Plant intracellular NLR receptors recognise pathogen interference to trigger immunity but how NLRs signal is not known. Enhanced disease susceptibility1 (EDS1) heterodimers are recruited by Toll-interleukin1-receptor domain NLRs (TNLs) to transcriptionally mobilise resistance pathways. By interrogating the Arabidopsis EDS1 ɑ-helical EP-domain we identify positively charged residues lining a cavity that are essential for TNL immunity signalling, beyond heterodimer formation. Mutating a single, conserved surface arginine (R493) disables TNL immunity to an oomycete pathogen and to bacteria producing the virulence factor, coronatine. Plants expressing a weakly active EDS1R493A variant have delayed transcriptional reprogramming, with severe consequences for resistance and countering bacterial coronatine repression of early immunity genes. The same EP-domain surface is utilised by a non-TNL receptor RPS2 for bacterial immunity, indicating that the EDS1 EP-domain signals in resistance conferred by different NLR receptor types. These data provide a unique structural insight to early downstream signalling in NLR receptor immunity.
Animal and plant innate immunity is governed by surface and intracellular receptors. Mammalian innate immune responses provide an initial barrier against microbial infection, with specific pathogen resistance being taken over by the adaptive immune system. By contrast, plants depend entirely on panels of germ line-encoded receptors1,2. Plant recognition of specific virulence factors (effectors) delivered by pathogens to host cells is mediated by intracellular nucleotide-binding domain/leucine-rich repeat (NLR) receptors in a process called effector-triggered immunity (ETI)2,3,4. NLR-effector recognition leads to induction anti-microbial defence pathways, often accompanied by localised host cell death at infection sites.
Two major plant NLR types are classified by their N-terminal domain architectures: CC-NLRs (or CNLs) have a coiled-coil (CC) domain and TIR-NLRs (TNLs) a Toll-interleukin 1 receptor (TIR) domain, which contribute to NLR activation and signalling1,4. A characteristic feature of ETI mediated by the different NLR types is amplification of a similar suite of defence pathways that are mobilised at a lower level by surface pattern-recognition receptors (PRRs) recognising pathogen-associated molecular patterns (PAMPs) in PAMP-triggered immunity (PTI)5,6,7,8,9,10. PTI pathways are often targeted by pathogen effectors to promote infection, and in ETI the transcriptional reestablishment and bolstering of immunity outputs is a major driver of resistance3,11. Numerous TFs contribute to ETI governed by CNL and TNL receptors3,12. Also, certain NLRs have nuclear functions3,13,14,15,16, suggesting that the path between NLR activation and gene expression reprogramming is in some cases short. However, the mechanisms by which NLRs converge on transcriptional defences are not known.
Another emerging ETI feature is the strong deployment of alternative (parallel) transcriptional branches, enabling the plant to compensate for disabling of particular host resistance sectors3,11,17,18,19. ETI buffering of defence pathways provides robustness against pathogen interference19. One important transcriptionally mobilised resistance sector against biotrophic pathogens is controlled by the stress hormone salicylic acid (SA)20,21. SA synthesis mediated by the enzyme isochorismate synthase1 (ICS1)22 is controlled by an ensemble of TFs operating within an intricate phytohormone network21,23. SA synthesis and signalling are targeted by pathogens of different classes, often using effectors to boost the SA-antagonising jasmonic acid (JA) hormone system24,25,26. Coronatine (COR) is a potent SA-antagonising virulence molecule produced by Pseudomonas bacteria which promotes disease by mimicking plant endogenous bioactive JA-isoleucine (JA-Ile)26,27. Like JA-Ile, COR signals by binding to nuclear F-box protein coronatine-insensitive1 (COI1)-jasmonate ZIM-domain (JAZ) coreceptors, which relieves JAZ repression of a basic helix-loop-helix (bHLH) TF, myelocytomatosis oncogene homolog2 (MYC2)28,29. MYC2 is a central TF for JA, ethylene and abscisic acid pathways orchestrating numerous stress outputs, including transcriptional dampening of SA accumulation28.
TNLs recognising different pathogen effectors signal via the nucleocytoplasmic, lipase-like protein enhanced disease susceptibility1 (EDS1), which mediates TNL-activated transcriptional reprogramming, resistance and host cell death30,31,32,33,34,35,36,37. Arabidopsis EDS1 forms separate heterodimer complexes with its sequence-related partners, phytoalexin deficient4 (PAD4) or senescence-associated gene101 (SAG101) to function in TNL ETI38,39,40,41,42,43. Analysis of the crystal structure of an EDS1-SAG101 heterodimer, and a structural homology-based model of EDS1-PAD4, showed that the partner N-terminal lipase-like (ɑ/β-hydrolase fold) domains provide a non-catalytic scaffold for binding and to promote contacts between the C-terminal ɑ-helical ‘EDS1-PAD4’ (EP) domains42,43. The EP-domains (PFAM:PF18117 [https://pfam.xfam.org/family/PF18117]) have no significant structural homologies outside the EDS1 family and their role in EDS1 signalling has not been established43.
Arabidopsis EDS1 and SA signalling pathways operate as genetically parallel, mutually reinforcing resistance sectors8,17,18,19,44. We reported that EDS1-PAD4 heterodimers, besides promoting ICS1 expression and SA accumulation, antagonise COR-stimulated MYC2 pathways in Arabidopsis TNL ETI against Pst bacteria19. PAD4 interacted with MYC2 in planta likely by indirect association. EDS1-PAD4 antagonism of the COR/JA MYC2-branch was nuclear, coincident with or after MYC2 release from COI1-JAZ nuclear complexes, and independent of EDS1-PAD4 promotion of the ICS1/SA-branch19. Hence we proposed a two-pronged EDS1 signalling mechanism in ETI for buffering SA immunity against genetic or pathogen interference19.
Here we interrogate the role of the functionally uncharacterised EDS1 EP-domain in Arabidopsis TNL ETI. We identify a positive surface lining a cavity created by the heterodimer, which is essential for pathogen resistance. In TNL immunity against Pst bacteria, the EP-domain confers rapid transcriptional reprogramming, which is needed for countering bacterial COR repression of a set of early immunity genes and for disease resistance. We find that the same EDS1 EP-domain surface is recruited by a CNL receptor (RPS2) for resistance against Pst bacteria, and in both TNL and CNL bacterial ETI responses, EDS1 signals via three genetically separable resistance sectors.
EDS1 EP-domain cavity residues mediate immune signalling
The Arabidopsis EDS1-SAG101 heterodimer crystal structure reveals a cavity formed by the partner EP-domains with nine EDS1 positively charged amino acids (three lysines, five arginines and a histidine) that are conserved across seed plants (Supplementary Fig. 1A, B)43. A similar distribution of positive residues was found in a homology-based structural model of the EDS1-PAD4 heterodimer43. We selected five solvent accessible EDS1 lysine (K) and two arginine (R) residues that are not part of the heterodimer interface and mutated these individually to alanines (Supplementary Fig. 1A, C). In a yeast 2-hybrid (Y2H) assay, the EDS1 EP-domain amino acid exchanges did not interfere with EDS1-PAD4 interaction compared to the EDS1LLIF lipase-like domain mutant which binds PAD4 very weakly (Supplementary Fig. 1D)43. Constructs of wild-type EDS1 cDNA (cEDS1) and EDS1 EP-domain variants under the EDS1 native promoter and tagged with yellow fluorescent protein (YFP) were transformed into the Arabidopsis Col eds1-2 null mutant6,43. Primary (T1) transformants for each construct were then inoculated with the oomycete pathogen Hyaloperonospora arabidopsidis (Hpa) isolate EMWA1 to test for TNL (RPP4) immunity phenotypes, measured against resistant and susceptible controls45. Several EP-domain cavity mutants had reduced RPP4 resistance (Supplementary Fig. 1C). One mutant, EDS1R493A, located in the centre of the cavity (Fig. 1a) was chosen for in-depth analysis because it displayed similarly high Hpa EMWA1 disease susceptibility as eds1-2. Consistent with the T1 TNL immunity phenotypes, two independent homozygous cEDS1R493A transgenic lines (R493A#1 and #2) were fully susceptible to Hpa isolate CALA2 (Fig. 1b), recognised in Col by RPP2A and RPP2B TNL genes46, and the bacterial pathogen Pst AvrRps4 recognised by the nuclear TNL pair RRS1S-RPS431,47 (Fig. 1c). These data show that EDS1R493A compromises immunity governed by TNLs against an oomycete and bacterial pathogen, without breaking EDS1 heterodimer formation.
