Abstract
Arabidopsis thaliana mitogen-activated protein kinases 3 and 6 (MPK3/6) are activated transiently during pathogen-associated molecular pattern-triggered immunity (PTI) and durably during effector-triggered immunity (ETI). The functional differences between these two kinds of activation kinetics and how they coordinate the two layers of plant immunity remain poorly understood. Here, by suppressor analyses, we demonstrate that ETI-mediating nucleotide-binding domain leucine-rich repeat receptors (NLRs) and the NLR signaling components NDR1 and EDS1 can promote the salicylic acid sector of defense downstream of MPK3 activity. Moreover, we provide evidence that both sustained and transient MPK3/6 activities positively control the expression of several NLR genes, including AT3G04220 and AT4G11170. We further show that NDR1 and EDS1 contribute to the up-regulation of these two NLRs in both an ETI and a PTI context. Remarkably, whereas in ETI MPK3/6 activities are dependent on NDR1 and EDS1, they are not in PTI, suggesting crucial differences in the two signaling pathways. Finally, we demonstrate that expression of the NLR AT3G04220 is sufficient to induce expression of defense genes from the salicylic acid branch. Overall, this study expands our knowledge of MPK3/6 functions during immunity and provides new insights into the intricate interplay of PTI and ETI.
Introduction
The plant defense responses to pathogens are usually viewed as a two-layered system (Jones and Dangl, 2006). In the first layer, cell surface-localized pattern-recognition receptors (PRRs) recognize conserved pathogen-associated molecular patterns (PAMPs) and then elicit PAMP-triggered immunity (PTI). The family of PRRs comprises leucine-rich repeat (LRR) receptor kinases (LRR-RKs) and LRR receptor proteins (LRR-RPs), which differ in their mechanisms of signal transduction but are considered to converge towards conserved PTI responses (Zipfel, 2014). In the second layer of immunity, pathogen effectors, which are secreted to counteract plant defense responses and to favor plant susceptibility, are recognized by intracellular nucleotide-binding domain LRR receptors (NLRs), giving rise to effector-triggered immunity (ETI) (Cui et al., 2015). Although signaling events involved in PTI are relatively well known, including receptor-like cytoplasmic kinases (RLCKs), mitogen-activated protein kinases (MAPKs), calcium-dependent protein kinases (CDPKs), and reactive oxygen species (Bigeard et al., 2015), the signaling mechanisms of ETI remain more elusive. The existence of two main classes of NLRs, containing in their N-terminal part either a coiled-coil (CC) domain for CC-NLR (CNL) or a Toll and interleukin-1 receptor (TIR) domain for TIR-NLR (TNL), suggests that ETI signaling might be processed through two distinct pathways depending on the type of NLR involved. For instance, initial reports supported the notion that Non-Race Specific Disease Resistance 1 (NDR1) mediates CNL-ETI while Enhanced Disease Susceptibility 1 (EDS1) contributes to TNL-ETI (Aarts et al., 1998). However, further studies undermined this conception by revealing that CNL-ETI could be independent of NDR1 (Day et al., 2006; Kapos et al., 2019), that EDS1 could play a role in CNL-ETI (Venugopal et al., 2009; Bhandari et al., 2019), and that CNLs and TNLs could cooperate (Wu et al., 2019).
In addition to this, the notion that there is a strict dichotomy between PTI and ETI, which would be consecutive in time and would each represent a specific kind of immunity, has been regularly challenged. Studies showing that PTI and ETI not only share numerous signaling components but also lead to similar gene reprogramming progressively built a model in which PTI and ETI are continuously linked, with connections allowing sophisticated and extensive modulations of plant defense responses (Tsuda and Katagiri, 2010; Peng et al., 2018; Lu and Tsuda, 2021; Yuan et al., 2021b). For instance, using a combination of mutants in key defense genes, it was shown that the Arabidopsis immune network is composed of four main sectors—three sectors mediated by the phytohormones jasmonate, ethylene, and salicylic acid (SA), plus the lipase-like Phytoalexin Deficient4 sector—and that these four sectors are essential for both PTI and ETI, although they are differently articulated in the PTI and ETI contexts, enabling synergistic effects during PTI and rather robust responses during ETI (Tsuda et al., 2009). More recently, it has also been discovered that PTI responses are required for the optimization of ETI responses and that, in return, ETI responses promote the accumulation of PTI actors, thereby demonstrating a mutual potentiation of the two kinds of immunity (Ngou et al., 2021; Yuan et al., 2021a). Despite these significant advances, the question of the molecular mechanisms underlying the crosstalk between PTI and ETI remains one of the most exciting in the field of plant–microbe interactions (Harris et al., 2020).
