S557A-GFP, and 35S:npr1^S557D-GFP plants 24 hours after mock, 0.1 mM, or 1 mM SA treatment. UBQ5 expression was used for normalization in all RT-qPCR experiments.

(H and I) Representative image (H) of and NPR1 protein levels (I) of 3.5-week-old WT, npr1-2, and independent NPR1pro:NPR1-eYFP, NPR1pro:npr1^S557A-eYFP, and NPR1pro:npr1^S557D-eYFP transgenic lines. Scale bar, 5 cm. Samples in (I) were collected 16 hours after mock or 0.1 mM SA treatment.

(J) Basal and SA-induced resistance in WT, npr1-2, NPR1pro:NPR1-eYFP (NPR1), NPR1pro:npr1^S557A-eYFP (npr1^S557A), and NPR1pro:npr1^S557D-eYFP (npr1^S557D) plants treated with mock or 1 mM SA 24 hours before inoculation with Psm (OD₆₀₀ = 0.001). n = 8.

(K) PR1 gene expression. RT-qPCR was performed in 3.5-week-old WT, npr1-2, NPR1pro:NPR1-eYFP (NPR1), NPR1pro:npr1^S557A-eYFP (npr1^S557A), and NPR1pro:npr1^S557D-eYFP (npr1^S557D) plants 24 hours after mock, 0.1 mM SA, or 1 mM SA treatment.

All data in the manuscript are presented as mean ± s.e.m. Individual columns were compared using one-way ANOVA with Tukey’s post hoc (E, F, and J) or two-tailed Student’s t-test (G and K), and the interactions were tested using two-way ANOVA (F and J). For all figures, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., not significant; different lowercase letters indicate significant differences between groups by Tukey’s post hoc at p < 0.05. See also Figure S1 and Table S1.

To investigate how pSer-557 contributes to NPR1 function, we examined the npr1^S557A-GFP and npr1^S557D-GFP variants in the npr1-2 background. While no obvious growth phenotype was observed in npr1^S557A-GFP, npr1^S557D-GFP lines exhibited stunted growth correlating with npr1^S557D levels (Figures 1C and 1D), suggesting that npr1^S557D might be partially auto-active. Therefore, we examined the basal resistance of NPR1-GFP, npr1^S557A-GFP, and npr1^S557D-GFP to Psm (OD₆₀₀ = 0.0002) and found that, similar to npr1-2, npr1^S557A-GFP was more susceptible than NPR1-GFP (Figures 1D and 1E). In contrast, npr1^S557D-GFP exhibited resistance correlating with npr1^S557D-GFP levels. We then tested the mutants for SA-induced resistance (Psm; OD₆₀₀ = 0.001). While npr1^S557A-GFP plants lacked SA-induced defense, npr1^S557D-GFP plants exhibited enhanced basal resistance but reduced sensitivity to SA (Figure 1F). We next examined the impact of pSer-557 on NPR1-mediated transcriptional activation, using PR1 as a marker. SA-induced expression of PR1 was abolished in npr1^S557A-GFP, whereas npr1^S557D-GFP exhibited elevated basal PR1 that was further induced by SA (Figure 1G). These results support our hypothesis that pSer-557 is necessary, but only partially sufficient for NPR1 activation by SA. Native promoter-driven NPR1-eYFP and its Ser-557 mutant lines yielded similar results (Figures 1H–1K), confirming that NPR1 activity is primarily controlled through PTMs²².

Elucidating the PTM cascade in SA-mediated activation of NPR1

To map the spatial and temporal dynamics of pSer-557 among NPR1 modifications, we first performed fractionation and found that Ser-557 phosphorylation occurs predominantly in the nucleus (Figure 2A). We next investigated temporal relationships between pSer-557 and other critical NPR1 phosphorylation sites: Ser-55/59 phosphorylation suppresses NPR1 activity^23,57, whereas Ser-11/15 phosphorylation promotes its activity and turnover through ubiquitination²⁶. We monitored changes in pSer-55/59, pSer-557, and pSer-11/15 following SA treatment using phospho-specific antibodies^23,26, with α-pS55/59 detecting a broad band when probing plant extracts²³. We observed that Ser-55/59 dephosphorylation began within 30 minutes of SA treatment (Figures 2B and 2C), followed by Ser-557 phosphorylation at two hours. Ser-11/15 phosphorylation began at four hours and peaked at approximately eight hours. These results suggest a PTM cascade during SA-induced NPR1 activation.

Figure 2. NPR1 Ser-557 phosphorylation is a pivotal step in the SA-induced PTM cascade that activates defense gene expression.

(A) Subcellular localization of NPR1 Ser-557 phosphorylation. Seven-day-old 35S:NPR1-GFP seedlings were treated with mock or 0.3 mM SA for 6 hours. After subcellular fractionation, nuclear and cytoplasmic proteins were blotted with α-pS557. Total NPR1 levels were detected using α-GFP, and fraction purity was determined using α-UGPase and α-Histone H3. *, nonspecific band.

(B and C) Time course of SA-induced NPR1 phosphorylation. 35S:NPR1-GFP and npr1-2 seedlings were treated with 0.3 mM SA for the indicated durations. NPR1 phosphorylation was detected using α-pS55/59, α-pS11/15, and α-pS557 (B) and quantified by normalizing to the total NPR1 levels (α-GFP) (C). For each blot, the band with the highest relative phosphorylation level was set to 1. n = 3.

(D) Hierarchy in NPR1 PTMs. 35S:NPR1-GFP, 35S:npr1^S55/59A-GFP, 35S:npr1^S55/59D-GFP, 35S:npr1^sim3-GFP, and npr1-2 seedlings were treated with mock or 0.3 mM SA for 6 hours. Proteins were analyzed using α-pS557, α-pS11/15, and α-GFP.

(E and F) Analysis of Ser-55/59 and Ser-11/15 phosphorylation in Ser-557 phospho-mutant lines. 35S:NPR1-GFP, independent 35S:npr1^S557A-GFP, and 35S:npr1^S557D-GFP transgenic lines, and npr1-2 seedlings were treated as in (D). Proteins were analyzed with α-pS55/59 (E), α-pS11/15 (F), and α-GFP.

(G and H) Quantification of Ser-11/15 phosphorylation in Ser-557 mutants from four independent experiments. SA-induced fold changes in Ser-11/15 phosphorylation of NPR1-GFP, npr1^S557A-GFP, and npr1^S557D-GFP relative to mock were calculated after normalizing each band intensity to its loading control (G). SA-induced Ser-11/15 phosphorylation in npr1^S557D-GFP relative to that observed in NPR1-GFP was calculated (H). n = 4.

(I) Analysis of Ser-557 phosphorylation in Ser-11/15 phospho-mutant lines. 35S:NPR1-GFP, 35S:npr1^S11/15A-GFP, 35S:npr1^S11/15D-GFP, and npr1-2 seedlings were treated and analyzed as in (D) using α-pS557 and α-GFP.

(J) Analysis of NPR1 ubiquitination in Ser-557 phospho-mutant lines. 35S:NPR1-GFP, 35S:npr1^S557A-GFP, and 35S:npr1^S557D-GFP seedlings were treated as in (D). Ubiquitinated proteins were immunoprecipitated using Ubiquitin-trap. NPR1 was detected using α-GFP. The intensity of the upper, ubiquitinated fraction (>90 kDa) was normalized to total NPR1 levels.

Individual columns were compared using one-way ANOVA with Tukey’s post hoc test (C) or two-tailed Student’s t-test (G and H). Total protein loading for (B, D-F, and I) was estimated using Ponceau S staining. See also Figure S1.

We next investigated the sequential dependency of PTMs during NPR1 activation genetically. It is known that npr1^S55/59A induces pSer-11/15 without SA treatment, while npr1^S55/59D fails to induce either pSer-11/15 or SAR²³. We observed that npr1^S55/59A triggered Ser-557 phosphorylation under mock conditions, which was further enhanced by SA, while npr1^S55/59D failed to induce pSer-557 even after SA treatment (Figure 2D). Conversely, neither the npr1^S557A nor npr1^S557D mutation affected Ser-55/59 dephosphorylation (Figure 2E). These results indicated that pSer-557 occurs downstream of and is dependent on Ser-55/59 dephosphorylation. We also found that while the SUMOylation mutant npr1^sim3 abolished pSer-11/15, it had no significant effect on pSer-557, indicating that NPR1 SUMOylation occurs downstream of or independent of pSer-557 (Figure 2D). SA-induced pSer-11/15, which peaks later than pSer-557, was blocked by npr1^S557A but enhanced by npr1^S557D as determined by quantification (Figures 2F–2H), indicating that pSer-11/15 depends on pSer-557. This epistatic relationship was further confirmed by normal pSer-557 induction in npr1^S11/15A and npr1^S11/15D (Figure 2I). However, Ser-11/15 was not constitutively phosphorylated in npr1^S557D-GFP under mock conditions (Figure 2F), indicating that although pSer-557 is necessary, it is not sufficient to promote pSer-11/15, which likely requires an unknown SA-dependent kinase.

Ser-557 phosphorylation regulates NPR1 activity by modulating its interactions with partners

We next tested whether pSer-557 alters SA-mediated NPR1 nuclear translocation and found that neither the npr1^S557A nor npr1^S557D mutation had a significant effect (Figure S1D). Interestingly, even without SA treatment, npr1^S557D-GFP formed nuclear SA-induced NPR1 biomolecular condensates (nSINCs), normally formed by NPR1-GFP after SA treatment¹⁹. We previously found that nSINC formation is crucial for NPR1-mediated recruitment of TFs and other components¹⁹. Therefore, we tested whether pSer-557 enhances NPR1’s interaction with its transcriptional partners using co-immunoprecipitation (co-IP). We found that interaction with TGA3⁵⁸ was reduced in npr1^S557A but enhanced in npr1^S557D (Figure S1E). Similar changes were observed for interaction with CDK8, a mediator component important for NPR1 function⁵⁹ (Figure S1F).

Sequential NPR1 ubiquitination is required for full activation and subsequent turnover^26,27. As pSer-557 primes Ser-11/15 phosphorylation, which is required for NPR1 ubiquitination and turnover, we investigated whether pSer-557 promotes NPR1 function by enhancing ubiquitination. We found that pSer-557 enhanced NPR1 interaction with the E3 ubiquitin ligase CUL3A (Figure S1G). This interaction depended on pSer-11/15, as the impaired interaction between npr1^S557A and CUL3A was restored in npr1^{S557A_S11/15D}. Consistently, npr1^S557D and npr1^S557A exhibited enhanced and reduced ubiquitination, respectively, likely reflecting their respective pSer-11/15 levels (Figure 2J).