Further examination of the EDS1-SAG101 and EDS1-PAD4 structures revealed a SAG101 and PAD4 EP-domain surface resembling the EDS1R493A patch but facing away from the cavity (Fig. 1d). We mutated PAD4 arginine 420 (PAD4R420A) which is aligned at the sequence level and predicted to form similar interactions with neighbouring residues as EDS1R493, but on the external PAD4 surface (Fig. 1d). YFP-tagged wild-type cPAD4 and PAD4R420A under the PAD4 native promoter were transformed into a Col pad4-1 sag101-3 mutant, in which loss of SAG101 is compensated for by PAD441,43. In RPP4 (Hpa EMWA1) infection assays of T1 transgenic lines, PAD4R420A was as resistant as cPAD4 and Col, whereas the EDS1 R493A#1, eds1-2 and pad4-1 sag101-3 plants were susceptible (Fig. 1e). These data suggest the location of positively charged residues such as EDS1R493 within the EP-domain cavity is critical for TNL immunity.
We tested whether EDS1R493A retains interaction with PAD4 in plant cells by performing transient expression and immunoprecipitation (IP) assays in eds1-2 pad4-1 sag101-3 protoplasts19. EDS1R493A fused to a FLAG tag interacted with PAD4-YFP as strongly as wild-type EDS1-FLAG, whereas a non-interacting EDS1LLIF-FLAG variant did not bind PAD4-YFP (Supplementary Fig. 2A). YFP-tagged cEDS1R493A protein in R493A#1 and R493A#2 had a similar nucleocytoplasmic distribution as YFP-cEDS1 at 24 hpi with Pst AvrRps4 (Supplementary Fig. 2B), indicating that loss of TNL immunity is not due to failed EDS1 nuclear accumulation35,36. Confocal microscopy imaging of multiple samples showed that the YFP-cEDS1R493A nucleocytoplasmic fluorescence signal was lower than for YFP-cEDS1 (Supplementary Fig. 2B). Also, R493A lines #1 and #2 accumulated less total EDS1 protein compared to cEDS1 (Supplementary Fig. 2C). We tested whether R493A susceptibility could be due to low protein accumulation by selecting independent homozygous transgenic lines (#1 and #2) for two other YFP-cEDS1 EP-domain mutants: K487A and K387A, which had displayed full RPP4 resistance in the T1 seedling assays (Supplementary Fig. 1C) and similar or lower basal EDS1 protein accumulation than R493A lines #1 and #2 (Supplementary Fig. 2C). The K487A and K387A lines were fully resistant to Pst AvrRps4 (Supplementary Fig. 2D) and Hpa EMWA1 (Supplementary Fig. 2E). Additionally, transgenic eds1-2 lines expressing genomic EDS1-YFP or EDS1R493A-YFP (denoted gEDS1 and gR493A) were selected because gEDS1 is generally more highly expressed than cEDS136. Although gEDS1 and gR493A accumulated to similar levels as cEDS1 in mock-treated tissues and after Pst AvrRps4 infection (Supplementary Fig. 3A), the gR493A transgenic lines were fully susceptible to Pst AvrRps4 (Supplementary Fig. 3B). The cR493A and gR493A lines were similarly defective in Pst AvrRps4 induced expression of EDS1-dependent defence marker genes6,35 PAD4, PBS3, ICS1 and FMO1, and in repression of MYB34, measured by qRT-PCR at 8 hpi (Supplementary Fig. 3C). Also, cR493A and gR493A lines displayed full susceptibility to Hpa EMWA1 infection (Supplementary Fig. 3D). These data show that impaired TNL RRS1S RPS4 and RPP4 resistance in the R493A lines is not a consequence of low EDS1R493A protein accumulation. We concluded that EDS1R493 lining the EP-domain cavity confers an important signalling property on the EDS1-PAD4 heterodimer.
EDS1R493A delays TNL transcriptional reprogramming
TNL/EDS1 bacterial immunity divides into two mutually reinforcing resistance branches: one requiring SA synthesised by EDS1-induced ICS1, the other independent of SA and involving EDS1 antagonism of MYC217,19. We determined whether SA accumulation is altered in R493A by measuring free (active) SA in leaves of Col, eds1-2 and cEDS1 and R493A plants at 0, 8 and 24 hpi with Pst AvrRps4. At 8 hpi, SA levels in R493A lines #1 and #2 remained low, resembling the eds1-2 null mutant (Fig. 2a). At 24 hpi, the R493A lines but not eds1-2, recovered SA accumulation to similar levels as cEDS1 and Col (Fig. 2a). Low SA accumulation at 8 hpi correlated with reduced expression of the SA-marker gene PR1 (pathogenesis related1) in R493A compared to cEDS1 and Col at 24 hpi (Fig. 2b). These data show that EDS1R493A is not a complete loss-of-function mutation but instead slow to mobilise the SA branch of TNL immunity against Pst AvrRps4 bacteria. Therefore, one function of the EDS1 EP-domain, which is compromised in R493A, is to promote SA immunity.
We interrogated the EDS1R493A defect in TNL (RRS1S RPS4) transcriptional reprogramming further using RNA-seq analysis. Four-week-old Col, eds1-2, cEDS1 and R493A#1 plants were infiltrated with Pst AvrRps4 and three independent biological replicates for each line analysed at 0, 8 and 24 hpi (see Methods). At 8 hpi there were only 12 DEGs between R493A and eds1-2 (Fig. 2c (pink boxes) and Supplementary Fig. 4A; Supplementary Table 1). These included EDS1 and PBS3. Therefore, at the level of RRS1S RPS4-triggered SA accumulation and transcriptional reprogramming, R493A behaves like the eds1-2 null mutant at 8 hpi (Fig. 2a, c and Supplementary Fig. 4A). At 24 hpi, R493A had a markedly different expression profile to eds1-2 (Fig. 2c and Supplementary Fig. 4A) with 2053 DEG. R493A did not fully recover at 24 hpi as there were 576 DEG between R493A and cEDS1 compared with 5993 DEG between eds1-2 and cEDS1 (Fig. 2c and Supplementary Fig. 4A). These gene expression profiles show that EDS1R493A fails to mobilise TNL transcriptional reprogramming at 8 hpi but recovers substantially at 24 hpi. Based on the R493A bacterial disease susceptibility phenotype (Fig. 1c and Supplementary Fig. 2D), we concluded that recovery of gene expression in R493A lines at 24 hpi is too late to halt pathogen growth. Hence, amino acid R493 in the EDS1 EP-domain lining the heterodimer cavity is critical for timely TNL transcriptional defence reprogramming in response to Pst AvrRps4.
EDS1R493A fails to antagonise bacterial COR-stimulated MYC2 in ETI
One function of the EDS1-PAD4 heterodimer in RRS1S RPS4 ETI is to inhibit COR stimulation of MYC2-regulated JA pathway genes independently of ICS119. Because R493A plants displayed a general delay in gene expression reprogramming (Fig. 2c) and in accumulation of SA (Fig. 2a), we tested whether EDS1R493A is defective in antagonising COR-promoted bacterial infection. For this, Pst AvrRps4 or the COR-deficient Pst ∆cor AvrRps4 strain were infiltrated into Col, eds1-2, cEDS1 and R493A leaves and bacterial growth measured at 3 dpi. As expected, Col and cEDS1 were equally resistant to Pst AvrRps4 and Pst ∆cor AvrRps4, reflecting EDS1 antagonism of COR-stimulated bacterial growth in RRS1S RPS4 ETI (Fig. 3a). The eds1-2 mutant was susceptible to both Pst AvrRps4 strains but supported 1.5-log lower Pst ∆cor AvrRps4 growth compared to Pst AvrRps4, indicating both COR-stimulated and COR-independent bacterial growth promotion in the absence of EDS1 (Fig. 3a)19. Notably, while R493 A lines displayed high susceptibility to Pst AvrRps4, they were as resistant as Col and cEDS1 to Pst ∆cor AvrRps4 (Fig. 3a). Therefore, loss of R493A resistance to Pst AvrRps4 is due to its failure to antagonise bacterial COR. We checked whether COR activity in R493A signals via MYC2 by crossing cEDS1 and R493A#1 into an eds1-2 myc2-3 mutant background. Whereas R493A (in eds1-2) was as susceptible as eds1-2, R493A eds1-2 myc2-3 was as resistant as Col and cEDS1 (Fig. 3b), indicating that EDS1R493A recovers resistance to Pst AvrRps4 when MYC2 is mutated. Hence, EDS1R493A in R493A#1 has a fully restored resistance function in the absence of COR or MYC2. These data expose a defect of EDS1R493A in counteracting COR-dependent, MYC2-promoted bacterial growth in TNL ETI. With respect to Pst AvrRps4, therefore, EDS1R493A loss of resistance is conditional on COR stimulation of the MYC2 JA signalling branch.