MAPKs are essential signaling components that allow plants to integrate various cues from biotic and abiotic stresses or developmental programs into appropriate cellular responses (Colcombet and Hirt, 2008). A canonical MAPK cascade encompasses a MAPK kinase kinase (MAP3K), a MAPK kinase (MAP2K or MKK), and a MAPK (or MPK), which are activated in a serial manner by phosphorylation (Ichimura et al., 2002). Active MAPKs can subsequently phosphorylate specific substrates on specific sites, thereby translating signal inputs into functional outputs (Dóczi and Bögre, 2018). In the context of immunity, two MAPK cascades have been particularly well characterized. The rapid and transient activation of both the MAP3K3/5-MKK4/5-MPK3/6 and MEKK1-MKK1/2-MPK4 cascades upon PAMP perception leads to gene reprogramming that is instrumental in mounting successful PTI responses (Asai et al., 2002; Frei dit Frey et al., 2014; Gao et al., 2008; Sun et al., 2018). Recently, it has been shown that MPK3/6 could also be activated in a sustained manner in response to effector recognition, and several roles, such as buffering of the SA sector of defense, promotion of camalexin production, or inhibition of photosynthesis, have been associated with this phenomenon (Tsuda et al., 2013; Xu et al., 2016; Su et al., 2018). Nonetheless, it is not clear so far whether these functions are specific to sustained MPK3/6 activation or simply the extension of processes already controlled by transient MPK3/6 activation. Similarly, the question whether sustained MPK3/6 activation constitutes a general feature of ETI or is restricted to the recognition of particular effectors by particular NLRs remains debatable (Lang and Colcombet, 2020).
Here, starting with an analysis of the regulation and functions of sustained activation of MPK3/6 during ETI, we came to the finding that MPK3/6 activities could bridge PTI and ETI by positively controlling the SA sector of defense through the expression of some NLR genes. We also showed that the ETI-regulating proteins NDR1 and EDS1 are involved in this process. Altogether, our results unveil a novel intricate interplay between PTI and ETI components.
Materials and methods
Plasmid constructs
The AT3G04220 coding sequence was amplified from Arabidopsis cDNA using the primers CACCATGGATTCTTCTTTTTTAC and GCATTTATAAAACTTCAATCTCTTG. Sequencing revealed a 27 bp insertion after nucleotide 1935 in comparison to the reference sequence from TAIR10. This new sequence was then introduced by digestion/ligation between the XhoI and StuI restriction sites in a dexamethasone-inducible expression vector (Gao et al., 2013).
Plant materials and growth conditions
All plants from this study are in the Columbia background. The rps2rpm1 (Nobori et al., 2018), rps4-2 (Saucet et al., 2015), eds1-2 (Bartsch et al., 2006), ndr1-1 (Century et al., 1995), mpk3-1 (Zhao et al., 2014), mpk6-4 (Xu et al., 2008), mkk4-18/mkk5-18 (Li et al., 2018), and K3CA-2 and K3WT-1 (Genot et al., 2017) backgrounds were described previously. The snc1-11 (SALK_116460), at4g11170 (SALK_007034) and at3g04220 (GABI_290D03) lines were purchased from the Eurasian Arabidopsis Stock Centre and homozygous plants were selected by genotyping using LBb1.3 (SALK), o849 (GABI), TGGTGATTCCGATTTTCTTCCAC and TCTGTTGCTTTAACCTTTGCTCC (snc1-11), TTTAGCGGTCAACACGAAAAC and CCAAAATTGAAAATAGAGAACCC (at4g11170), and GTCGTCTTTATCTCTCACGCG and GAAGGGCCTCTTCATAGTTGG (at3g04220) primers. The DEX-AT3G04220/at3g04220 line was obtained by floral dip, and transformed plants were selected on hygromycin. The K3CA-2/ndr1-1, K3CA-2/eds1-2, and K3CA-2/snc1-11 lines were obtained by crosses, and homozygous plants were selected by genotyping and segregation analysis. All plants were grown in growth rooms at 20 °C in short-day conditions (8 h light/16 h dark) at 60% relative humidity and under a light intensity of ~150 μmol m–2 s–1.