SnRK1 directly phosphorylates NPR1 at Ser-557

To identify the kinase phosphorylating Ser-557, we focused on the SnRK kinase family, known for its roles in stress responses^60–63. Crucially, the Ser-557 site matches the consensus sequence for SnRK1 substrates, Ф-x-Basic-xx-Ser-xxx-Ф (x, any residue; Ф, hydrophobic)^64,65. We performed an in vitro phosphorylation screen with Nicotiana benthamiana-expressed SnRK kinases and found that SnRK1α2, a catalytic subunit of SnRK1, could phosphorylate Ser-557 (Figure S2A). As Arabidopsis expresses SnRK1α1 and SnRK1α2, we further tested them in vitro. Both in vitro-synthesized SnRK1α1 and SnRK1α2 phosphorylated Ser-557 at low levels, consistent with their low catalytic activity, indicated by moderate phosphorylation of the conserved T-loop (Thr-175 in SnRK1α1 and Thr-176 in SnRK1α2)^66,67 detected using α-pAMPK (Figure 3A). Ser-557 phosphorylation was further enhanced by SnRK1-ACTIVATING KINASE 1 (SnAK1), which alone did not phosphorylate Ser-557. Moreover, the phospho-dead npr1^S557A mutation diminished the overall phosphorylation detected using a general α-pS/pT phospho-antibody, indicating that Ser-557 is the primary SnRK1α target. Mass spectrometry confirmed that only pSer-557 was increased with activated SnRK1α, while other detectable phospho-Ser/Thr residues showed minimal changes (Figures 3B and S2B).

Figure 3. SnRK1 phosphorylates NPR1 at Ser-557 in response to SA induction.

(A) In vitro phosphorylation assays of recombinant NPR1. Ser-557 and total phosphorylation levels were detected using α-pS557 and α-pS/pT, respectively. In vitro SnRK1 activity was estimated by measuring phosphorylation of its T-loop using α-pAMPK, and total SnRK1/SnAK1 proteins were detected using α-FLAG. Total NPR1 was stained using Coomassie Brilliant Blue (CBB). *, nonspecific band.

(B) LC-MS/MS analysis of in vitro phosphorylated GST-NPR1 by SnRK1. Extracted ion chromatograms (XICs) of MS2 fragment ions over retention time for the most confidently detected peptide (AFS[Phos]EDNLELGNSSLTDSTSSTSK) are shown for each group.

(C) Interaction of GFP-fused SnRK1α with NPR1-mCherry or HA-mCherry transiently expressed in N. benthamiana, evaluated 6 hours after 1 mM SA treatment.

(D) Interaction between endogenous SnRK1α and NPR1-GFP or GFP in transgenic Arabidopsis seedlings treated with water or 0.3 mM SA for 6 hours. SnRK1α intensities for the mock samples were set to 1 for normalization.

(E) In vitro interactions between SnRK1α, SnAK1, and GST-fused NPR1.

(F) Effect of snrk1 mutations on Ser-557 phosphorylation. NPR1-GFP was transiently expressed in WT and snrk1α1^−/− snrk1α2^+/− (sesquiα2-1) leaf-derived protoplasts, followed by treatment with mock or 0.1 mM SA for 6 hours. NPR1-GFP was immunoprecipitated and analyzed with α-pS557 and α-GFP. Relative phosphorylation levels were calculated after normalization to total NPR1-GFP, with the mock-treated WT sample set to 1.

(G) Effect of snrk1 mutations on Ser-557 phosphorylation of endogenous NPR1. 3.5-week-old WT and sesquiα2-1 plants were treated with mock or 1 mM SA 8 hours before collection. NPR1 protein was immunoprecipitated using α-NPR1 and analyzed using α-pS557 and α-NPR1. Relative phosphorylation levels were calculated as in (F). *, IgG heavy chain band from IP.

(H) Effect of snrk1 mutations on endogenous NPR1 Ser-557 phosphorylation during biological SAR induction. 3.5-week-old WT and sesquiα2-1 plants were inoculated with MgCl₂ or Psm avrRpt2 (OD₆₀₀ = 0.01) in two local leaves, followed by collection of three systemic leaves after 2 days. NPR1 was enriched by α-NPR1 IP and analyzed using α-pS557 and α-NPR1. Relative phosphorylation levels were calculated as in (F). *, IgG heavy chain band from IP.

(I) Effect of snrk1 mutations on Ser-557 phosphorylation in stable transgenic lines. NPR1-eYFP seedlings in npr1-7 (npr1) or npr1 sesquiα2-1 (npr1 α2–1) backgrounds were treated with mock or 0.3 mM SA for 6 hours. Proteins were analyzed using α-pS557 and α-GFP. Relative phosphorylation levels were calculated as in (F). *, nonspecific band.

(J) Requirement of SA biosynthesis for Ser-557 phosphorylation in systemic tissue. 3.5-week-old WT and SA-deficient sid2-2 plants were inoculated with Psm avrRpt2 (OD₆₀₀ = 0.01) in local leaves, followed by collection of systemic leaves at the indicated time points. NPR1 protein was enriched through α-NPR1 IP and analyzed using α-pS557 and α-NPR1. Relative phosphorylation levels were calculated by normalizing to the corresponding total protein levels, with the WT sample at 12 hpi set to 1. See also Figure S2.

Co-IP in N. benthamiana and transgenic Arabidopsis confirmed interactions between both SnRK1α subunits and NPR1 in vivo (Figure 3C), which were enhanced by SA treatment (Figure 3D), likely due to increased NPR1 nuclear translocation. Pull-down assays further showed that SnRK1α1 and SnRK1α2 interact with NPR1 in vitro independently of their phosphorylation by SnAK1 (Figure 3E). To confirm that SnRK1 phosphorylates Ser-557 in vivo, we treated NPR1-GFP-expressing protoplasts from wild-type (WT; Col-0) and snrk1α1^−/− snrk1α2^+/− (sesquiα2-1) with mock or SA, as the double snrk1α1 snrk1α2 knockout is lethal^68,69. We found that sesquiα2-1 was defective in SA-induced pSer-557, suggesting that SnRK1 phosphorylates Ser-557 in vivo (Figure 3F). Endogenous NPR1 pSer-557 was also impaired in sesquiα2-1 following SA treatment (Figure 3G) and in systemic tissues after SAR induction by Psm avrRpt2 (Figure 3H). Similarly, the snrk1 mutation reduced SA-induced pSer-557 in NPR1-overexpressing seedlings (Figure 3I), while systemic Ser-557 phosphorylation following SAR induction required de novo SA biosynthesis (Figure 3J). Together, these data establish SnRK1 as the bona fide Ser-557 kinase during SAR.

To investigate whether phosphorylation of NPR1 by SnRK1 is conserved across plant species, a phylogenetic analysis was performed, revealing a largely conserved Ser/Thr at the position corresponding to AtNPR1 Ser-557 (Figure S2C). Both tomato (Solanum lycopersicum) and sorghum (Sorghum bicolor) NPR1 orthologs containing the conserved Ser/Thr were phosphorylated by their corresponding SnRK1 orthologs activated by GEMINIVIRUS REP-INTERACTING KINASE (GRIK/SnAK) orthologs (Figures S2D and S2E). In contrast, OsNH1, the rice (Oryza sativa) NPR1 ortholog with histidine at this position, was not phosphorylated by OsSnRK1 (Figure S2F).

SnRK1 is required for SA- and NPR1-mediated immunity

We next examined SAR marker gene expression in WT and two sesqui mutants by qPCR and found that both sesqui mutants were defective in PR1 and WRKY38 induction (Figures 4A and 4B). To comprehensively investigate the role of SnRK1 in SA- and NPR1-mediated immunity, we first performed QuantSeq analysis in WT, npr1-2, npr1^S557A, and sesqui mutants after mock or SA treatments. Under mock conditions, all genotypes exhibited similar transcriptional patterns (Figures 4C and 4D), as SnRK1 retains partial functionality in the sesqui mutants^68,69, and NPR1 is quiescent without stimulus²⁴. SA treatment revealed 3,666 SA-responsive differentially expressed genes in WT, compared to 1,865 in npr1-2, 2,241 in npr1^S557A, and only 845 and 645 in sesquiα2-1 and sesquiα2–2, respectively (|log2foldchange| > 1, adjusted p-value < 0.05) (Figures 4E and 4F; Table S2). Gene Ontology (GO) Biological Process (BP) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of SA-induced, Ser-557- and SnRK1-dependent genes showed enrichments for defense responses and protein secretory pathways (Figures 4G and 4H), which are controlled by NPR1 and essential for SAR establishment¹⁰. Conversely, SA-repressed, Ser-557- and SnRK1-dependent genes were predominantly associated with growth (Figure 4I). We also observed that sesqui mutants exhibited compromised basal and SA-induced resistance (Figure 4J), similar to npr1^S557A. To determine whether the susceptibility of sesquiα2-1 resulted from reduced pSer-557, we transformed WT NPR1 or npr1^S557D into sesquiα2-1 npr1. Whereas WT NPR1 failed to restore resistance, npr1^S557D rescued the disease susceptibility phenotype (Figure 4K), supporting the conclusion that SnRK1 promotes immunity primarily through Ser-557 phosphorylation-dependent activation of NPR1.

Figure 4. SnRK1 mediates SA-triggered immunity through NPR1.

(A and B) Defense gene expression. Relative expression levels of PR1 (A) and WRKY38 (B) were measured by RT-qPCR in 3.5-week-old WT, npr1-2, sesquiα2-1, and sesquiα2–2 plants 24 hours after water or 1 mM SA treatment.

(C and D) Global transcriptome profiling by QuantSeq. Principal component analysis (PCA) (C) and hierarchical clustering (D) were performed using gene expression data from 3.5-week-old WT, npr1-2, sesquiα2-1, and sesquiα2–2 plants 24 hours after treatment with water or 1 mM SA.

(E and F) Venn diagrams comparing upregulated (E) and downregulated (F) differentially expressed genes across genotypes. Groups of interest are labeled with bold, white numbers.

(G and H) Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses of WT-specific upregulated genes. GO biological process (BP) (G) and KEGG pathway (H) enrichment analyses were performed on the 1,027 genes exclusively upregulated in WT (E).

(I) GO BP enrichment analysis of the 901 WT-specific downregulated genes (F).

(J) Basal and SA-induced resistance in WT, sesquiα2-1, sesquiα2–2, and npr1-2 plants. 3.5-week-old plants were treated with mock or 1 mM SA 24 hours before inoculation with Psm (OD₆₀₀ = 0.001). n = 8.

(K) Genetic interaction between sesqui and npr1^S557D mutations. Left: Bacterial growth in WT, sesquiα2-1, npr1-7, npr1 sesquiα2-1 (npr1 α2–1) plants, and independent transformants of 35S:NPR1-eYFP and 35S:npr1^S557D-eYFP in the npr1 α2–1 background. Plants were inoculated with Psm (OD₆₀₀ = 0.0002). n = 8. Right: Relative protein levels were calculated by normalizing band intensities to β-Tubulin.

Individual columns were compared using one-way ANOVA with Tukey’s post hoc test (J and K), and the interactions were tested using two-way ANOVA (A, B, and J).

SA coordinates NPR1-independent metabolism and NPR1-dependent immunity through activation of SnRK1

To investigate whether SA-induced NPR1 Ser-557 phosphorylation requires SnRK1 activation in addition to NPR1 nuclear translocation, we examined lines expressing npr1^C82A-GFP and npr1^C216A-GFP with constitutive nuclear localization²⁴. We found that constitutive nuclear localization alone was insufficient to induce Ser-557 phosphorylation (Figure S3A), suggesting that this modification requires an additional SA-dependent mechanism.