As EDS1R493A compromised TNL (RPP4 and RPP2A, B) immunity to Hpa (Fig. 1b, e and Supplementary Fig. 2E, 3D), we tested whether there is also recovery of RPP4 resistance to Hpa EMWA1 in the eds1-2 myc2-3 transgenic lines. Col, myc2-3, cEDS1 in eds1-2 and cEDS1 in eds1-2 myc2-3 expressed full RPP4 resistance after quantifying EMWA1 sporulation on leaves (Fig. 3c). There were similar high levels of EMWA1 sporulation in eds1-2, eds1-2 myc2-3, R493A eds1-2 and R493A eds1-2 myc2-3 (Fig. 3c). These data show that increased susceptibility to Hpa EMWA1 in R493A is not via MYC2. We concluded that the EDS1 EP-domain and associated heterodimer cavity have broader functions in TNL immunity than antagonising MYC2.
TNL/EDS1 signalling delay is exploited by bacterial COR
Because COR-activated MYC2 represses SA accumulation19,26,27 and EDS1R493A delays SA accumulation at 8 hpi with Pst AvrRps4 (Fig. 2a), we tested whether R493A resistance to Pst ∆cor AvrRps4 (Fig. 3a) is due to restored SA. Col and cEDS1 accumulated similar free SA levels at 8 hpi with Pst AvrRps4 and higher SA in response to Pst ∆cor AvrRps4, consistent with COR dampening SA accumulation (Fig. 4a). The eds1-2 null mutant failed to accumulate SA in response to either strain (Fig. 4a), fitting with EDS1 promotion of ICS1 expression and SA independently of its suppression of MYC2 in TNL (RRS1S RPS4) immunity19. Strikingly, R493A lines failed to accumulate SA in response to Pst AvrRps4 and Pst ∆cor AvrRps4 (Fig. 4a). Therefore, delayed SA accumulation in R493A is not caused by a failure to antagonise bacterial COR. These data suggest that antagonism of COR/MYC2 signalling and promotion of SA accumulation are distinct properties of the EDS1-PAD4 heterodimer EP-domain, and more precisely EDS1R493 lining the EP cavity, in TNL ETI to Pst AvrRps4.
Next we explored which gene expression sectors affected by bacterial COR might explain the compromised TNL ETI in R493A by performing RNA-seq analysis at 8 and 24 hpi with Pst ∆cor AvrRps4. To compare between Pst AvrRps4 and Pst ∆cor AvrRps4 RNA-seq experiments, gene expression data for all lines was normalised against the respective Col control within each treatment (see Methods). A multi-dimensional scaling (MDS) plot shows that cEDS1 and Col expression profiles clustered together at 8 and 24 hpi for both treatments (Fig. 4b). At 8 hpi, R493A clustered away from cEDS1 and eds1-2 with Pst ∆cor AvrRps4 but close to eds1-2 with Pst AvrRps4 (Fig. 4b, coloured circles). At 24 hpi, the R493A transcriptome was similar to cEDS1 with Pst ∆cor AvrRps4 (filled triangles) but distinct from cEDS1 with Pst AvrRps4 (filled squares) (Fig. 4b). This analysis shows that EDS1R493A causes a general delay in TNL gene expression reprogramming which is exacerbated by bacterial COR. The results underscore importance of the EDS1 EP-domain for rapid transcriptional mobilisation of multiple pathways besides blocking bacterial COR actions in TNL ETI. We compared our transcriptome data at 8 hpi with publicly available SA-responsive and JA-responsive transcriptomes8,48. This analysis showed that R493A exhibits lower expression of genes regulated by ICS1 compared to cEDS1 regardless of bacterial COR status (Supplementary Fig. 4B). In contrast, genes that are repressed by JA are reduced in R493A with Pst AvrRps4 but not with Pst ∆cor AvrRps4 (Supplementary Fig. 4B). Thus, the R493A transcriptome reflects both slow mobilisation of SA pathways and defective antagonism of COR/MYC2-regulated JA pathways.
We searched for DEG between Pst AvrRps4 and Pst ∆cor AvrRps4 treatments in the different lines at 8 and 24 hpi. After hierarchical clustering, a derived expression heatmap emphasised the extent of delay in R493A at 8 hpi compared to cEDS1, in the presence and absence of bacterial COR (Fig. 4c). Only 75 (8 hpi) and 3 (24 hpi) DEG spread across different clusters were found for cEDS1, consistent with EDS1 effectively antagonising COR in TNL ETI. While R493A was slow in ETI transcriptional reprogramming against both Pst AvrRps4 and Pst ∆cor AvrRps4 (Fig. 4b), there remained 1009 (8 hpi) and 210 (24 hpi) DEG between these treatments in R493A (Supplementary Table 2), indicating that the transcriptional delay in R493A is compounded by COR. Because R493A is susceptible to Pst AvrRps4 but resistant to Pst ∆cor AvrRps4 infection (Fig. 3a), we concluded that EDS1 early interference with COR-dependent expression changes (before or at 8 hpi) is important for inhibiting bacterial growth in the TNL immune response.
COR represses a set of immunity-related genes in R493A
One expression cluster (#17) stood out in the above analysis because it contains EDS1-dependent DEG at 8 hpi that are more highly expressed in R493A with Pst ∆cor AvrRps4 compared to Pst AvrRps4 (Fig. 4c, Supplementary Data 1), suggesting there is COR repression of a set of genes in R493A but not in cEDS1 (which counters COR effects) in TNL immunity. Cluster #17 comprises 383 genes associated with Gene Ontology (GO) terms phosphorylation (37/376 eg. MPK3, MPK2), cell death (18/376 eg. RPS2, NPR1, CPR5) and defence responses (73/376 eg. RPS4, RPP4, ADR1-L2) (Supplementary Data 2). Notable members of cluster #17 are functionally defined NLR (TNL and CNL) and WRKY family TF genes (Fig. 4c, Supplementary Data 1).
We found a significant overlap between cluster #17 genes and genes regulated by the SA analogue Benzothiadiazole (BTH)25 or JA48 (Fig. 4d; Supplementary Data 3). This comparison suggests that early in bacterial infection, TNL/EDS1 signalling protects a large number of SA-promoted/JA-repressed (169/383) genes but also a set of SA/JA-unrelated immunity genes (82/383) from bacterial COR repression. One third (118/383) cluster #17 genes overlap with a set of ETI-associated genes extracted from transcriptomic analyses of the Arabidopsis RPS2 (CNL) response to AvrRpt2 expressed in planta or delivered by Pst bacteria9,49,50 (Supplementary Fig. 4C). Therefore, cluster #17 likely contains some new ETI-related genes. We did not detect significant enrichment of particular cis-regulatory elements in the promoters of cluster #17 genes (by MEME, http://meme-suite.org), suggesting that these genes are controlled by multiple TFs.
In summary, comparing cEDS1, eds1-2 and R493A transcriptomes between Pst AvrRps4 and Pst ∆cor AvrRps4 treatments uncovered a set of TNL/EDS1-controlled genes (cluster #17) whose repression in response to bacterial COR at 8 hpi is inadequately blocked by EDS1R493A. Therefore, R493A highlights a role of the EDS1 heterodimer EP-domain in countering bacterial COR repression of host immunity genes within the first 8 h of TNL immunity.