Plant treatments and bacterial infections
All chemical treatments and bacterial infections were performed on 1.5-month-old plants by syringe infiltration. The PAMP flg22 and the steroids estradiol and dexamethasone were used at 1 µM, 10 µM, and 5 µM in 10 mM MgCl2. The Pseudomonas syringae pv. tomato DC3000 wild-type (WT), AvrRpt2, AvrRpm1, AvrRps4 (Aarts et al., 1998), and the non-polar hrcC- (Peñaloza-Vázquez et al., 2000) strains were described previously. The bacteria were grown on solid NYGA medium (0.5% bactopeptone, 0.3% yeast extract, 2% glycerol, 1.5% agar) and liquid Luria-Bertani medium supplemented with the appropriate antibiotics (50 µg ml–1 rifampicin for WT and hrcC-, and 50 µg ml–1 rifampicin + 25 µg ml–1 kanamycin for AvrRpt2, AvrRpm1, and AvrRps4).
Fresh cultures of bacteria were washed and resuspended in 10 mM MgCl2 at a final OD600=0.015 for RNA and protein analyses, and at a final OD600=0.005 for pathoassays. Technical repeats were typically constituted from punches of leaves taken from at least two different plants. Bacterial load was quantified by counting the colony-forming units.
Protein methods
For immunoblotting, proteins were extracted in a non-denaturing buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 0.1% NP40, 5 mM EGTA, 0.1 mM DTT) or in non-denaturing Lacus buffer (15 mM EGTA, 15 mM MgCl2, 75 mM NaCl, 1% Tween, 25 mM Tris–HCl, pH 7.5, 1 mM DTT) in the presence of inhibitors of proteases and phosphatases, and then quantified by Bradford assay. Approximately 10 µg of total proteins were loaded on SDS-PAGE gels. The antibodies used were anti-pTpY (Cell Signaling 4370L), anti-MPK3 (Sigma M8318), anti-MPK6 (Sigma A7104), anti-H3 (Abcam Ab1791), and anti-PEPC (ThermoFisher 4100-4163) at 1:10 000 dilution.
For immunoprecipitation, proteins were extracted in a Lacus non-denaturing buffer in the presence of inhibitors of proteases and phosphatases, and then quantified by Bradford assay. A 100 µg sample of total proteins was mixed with 20 µl of sepharose beads (GE Healthcare) and 0.5 µl of anti-myc (Sigma C3956) antibody and incubated for 2 h with gentle shaking at 4 °C. Then, the immunoprecipitate was washed twice in SUC1 buffer (50 mM Tris–HCl, pH 7.4, 250 mM NaCl, 5 mM EGTA, 5 mM EDTA, 0.1% Tween) and twice in Kinase buffer (20 mM HEPES, pH 7.5, 15 mM MgCl2, 5 mM EGTA, 1 mM DTT).
For kinase assays, immunoprecipitates were resuspended in 15 µl of kinase buffer containing 0.1 mM ATP, 1 mg ml–1 myelin basic protein as substrate, and 2 µCi ATP [γ-33P]. After 30 min of reaction at room temperature samples were loaded on SDS-PAGE gel. Then the gels were dried and revealed using an Amersham™ Typhoon™ imager.
For nucleocytoplasmic fractioning, proteins were extracted in Honda buffer (2.5 % Ficoll type 400, 5 % Dextran MW 35-45 k, 0.4 M sucrose, 25 mM Tris-HCl pH 7.5, 10 mM MgCl2, 5 mM DTT) in presence of inhibitors of proteases and phosphatases. After 15 min of incubation on ice in the presence of 0.5% Triton X-100, an aliquot, corresponding to the total fraction, was collected. After centrifugation at 1500 g for 5 min at 4 °C, an aliquot of the supernatant, corresponding to the nuclei-depleted fraction, was collected. After washing with Honda buffer + 0.1% Triton X-100, the pellet was resuspended in Honda buffer, and this sample corresponded to the nuclei-enriched fraction.