Since SA treatment caused only a moderate increase in SnRK1α1 and SnRK1α2 transcripts (Figure S3B) and had little effect on protein levels (Figure S3C) or subcellular localization (Figure S3D), we hypothesized that SA induces pSer-557 by enhancing SnRK1 activity. Indeed, SA-induced genes in our QuantSeq data significantly overlapped with those induced by transient SnRK1α1 overexpression³⁹ (34.9%), with overlapping genes enriched in sugar starvation response and amino acid catabolism, suggesting that SA induces metabolic perturbations characteristic of SnRK1 activation (Figures S4A–S4C).

To assess SA-induced SnRK1 activity in vivo, we measured nuclear phosphorylation of a SnRK1α-activity reporter, NLS-ratACC (Acetyl-CoA Carboxylase 1 peptide)-GFP-HA containing a SnRK1/AMPK phosphorylation target sequence⁷⁰. We observed significant ACC phosphorylation in WT beginning 2 hours after SA treatment (Figures 5A and 5B), whereas this response was absent in snrk1α1 (Figures 5C and 5D). Consistent with SnRK1 activation, the stress-responsive SnRK1 marker genes DARK INDUCIBLE 2 (DIN2), DIN6, and SENESCENCE 5 (SEN5)^39,47 were also similarly induced by SA in mature plants (Figure 5E).

Figure 5. SA modulates SnRK1 activity by metabolic reprogramming.

(A-D) Dynamics of SA-induced SnRK1 activation. Seedlings containing the nuclear SnRK1 activity reporter 35S:NLS-ratACC-GFP-HA in WT (A and B) or the snrk1α1 (C and D) background were treated with mock, 0.1 mM SA, or 0.3 mM SA for the indicated durations. Nuclear SnRK1-dependent phosphorylation of the reporter was detected using α-pACC and normalized to total protein levels determined using α-GFP (A and C). The mock-treated sample for each time point was set to 1 (B and D). n = 4.

(E) Expression of SnRK1 target genes DIN2, DIN6, and SEN5 in response to SA measured by RT-qPCR. 3.5-week-old WT plants were treated with mock, 0.1 mM SA, or 1 mM SA 24 hours before collection.

(F) Dynamics of SA-induced expression of SnRK1- and NPR1-target genes DIN2 and PR1, respectively. 35S:NPR1-GFP seedlings were treated as in Figure 2B, and gene expression was measured by RT-qPCR.

(G) Transient SnRK1 overexpression facilitates Ser-557 phosphorylation. Protoplasts derived from leaves of 35S:NPR1-GFP and 35S:npr1^C82A-GFP were transiently transfected with mock, SnRK1α1-mRFP, or the kinase-dead mutant snrk1α1^K48M-mRFP. Proteins were analyzed using α-pS557, α-GFP, and α-mRFP. *, nonspecific band.

(H) Effect of darkness-induced SnRK1 activation on Ser-557 phosphorylation. 35S:NPR1-GFP, 35S:npr1^C82A-GFP, and 35S:npr1^C216A-GFP seedlings were kept under light or darkness for 6 hours. Proteins were analyzed using α-pS557 and α-GFP. *, nonspecific band.

(I and J) Effects of trehalose 6-phosphate (Tre6P), a SnRK1 inhibitor, on Ser-557 phosphorylation (I) and PR1 expression (J). 35S:NPR1-GFP seedlings grown in sucrose-free ½ MS medium were pretreated with mock or 1 mM Tre6P for 2 hours and subsequently treated with mock or 0.3 mM SA for 6 hours. Relative Ser-557 phosphorylation levels were determined using α-pS557 and α-GFP, whereas PR1 expression levels were determined by RT-qPCR. *, nonspecific band.

(K-N) Sugar measurements following SA treatment and biological SAR induction. Levels of Tre6P (K), glucose 6-phosphate (Glc6P; L), trehalose (M), and glucose (N) were measured in 3.5-week-old plants 2 hours and 8 hours after mock or 1 mM SA treatment, or in systemic tissues 24 hours after local SAR induction with Psm avrRpt2 (OD₆₀₀ = 0.01) or MgCl₂. n = 4.

(O) Expression of Class I and Class II TPS genes following SA treatment. RT-qPCR was performed in 3.5-week-old WT plants 8 hours after mock or 1 mM SA treatment to measure expression of TPS1 (Class I) as well as TPS5, TPS6, and TPS7 (Class II).

(P) Expression of all 10 TPP genes following SA treatment. The experiment was performed as in (O).

(Q) Expression of TPS1 and TPS5 genes in seedlings was measured after mock or 0.3 mM SA treatment for 2 hours and quantified as in (O).

Individual columns were compared using two-tailed Student’s t-test (B, D, E, F, and J-Q) and the interactions were tested using two-way ANOVA (J). Total protein loading for (A, C, H, and I) was estimated using Ponceau S staining. See also Figures S3 and S4.

We then investigated the timing of SnRK1 and NPR1 activation and found a strong temporal correlation among DIN2 induction, PR1 increase, and Ser-557 phosphorylation, all of which began 2 hours after SA treatment (Figures 2B, 2C, and 5F). We next tested whether SnRK1 overexpression, known to mimic its activation³⁹, promotes pSer-557. Transient overexpression of WT SnRK1α1, but not the kinase-dead (K48M) mutant, induced pSer-557 in npr1^C82A-GFP but not in NPR1-GFP (Figure 5G), indicating that both NPR1 nuclear localization and SnRK1 activation are necessary for pSer-557.

To confirm that SnRK1 activation promotes pSer-557, we activated SnRK1 with darkness, a well-established trigger of SnRK1 activity³⁹. Consistent with the SnRK1α1 overexpression results, darkness increased pSer-557 in npr1^C82A and npr1^C216A, but not WT NPR1 (Figure 5H). Conversely, applying the SnRK1 inhibitor trehalose-6-phosphate (Tre6P)⁷¹ inhibited SA-induced pSer-557 (Figure 5I) and dampened PR1 induction (Figure 5J).

We then investigated mechanisms of SA-mediated SnRK1 activation. First, we examined SnRK1 T-loop phosphorylation, which correlates with AMPK activity in mammals⁷², but not clearly with in planta SnRK1 activity^39,48. We found no consistent effect of SA on T-loop phosphorylation (Figure S4D), ruling this out as the activation mechanism. In contrast to AMPK, whose activity is allosterically activated by increased AMP levels⁷³, SnRK1 appears to be insensitive to AMP and ADP^74,75. Instead, sugar phosphate levels, especially Tre6P and glucose-6-phosphate (Glc6P), are inversely correlated with SnRK1 activity in vivo^43,44. We used liquid chromatography-mass spectrometry to measure glucose, trehalose, and their phosphates in response to both SA treatment and avrRpt2-induced SAR. Glucose and sugar phosphate levels decreased, whereas trehalose levels remained unchanged (Figures 5K–5N). Tre6P is synthesized by Trehalose-6-Phosphate Synthase (TPS) and converted to trehalose by Trehalose-6-Phosphate Phosphatase (TPP)⁷⁶. In Arabidopsis, 11 TPSs are grouped into Class I (TPS1–4) and Class II (TPS5–11), with TPS1 as the predominant Tre6P-producing enzyme⁷⁷ and TPS5–7 having crucial regulatory roles⁷⁸. We found that SA suppressed TPS transcription (Figure 5O), while generally enhancing TPP expression (Figure 5P). Notably, changes in key genes TPS1 and TPS5 were evident 2 hours after SA treatment (Figure 5Q). SA also lowered intracellular ATP levels (Figure S4E), suggesting a broader effect on cellular energy status. Altogether, these results indicate that SA induces SnRK1 activity through energy depletion³⁹. Finally, we used the ASP-SPARK reporter line to assess SnRK1 activity at different SA concentrations by monitoring the condensation of a fluorescently tagged peptide and a phospho-binding domain⁷⁹. We found that 20 μM SA was sufficient to activate SnRK1 (Figure S4F).

SA inhibits TOR-mediated phosphorylation of NPR1 at Ser-55/59 in a SnRK1-dependent manner

To investigate SnRK1-dependent transcriptional changes induced by SA, we performed GO and KEGG analyses on the 358 NPR1-independent but SnRK1-dependent SA-induced genes (Figure 4E). These genes were enriched in redox- and abiotic stress-related processes. Alternative respiration was the top GO term (Figure 6A), whereas amino acid metabolism was the most enriched KEGG pathway (Figure 6B). In contrast, SnRK1-dependent downregulated genes were related to growth and metabolism (Figures S5A and S5B). Together, our findings suggest that SA plays a dual role in orchestrating immune responses and modulating primary metabolism through SnRK1.

Figure 6. SA inhibits TOR activity partially through SnRK1.

(A and B) GO and KEGG enrichment analyses of SA-induced genes that are dependent on SnRK1 but independent of NPR1. GO BP enrichment (A) and KEGG pathway enrichment (B) analyses of the 358 transcripts in Figure 4E upregulated by SA in WT, npr1^S557A-GFP, and npr1-2, but not in sesquiα2-1 or sesquiα2–2.

(C) SUnSET assay of nascent protein synthesis in response to SA in WT, npr1-2, sesquiα2-1, sesquiα2–2, and npr1^S557A-GFP plants. Four 5-mm leaf disks from 3.5-week-old plants were floated in water or 1 mM SA for the indicated durations. Puromycin labeling was initiated 45 min before sample collection and detected using α-puromycin.

(D) Time course of TOR-mediated RPS6 phosphorylation following SA treatment. WT seedlings were treated with 0.3 mM SA for the indicated durations or 10 μM Torin 2 (T) for 2 hours. Proteins were analyzed using α-RPS6, α-pRPS6 (Ser-237), and α-pRPS6 (Ser-240) (left). Relative phosphorylation levels were calculated by normalizing to the corresponding RPS6 levels, with mock samples set to 1 (right). n = 3.

(E) Effects of varying concentrations of SA on RPS6 phosphorylation. WT seedlings were treated with mock; 0.3, 0.5, and 1 mM SA; or 10 μM Torin 2 (T) for 2 hours. Protein analysis and quantification were performed as in (D).

(F and G) Effects of SA (F) and SAR induction (G) on SnRK1 activity and TOR-mediated RPS6 phosphorylation. 3.5-week-old WT plants carrying 35S:NLS-ratACC-GFP-HA were treated with mock or 1 mM SA for 24 hours (F) or subjected to biological SAR induction as in Figure 3H (G) before analysis using α-pACC, α-GFP, α-pRPS6 (Ser-237), α-pRPS6 (Ser-240), and α-RPS6. Relative phosphorylation levels were calculated after normalizing to corresponding protein levels, with mock samples set to 1.

(H) Role of SnRK1 in inhibiting TOR-mediated RPS6 phosphorylation during SA induction. WT and sesquiα2-1 seedlings were treated with 0.3 mM SA for the indicated durations or 10 μM Torin 2 (T) for 2 hours. Western blotting (left) and RPS6 phosphorylation analysis (right) were performed as described in (D). n = 3.

Individual columns were compared using one-way ANOVA with Tukey’s post hoc test (D and E) or two-tailed Student’s t-test (H). Total protein loading for (C-E and H) was estimated using Ponceau S staining. See also Figure S5.