A positive charge at EDS1R493 is essential for TNL immunity
We have shown that mutations of positively charged residues in the EDS1 EP cavity impair TNL immunity (Supplementary Fig. 1C and Fig. 1b, c). To test the importance of the charge at R493, we generated positively (lysine, cEDS1R493K) and negatively (glutamate, cEDS1R493E) charged variants of EDS1R493. Like EDS1R493A, YFP-tagged EDS1R493K and EDS1R493E displayed wild-type nucleocytoplasmic localisation in transient expression assays in N. benthamiana (Supplementary Fig. 5A). The FLAG-tagged EDS1R493 variants interacted with PAD4-YFP in IP experiments (Supplementary Fig. 5B), consistent with the charge at the EP cavity not affecting heterodimer formation. Additionally, the EDS1R493 mutations did not alter interactions between EDS1/PAD4-YFP complexes and StrepII-HA(SH)-tagged MYC2 in IPs of transiently expressed proteins (Supplementary Fig. 5C), showing that R493 does not affect EDS1/PAD4 association with MYC2. In these assays, EDS1 but not EDS1LLIF (which fails to bind PAD4 stably) decreased PAD4–MYC2 IP signals (Supplementary Fig. 5C), consistent with earlier evidence that PAD4–MYC2 interaction is competed by EDS1 in IPs and that PAD4–MYC2 binding alone does not explain antagonism of MYC2 in TNL (RRS1S RPS4) ETI19.
Two independent homozygous transgenic lines (#1 and #2) were selected in Arabidopsis eds1-2 for cEDS1R493K and cEDS1R493E. In these lines, YFP-tagged EDS1R493K protein accumulated to similar or slightly higher levels than EDS1R493A whereas EDS1R493E accumulation was lower (Supplementary Fig. 6A, B). There was increased accumulation of all EDS1 forms in response to Pst AvrRps4 at 24 hpi. In RRS1S RPS4 resistance to Pst AvrRps4 (Supplementary Fig. 6A, B) and RPP4 resistance to Hpa EMWA1 (Fig. 5a), the R493E lines were as susceptible as eds1-2 and R493 A, whereas R493K and cEDS1 lines were as resistant as Col. Therefore, a positive charge at EDS1 amino acid 493 rather than arginine per se is required for TNL immunity.
Because R493A susceptibility to Pst AvrRps4 was conditional on COR (Fig. 3a), we tested TNL (RRS1S RPS4) responses of R493K and R493E to Pst AvrRps4 and Pst ∆cor AvrRps4. Measured against Col, cEDS1 and eds1-2, R493K was fully resistant and R493E fully susceptible to both Pst AvrRps4 strains, whereas R493A#1 was susceptible only to Pst AvrRps4 (Fig. 5b). Free SA accumulation in the R493K and R493E lines at 8 and 24 hpi with Pst AvrRps4 or Pst ∆cor AvrRps4 mirrored, respectively, cEDS1 and eds1-2 (Fig. 5c).
We next measured the expression of marker genes (EDS1, PAD4 and ICS1) for the EDS1/PAD4-induced SA immunity branch at 8 hpi with Pst AvrRps4 and Pst ∆cor AvrRps4 in the R493K and R493E lines compared to Col-0, eds1-2, cEDS1 and R493A#1 (Supplementary Fig. 6C). The disease resistance, SA accumulation and gene expression phenotypes show that R493K phenocopies wild-type EDS1, and R493E the eds1-2 null mutant. We performed qRT-PCR analysis of MYC2-branch JA response marker genes SA methyl transferase 1 (BSMT1), JAZ10 and Vegetative Storage Protein 1 (VSP1) in the R493 variants at 24 hpi with Pst AvrRps419. Here, R493K behaved like wild-type cEDS1 and R493E like eds1-2 (Supplementary Fig. 6D). By contrast, R493A repressed BSMT1, JAZ10 and VSP1 MYC2-branch genes nearly as strongly as cEDS1 (or Col) (Supplementary Fig. 6D). Hence, EDS1R493A antagonism of the COR-stimulated MYC2-branch at a late time point (24 hpi) is insufficient for restricting Pst AvrRps4 infection in ETI (Fig. 3a). We concluded that a positive charge at EDS1R493 is critical for timely defence gene expression changes and for effective RRS1S RPS4-triggered immunity.
EDS1R493A weak function is responsive to TNL activation
EDS1/PAD4 complexes confer post-infection basal immunity to virulent Pst DC3000 in the absence of TNL-effector recognition39,40,42,43. We therefore tested whether EDS1R493A restriction of Pst DC3000 growth is conditional on bacterial COR by infiltrating leaves of different EDS1R493 variants with Pst DC3000 or Pst ∆cor. Col, cEDS1 and R493K expressed similar resistance to each bacterial strain, with ~10-fold higher Pst DC3000 growth than Pst ∆cor at 3 dpi (Fig. 6a). The eds1-2, R493A and R493E lines were similarly susceptible to each strain, with ~100-fold higher Pst DC3000 growth compared to Pst ∆cor (Fig. 6a). Therefore, R493A does not recover basal resistance to Pst ∆cor in contrast to the recovered TNL resistance to Pst ∆cor AvrRps4 (Fig. 5b). These data show that a positive charge at EDS1R493 is essential for both TNL ETI and basal immunity against Pst bacteria, and that EDS1R493A is partially equipped for TNL ETI but unequipped for basal immunity regardless of bacterial COR repressive effects.
CNL receptor RPS2 signals via the EDS1 EP-domain in ETI
The ICS1/SA pathway partially compensates for eds1 loss-of-function in TNL ETI mediated by RRS1S RPS4 and fully for eds1 in CNL ETI mediated by RPS2 or HRT (HR to turnip crinkle virus)17,44. We removed the ICS1/SA sector to expose an EDS1-only resistance function by crossing cEDS1 and R493A#1 transgenic lines into an eds1 ics1 (eds1-2 sid2-1) mutant background. Growth of Pst ∆cor AvrRps4 and Pst AvrRpt2 bacteria was then quantified as a measure, respectively, of TNL (RRS1S RPS4) and CNL (RPS2) immunity. As expected, cEDS1 and R493A (in eds1-2) conferred full TNL immunity to Pst ∆cor AvrRps4, as did cEDS1 in eds1-2 sid2-1 (Fig. 6b), consistent with EDS1 compensating for loss of ICS1-generated SA in TNL ETI17. By contrast, Pst ∆cor AvrRps4 growth in R493A eds1-2 sid2-1 was intermediate between cEDS1 eds1-2 sid2-1 and eds1-2 sid2-1 (Fig. 6b). Therefore, ICS1-generated SA compensates fully for R493A weak function in RRS1S RPS4 ETI.
In the Pst AvrRpt2 growth assays, R493A in eds1-2 sid2-1 also exhibited intermediate susceptibility between cEDS1 eds1-2 sid2-1 and eds1-2 sid2-1 (Fig. 6c). These data show that the same EDS1 EP-domain surface functions in bacterial resistance mediated by a TNL receptor pair and a CNL receptor, and that in both ETI systems the weak resistance signalling activity of EDS1R493A is buffered by ICS1/SA. Significantly, after stripping away SA and COR/JA effects, there was residual EDS1-controlled resistance in R493A eds1-2 sid2-1 compared to eds1-2 sid2-1 plants (Fig. 6b). Because this remaining resistance in R493A is independent of SA/COR effects in TNL ETI, we propose it is a further resistance output (sector X) regulated by the EDS1 EP domain (Fig. 7).
Arabidopsis EDS1 heterodimers, formed principally by the N-terminal lipase-like domains, are required for TNL-triggered ETI and basal immunity against host-adapted bacterial (Pst) and oomycete (Hpa) pathogens39,40,41,43,51. Current evidence suggests that the partner lipase-like domains with characteristic ɑ/β-hydrolase folds stabilise the heterodimer, bringing together the essential ɑ-helical EP-domains to form a cavity42,43. Here, using structure-guided mutants and reductionist genetic approaches, we identify conserved positively charged residues (K440/441, K478 and R493) lining the EDS1 EP-domain cavity which control TNL immunity signalling beyond heterodimer formation. Analysis of EDS1R493A defects in TNL ETI against COR-producing bacteria shows that the EP-domain ensures rapid transcriptional mobilisation of host immune response pathways and timely blocking of bacterial COR-mediated repression of EDS1-induced immunity genes. We further establish that the same EDS1 EP-domain surface lining the cavity is utilised by a CNL receptor (RPS2) in ETI, and as in the TNL bacterial response, CNL resistance is made up of three genetically separable resistance sectors.