RNA methods
RNA was extracted using a Nucleospin™ RNA Plus Kit (Macherey-Nagel) according to the manufacturer’s instructions and quantified with a Nanodrop spectrophotometer. Typically, 1 µg of total RNA was used to perform reverse transcription (RT) reactions, using SuperScript™ II Reverse Transcriptase (Invitrogen) and following the manufacturer’s instructions. Quantitative PCRs (qPCRs) were carried out with a LightCycler® 480 System (96 wells), using LightCycler® 480 SYBR Green I Master mix (Roche), and following the manufacturer’s standard instructions. ACT2 (AT3G18780) was used as an internal reference to calculate relative expression. Occasionally, SAND (AT2G28390) was used as an internal reference to verify that the results were not biased by the choice of ACT2. The primers used for qPCR are listed in Supplementary Table S1.
Results
Sustained MAPK activation is characteristic of RPS2/RPM1-mediated ETI responses, concerns mostly MPK3, and leads to nuclear accumulation of MPK3
Sustained MPK3/6 activities have been reported in response to pathogen effectors (Tsuda et al., 2013; Su et al., 2018; Wang et al., 2018). Yet, whether these activations represent a general feature of ETI remains controversial (Cui et al., 2017; Lang and Colcombet, 2020; Ngou et al., 2020). To get a better understanding of the question, we compared the pattern of MPK3/6 activities at late time points [5 hours post infection (hpi) and 8 hpi] in Col-0 plants before and after infiltration with various Pseudomonas syringae pv. tomato DC3000 strains (hereafter referred to as Pst), differing in their ability to stimulate plant immunity, or mock infiltration. The Pst WT strain expresses dozens of effectors but does not elicit a strong ETI response, in contrast to the Pst AvrRpt2, AvrRpm1, and AvrRps4 strains which express the eponymous effectors, while the Pst hrcC- strain is impaired in the effector translocation machinery and triggers only PTI responses. As shown in Fig. 1A, the activities of MPK3/6 were highest in response to Pst AvrRpt2 and AvrRpm1. We also observed a sustained, but notably weaker, activation in response to Pst AvrRps4. Finally, the samples infiltrated with Pst WT and hrcC- did not show an activation signal different from the mock-treated samples. Furthermore, we noticed that sustained MAPK activation concerned mostly MPK3 compared with MPK6 (Fig. 1A). In addition, sustained MPK3 activation was correlated with a concomitant increase in the amount of MPK3 proteins, whereas the amount of MPK6 remained globally unchanged (Fig 1A). However, this accumulation of MPK3 was not sufficient to explain the increase in activation, which was of considerably higher amplitude.
Sustained MAPK activation is characteristic of RPS2/RPM1-mediated ETI responses, concerns mostly MPK3, and leads to nuclear accumulation of MPK3. The activities and quantities of MPK3/6 were analyzed through immunoblotting with anti-pTpY, anti-MPK3, and anti-MPK6 antibodies in Col-0 in response to various Pst strains (A), in Col-0 and rps2rpm1 backgrounds in response to Pst AvrRpt2 and AvrRpm1 (B), and in the total fraction (TF), nuclei-depleted fraction (ND), and nuclei-enriched fraction (NE) from Col-0 plants in response to mock treatment or Pst AvrRpt2 (8 hpi) (C). Coomassie staining of blots served as a loading control (LC). pMPK3 and pMPK6 represent the phosphorylated active forms of MPK3 and MPK6. Immunoblottings with anti-H3 and anti-PEPC antibodies served as markers for the nuclear and cytoplasmic fractions, respectively. Values (means ±SD) indicate MPK3 fold change between mock and AvrRpt2 samples, and were calculated from three independent experiments using ImageJ software from non-saturated immunoblot pictures. All experiments were repeated at least three independent times, with similar results.
To confirm that the observed effects were really due to the recognition of the effectors, we compared the Col-0, rps2rpm1, and rps4-2 backgrounds. The CNLs RPS2 and RPM1 guard the RPM1-interacting protein 4 (RIN4) against modifications caused by AvrRpt2 and AvrRpm1, respectively (Belkhadir et al., 2004), while RPS4 contributes to the recognition of AvrRps4 (Saucet et al., 2015). Results indicated that both the activation of MPK3/6 and the accumulation of MPK3 were absent in rps2rpm1 in response to Pst AvrRpt2 and AvrRpm1 (Fig. 1B). In contrast, no clear differences could be detected between Col-0 and rps4-2 in response to Pst AvrRps4 (Supplementary Fig. S1), suggesting that the sustained MPK3/6 activation and MPK3 accumulation caused by this effector depends on other or additional receptors.