Enrichment of amino acid metabolic pathways among SA-induced, SnRK1-dependent transcripts suggests potential changes in translation. To assess such changes, we performed the surface sensing of translation (SUnSET) assay^80,81, which measures incorporation of puromycin, an aminoacyl-tRNA analog, into nascent peptides by immunoblotting. We observed a reduction in global translation within 2 hours of SA treatment in a SnRK1-dependent but NPR1-independent manner (Figure 6C).

These findings prompted us to investigate TOR, another central nutrient-sensing kinase that regulates translation and acts antagonistically to SnRK1³⁸. Reduced glucose levels, which were also observed after SA treatment (Figure 5N), have been associated with TOR inhibition³⁷. Phosphorylation of P70-S6 kinase (S6K) and its downstream target ribosomal protein S6 (RPS6) are well-established readouts of TOR activity^37,82,83. SA rapidly reduced S6K (Figures S5C and S5D) and RPS6 (Figures 6D and 6E) phosphorylation in seedlings, resembling the effect of the TOR inhibitor Torin 2. This reduction in TOR activity was accompanied by increased SnRK1 activity, as measured using the NLS-ratACC-GFP-HA reporter, both in SA-treated mature plants (Figure 6F) and in systemic tissues following SAR induction (Figure 6G).

Since SA inhibits translation in a SnRK1-dependent manner, we examined whether SA-induced TOR inhibition also requires SnRK1. We found that the SA-induced decrease in RPS6 phosphorylation was largely abolished in sesquiα2-1 (Figure 6H), suggesting that SA-mediated TOR inhibition is SnRK1-dependent.

We next tested whether SA-mediated TOR inhibition affects NPR1 PTMs. Since NPR1 is phosphorylated at Ser-55/59 in the cytosol in its quiescent state and SA rapidly decreases this phosphorylation, we hypothesized that TOR may regulate pSer-55/59. Supporting this hypothesis, the conserved Ser-55/59 region matches the TOR/mTOR substrate consensus motif^84,85, with a proline at +1 relative to Ser-59 (Figure S5E). Using α-pS55/59, we observed TOR-dependent Ser-55/59 phosphorylation in vitro, which was suppressed by Torin 2 (Figure S5F). To test whether pSer-55/59 is TOR-sensitive in vivo, we treated NPR1-GFP transgenic seedlings and NPR1-GFP-transfected protoplasts with Torin 2 or the TOR-specific inhibitor AZD-8055^37,86. Both inhibitors abolished pSer-55/59 (Figures 7A and S5G), with the time-course of AZD-8055 treatment showing pSer-55/59 reduction within 1 hour (Figure 7B), consistent with the kinetics of SA-induced dephosphorylation (Figures 2B and 2C). The dependency of pSer-55/59 on TOR was further validated using an estradiol-inducible TOR-silencing line (tor-es)^35,36 (Figure 7C), while co-immunoprecipitation revealed an interaction between endogenous NPR1 and TOR (Figure 7D). To connect TOR activity and its interaction with NPR1 to pSer-55/59, we immunoprecipitated NPR1-GFP at intervals over 2 hours after SA treatment and simultaneously measured RPS6 phosphorylation as a readout of TOR activity and pSer-55/59 in the lysates. Within 30 minutes, SA reduced TOR activity, abolished TOR-NPR1 interaction, possibly due to NPR1 nuclear translocation, and decreased pSer-55/59 (Figure 7E).

Figure 7. TOR mediates phosphorylation of NPR1 at Ser-55/59 to inhibit its activity under optimal conditions.

(A and B) Effect of TOR inhibition on Ser-55/59 phosphorylation. (A) 35S:NPR1-GFP seedlings were treated with mock, 0.3 mM SA (6 hours), the TOR inhibitor Torin 2 (10 μM; 1 or 3 days), or AZD-8055 (2 μM; 1 or 3 days). (B) 35S:NPR1-GFP seedlings were treated with 2 μM AZD-8055 for the indicated durations. Phosphorylated and total NPR1 were detected using α-pS55/59 and α-GFP.

(C) Effect of estradiol-induced silencing of TOR on Ser-55/59 phosphorylation. NPR1-GFP was transiently expressed in protoplasts derived from leaves of WT or the estradiol-inducible TOR-silencing line (tor-es) treated overnight with either DMSO (mock) or 5 μM β-estradiol. Western blotting was performed as in (B).

(D) Co-immunoprecipitation of endogenous TOR with NPR1-GFP in transgenic Arabidopsis lines expressing 35S:NPR1-GFP or 35S:GFP.

(E) Effects of SA on TOR activity and TOR-NPR1 interaction. 35S:NPR1-GFP seedlings were treated with mock or 0.3 mM SA for the indicated durations. The lysate was split into two portions: one was subjected to GFP immunoprecipitation followed by blotting with α-TOR and α-GFP; and the other was directly analyzed using α-TOR, α-pRPS6 (Ser-237), α-pRPS6 (Ser-240), α-RPS6, α-pS55/59, and α-GFP. Total protein loading was estimated using Ponceau S staining. *, nonspecific band.

(F) Basal and SA-induced resistance in tor mutants. 3.5-week-old WT, tor-es #1, and tor-es #2 were pretreated with mock or 10 μM β-estradiol 2 days before mock or 1 mM SA treatment. Psm (OD₆₀₀ = 0.001) was inoculated one day after mock or SA treatment. n = 8.

(G) Effect of TOR inhibition on PR1 gene expression. RT-qPCR was performed on 35S:NPR1-GFP, 35S:npr1^S55/59A-GFP, and 35S:npr1^S55/59D-GFP seedlings pre-treated with mock or 2 μM AZD-8055 and subsequently treated with mock or 0.3 mM SA for 8 hours.

(H) Effect of TOR inhibition on basal resistance. 3.5-week-old 35S:NPR1-GFP and 35S:npr1^S55/59D-GFP plants were pre-treated with mock or 2 μM AZD-8055 for 24 hours before inoculation with Psm (OD₆₀₀ = 0.0002). n = 8.

(I) Proposed model of SnRK1- and TOR-mediated coordination between metabolic status and SA-induced immunity through differential phosphorylation of NPR1. Under optimal growth conditions (left), TOR is active (bright green) and promotes phosphorylation of NPR1 at Ser-55/59, thereby maintaining NPR1 in the cytoplasm through oligomerization, whereas SnRK1 remains inactive (dim red) under nutrient-rich conditions. Upon immune activation (right), SA, through an unknown mechanism, reduces glucose, Glc6P, and Tre6P levels, leading to activation of SnRK1 (bright red) and inhibition of TOR (dim green), accompanied by reduced Ser-55/59 phosphorylation. SA-induced redox changes also facilitate NPR1 translocation into the nucleus where SnRK1 phosphorylates NPR1 at Ser-557. This phosphorylation partially activates NPR1, promotes nSINC formation and interactions with transcriptional regulators, and primes subsequent phosphorylation at Ser-11/15, which is required for both full-scale defense gene induction and NPR1 turnover via sequential ubiquitination (Ub).

Individual columns were compared using one-way ANOVA with Tukey’s post hoc test (F and H) or a two-tailed Student’s t-test (G), and the interactions were tested using two-way ANOVA (F-H).

See also Figure S5.

To examine the functional impact of TOR on plant defense, we treated two tor-es lines with estradiol, SA, or a combination of SA and estradiol. We found that TOR silencing enhanced basal immunity compared to WT (Figure 7F). Following SA treatment, however, TOR-silenced plants did not exhibit greater resistance than WT, likely because TOR knockdown mimics SA-induced NPR1 dephosphorylation at Ser-55/59.

To directly link TOR-mediated immunity to pSer-55/59, we treated NPR1-GFP, npr1^S55/59A, and npr1^S55/59D with mock, SA, or SA and AZD-8055. TOR inhibition by AZD-8055 further enhanced SA-induced PR1 expression in WT, but not in either Ser-55/59 phospho-mutant, indicating that TOR regulates PR1 expression through pSer-55/59 (Figure 7G). Consistently, AZD-8055 also enhanced resistance to Psm infection in WT, but not in npr1^S55/59D (Figure 7H).

DISCUSSION

Although several NPR1 PTMs have been identified, with effects ranging from the control of NPR1 subcellular localization^23,29 to the regulation of its interactions with TFs²³, the modifications and enzymes controlling NPR1’s transcriptional coactivator function are not fully understood. Here, we identify Ser-557 phosphorylation as a necessary step in activating NPR1 immune function through two primary mechanisms. pSer-557 promotes nSINC formation, which enhances NPR1’s interactions with transcriptional partners, including TGA3 and the mediator complex component CDK8, leading to partial transcriptional activation. Additionally, pSer-557 is required for SA-induced Ser-11/15 phosphorylation, thereby enhancing NPR1 interaction with CUL3A and facilitating its ubiquitination, leading to its full activation and subsequent turnover. Together, our data indicate that Ser-557 phosphorylation is necessary and likely an initial step in triggering transcriptional reprogramming and SAR (Figure 7I).

The identification of the two master nutrient regulators SnRK1 and TOR as pivotal kinases regulating NPR1 activity fills a gap in the signaling pathway between SA and NPR1 and establishes a direct connection between cellular metabolism and SA-mediated defense essential for coordinating plant growth and survival. Under optimal conditions, TOR-dependent phosphorylation of NPR1 at Ser-55/59 and oligomerization of NPR1 through intermolecular disulfide bonds represent two complementary mechanisms that keep NPR1 quiescent. Within an hour after SA treatment, SnRK1-dependent TOR inhibition occurs, and the interaction between NPR1 and TOR is disrupted, leading to rapid Ser-55/59 dephosphorylation. However, whether a phosphatase is actively involved remains unknown. Upon translocation into the nucleus, NPR1 is phosphorylated at Ser-557 by SnRK1, partially activating the protein and priming it for subsequent PTMs. These findings support a model in which NPR1 activation is a multilayered, stepwise process (Figure 7I).

As part of their role in coordinating growth and stress responses, SnRK1 and TOR have been shown to antagonize each other under biotic and abiotic stresses across species⁴⁸. Interestingly, acetylsalicylic acid (aspirin), a derivative of salicylic acid, activates AMPK in cancer cell lines while inhibiting mTOR signaling in a partially AMPK-dependent manner⁸⁷. AMPK-mTOR antagonism also occurs in mammalian immune responses^88,89. Moreover, differential phosphorylation of shared substrates by AMPK and mTOR is well documented. Under nutrient-rich conditions, mTOR prevents UNC-51-LIKE KINASE 1 from initiating autophagy by phosphorylating Ser-757, while AMPK activation during glucose starvation promotes autophagy through Ser-317/777 phosphorylation⁹⁰. This antagonistic regulation of autophagy is also present in plants, mediated by the direct phosphorylation of SnRK1 substrates^91,92. Notably, autophagy can also be induced by SA and is required for restricting cell death during infection through NPR1⁵⁷.