We focused on a single conserved, positively charged EP-domain residue, EDS1R493, because mutation of this to a neutral alanine (R493A) caused complete loss of TNL immunity to Pst AvrRps4 and two Hpa strains, EMWA1 and CALA2 (Fig. 1b, c, e), indicating its broad importance for TNL ETI. A positive EDS1R493 charge rather than the arginine per se is required for ETI because an R493K exchange behaved as wild-type EDS1, and R493E as the null eds1-2 mutant, in Pst AvrRps4 and Hpa EMWA1 infection assays (Fig. 5a, b). Thus, altering the charge at position R493 disables EDS1 heterodimer function without disturbing its interfaces (Fig. 5a, b, Supplementary Fig. 5B). Retained partner interactions of EDS1R493A, EDS1R493K and EDS1R493E variants (Supplementary Fig. 5A, B) and molecular modelling suggest that the charge differences do not destroy integrity of the EP-domain but rather modify ionic interactions in the cavity. Mutation of PAD4R420 with a highly similar sequence and structural environment to EDS1R493 but outside the cavity, did not compromise TNL ETI (Fig. 1d, e), suggesting that the location and charge of EDS1R493 within the EP-domain cavity are crucial for EDS1-PAD4 signalling. We anticipate that functional interactions between the heterodimer EP-domains and other proteins or molecules are determined by the charge status of EDS1R493 and other positive residues lining the cavity (Supplementary Fig. 1C). The R493A transgenic lines #1 and #2 accumulated less EDS1-YFP protein than a control cEDS1 line (Supplementary Fig. 2C, 3A, 6A, B). We established that R493A lower accumulation does not account for its defect in TNL immunity because EP-domain K487A and K387A mutant lines, as well as R493K line #1, with similar EDS1 protein levels as R493A, expressed full TNL immunity to Pst AvrRps4 and Hpa EMWA1 (Fig. 5a, b and Supplementary Fig. 2D, E, 6A, B). Additionally, EDS1R493A expressed from a genomic construct in gR493A lines was as abundant as cEDS1 and yet fully defective in TNL ETI (Supplementary Fig. 3). We cannot exclude that low accumulation of EDS1R493E protein with a negative charge at residue R493 (Supplementary Fig. 6A, B) is below the threshold needed for a working EDS1 pool, although in a previous study very low amounts of nuclear-enriched EDS1-YFP protein were found to be sufficient for TNL ETI against Pst AvrRps4 and Hpa EMWA136.
A striking feature of EDS1R493A is that loss of TNL immunity to Pst AvrRps4 is conditional on bacterial COR signalling via MYC2, as indicated by Pst AvrRps4 and Pst Δcor AvrRps4 growth differences in R493A transgenic lines and fully restored resistance to Pst AvrRps4 in an R493A eds1-2 myc2-3 background (Fig. 3a, b and Fig. 5b). By contrast, EDS1R493A behaves as a complete loss-of-function mutation in basal resistance to virulent Pst DC3000 or Pst Δcor (Fig. 6a). Thus, TNL (RRS1S RPS4) effector recognition and/or activation13,14,52 appears to equip EDS1 complexes, via the EP-domain positively charged surface, to block COR/MYC2-stimulated Pst growth (Figs. 3a, 5b).
Loss of TNL (RPP4) immunity to Hpa EMWA1 in R493A was not recovered by mutation of MYC2 (Fig. 3c). Therefore, EDS1 heterodimers via the EP-domains likely mobilise other anti-microbial pathways or processes against Hpa independently of antagonising MYC2 signalling. The Arabidopsis mutant combinations characterised here will help to identify which pathways or sectors are responsible for restricting Hpa growth in TNL ETI, building on earlier gene expression microrarray analyses of ETI responses to Hpa isolates53,54. A number of protein effectors delivered to plant host cells by Hpa, fungi and P. syringae bacteria suppress SA immunity by targeting JA signalling24,25,55,56,57. For example, P. syringae effectors HopX1 and HopZ1 derepress JA response genes independently of bacterial COR55,57. P. syringae effector HopBB1 activates a subset of JA outputs by increasing the binding between two JA response repressors, JAZ3 and a TF TCP14, leading to their COI1-mediated elimination25. Hpa effector HaRxL44 interacts with and degrades Mediator subunit 19 A to suppress SA signalling56. Since, EDS1 complexes effectively block bacterial COR virulence early in RRS1S RPS4 ETI (Figs. 3a, 5b) and in our RNA-seq analysis there is a significant overlap between TNL/EDS1-dependent DEG at 8 hpi and SA/JA-responsive genes (Supplementary Fig. 4B), we speculate that an important EDS1 heterodimer function in ETI is to preserve SA-based immunity against interference with the phytohormone network by effectors from multiple pathogens.
A further insight to TNL/EDS1 ETI against Pst AvrRps4 bacteria gained from this analysis is that perturbation of the EDS1 EP-domain in R493A lines causes a general delay in transcriptional reprogramming of defence pathways, which, although not caused by bacterial COR, renders the plant vulnerable to COR disease-promoting effects. This is seen most clearly at the level of free (active) SA accumulation at 8 hpi and 24 hpi (Figs. 2a, 4a) and in RNA-seq analyses of wild-type Col (or cEDS1), eds1-2 and R493A responses at 8 hpi and 24 hpi with Pst AvrRps4 and Pst ∆cor AvrRps4 (Figs. 2c, 4b, c). Partial recovery of gene expression changes in R493 A lines at 24 hpi with Pst AvrRps4 without restoration of RRS1S RPS4 immunity (Figs. 2, 4b) further emphasizes that there is a critical expression time-window for ETI to succeed. Hence, early TNL/EDS1 induction of the ICS1/SA resistance branch and probably numerous other anti-microbial outputs (before or at 8 hpi) is important for TNL ETI against COR-producing Pst AvrRps4 (Figs. 3a, 6b). This might also be the case for TNL-recognised Hpa strains (Fig. 1b, e), although the time-frame for transcriptional activation of ETI responses to Hpa strains appears to be more protracted53,54. It remains to be established whether timely defence gene expression reprogramming mediated by the EDS1 EP-domain is also necessary for TNL immunity against Hpa strains.
Based on the above data, we propose that effective TNL/EDS1 immunity involves a critical early step for rapid transcriptional activation of a broad set of defence pathways, rather than selectively mobilising resistance outputs. This makes sense because ETI and basal immune responses differ principally in timing and amplitude of transcriptional reprogramming, while the topology of co-expression networks appears to be quite stable5,8,9,49,58. Moreover, ETI specifically triggered by one pathogen effector or strain confers broad-spectrum immunity to a range of pathogen types59.
By comparing Pst AvrRps4 and Pst ∆cor AvrRps4 transcriptomes in a weakly active EDS1R493A mutant we have uncovered an interesting set of immune-related DEG (cluster #17) (Fig. 4c). These genes would not have emerged from analysis of cEDS1 and null eds1-2 mutant responses alone because they are repressed at 8 hpi only with Pst AvrRps4 in R493A (Fig. 4c). Cluster #17 contains a number of functionally defined NLR receptor and WRKY TF genes, and one member (ADR1-L2) of a family of conserved helper NLRs (Fig. 4c), which contribute to ETI60,61,62. Most of the cluster #17 genes are not represented in ‘ETI-related’ gene sets extracted from RPS2 (CNL) transcriptomic analyses (Supplementary Fig. 4C)9,49. Therefore, cluster #17 might represent a new level of plant host gene expression control in TNL ETI in which key immunity components are protected from pathogen repression. It is tempting to speculate that susceptibility of R493A to Hpa (Fig. 1b, c) is due in part to a failure to block Hpa effectors from targeting some or all of these genes for repression.