To determine whether sustained MPK3 activity and accumulation could affect its subcellular localization, we quantified the abundance of MPK3 protein in nuclear and cytoplasmic fractions. The results (Fig. 1C) show that both fractions contained more MPK3 in response to Pst AvrRpt2 infiltration than in response to mock infiltration, although the abundance of MPK3 was very low in the nuclear fraction relative to the cytoplasmic fraction. Moreover, the nuclear fraction was considerably more enriched (>10-fold) than the cytoplasmic fraction (~2-fold), clearly indicating that in response to Pst AvrRpt2, MPK3 accumulates in the nucleus.
NDR1 and EDS1 contribute to sustained but not transient MAPK activation
Both NDR1 and EDS1 are involved in the ETI signaling caused by AvrRpt2 recognition (Aarts et al., 1998; Day et al., 2006; Venugopal et al., 2009; Bhandari et al., 2019). To determine whether there is a link between these two regulators and sustained MPK3/6 activities, we measured MPK3/6 activities in the ndr1-1 and eds1-2 backgrounds. As shown in Fig. 2A, there was a significant decrease in MPK3/6 activities in the two mutants relative to Col-0, demonstrating that NDR1 and EDS1 act upstream of these MAPKs and contribute to their activation. Since we observed a concomitant decrease in the abundance of MPK3 protein, we quantified normalized blot signals from independent experiments. We concluded that the decrease in MPK3 activation in the two backgrounds is not due to the lower protein abundance, and also that the contribution of NDR1 to MPK3 activation is greater than that of EDS1 (Fig. 2B). Moreover, through nucleocytoplasmic fractioning, we found that the AvrRpt2-mediated nuclear enrichment of MPK3 was impaired in ndr1-1 and eds1-2 (Fig. 2C, D).
NDR1 and EDS1 contribute to sustained but not transient MAPK activation. (A) The activities and quantities of MPK3/6 were analyzed through immunoblotting with anti-pTpY, anti-MPK3, and anti-MPK6 antibodies in different genetic backgrounds in response to Pst AvrRpt2. (B) Ratio of MPK3 activity in the ndr1-1 and eds1-2 backgrounds relative to Col-0 in response to Pst AvrRpt2. Quantification was performed using ImageJ software from non-saturated immunoblot pictures of three independent replicates. Values of MPK3 activity were normalized by values of MPK3 abundance. The graph was drawn with the boxplot function in R. The box represents the 50% of the central data and the line represents the median. The error bars represent the range of the data. (C–F) Activities and quantities of MPK3/6 were analyzed through immunoblotting with anti-pTpY, anti-MPK3, and anti-MPK6 antibodies in response to mock or Pst AvrRpt2 (8 hpi) in the total fraction (TF), nuclei-depleted fraction (ND), and nuclei-enriched fraction (NE) from ndr1-1 (C) and eds1-2 (D), or in response to estradiol in the XVE-AvrRpt2 backgrounds (E), and in response to flg22 in Col-0, ndr1-1, and eds1-2 at intervals from 0 min to 60 min (F). Coomassie staining of blots served as a loading control (LC). pMPK3 and pMPK6 represent the phosphorylated active forms of MPK3 and MPK6. For (C) and (D), values (means ± half of the data range) indicate MPK3 fold change between mock and AvrRpt2 samples, and were calculated from two independent experiments using ImageJ software from non-saturated immunoblot pictures. Immunoblottings with anti-H3 and anti-PEPC antibodies served as markers for the nuclear and cytoplasmic fractions, respectively. All other experiments were repeated at least three independent times, with similar results.
To consolidate the roles of NDR1 and EDS1 upstream of sustained MPK3/6 activities, we crossed the ndr1-1 and eds1-2 lines with the XVE-AvrRpt2 line, which allows direct expression of the AvrRpt2 effector in the plant cell through an estradiol-inducible system (Tsuda et al., 2013) (Supplementary Fig. S2). Again, we could show that sustained MPK3/6 activities elicited by the expression of AvrRpt2 were significantly compromised in the absence of functional NDR1 and EDS1, with a higher contribution of NDR1 compared with EDS1 (Fig. 2E).