SnRK1/AMPK responds to a wide range of stresses, suggesting that diverse signals converge on this master kinase, with energy deficit serving as a unifying trigger for its activation³⁹. In mammals, AMPK is activated by an increased AMP-to-ATP ratio following energy depletion⁹³. However, plant SnRK1 orthologs are insensitive to AMP and ADP⁷⁵. Instead, Tre6P and Glc6P inversely correlate with in planta SnRK1 activity^44,71,76, which we also observed after SAR induction. Consistent with reduced Tre6P, SA treatment antagonistically regulates expression of the TPS/TPP module, disrupting trehalose-Tre6P homeostasis. This is consistent with a recent report that SA decreases intracellular sugar levels⁹⁴ and supports another study showing that SA lowers Tre6P levels during heat stress in maize by altering TPS/TPP activities⁹⁵. SA also reduces ATP levels, likely by inducing alternative respiration. Together, these observations suggest that SA-induced sugar/energy deprivation is the likely trigger for SnRK1 activation.

As a central regulator of cellular stress responses, the SnRK1/AMPK/SNF1 family coordinates adaptation to adverse conditions through enzymatic regulation and transcriptional reprogramming^45,96. NPR1 phosphorylation by SA-activated SnRK1 is an integral step in defense. Moreover, SA-induced, SnRK1-dependent genes are enriched in sugar, amino acid, and fatty acid catabolism, emphasizing the importance of alternative energy sources and metabolites in response to biotic stress. Interestingly, a recent study showed that sugar depletion prior to PTI induction represses SA signaling through SnRK1, likely because prolonged energy deprivation compromises the metabolic state and cellular functions required for effective defense activation⁹⁷. In contrast, without prior sugar starvation, SA induces a transient metabolic reprogramming that activates SnRK1, promoting NPR1 phosphorylation and robust immune responses.

We further show that SA inhibits the other master metabolic regulator, TOR, in a SnRK1-dependent manner, leading to reduced cap-dependent translation as part of the growth-to-defense transition. Our findings align with a recent study showing that SA represses Arabidopsis growth under nutrient-rich conditions by interfering with TOR signaling⁹⁸, and provide mechanistic insight into the previously reported SA-triggered translation inhibition^99–101.

In summary, this study reveals three interconnected layers through which SA coordinates plant physiology (Figure 7I). First, upon pathogen challenge, SA depletes sugar phosphates and activates SnRK1, which inhibits TOR signaling to trigger the general growth-to-defense transition. Second, the rebalancing of activities between these two master kinases shifts NPR1 phosphorylation from Ser-55/59 to Ser-557. Third, this phosphorylation switch activates NPR1 and primes it for subsequent PTMs, thereby coordinating the plant immune response with the cellular metabolic state.

Limitations of the study

It remains unclear how SA lowers intracellular sugar and sugar phosphate levels. While multiple proteins have been reported to bind SA, including NPR1 and its paralogs NPR3/4^102–105, the primary target through which SA regulates metabolism is unknown. While we found that SA inhibits TOR partly through SnRK1 and significantly lowers glucose levels, it remains elusive how these signals are integrated by TOR, and the full metabolic impact of SA in balancing SnRK1 and TOR has yet to be elucidated. Finally, structural studies of NPR1 in complex with its interacting partners will be required to fully explain how pSer-557 activates NPR1 and induces SAR.

STAR★METHODS

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Plants

All Arabidopsis thaliana plants used in this study were in the Col-0 ecotype background. Throughout the text and figures, “WT” refers to Col-0 plants, and “mature plants” refers to 3.5-week-old plants and “seedlings” refers to 7-day-old plants. Arabidopsis mutants npr1-2⁹ and transgenic lines expressing NPR1-GFP²⁴, npr1^sim3-GFP²³, npr1^C82A-GFP²⁴, npr1^C216A-GFP²⁴, npr1S11/15A-GFP26, npr1S11/15D-GFP26, npr1S55/59A-GFP23, and npr1S55/59D-GFP23 in the npr1-2 background were described previously. The sesquiα2-1 and sesquiα2–2 lines⁶⁸ were provided by Dr. Elena Baena-Gonzalez, SnRK1α1-GFP¹⁰⁶ by Dr. Csaba Koncz, and ASP-SPARK⁷⁹ by Drs. Steffen Vanneste, Jelle Van Leene, and Wouter Smagghe. The estradiol-inducible tor silencing tor-es lines (CS6982 and CS69830)³⁵ were obtained from the Arabidopsis Biological Resource Center. Transgenic Arabidopsis lines expressing npr1^S557A-GFP or npr1^S557D-GFP in the npr1-2 background were generated by floral dipping¹⁰⁷, and plants carrying single insertions and homozygous for the transgenes were used. The NPR1-GFP and mutant variant lines in the sesquiα2-1 npr1 background were generated by first introducing a CRISPR/Cas9-mediated npr1 knock-out into the sesquiα2-1 background, followed by Agrobacterium-mediated floral dip transformation¹⁰⁷ with 35S:NPR1-GFP or its mutant variants. The NLS-ratACC-GFP-HA line in the background of snrk1α1 was generated by CRISPR/Cas9-based knock-out of snrk1α1 in the background of NLS-ratACC-GFP-HA #4.2¹⁰⁸. The CRISPR/Cas9 constructs were designed following the published procedure¹⁰⁹ and the target sites were chosen by the CHOPCHOP v3 online tool for snrk1α1¹¹⁰ or based on a previous publication for npr1¹⁰⁴.

A. thaliana and N. benthamiana seeds were stratified at 4 °C for two days before being sown in soil (ProMix PGX) for plant growth under 12 h/12 h light/dark cycles with a light intensity of 80–100 μmol m⁻² s⁻¹, temperature of 22 °C and relative humidity of 65%. For the selection of sesqui individuals from the segregating sesqui progeny, seeds were selected against glufosinate-ammonium (BASTA; 10 mg/L). Resistant individuals were then transferred to newly prepared soil or non-BASTA-containing medium for experiments. For the selection of npr1^S557A-GFP or npr1^S557D-GFP lines, the plants were grown on 1/2 Murashige and Skoog (MS) media with 1% sucrose and 50 mg/L kanamycin.

Bacterial strains

Psm ES4326 and Psm ES4326/avrRpt2 were grown for two days on King’s Broth (KB) plates with appropriate antibiotics at 30 °C. Agrobacterium tumefaciens strain GV3101 was grown in Luria-Bertani (LB) medium at 30 °C.