Analysis of EDS1R493A reinforces a two-pronged EDS1 transcriptional mechanism in RRS1S RPS4 ETI–(i) promoting ICS1 generated SA and (ii) blocking COR/MYC2 suppression of SA immunity19 (Fig. 7). Delayed free SA accumulation in R493A plants at 8 hpi with Pst AvrRps4 and Pst ∆cor AvrRps4 (Fig. 4a) highlights a failure of this weak EDS1 allele to rapidly mobilise the SA-branch independently of COR. Moreover, early antagonism of COR/MYC2 signalling is important in RRS1S RPS4 ETI because EDS1R493A is fully susceptible to Pst AvrRps4 despite being able to suppress COR/MYC2 marker genes almost as well as cEDS1 later at 24 hpi (Figs. 2c, 4c). It is significant that EDS1R493A retains a portion of TNL resistance to Pst ∆cor AvrRps4 after removal of the ICS1/SA branch in an eds1-2 sid2-1 double mutant (Fig. 6b). We therefore add a third EDS1 TNL immunity sector (denoted X in Fig. 7) which is independent of SA and JA/COR crosstalk. By testing R493A eds1-2 sid2-1 plants in ETI conferred by CNL receptor RPS2 recognising Pst AvrRpt244,50 (Fig. 6c), it becomes clear that the EDS1 EP-domain cavity surface signals in bacterial ETI triggered not only by TNL receptors but at least one CNL receptor type. This has implications for understanding plant NLR receptor molecular functions because the N-terminal TIR and CC domains are structurally different and therefore unlikely to be solely responsible for signalling via the EDS1 EP-domain in bacterial ETI4.
Plant materials, growth conditions and pathogen strains
All mutants and lines are in Arabidopsis accession Col-0. The mutants eds1-2, sid2-1, eds1-2 sid2-1, myc2-3, pad4-1 sag101-3, eds1-2 myc2-3, as well as YFP-cEDS1 and gEDS1-YFP eds1-2 transgenic lines were previously described19,35,43. Pseudomonas syringae pv. tomato (Pst) strain DC3000, Pst ∆cor, Pst DC3000 AvrRps4 (Pst AvrRps4), DC3000 ∆cor AvrRps4 (Pst ∆cor AvrRps4) and DC3000 AvrRpt2 (Pst AvrRpt2) are described17. Plants were grown on soil in controlled environment chambers under a 10 h light regime (150-200 μE/m2s) at 22 °C and 60% relative humidity.
Pathogen infection assays
For bacterial growth assays, Pst AvrRps4 or Pst ∆cor AvrRps4 (OD600 = 0.0005) in 10 mM MgCl2 were hand-infiltrated into leaves of 4-week-old plants and bacterial titres measured at 4 h post infiltration (day 0) and day 3 as described41. Each biological replicate consists of three leaf disks from different plants and data shown in each experiment is compiled from 3–4 biological replicates. Statistical analysis was performed using one-way ANOVA with multiple testing correction using Tukey’s HSD (p < 0.005).
For gene expression and protein accumulation assays, leaves from 4-week-old plants were hand-infiltrated with bacteria (OD600 = 0.005) and samples taken at indicated time points. For measuring protein accumulation, samples were pooled from at least three different plants. For gene expression analysis by qRT-PCR, four or more leaves from different plants were pooled as one biological replicate and two biological replicates were used in each independent experiment. Statistical analysis was performed by Student’s t-test with multiple testing correction using Bonferroni method (p < 0.05).
Hpa isolates EMWA1 and CALA2 were sprayed onto 2–3 week-old plants at 4 × 104 spores/ml dH2O. Plant host cell death and Hpa infection structures were visualised under a light microscope after staining leaves with lactophenol trypan blue as described63. T1 complementation assays of Arabidopsis transgenic lines were performed as previously described45 and Hpa-infected seedlings rescued by spraying with Ridomil Gold (Syngenta). To quantify Hpa sporulation on leaves, three pots of each genotype were infected and treated as biological replicates. Plants were harvested at 6 dpi, their fresh weight determined, and conidiospores suspended in 5–10 ml dH2O and counted under the microscope using a Neubauer counting chamber.
RNA isolation, library preparation and sequencing
For RNA-seq experiments, Col, eds1-2, cEDS1 and R493A#1 4-week-old plants were infiltrated with Pst AvrRps4 or Pst ∆cor AvrRps4 using the same bacterial titre as for gene expression assays. To randomise samples and reduce variation, total RNA was isolated from four individual plants per genotype (three infected leaves per plant) and pooled as one biological replicate. Each biological replicate was from an independent experiment. Total RNA was purified with an RNeasy Plant Mini Kit (Qiagen) according to manufacturer’s instructions. RNA-seq libraries were prepared from 1 µg total RNA according to TruSeq RNA sample preparation v2 guide (Illumina). Library construction and RNA sequencing was done by the Max-Planck Genome Centre (MPIPZ, Cologne), and produced 21–32 million 100 bp single-end reads per sample. RNA-seq reads were mapped to the annotated genome of Arabidopsis thaliana (TAIR10) using TOPHAT2 (a = 10, g = 10)64 and transformed into a read count per gene per sample using the htseq-count script (s = reverse, t = exon) in the package HTSeq65. Genes with <100 reads across samples were discarded. Count data from the remaining genes were TMM-normalised and log2 transformed using functions ‘calcNormFactors’ (R package EdgeR66) and ‘voom’ (R package limma67). To analyse differential gene expression over time between the different genotypes and treatments, for each analysis we fitted a linear model to the respective log2-transformed count data using the function lmFit (R package limma67) and subsequently performed moderated t-tests for specific comparisons of interest. In all cases, the resulting p-values were adjusted for false discoveries due to multiple hypothesis testing via the Benjamini-Hochberg procedure. For each comparison, we extracted a set of significantly differentially expressed genes between the tested conditions (adjusted p-value ≤ 0.05, |log2FC| ≥ 1). RNA-seq experiments for Pst AvrRps4 and Pst ∆cor AvrRps4 were performed in separate batches and therefore normalised to Col for the respective treatments to negate potential batch effects. The normalised values were used to generate a heatmap with hierarchial clustering. Circos plot was created using the R package ‘Circlize’68, to show the overlap of cluster #17 with other datasets.
Total RNA was extracted using a Plant RNA kit (Bio-budget). Five hundred nanogram total RNA was used for cDNA synthesis (quanta bio) and qRT-PCR analysis was performed using SYBR green master mix. The housekeeping gene GapDH was used as reference.
The pENTR/D-TOPO-cEDS1, pENTR/D-TOPO-gEDS1 and pENTR/D-TOPO cPAD4 vectors used for site-directed mutagenesis are previously described36,43. Site-directed mutagenesis on the entry vectors was performed according to the QuikChange II site-directed mutagenesis manual (Agilent). Mutated entry clones were verified by sequencing and recombined into a pAM-PAT-based binary vector backbone by LR reaction.
Generation of Arabidopsis transgenic plants
Stable transgenic lines were generated by transforming binary expression vectors into Arabidopsis null mutants eds1-2 or pad4-1 sag101-3, as indicated, using Agrobacterium-mediated floral dipping.
Yeast two-hybrid assays
Yeast 2-hybrid (Y2H) assays were performed using the Matchmaker system (Clontech) with strain AH109. Gateway cassettes were cloned into pGADT7 and pGBKT7 plasmids. Newly made PAD4 and EDS1 site-directed EP-domain mutants and the EDS1LLIF variant43 were recombined into these plasmids by LR reaction. pGAD-containing and pGBK-containing co-transformants were selected on plates lacking leucine and tryptophan (-LW). Single colonies were re-streaked on plates additionally lacking histidine and adenine (-LWHA) to monitor reporter activation. Yeast growth was recorded after 2–5 days incubation at 30 °C.
Transient expression in Arabidopsis protoplasts
Leaf mesophyll protoplasts were prepared from 4-week-old eds1-2 pad4-1 sag101-3 plants and transfected with plasmid DNA69. After transfection, protoplasts were incubated at room temperature under weak light (1.5 μE/m2s) for 16 h. Protoplasts were harvested and IPs performed as described below.