Since NDR1 and EDS1 are instrumental in the sustained activation of MPK3/6, we were curious to see whether they also contribute to the transient activation of MPK3/6. To test this, we infiltrated Col-0, ndr1-1, and eds1-2 leaves with the PAMP flg22 and quantified MPK3/6 activities at 15, 30, and 60 min post infiltration. However, in this experimental setup, we could not detect any obvious difference between the three genotypes (Fig. 2F). As prior to this we made sure that mock-treated plants do not display PAMP-unrelated MAPK activation (Supplementary Fig. S3), our results demonstrate that NDR1 and EDS1 are not involved in transient flg22-mediated MPK3/6 activation.
Disturbed MAPK activities result in resistance/susceptibility phenotypes
In an attempt to understand the impact of sustained MPK3/6 activities on plant defense responses and plant resistance to pathogens, we performed pathoassays with Pst AvrRpt2 in different plant backgrounds displaying modifications in the patterns of MPK3/6 activation. The K3CA-2 line is a gain-of-function line that expresses a mutated form of MPK3 under the control of the endogenous promoter. This line exhibits a higher basal level of MPK3 activity (Genot et al., 2017; Lang et al., 2017) and also a stronger sustained activation of MPK3 upon Pst AvrRpt2 infiltration compared with a K3WT-1 line expressing a WT form of MPK3 (Supplementary Fig. S4A). The single mutants mpk3-1 and mpk6-4 are defective in the respective MAPKs (Xu et al., 2008; Zhao et al., 2014), yet measurements of their sustained activities upon AvrRpt2 recognition revealed mild effects. As sustained MPK6 activation in response to Pst AvrRpt2 is weak, the mpk6-4 loss of function mutant shows little change in the strong MPK3 activation, while in mpk3-1, we observed a drastic increase in the level of MPK6 activation, which should compensate for the absence of MPK3 by an unknown mechanism (Supplementary Fig. S4B). Finally, the recently characterized mkk4-18/mkk5-18 line (hereafter referred to as mkk4mkk5) (Li et al., 2018) harbors a weak allele of MKK4 and a loss-of-function allele of MKK5, two genes coding for the MAP2Ks acting upstream of MPK3/6. The mkk4mkk5 line consistently shows a lower level of MPK3/6 activation in response to both Pst AvrRpt2 and flg22 (Supplementary Fig. S4C, D).
In line with their respective patterns of MAPK activation, the K3CA-2 line appears to be more resistant to Pst AvrRpt2 infiltration than WT controls, whereas the mkk4mkk5 line is more sensitive than Col-0, although it is not as sensitive as the ndr1-1 line (Fig. 3). In addition, we observed that the mpk3-1 (but not mpk6-4) and eds1-2 lines have a sensitivity that is intermediate between that of Col-0 and mkk4mkk5 (Fig. 3).
Disturbed MAPK activities result in resistance/susceptibility phenotypes. Plants were infiltrated with same amount of Pst AvrRpt2 and bacterial load was measured 3 d later. The dotplot and histogram showing mean values and SDs for different genotypes were generated from data of at least three independent replicates. Different letters above the bars indicate statistically significant differences (Kruskal–Wallis test followed by a non-parametric Tukey post-hoc test, P<0.05, 22<n<25). CFU, colony-forming units.
Overall, our data demonstrate that MPK3/6 activities are important for resistance phenotypes. They are also consistent with our findings that sustained MPK3/6 activities concern chiefly MPK3, and are dependent to different extents on NDR1 and EDS1. Yet, it must be borne in mind that the K3CA-2 and mkk4mkk5 lines we used are affected not only in the pattern of sustained MPK3/6 activation but also in the pattern of transient activation. We therefore could not rule out the possibility that the phenotypes we obtained for these lines are due to modifications not in sustained but in transient MPK3/6 activation. As a matter of fact, the K3CA-2 and mkk4mkk5 lines display similar resistance/susceptibility phenotypes in response to Pst AvrRps4 infiltration (Supplementary Fig. S5), even if this strain does not provoke a strong sustained MPK3/6 activation (Fig. 1A, Supplementary Fig. S1).