METHOD DETAILS

Plasmids

Plasmid information is presented in the Key Resources Table. The full-length cDNAs of NPR1, all NPR1 mutants, SnRK1α, SnRK3.8, SnRK3.22, CDK8, TGA3, and CUL3A constructs were amplified via Phusion Flash High-Fidelity DNA polymerase (Thermo Scientific) and cloned into the Gateway donor vector pDONR207 by BP recombination to obtain the entry clones. For plant expression, the clones were transferred to the destination vectors pBI221¹¹¹ or pK7FWG2²³ for C-terminal tagging via LR recombination. For expression in bacteria, NPR1 and its mutant clones were transferred into the gateway destination vector pDEST15 via LR recombination for N-terminal GST-tagging. All constructs were verified by Sanger or nanopore sequencing before use.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse monoclonal anti-GFP (JL8)	Clontech	Cat. #632381; RRID: AB_2313808
Mouse monoclonal anti-RFP (6G6)	Chromotek	Cat. #6g6; RRID: AB_2631395
Mouse monoclonal anti-β-Tubulin (BT7R)	Invitrogen	Cat. #MA5-16308; RRID: AB_2537819
Rabbit polyclonal anti-p(Ser/Thr) Phe	Cell Signaling Technology	Cat. #9631; RRID: AB_330308
Rabbit monoclonal anti-pAMPKα (Thr172) (40H9)	Cell Signaling Technology	Cat. #2535; RRID: AB_331250
Mouse monoclonal anti-Puromycin (12D10)	Sigma-Aldrich	Cat. #MABE343; RRID: AB_2566826
Mouse monoclonal anti-FLAG	Sigma-Aldrich	Cat. #F1804; RRID: AB_262044
Rabbit polyclonal anti-RPS6	Enganti et al. (2017)112	N/A
Rabbit polyclonal anti-pRPS6 (Ser-237)	Enganti et al. (2017)112	N/A
Rabbit polyclonal anti-pRPS6 (Ser-240)	Enganti et al. (2017)112	N/A
Rabbit polyclonal anti-Ribosomal S6 kinase 1/2	Agrisera	Cat. #AS121855; RRID: AB_2832919
Rabbit polyclonal anti-pS6K1 (Thr-449)	Abcam	Cat. #ab207399; RRID: AB_2935691
Rabbit polyclonal anti-pAcetyl-CoA Carboxylase (Ser-79)	Cell Signaling Technology	Cat. #3661; RRID: AB_330337
Rabbit polyclonal anti-AKIN10 (SnRK1α1)	Agrisera	Cat. #AS10919, RRID: AB_10754154
Rabbit polyclonal anti-AKIN11 (SnRK1α2)	Agrisera	Cat. #AS10920; RRID: AB_10748837
Rabbit polyclonal anti-UGPase	Agrisera	Cat. #AS05086; RRID: AB_1031827
Rabbit Polyclonal anti-Histone H3	Invitrogen	Cat. #PA5-31954; RRID: AB_2549427
Rabbit polyclonal anti-AtNPR1	Mou et al. (2003)24	N/A
Rabbit polyclonal anti-pAtNPR1 (Ser-557)	This study	N/A
Rabbit polyclonal anti-pAtNPR1 (Ser-11/15)	Spoel et al. (2009)26	N/A
Rabbit polyclonal anti-pAtNPR1 (Ser-55/59)	Saleh et al. (2015)23	N/A
Rabbit polyclonal anti-TOR (N)	Abiocode	Cat. #R2854-1
Anti-GST Antibody, HRP Conjugated	GE Healthcare	Cat. #RPN1236; RRID: AB_771429
Anti-HA Antibody, HRP Conjugated	Cell Signaling Technology	Cat. #2999; RRID: AB_1264166
Anti-Ubiquitin coupled to magnetic agarose	Chromotek	Cat. #utma; RRID: AB_2943578
Anti-GFP coupled to magnetic agarose	Chromotek	Cat. #gtma; RRID: AB_2631358
Anti-FLAG M2 coupled to magnetic agarose	Sigma-Aldrich	Cat. #M8823; RRID: AB_2637089
Bacterial and virus strains
Agrobacterium tumefaciens strain GV3101	N/A	N/A
Pseudomonas syringae pv. maculicola ES4326	Dong et al. (1991)122	N/A
Pseudomonas syringae pv. maculicola ES4326/avrRpt2	Dong et al. (1991)122	N/A
Escherichia coli strain BL21 (DE3)	New England Biolabs	Cat. #C2527H
Chemicals, peptides, and recombinant proteins
3× FLAG peptide	Sigma-Aldrich	Cat. #F4799
MG132	Sigma-Aldrich	Cat. #M8699; CAS #1211877-36-9
Sodium salicylate	Sigma-Aldrich	Cat. #S3007; CAS #54-21-7
β-Estradiol	Sigma-Aldrich	Cat. #E2758; CAS #50-28-2
TRIzol Reagent	Invitrogen	Cat. #E15596026
Protease Inhibitor Cocktail	Roche	Cat. #11836170001
PhosSTOP	Roche	Cat. #4906845001
Trehalose 6-phosphate dipotassium salt	Santa Cruz	Cat. #sc-216004; CAS #136632-28-5
Lambda Protein Phosphatase	New England Biolabs	Cat. #P0753
DynaGreen Protein A/G Magnetic Beads	Invitrogen	Cat. #80104G
Pierce Glutathione Magnetic Agarose Beads	ThermoFisher	Cat. #78601
Critical commercial assays
FastStart Universal SYBR Green Master Kit	Roche	Cat. #04913850001
ATP Bioluminescent Assay Kit	Sigma-Aldrich	Cat. #FLAA
Next Generation Cell Free Protein Expression Kit (Wheat Germ)	Sigma-Aldrich	Cat. #CFPS700
QuantSeq 3’ mRNA-Seq V2 Library Prep Kit FWD with UDI 12 nt Sets A1-A4	Lexogen	Cat. #193.384
Experimental models: Organisms/strains
Arabidopsis: Col-0	N/A	N/A
Arabidopsis: 35S:GFP	This paper	N/A
Arabidopsis: npr1-2	Cao et al. (1997)9	N/A
Arabidopsis: npr1-7	Ding et al. (2018)104	N/A
Arabidopsis: 35S:NPR1-GFP/npr1-2	Mou et al. (2003)24	N/A
Arabidopsis: 35S:NPR1-eYFP/npr1-7	This paper	N/A
Arabidopsis: 35S:NPR1-eYFP/sesquiα2-1 npr1	This paper	N/A
Arabidopsis: 35S:npr1S557D-eYFP/sesquiα2-1 npr1	This paper	N/A
Arabidopsis: 35S:npr1S557A-GFP/npr1-2	This paper	N/A
Arabidopsis: 35S:npr1S557D-GFP/npr1-2	This paper	N/A
Arabidopsis: 35S:npr1S11/15A-GFP/npr1-2	Spoel et al. (2009)26	N/A
Arabidopsis: 35S:npr1S11/15D-GFP/npr1-2	Spoel et al. (2009)26	N/A
Arabidopsis: 35S:npr1S55/59A-GFP/npr1-2	Saleh et al. (2015)23	N/A
Arabidopsis: 35S:npr1S55/59D-GFP/npr1-2	Saleh et al. (2015)23	N/A
Arabidopsis: 35S:npr1sim3-GFP/npr1-2	Saleh et al. (2015)23	N/A
Arabidopsis: 35S:npr1C82A-GFP/npr1-2	Mou et al. (2003)24	N/A
Arabidopsis: 35S:npr1C216A-GFP/npr1-2	Mou et al. (2003)24	N/A
Arabidopsis: NPR1pro:NPR1-eYFP/npr1-2	This paper	N/A
Arabidopsis: NPR1pro:npr1S557A-eYFP/npr1-2	This paper	N/A
Arabidopsis: NPR1pro:npr1S557D-eYFP/npr1-2	This paper	N/A
Arabidopsis: APA1pro:ASP-SPARK	Safi et al. (2023)79	N/A
Arabidopsis: 35S:SnRK1α1-GFP	Bitrián et al. (2011)106	N/A
Arabidopsis: 35S:NLS-ratACC-GFP-HA	Muralidhara et al. (2021)70	N/A
Arabidopsis: 35S:NLS-ratACC-GFP-HA/snrk1α1	This paper	N/A
Arabidopsis: tor-es #1	Xiong and Sheen. (2012)35	N/A
Arabidopsis: tor-es #2	Xiong and Sheen. (2012)35	N/A
Arabidopsis: sesquiα2-1	Belda-Palazon et al. (2020)68	N/A
Arabidopsis: sesquiα2-2	Belda-Palazon et al. (2020)68	N/A
Arabidopsis: sesquiα2-1 npr1	This paper	N/A
Arabidopsis: sid2-2	Dewdney et al. (2001)123	N/A
Nicotiana benthamiana	N/A	N/A
Oligonucleotides
Primers see Table S4
Recombinant DNA
pEarleyGate202-SnRK1α2-FLAG	This paper	N/A
pEarleyGate202-SnRK3.8-FLAG	This paper	N/A
pEarleyGate202-SnRK3.22-FLAG	This paper	N/A
pK7FWG2-35S:NPR1-GFP	Saleh et al. (2015)23	N/A
pK7FWG2-35S:npr1S557A-GFP	This paper	N/A
pK7FWG2-35S:npr1S557D-GFP	This paper	N/A
pK7FWG2-35S:npr1S55/59A-GFP	Saleh et al. (2015)23	N/A
pK7FWG2-35S:npr1S55/59D-GFP	Saleh et al. (2015)23	N/A
pK7FWG2-35S:npr1S11/15A-GFP	Saleh et al. (2015)23	N/A
pK7FWG2-35S:npr1S557A_S11/15D-GFP	This paper	N/A
pK7FWG2-35S:npr1S11/15D-GFP	Saleh et al. (2015)23	N/A
pK7FWG2-35S:npr1sim3-GFP	Saleh et al. (2015)23	N/A
pGWB515-3HA-TGA3	This paper	N/A
pEarleyGate100-35S:NPR1-mCherry	This paper	N/A
pEarleyGate100-35S:HA-mCherry	Gu et al. (2016)124	N/A
pEarleyGate201-35S:CUL3A-HA	This paper	N/A
pEarleyGate100-35S:CDK8-mCherry	This paper	N/A
pBI221-35S:SnRK1α1-GFP	This paper	N/A
pBI221-35S:SnRK1α1-mRFP	This paper	N/A
pBI221-35S:snrk1α1K48M-mRFP	This paper	N/A
pBI221-35S:SnRK1α2-GFP	This paper	N/A
pBI221-35S:NPR1-GFP	This paper	N/A
pDEST15-GST-AtNPR1	Wang et al. (2020)103	N/A
pDEST15-GST-SbNPR1	This paper	N/A
pDEST15-GST-OsNH1	This paper	N/A
pDEST15-GST-SlNPR1	This paper	N/A
pCambia1300-SnRK1α1-YFP	This paper	N/A
pCambia1300-SnRK1α2-YFP	This paper	N/A
Codon-Optimized DNA see Table S3
Software and algorithms
ImageJ (FIJI)	Schindelin et al. (2012)114	RRID: SCR_002285
GraphPad Prism 9	GraphPad	RRID: SCR_002798
Scaffold 5	Proteome Software	RRID: SCR_014321
R	N/A	RRID: SCR_001905
Photoshop 2023	Adobe	RRID: SCR_014199
Illustrator 2023	Adobe	RRID: SCR_010279
CHOPCHOP v3	Labun et al. (2019)110	RRID: SCR_015723
ShinyGO 0.82	Ge et al. (2020)125	RRID: SCR_019213
Venn Diagrams	N/A	http://bioinformatics.psb.ugent.be/webtools/Venn/

Chemical treatments

To inhibit TOR activity, 10 μM Torin 2 or 2 μM AZD-8055 was dissolved in DMSO before application. To silence TOR in the tor-es plants, 10 μM β-estradiol was applied by spraying. To silence TOR in tor-es-derived protoplasts, 5 μM β-estradiol was added to the W5 medium during incubation. For SA treatment, the indicated concentration of SA was either sprayed onto mature plants or added to liquid media for seedlings and protoplasts.

LC-MS/MS and data analysis

Mature Arabidopsis plants expressing NPR1-GFP were sprayed with 0.5 mM SA and returned to the growth chamber for 24 h. Total protein was extracted from 5 g of leaf material using a buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, 0.2% Nonidet P-40, PhosSTOP phosphatase inhibitor cocktail (Roche), EDTA-free protease inhibitor cocktail (Roche), and 50 μM MG115. The extracts were clarified by centrifugation at 16,595 rcf for 15 min, and the soluble protein fraction was collected into a new tube. NPR1-GFP was purified from the extracts using GFP-Trap agarose beads (Chromotek) following the manufacturer’s instructions. Immunopurified NPR1-GFP was digested directly on GFP-Trap beads using either trypsin or Glu-C and the peptides were analyzed by UPLC-MS/MS at the Duke Proteomics Core Facility. The data in Figure 3B were acquired using data-independent acquisition (DIA) on an Orbitrap Astral and analyzed using Biognosys Spectronaut v19. The figure displays MS2 fragment ion extracted ion chromatograms (XICs) over retention time for the indicated peptides. Retention times were aligned using Spectronaut’s iRT-based prediction.

Pathogen infection assay

Psm ES4326 was grown for two days at 30 °C on KB plates supplemented with 100 μg/mL streptomycin before being resuspended in 10 mM MgCl₂ and diluted to the indicated optical density. To examine SA-induced resistance, 1 mM SA or H₂O was sprayed onto the plants one day prior to bacterial infiltration. The Psm ES4326 inoculum was pressure-infiltrated into two mature leaves per plant using a needleless syringe, with 12 plants used per genotype/treatment. At 3 dpi, sixteen infected leaf disks (two leaf disks pooled per biological replicate; n = 8) were collected from the inoculated leaves, ground in 500 μL 10 mM MgCl₂ using a Geno/Grinder (Cole-Parmer), serially diluted, and plated. Colonies were counted 2 days later.

Induction of SAR by avirulent pathogen

For SAR induction, Psm ES4326 carrying the effector gene avrRpt2 (Psm ES4326/avrRpt2) at an OD₆₀₀ of 0.01, or 10 mM MgCl₂ (mock), was infiltrated into two leaves per plant. Systemic leaves were collected at 1 day (sugar analysis) or 2 days (other assays) post inoculation. The ground tissue was processed for western blotting, mass-spectrometry, or sugar analysis as described.

Western blotting and antibodies

Eight 5-mm leaf disks from 3.5-week-old plants or ten 7-day-old seedlings were collected, flash-frozen in liquid N₂, and pulverized using a Geno/Grinder and homogenized on ice in 60 μL extraction buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 0.2% Triton X-100, 0.2% Nonidet P-40, 50 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)]. Homogenates were centrifuged at 16,595 rcf for 15 min at 4 °C and 60 μL of the supernatant was transferred to a new centrifuge tube and mixed with an equal volume of 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol. Protein samples were heated at 70 °C for 15 min, separated in a 4–12% Bis-Tris SDS-PAGE gel (Invitrogen), and transferred to nitrocellulose membranes using iBlot 3 (Invitrogen).

For detection of NPR1 phosphorylation, blots were probed with phospho-specific α-pS11/15²⁶, α-pS55/59²³, and α-pS557 antibodies, all of which were generated by ProteinTech Group. α-pS557 was generated using the antigenic peptide (Cys-ETLKKAF[pS]EDNLE) and the corresponding non-phosphorylated peptide (Cys-ETLKKAFSEDNLE). The specificity of the α-pS557 antibody was validated using mock-treated sample, the npr1-2 and npr1^S557A mutants, as well as immunoprecipitation prior to western blotting as summarized in Table S1. For normalization, total NPR1-GFP protein was measured using α-GFP (Takara Bio). Equal loading was confirmed by the Ponceau S staining.

For detection of the phosphorylated endogenous NPR1, 1.5 g leaf tissue was collected from 3.5-week-old plants, homogenized with 3 mL extraction buffer, and the supernatant was incubated overnight with 20 μL α-NPR1 antibody²⁴ at 4 °C with gentle rocking. Then, 30 μL of Dynabeads Protein A/G (Invitrogen) was added and the samples were incubated for 1 hour at 4 °C. The beads were then washed 3 times with 1 mL of extraction buffer. Protein samples were heated at 70 °C for 10 min before being subjected to western blotting. Mouse anti-rabbit IgG (Light-chain specific) (D4W3E) monoclonal antibody (Cell Signaling Technology; Figure 3J) or VeriBlot for IP Detection Reagent (HRP) (Abcam; Figures 1B, 3G, and 3H) was used in place of the conventional HRP-conjugated α-rabbit secondary antibody.