Protein extraction, immunoprecipitation, immunoblotting
Total leaf extracts or protoplasts were processed in extraction buffer (50 mM Tris pH 7.5, 150 mM NaCl, 10% (v/v) glycerol, 2 mM EDTA, 5 mM DTT, protease inhibitor (Roche, 1 tablet per 50 ml, 0.1% Triton). Lysates were centrifuged for 15 min, 12,000 rpm at 4 °C. 50 µl of supernatant was used as input sample. Immunoprecipitations (IPs) were conducted by incubating the input sample with 12 μl GFP-TrapA beads (Chromotek) for 2 h at 4 °C. Beads were collected by centrifugation at 2000 rpm, 1 min at 4 °C. Beads were washed three times in extraction buffer and boiled at 95 °C in 2× Laemmli buffer for 10 min. Proteins were separated by SDS-PAGE and analysed by immunoblotting. Antibodies used were α-GFP (1:5000; Sigma Aldrich, 11814460001), α-HA (1:5000; Sigma Aldrich, 11867423001), α-FLAG (1:5000; Sigma Aldrich, F3165). Secondary antibodies coupled to Horseradish Peroxidase (HRP) were used for protein detection on blots (1:5000; Sigma Aldrich, 005000000005295191).
Free SA was quantified from leaf tissues (70–200 mg fresh weight), of 4-week-old plants using a chloroform/methanol extraction and analysed by gas chromatography coupled to a mass spectrometer (GC-MS, Agilent)70. Statistical analysis was performed by Student’s t-test with multiple testing correction using the Bonferroni method (p < 0.05).
Three independent experiments were performed for all assays unless otherwise indicated. The number of biological replicates used for each experiment shown isindicated in the corresponding figure legend. Statistical significance was determined either using a Student’s t test or one-way ANOVA. Bonferroni or Tukey’s HSD Posthoc tests for multiple variate analysis were applied as appropriate.
Further information on experimental design is available in the Nature Research Reporting Summary linked to this article.
Source data are provided as a Source Data file. The RNA-seq data are deposited in the National Center for Biotechnology Information Gene Expression Omnibus (GEO) database with accession number GSE116269.
Jacob, F., Vernaldi, S. & Maekawa, T. Evolution and conservation of plant NLR functions. Front. Immunol. 4, 297 (2013).
Jones, J. D. G., Vance, R. E. & Dangl, J. L. Intracellular innate immune surveillance devices in plants and animals. Science 354, aaf6395 (2016).
Cui, H., Tsuda, K. & Parker, J. E. Effector-triggered immunity: from pathogen perception to robust defense. Annu. Rev. Plant Biol. 66, 487–511 (2015).
Zhang, X., Dodds, P. N. & Bernoux, M. What do we know about NOD-like receptors in plant immunity? Annu. Rev. Phytopathol. 55, 205–229 (2017).
Tao, Y. et al. Quantitative nature of arabidopsis responses during compatible and incompatible interactions with the bacterial pathogen Pseudomonas syringae. Plant Cell 15, 317–330 (2003).
Bartsch, M., Bautor, J. & Parker, J. E. Salicylic acid–independent ENHANCED DISEASE SUSCEPTIBILITY1 signaling in arabidopsis immunity and cell death is regulated by the monooxygenase FMO1 and the nudix hydrolase NUDT7. Plant Cell 18, 1038–1051 (2006).
Navarro, L. et al. The transcriptional innate immune response to flg22. Interplay and overlap with Avr gene-dependent defense responses and bacterial pathogenesis 1. Plant Physiol. 135, 1113–1128 (2004).
Tsuda, K., Sato, M., Stoddard, T., Glazebrook, J. & Katagiri, F. Network properties of robust immunity in plants. PLoS. Genet. 5, e1000772 (2009).
Mine, A. et al. The defense phytohormone signaling network enables rapid, high-amplitude transcriptional reprogramming during effector-triggered immunity. Plant Cell 30, 1199–1219 (2018).
Jacob, F. et al. A dominant-interfering camta3 mutation compromises primary transcriptional outputs mediated by both cell surface and intracellular immune receptors in Arabidopsis thaliana. New Phytol. 217, 1667–1680 (2018).
Tsuda, K. & Somssich, I. E. Transcriptional networks in plant immunity. New Phytol. 206, 932–947 (2015).
Birkenbihl, R. P., Liu, S. & Somssich, I. E. Transcriptional events defining plant immune responses. Curr. Opin. Plant Biol. 38, 1–9 (2017).
Le Roux, C. et al. A receptor pair with an integrated decoy converts pathogen disabling of transcription factors to immunity. Cell 161, 1074–1088 (2015).
Sarris, P. F. et al. A plant immune receptor detects pathogen effectors that target WRKY transcription factors. Cell 161, 1089–1100 (2015).
Fenyk, S. et al. The potato nucleotide-binding leucine-rich repeat (NLR) immune receptor Rx1 is a pathogen-dependent DNA-deforming protein. J. Biol. Chem. 290, 24945–24960 (2015).
Townsend, P. D. et al. The intracellular immune receptor Rx1 regulates the DNA-binding activity of a Golden2-like transcription factor. J. Biol. Chem. 293, 3218–3233 (2018).
Cui, H. et al. A core function of EDS1 with PAD4 is to protect the salicylic acid defense sector in Arabidopsis immunity. New Phytol. 213, 1802–1817 (2017).
Tsuda, K. et al. Dual regulation of gene expression mediated by extended MAPK activation and salicylic acid contributes to robust innate immunity in Arabidopsis thaliana. PLoS Genet. 9, e1004015 (2013).
Cui, H. et al. Antagonism of transcription factor MYC2 by EDS1/PAD4 complexes bolsters salicylic acid defense in Arabidopsis effector-triggered immunity. Mol. Plant 11, 1053–1066 (2018).
Fu, Z. Q. & Dong, X. Systemic acquired resistance: turning local infection into global defense. Annu. Rev. Plant Biol. 64, 839–863 (2013).
Seyfferth, C. & Tsuda, K. Salicylic acid signal transduction: the initiation of biosynthesis, perception and transcriptional reprogramming. Front. Plant Sci. 5, 1–10 (2014).
Wildermuth, M. C., Dewdney, J., Wu, G. & Ausubel, F. M. Isochorismate synthase is required to synthesize salicylic acid for plant defence. Nature 414, 562–565 (2001).
Pieterse, C. M. J., Van der Does, D., Zamioudis, C., Leon-Reyes, A. & Van Wees, S. C. M. Hormonal modulation of plant immunity. Annu. Rev. Cell Dev. Biol. 28, 489–521 (2012).
Kazan, K. & Lyons, R. Intervention of phytohormone pathways by pathogen effectors. Plant Cell 26, 2285–2309 (2014).
Yang, L. et al. Pseudomonas syringae Type III effector HopBB1 promotes host transcriptional repressor degradation to regulate phytohormone responses and virulence. Cell Host Microbe 21, 156–168 (2017).
Zheng, X.-Y. et al. Coronatine promotes Pseudomonas syringae virulence in plants by activating a signaling cascade that inhibits salicylic acid accumulation. Cell Host Microbe 11, 587–596 (2012).
Brooks, D. M., Bender, C. L. & Kunkel, B. N. The Pseudomonas syringae phytotoxin coronatine promotes virulence by overcoming salicylic acid-dependent defences in Arabidopsis thaliana. Mol. Plant Pathol. 6, 629–639 (2005).
Kazan, K. & Manners, J. M. MYC2: The master in action. Mol. Plant 6, 686–703 (2013).
An, C. et al. Mediator subunit MED25 links the jasmonate receptor to transcriptionally active chromatin. Proc. Natl Acad. Sci. USA 114, E8930–E8939 (2017).
Wiermer, M., Feys, B. J. & Parker, J. E. Plant immunity: the EDS1 regulatory node. Curr. Opin. Plant Biol. 8, 383–389 (2005).
Wirthmueller, L., Zhang, Y., Jones, J. D. G. & Parker, J. E. Nuclear accumulation of the arabidopsis immune receptor RPS4 is necessary for triggering EDS1-dependent defense. Curr. Biol. 17, 2023–2029 (2007).
Heidrich, K. et al. Arabidopsis EDS1 connects pathogen effector recognition to cell compartment-specific immune responses. Science 334, 1401–1404 (2011).