NLR and NLR signaling contribute to the SA sector of defense downstream of MPK3 activation
The finding that MPK3 accumulates in the nucleus in response to Pst AvrRpt2 prompted us to look at the genes whose expression is controlled by MPK3/6. In previous work, we established that the expression of K3CA-2 leads to the up-regulation of numerous NLR genes, and assumed that this up-regulation could be responsible for the SA-dependent autoimmune phenotype of K3CA-2 (Lang et al., 2017). This hypothesis was confirmed by the fact that mutation in the CNL SUMM2 partly rescues the K3CA-2 phenotype (Genot et al., 2017). To extend these findings, we crossed K3CA-2 with the ndr1-1 and eds1-2 lines and also with the snc1-11 line, which is impaired in the function of SNC1, a TNL that is up-regulated in the K3CA-2 transcriptome (Genot et al., 2017; Lang et al., 2017). As shown in Fig. 4A, the developmental phenotype of K3CA-2 is rescued partially by the snc1-11 and ndr1-1 mutations, and fully by the eds1-2 mutation. Next, we analyzed the expression levels of PR1, SID2, and PBS3, three different marker genes for the SA sector of defense. PR1 is a characteristic SA-responsive gene (Tsuda et al., 2013), while SID2 and PBS3 code for enzymes involved in the synthesis of SA (Huang et al., 2020). In agreement with the developmental phenotypes, we found that the expression of these genes in K3CA-2 is reduced mildly by the snc1-11 and ndr1-1 mutations, and drastically by the eds1-2 mutation (Fig. 4B). Remarkably, when we analyzed the MPK3 activity and abundance in the different lines, we observed that they remained largely the same in K3CA-2, K3CA-2/snc1-11, and K3CA-2/ndr1-1, but were considerably lower in K3CA-2/eds1-2 (Fig. 4C). Overall, these findings confirm that NLRs such as SNC1 and NLR signaling components such as NDR1 can act downstream of MPK3 activity to promote the SA sector of defense. The K3CA-2 suppressor approach also highlights, in a more obvious fashion than in response to AvrRpt2 recognition (Fig. 2A, B, E), the original regulatory role of EDS1, which appears to be essential to achieve and maintain sufficiently high levels of active MPK3 to trigger the SA pathway of defense. Incidentally, such a role, which places EDS1 upstream of MPK3/6, is also compatible with the idea that EDS1 could drive a positive feedback loop downstream of MPK3/6 to ensure their sustainable activation.
NLR and NLR signaling contribute to the SA sector of defense downstream of MPK3 activation. (A) Developmental reversion of the K3CA-2 phenotype by ndr1-1, snc1-11, and eds1-2. A representative image of 1.5-month-old plants (left) and fresh weight (FW) of 1.5-month-old plants (right) are presented. The graph was drawn with the boxplot function in R. The box represents the 50% of the central data and the line represents the median. The error bars represent the range of the data. Circles represent outlier data. Different letters above the boxes indicate statistically significant differences (Kruskal–Wallis test followed by a non-parametric post-hoc Tukey test, P<0.05, n=19). (B) RT–qPCR experiments showing the relative expression levels of SA-related PR1, PBS3, and SID2 genes in different genetic backgrounds in 1.5-month-old plants. Data represent the means ±SD calculated from three technical repeats. Different letters above the bars indicate statistically significant differences (Kruskal–Wallis test followed by a non-parametric post-hoc Tukey test, P<0.05, n=3). Experiments were repeated two independent times with similar results. (C) The activity and quantity of K3CA in different genetic backgrounds in 1.5-month-old plants were measured by anti-myc immunoprecipitation followed by a kinase assay, and immunoblotting with anti-myc antibodies. Coomassie staining of the blot served as a loading control (LC). Experiments were repeated at least two independent times, with similar results.