For the detection of SnRK1α1 phosphorylation in the 35S:SnRK1α1-GFP line, an α-pAMPKα (Thr-172) (Cell Signaling Technology) antibody was used. For the detection of phosphorylation of the nuclear ratACC-GFP-HA reporter protein, an α-pACC (Ser-79) (Cell Signaling Technology) antibody was used.

For the detection of S6K and RPS6 phosphorylation, the blots were probed with α-S6K1/2 (Agrisera), α-pS6K (Thr-449) (Abcam), α-RPS6¹¹², α-pRPS6 (Ser-237)¹¹², and α-pRPS6 (Ser-240)¹¹² antibodies.

Quantitative real-time PCR

Total RNA was isolated from seedlings or mature leaf tissues using the TRIzol reagent (ThermoFisher) according to the manufacturer’s instructions. DNA was removed by DNase I treatment (ThermoFisher) before cDNA synthesis using SuperScript IV cDNA Synthesis System (ThermoFisher). Quantitative PCR was performed with diluted cDNA (1:6 dilution with water) and the FastStart Universal SYBR Green Master Kit (Roche) on a Mastercycler ep real-plex (Eppendorf).

Nuclear-cytoplasmic fractionation

Approximately 0.1 g of ground Arabidopsis seedling tissues were dissolved in nuclei isolation buffer [20 mM Tris-Cl (pH 7.5), 25% glycerol, 20 mM KCl, 2.5 mM MgCl2, 2 mM EDTA, 250 mM sucrose, 50 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)] on ice and passed through a 70-μm cell strainer (Corning). The collected extracts were centrifuged at 1,200 g for 10 min at 4 °C. Supernatants were collected as the cytosolic fraction and pellets were suspended in nuclei washing buffer (nuclei isolation buffer supplemented with 0.08 % Triton X-100) by gentle tapping and centrifuged at 1,500 g for 10 min at 4 °C. Both supernatant and pellet fractions were mixed with an equal volume of 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol followed by heating at 70 °C for 15 min. The supernatant and pellet fractions were analyzed by western blotting using α-GFP (Takara Bio), α-pS557, α-Histone H3 (Invitrogen; nuclear protein), and α-UGPase (Agrisera; cytoplasmic protein) antibodies.

Transient gene expression in N. benthamiana

For protein expression in N. benthamiana, the Agrobacterium tumefaciens strain GV3101 carrying plant binary vectors was used as described previously^57,113. Agrobacterium was cultured overnight at 30 °C in 5 mL LB medium supplemented with appropriate antibiotics: gentamycin (50 μg/mL), rifampicin (25 μg/mL), and spectinomycin (100 μg/mL) or kanamycin (50 μg/mL). The overnight culture was centrifuged at 3,000 rcf for 5 min and resuspended in the infiltration buffer containing 10 mM MgCl₂, 10 mM MES (pH 5.7), and 200 μM acetosyringone (Sigma). The Agrobacteria inocula was incubated at room temperature for 2 hours and diluted to OD₆₀₀ = 0.2 for infiltration.

Confocal laser scanning microscopy

Imaging of proteins expressed in N. benthamiana was performed 30 hours after Agrobacteria infiltration and 6 hours after mock or SA treatment. For Arabidopsis protoplasts and transgenic Arabidopsis roots, images were taken 2 hours (ASP-SPARK) or 5–6 hours (other experiments) post-treatment. Arabidopsis seedlings were stained with propidium iodide (PI; 10 μg/mL) by submerging them in the staining solution for 1 min in the dark before imaging. All images were captured with the Zeiss 880 airyscan inverted confocal laser scanning microscope using a 40x/1.2 water correction objective. The scanning resolution was set to 1,024 × 1,024 pixels. A 488-nm argon laser was used to excite GFP and a 505–530 nm band pass filter was used for the detection of the emission signal. A 561-nm diode laser was used to excite mCherry, and the emission was collected with a 575–615 nm band pass filter. PI was excited with the 488 nm argon laser and emission was detected with a 590–620 nm band pass filter. Z-stack imaging was performed with 1 μm intervals to cover the entire N. benthamiana epidermal cell. A max intensity Z-projection was performed in the Fiji (ImageJ) software¹¹⁴ on the image stack to generate the final image. The relative GFP fluorescence intensity between the nucleus and cytosol was analyzed using Fiji.

In vitro phosphorylation assays

Phosphorylation of AtNPR1 by SnRK1/SnAK1

For Figure S2A, FLAG-tagged kinases were expressed in N. benthamiana for 24 hours after Agrobacterium infiltration. The leaf tissue was homogenized in 3 mL extraction buffer [25 mM HEPES (pH 7.5), 150 mM NaCl, 0.5% Triton X-100, 5 mM DTT, and EDTA-free protease inhibitor cocktail (Roche)]. After centrifugation at 16,595 rcf for 15 min at 4 °C, the supernatant was transferred to a new tube, mixed with 10 μL α-FLAG M2 Magnetic Beads (Sigma), and incubated at 4 °C for 3 hours with gentle rocking. Beads were washed four times with 600 μL extraction buffer.

For Figure 3A, GST-NPR1 and its S557A variant were expressed in E. coli and purified using glutathione sepharose (GE Healthcare) followed by ÄKTA go Fast Protein Liquid Chromatography (FPLC) (Cytiva) following the manufacturer’s protocol. FLAG-tagged SnRK1 and SnAK1 were synthesized using a wheat germ-based translation system (Sigma), and equal amounts of synthesized kinases were confirmed by western blotting. The wheat germ extract containing FLAG-tagged kinase proteins was mixed with 10 μL α-FLAG M2 Magnetic Beads (Sigma) and incubated at 4 °C for 3 hours with gentle rocking. Beads were washed four times with 600 μL extraction buffer.

To perform an in vitro kinase assay, 1 μg of GST-NPR1 or GST-npr1^S557A was incubated with the immunoprecipitated kinases on beads in the kinase buffer [20 mM HEPES (pH 7.5), 500 μM CaCl₂, 10 mM MgCl₂, 2 mM DTT, 100 μM ATP] at 30 °C for 30 min with shaking at 850 rpm. For the lambda phosphatase treatment, the kinase reaction mixture was split into two halves, mixed with 1/8 volume of MnCl₂, 1/8 volume of Protein MetalloPhosphatase (PMP) buffer (NEB), and 1/40 volume of lambda phosphatase (NEB) or water. The mixtures were then incubated at 30 °C for 30 min with shaking at 1,250 rpm, mixed with 1/3 volume of 4x LDS sample buffer (Invitrogen), and heated at 85 °C for 10 min before western blotting analysis. The SDS-PAGE was stained with Coomassie Brilliant Blue (CBB) to estimate total GST-NPR1 protein. The membranes were blotted with α-FLAG (Sigma) and α-pAMPK (Thr-172) (Cell Signaling Technology) antibodies for the detection of total and phosphorylated kinases, respectively, and with α-pS557 and α-p(Ser/Thr) Phe (Cell Signaling Technology) antibodies for the detection of phosphorylated NPR1.

Phosphorylation of NPR1 homologs in crops by SnRK1/GRIK1

E. coli-codon-optimized protein coding sequences of NPR1, SnRK1, and GRIK1 from Sorghum bicolor (sorghum), Solanum lycopersicum (tomato), and Oryza sativa Japonica Group (Japanese rice) were synthesized by GenScript (Table S3). GST-NPR1 variants were expressed in E. coli and purified using glutathione sepharose (GE Healthcare). FLAG-tagged SnRK1 and GRIK1 were synthesized using a wheat germ-based translation system (Sigma), and similar amounts of synthesized kinases were confirmed by western blotting. The kinase reaction steps were identical to those described in the in vitro phosphorylation assays of AtNPR1 by SnRK1.

Phosphorylation of AtNPR1 by TOR

TOR was immunoprecipitated from 7-day-old WT seedlings (approximately 120×/sample). Tissues were homogenized on ice in 2 mL IP buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Nonidet P-40, 3 mM DTT, 50 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)]. Homogenates were centrifuged at 16,595 rcf for 15 min at 4 °C and supernatants were transferred to new centrifuge tubes, mixed with 20 μL α-TOR antibody (Abiocode), and incubated overnight at 4 °C with gentle rocking. Then, 100 μL of Dynabeads Protein A/G (Invitrogen) was added and the samples were incubated for 1 hour. The beads were then washed three times with 1 mL wash buffer [25 mM HEPES (pH 7.5), 150 mM NaCl, 0.5% Triton X-100, 5 mM DTT, 50 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)]. For in vitro kinase assay, 1 μg of GST-NPR1 or GST-npr1^S55/59D was incubated with the immunoprecipitated TOR on beads as described above. For Torin 2 treatment, 1 μM Torin 2 was added to the reaction mixture before incubation. The reaction was stopped by adding 1/3 volume of 4x LDS sample buffer (Invitrogen) and heating at 85 °C for 10 min. The samples were subjected to western blotting analysis with α-pS55/59²³ antibody.

In vivo phosphorylation assays

To detect phosphorylation of AtNPR1 by SnRK1 and TOR in vivo, a combination of mesophyll protoplast transformation and whole-plant treatment assays were used.

SnRK1-mediated phosphorylation of AtNPR1

Mesophyll protoplasts were isolated from 3.5-week old WT and sesquiα2-1 leaves and transformed with pBI221–35S:NPR1-GFP as previously described¹¹⁵. The resulting protoplasts were incubated overnight at room temperature before being treated with mock or 0.1 mM SA for 6 hours. Protoplasts were then lysed in extraction buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 0.2% Triton X-100, 0.2% Nonidet P-40, 50 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)] and incubated overnight with 15 μL GFP-trap magnetic agarose (Chromotek) at 4 °C with gentle rocking. Beads were washed four times with 1 mL IP buffer and proteins were eluted by heating at 70 °C in the 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol. The immunoprecipitated samples were analyzed by western blotting using α-GFP (Takara Bio) and α-pS557 antibodies.

For the detection of Ser-557 phosphorylation in mature plants (Figures 3G and 3H), 1.5 g leaf tissue was collected 8 hours after mock or SA treatment. Tissue was homogenized with 3 mL of the extraction buffer, and the supernatant was incubated overnight with 20 μL α-NPR1 antibody at 4 °C with gentle rocking. Then, 30 μL of Dynabeads Protein A/G (Invitrogen) was added, and the samples were incubated for 1 hour at 4 °C. The beads were subsequently washed three times, and bound proteins were eluted before western blotting.

TOR-mediated phosphorylation of AtNPR1

For the detection of NPR1 phosphorylation at Ser-55/59, 35S:NPR1-GFP seedlings grown vertically on 1/2 MS plate (with 1% sucrose) were transferred on day 4 (for 3-day treatment) or day 6 to 1/2 MS liquid media (with 1% sucrose). Either 10 μM Torin 2 or 2 μM AZD-8055 was added for the indicated treatment durations and an equivalent volume of DMSO was added to the mock samples. For the SA-treated samples, 0.3 mM SA was added to the media 6 hours before sample collection. Tissue was homogenized in 60 μL extraction buffer followed by centrifugation at 16,595 rcf for 15 min at 4 °C.