Bhattacharjee, S., Halane, M. K., Kim, S. H. & Gassmann, W. Pathogen effectors target arabidopsis EDS1 and alter its interactions with immune regulators. Science 334, 1405–1408 (2011).
Gao, F., Dai, R., Pike, S. M., Qiu, W. & Gassmann, W. Functions of EDS1-like and PAD4 genes in grapevine defenses against powdery mildew. Plant Mol. Biol. 86, 381–393 (2014).
García, A. V. et al. Balanced nuclear and cytoplasmic activities of EDS1 are required for a complete plant innate immune response. PLoS Pathog. 6, e1000970 (2010).
Stuttmann, J. et al. Arabidopsis thaliana DM2h (R8) within the Landsberg RPP1-like resistance locus underlies three different cases of EDS1-conditioned autoimmunity. PLoS Genet. 12, 1–29 (2016).
Xu, F. et al. Autoimmunity conferred by chs3-2D relies on CSA1, its adjacent TNL-encoding neighbour. Sci. Rep. 1, 1–7 (2015).
Zhou, N., Tootle, T. L., Tsui, F., Klessig, D. F. & Glazebrook, J. PAD4 functions upstream from salicylic acid to control defense responses in Arabidopsis. Plant Cell 10, 1021–1030 (1998).
Jirage, D. et al. Arabidopsis thaliana PAD4 encodes a lipase-like gene that is important for salicylic acid signaling. Proc. Natl Acad. Sci. USA 96, 13583–13588 (1999).
Feys, B. J., Moisan, L. J., Newman, M. & Parker, J. E. Direct interaction between the Arabidopsis disease resistance signaling proteins, EDS1 and PAD4. EMBO J. 20, 5400–5411 (2001).
Feys, B. J. et al. Arabidopsis SENESCENCE-ASSOCIATED GENE101 stabilizes and signals within an ENHANCED DISEASE SUSCEPTIBILITY1 complex in plant innate immunity. Plant Cell 17, 2601–2613 (2005).
Rietz, S. et al. Different roles of Enhanced Disease Susceptibility1 (EDS1) bound to and dissociated from Phytoalexin Deficient4 (PAD4) in Arabidopsis immunity. New Phytol. 191, 107–119 (2011).
Wagner, S. et al. Structural basis for signaling by exclusive EDS1 heteromeric complexes with SAG101 or PAD4 in plant innate immunity. Cell Host Microbe 14, 619–630 (2013).
Venugopal, S. C. et al. Enhanced disease susceptibility 1 and salicylic acid act redundantly to regulate resistance gene-mediated signaling. PLoS Genet. 5, e1000545 (2009).
Stuttmann, J. et al. Perturbation of Arabidopsis amino acid metabolism causes incompatibility with the adapted biotrophic pathogen Hyaloperonospora arabidopsidis. Plant Cell 23, 2788–2803 (2011).
Sinapidou, E. et al. Two TIR:NB:LRR genes are required to specify resistance to Peronospora parasitica isolate Cala2 in Arabidopsis. Plant J. 38, 898–909 (2004).
Saucet, S. B. et al. Two linked pairs of Arabidopsis TNL resistance genes independently confer recognition of bacterial effector AvrRps4. Nat. Commun. 6, 1–12 (2015).
Hickman, R. et al. Architecture and dynamics of the jasmonic acid gene regulatory network. Plant Cell 29, 2086–2105 (2017).
Hatsugai, N. et al. A plant effector‐triggered immunity signaling sector is inhibited by pattern‐triggered immunity. EMBO J. 36, 2758–2769 (2017).
Kunkel, B. N., Bent, A. F., Dahlbeck, D., Innes, R. W. & Staskawicz, B. J. RPS2, an Arabidopsis disease resistance locus specifying recognition of Pseudomonas syringae strains expressing the avirulence gene avrRpt2. Plant Cell 5, 865–875 (1993).
Parker, J. E. et al. Characterization of edsl, a mutation in Arabidopsis suppressing resistance to Peronospora parasitica specified by severa1 different RPP genes. Plant Cell 8, 2033–2046 (1996).
Williams, S. J. et al. Structural basis for assembly and function of a heterodimeric plant immune receptor. Science 344, 299–303 (2014).
Eulgem, T., Weigman, V. J., Chang, H., Mcdowell, J. M. & Holub, E. B. Gene expression signatures from three genetically separable resistance gene signaling pathways for downy mildew resistance. Plant Physiol. 135, 1129–1144 (2004).
Wang, W. et al. Timing of plant immune responses by a central circadian regulator. Nature 470, 110–115 (2011).
Gimenez-Ibanez, S. et al. The bacterial effector HopX1 targets JAZ transcriptional repressors to activate jasmonate signaling and promote infection in Arabidopsis. PLoS Biol. 12, e1001792 (2014).
Caillaud, M.-C. et al. A downy mildew effector attenuates salicylic acid-triggered immunity in Arabidopsis by interacting with the host mediator complex. PLoS Biol. 11, e1001732 (2013).
Jiang, S. et al. Bacterial effector activates jasmonate signaling by directly targeting JAZ transcriptional repressors. PLoS Pathog. 9, e1003715 (2013).
Kim, Y. et al. Mechanisms underlying robustness and tunability in a plant immune signaling network. Cell Host Microbe 15, 84–94 (2014).
Rentel, M. C., Leonelli, L., Dahlbeck, D., Zhao, B. & Staskawicz, B. J. Recognition of the Hyaloperonospora parasitica effector ATR13 triggers resistance against oomycete, bacterial, and viral pathogens. Proc. Natl Acad. Sci. USA 105, 1091–1096 (2008).
Bonardi, V. et al. Expanded functions for a family of plant intracellular immune receptors beyond specific recognition of pathogen effectors. Proc. Natl Acad. Sci. USA 108, 16463–16468 (2011).
Schön, M. et al. Analyses of wrky18 wrky40 plants reveal critical roles of SA/EDS1 signaling and indole-glucosinolate biosynthesis for Golovinomyces orontii resistance and a loss-of resistance towards Pseudomonas syringae pv. tomato AvrRPS4. Mol. Plant Microbe. Interact. 26, 758–767 (2013).
Dong, O. X. et al. TNL-mediated immunity in Arabidopsis requires complex regulation of the redundant ADR1 gene family. New Phytol. 210, 960–973 (2016).
Muskett, P. R. et al. Arabidopsis RAR1 exerts rate-limiting control of R gene-mediated defenses against multiple pathogens. Plant Cell 14, 979–992 (2002).
Kim, D. et al. TopHat2: Accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 14, R36 (2013).
Anders, S., Pyl, P. T. & Huber, W. HTSeq-A Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015).
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2009).
Law, C. W., Chen, Y., Shi, W. & Smyth, G. K. Voom: precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 15, 1–17 (2014).
Gu, Z., Gu, L., Eils, R., Schlesner, M. & Brors, B. Circlize implements and enhances circular visualization in R. Bioinformatics 30, 2811–2812 (2014).
Yoo, S.-D., Cho, Y.-H. & Sheen, J. Arabidopsis mesophyll protoplasts: a versatile cell system for transient gene expression analysis. Nat. Protoc. 2, 1565–1572 (2007).
Straus, M. R., Rietz, S., Ver Loren van Themaat, E., Bartsch, M. & Parker, J. E. Salicylic acid antagonism of EDS1-driven cell death is important for immune and oxidative stress responses in Arabidopsis. Plant J. 62, 628–640 (2010).
We thank Kenichi Tsuda and Takaki Maekawa (MPIPZ Cologne) for helpful discussions. This work was supported by The Max-Planck Society and Deutsche Forschungsgemeinschaft (DFG) grants within SFB 635 and SFB 670 (JEP, DDB), SFB 680 (JEP, DL), and an International Max-Planck Research School (IMPRS) doctoral fellowship (PvB).
The authors declare no competing interests.
Journal peer review information: Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Bhandari, D.D., Lapin, D., Kracher, B. et al. An EDS1 heterodimer signalling surface enforces timely reprogramming of immunity genes in Arabidopsis. Nat Commun 10, 772 (2019). https://doi.org/10.1038/s41467-019-08783-0
Journal of Plant Biology (2021)
Cellular and Molecular Life Sciences (2020)