For the detection of Ser-55/59 phosphorylation of transiently expressed NPR1-GFP (for Figures 7C and S5G), mesophyll protoplasts isolated from 3.5-week-old WT or tor-es leaves were transformed with pBI221–35S:NPR1-GFP. The resulting protoplasts were incubated overnight at room temperature with or without 5 μM β-estradiol, 10 μM Torin 2, or 2 μM AZD-8055. On the second day, the W5 medium was removed from the samples after centrifugation, and 2x extraction buffer [100 mM Tris-HCl (pH 7.5), 300 mM NaCl, 10 mM EDTA, 0.4% Triton X-100, 0.4% Nonidet P-40, 100 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)] was added to the protoplasts at a 1:1 ratio. Homogenates were centrifuged at 16,595 rcf for 15 min at 4°C.

An equal volume of supernatant was mixed with 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol before being heated to 70 °C for 15 min. Protein samples were analyzed by western blotting with α-pS55/59²³, α-GFP (Takara Bio), and α-β-Tubulin (Invitrogen) antibodies.

Co-immunoprecipitation

Approximately 1 g of N. benthamiana leaves (Figures 3C, S1E–G) or 0.2 g of Arabidopsis seedlings (Figures 3D, 7D, and 7E) were flash-frozen in liquid N₂, pulverized using a Geno/Grinder, and then homogenized in 2 mL extraction buffer on ice [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 0.2% Triton X-100, 0.2% Nonidet P-40, 50 μM MG132, PhosSTOP phosphatase inhibitor cocktail (Roche), and EDTA-free protease inhibitor cocktail (Roche)]. Homogenates were cleared by centrifugation at 16,595 rcf for 15 min at 4 °C and supernatants were recovered. After centrifugation, 30 μL of the supernatants were transferred to new centrifuge tubes, mixed with an equal volume of 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol, and heated at 85 °C for 10 min as input. The remaining supernatants were incubated overnight with GFP-trap magnetic agarose (Chromotek) or HA-trap magnetic agarose (Chromotek) at 4 °C with gentle rocking. Beads were washed four times with 1 mL extraction buffer and proteins were eluted by heating in the 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol at 85 °C for 10 min. The input and the immunoprecipitated samples were analyzed by western blotting using α-GFP (Takara Bio), α-HA (Cell Signaling Technology), α-pS55/59²³, α-RFP (ChromoTek), α-SnRK1α1 (Agrisera), α-SnRK1α2 (Agrisera), and α-TOR (Abiocode) antibodies.

In vitro pull-down assays

FLAG-tagged proteins for in vitro pull-down assays were synthesized using a wheat germ-based translation system (Sigma). GST-tagged NPR1 was expressed in E. coli and purified using glutathione sepharose (GE Healthcare), followed by FPLC. Equal amounts (1 μg) of GST-NPR1 were mixed with wheat germ-extract containing corresponding kinases and 600 μL IP buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 0.5% Triton X-100, 0.2% Nonidet P-40, 2 mM DTT, and EDTA-free protease inhibitor cocktail (Roche)]. The mixture was rocked at 4 °C for 1.5 hours followed by the addition of 15 μL glutathione magnetic agarose (ThermoFisher) and rocked for another 4 hours. After IP, beads were washed four times for 15 min at 4 °C with IP buffer, proteins were then eluted by heating at 85 °C for 10 min in the 2x LDS sample buffer (Invitrogen) supplemented with 50 mM DTT and 5% β-mercaptoethanol. The input and pull-down samples were analyzed by western blotting using α-FLAG (Sigma), α-pAMPKα (Thr-172) (Cell Signaling Technology), and α-GST-HRP (GE Healthcare) antibodies. Total GST-NPR1 protein from pull-down samples was estimated by CBB staining.

Metabolite measurements

Sugar and sugar derivative measurements were performed following Lunn et al.¹¹⁶ with modifications. Samples and all reagents were maintained at or below 4 °C throughout the extraction procedure. Frozen plant material (100 mg FW) was homogenized using a tissue lyser at liquid nitrogen temperature and extracted vigorously with 1 mL of ice-cold CHCl₃/MeOH (3:7, v/v). The homogenate was transferred to −20 °C for 2 hours with occasional mixing.

After incubation, 500 μL of ice-cold water was added, and the mixture was vortexed and centrifuged at 5,000 rcf for 10 min at 4 °C. The aqueous MeOH supernatant was transferred to a 5 mL Eppendorf tube. The remaining phase was re-extracted with 500 μL of ice-cold water, vortexed, and centrifuged as described above, and the supernatant was combined with the first fraction. The pooled supernatant was back-extracted with 300 μL of CHCl₃ and centrifuged at 5,000 rcf for 10 min at 4 °C. The upper phase was transferred to a new 5 mL Eppendorf tube.

The resulting extract was dried under an N₂ stream at room temperature to remove MeOH, and the remaining aqueous phase was lyophilized overnight. The dried extract was resuspended in 50 μL of ice-cold water and centrifuged at 5,000 rcf for 10 min at 4 °C. The supernatant was transferred to LC vials for further analysis.

Samples were analyzed using an Orbitrap Exploris 120 mass spectrometer (ThermoFisher) coupled to a Vanquish Flex Binary UHPLC system (ThermoFisher) equipped with a ZIC-pHILIC column (150 × 2.1 mm, 5 μm particle size; EMD Millipore). 4 μL of the extract was injected using the following LC conditions: Buffer A consisted of 20 mM (NH₄)₂CO₃ with 0.1% NH₄OH, and Buffer B was CH₃CN. The chromatographic gradient was run at a flow rate of 0.150 mL/min as follows: 0–20 min, linear gradient from 80% to 20% B; 20–20.5 min, linear gradient from 20% to 80% B; 20.5–35 min, hold at 80% B.

MS analyses were performed in negative mode for all samples. MS1 data were acquired over an m/z range of 70–750 at a resolution of 60,000. For MS2 scans, the isolation window was set to m/z 1.5, and stepped collision energies of 30, 50, and 150 were applied, with fragment ions scanned at a resolution of 15,000. LC/MS data were processed using ThermoFisher FreeStyle software (version 1.8 SP2). Compounds were identified based on retention time and MS/MS spectra matched to authentic standards.

To quantify metabolite abundance, extracted ion chromatograms were generated using the [M-H] ion of each metabolite with 5 ppm mass accuracy. Standard curves prepared from authentic standards of Glc6P (Sigma-Aldrich), Tre6P (Cayman chemical), glucose (Sigma-Aldrich), and trehalose (European Pharmacopoeia) were used for absolute quantification.

Phylogenetic analysis

Orthologs of AtNPR1 across land plants were identified in a previous study¹⁶. Twenty-six species representing major clades in angiosperms were selected, including Eudicots (Rosids, Asterids and their sister clades) and Monocots, and their protein sequences were retrieved from Phytozome 13¹¹⁷. MAFFT implemented in the EMBL-EBI Job Dispatcher sequence analysis tool¹¹⁸ was used for multiple sequence alignment with default parameters. The results were visualized using AlignmentViewer (http://alignmentviewer.org).

QuantSeq and data analysis

Total RNA was extracted from leaf tissues using TRIzol reagent (ThermoFisher). 500 ng RNA was used to construct libraries using the Lexogen QuantSeq 3′ mRNA-Seq FWD kit, compatible with Illumina sequencing platforms. Sequencing was performed at the Duke Center for Genomic and Computational Biology on the NovaSeq X system, generating single-end 50 bp reads. Adapter sequences and poly(A) tails were removed using Trim Galore (https://github.com/FelixKrueger/TrimGalore). The resulting cleaned reads were aligned to the Arabidopsis reference genome (TAIR10) using the STAR aligner¹¹⁹. Transcript quantification was performed with HTSeq¹²⁰, and count normalization with differential expression analysis was conducted using DESeq2¹²¹. Transcripts were considered responsive to treatment or genotype if they exhibited a fold change greater than 2 and an adjusted p-value below 0.05. Heatmaps were generated in R using the pheatmap package.

ATP measurements

Arabidopsis seedlings (~0.1g) were weighed after treatment, flash-frozen in liquid N₂, and pulverized using a Geno/Grinder. Then, 5% trichloroacetic acid (2 mL per gram of tissue) was added to the ground sample, followed by vortexing and incubation at room temperature for 30 min. The samples were centrifuged at 21,130 rcf for 5 min, and the supernatants were diluted 50-fold with distilled water. ATP was measured using ATP bioluminescent Assay Kit (Sigma) following manufacturer’s protocol. The luminescence was measured by a Victor Nivo (Revvity) plate reader.

SUnSET assay

SUnSET assay was performed as described previously^80,81 with modifications. Briefly, four 5-mm leaf disks from 3.5-week-old plants were floated in water in 6-well plates. 1 mM SA was added at indicated times (0, 1, 2, or 4 hours) prior to puromycin labeling. 50 μM puromycin was applied to all samples for 45 min before the tissues were flash-frozen in liquid N₂. 80 μL 2x LDS-sample buffer (Invitrogen) was added to the ground tissue followed by heating at 95 °C for 10 min. After heating, the samples were centrifuged at 21,130 rcf for 5 min, and the supernatant was subjected to SDS-PAGE. The puromycin-labeled proteins were detected with an α-puromycin antibody (Sigma). Equal loading was confirmed by Ponceau S staining.

Quantification and statistical analysis

Data plotting and statistical analyses were performed using GraphPad Prism 9. For all statistical data, the values represent the mean, and the error bars show the standard error of the mean. Asterisks indicate statistical significance (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001).

Two-tailed Student’s t-test was used for pairwise comparisons, and one-way ANOVA followed by Tukey’s post hoc was used for multiple comparisons. Different lowercase letters indicate statistical significance tested between multiple groups by Tukey’s post hoc at p < 0.05. Interactions were tested using two-way ANOVA. All experiments, except the LC-MS/MS and QuantSeq, were repeated at least twice with similar results.

Supplementary Material

1

Table S1. Summary of all figures involving the use of the α-pS557 antibody.

2

Table S2. Differentially expressed genes (DEGs) after SA treatment.

3

Table S3. Codon-optimized DNA sequences.

4

Table S4. Primer information, related to STAR Methods.

5

Document S1. Figures S1–S5.

Highlights.

• Salicylic acid activates SnRK1 and inhibits TOR through metabolic reprogramming

• TOR phosphorylation of NPR1 Ser-55/59 suppresses immunity under optimal conditions

• SnRK1 phosphorylation of NPR1 Ser-557 promotes subsequent PTMs and induces defense

• Antagonistic NPR1 phosphorylation by TOR and SnRK1 links metabolism to immunity

Materials availability

All unique constructs and reagents in this study will be available from the lead contact upon completion of a Materials Transfer Agreement.

Data and code availability

This paper does not report original code. The RNA sequencing data have been deposited at Sequence Read Archive (SRA) as PRJNA1273185 while the proteomics data have been deposited at the PRoteomics IDEntifications (PRIDE) database as PXD064730. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.