The NLRs AT3G04220 and AT4G11170 are up-regulated in both ETI and PTI in a manner that is dependent on MPK3/6, EDS1, and NDR1
To get a deeper understanding of the NLR gene up-regulation mediated by MPK3/6 activities, we first compared our 20 candidate NLR genes from the K3CA-2 transcriptome (Genot et al., 2017; Lang et al., 2017) with genes up-regulated by the conditional expression of constitutively active forms of AtMKK4 (MKK4DD) and its ortholog NtMKK2 (MKK2DD) (Tsuda et al., 2013; Su et al., 2018), and found 12 genes commonly up-regulated in the three conditions. As the transcriptomic analyses were performed with plants of different ages, shortly after the expression of MKK2DD and MKK4DD (6 h and 24 h, respectively), the significant overlap we observed (12/20) tends to attest that NLR up-regulation is not a pleiotropic effect of K3CA-2 but rather a direct consequence of MPK3/6 activation. Then, we compared the 12 NLR genes with a list of 55 NLR genes whose expression was found to be induced in response to flg22 (Yu et al., 2013) and with a list of genes that were up-regulated in the 24 h after infiltration with the Pst AvrRpt2 and AvrRpm1 strains (Mine et al., 2018). Based on these comparisons, we identified seven NLR genes common to all conditions. These seven NLRs comprise five TNLs and two CNLs (Table 1), and represent NLR genes that are likely regulated by MPK3/6 during both PTI and ETI.
List of NLR genes up-regulated by K3CA-2, MKK4/5, flg22, AvrRpt2, and AvrRpm1
| Gene | Protein description |
|---|---|
| AT1G12290 | CC-NBS-LRR class, localized at the plasma membrane |
| AT4G11170 | TIR-NBS-LRR class (Resistance Methylated Gene 1, RMG1) |
| AT3G04220 | TIR-NBS-LRR class |
| AT5G41750 | TIR-NBS-LRR class |
| AT1G66090 | TIR-NBS-LRR class |
| AT1G15890 | CC-NBS-LRR class |
| AT1G57630 | TIR-domain-containing protein |
| Gene | Protein description |
|---|---|
| AT1G12290 | CC-NBS-LRR class, localized at the plasma membrane |
| AT4G11170 | TIR-NBS-LRR class (Resistance Methylated Gene 1, RMG1) |
| AT3G04220 | TIR-NBS-LRR class |
| AT5G41750 | TIR-NBS-LRR class |
| AT1G66090 | TIR-NBS-LRR class |
| AT1G15890 | CC-NBS-LRR class |
| AT1G57630 | TIR-domain-containing protein |
NLR genes commonly up-regulated by K3CA-2 (Genot et al., 2017; Lang et al., 2017), NtMKK2DD (Su, et al., 2018), AtMKK4DD (Tsuda et al., 2013), flg22 treatment (Yu et al., 2013), and Pst AvrRpt2 and AvrRpm1 infiltrations (Mine et al., 2018). Descriptions were retrieved from TAIR (https://www.arabidopsis.org).
List of NLR genes up-regulated by K3CA-2, MKK4/5, flg22, AvrRpt2, and AvrRpm1
| Gene | Protein description |
|---|---|
| AT1G12290 | CC-NBS-LRR class, localized at the plasma membrane |
| AT4G11170 | TIR-NBS-LRR class (Resistance Methylated Gene 1, RMG1) |
| AT3G04220 | TIR-NBS-LRR class |
| AT5G41750 | TIR-NBS-LRR class |
| AT1G66090 | TIR-NBS-LRR class |
| AT1G15890 | CC-NBS-LRR class |
| AT1G57630 | TIR-domain-containing protein |
| Gene | Protein description |
|---|---|
| AT1G12290 | CC-NBS-LRR class, localized at the plasma membrane |
| AT4G11170 | TIR-NBS-LRR class (Resistance Methylated Gene 1, RMG1) |
| AT3G04220 | TIR-NBS-LRR class |
| AT5G41750 | TIR-NBS-LRR class |
| AT1G66090 | TIR-NBS-LRR class |
| AT1G15890 | CC-NBS-LRR class |
| AT1G57630 | TIR-domain-containing protein |
NLR genes commonly up-regulated by K3CA-2 (Genot et al., 2017; Lang et al., 2017), NtMKK2DD (Su, et al., 2018), AtMKK4DD (Tsuda et al., 2013), flg22 treatment (Yu et al., 2013), and Pst AvrRpt2 and AvrRpm1 infiltrations (Mine et al., 2018). Descriptions were retrieved from TAIR (https://www.arabidopsis.org).
Next, we compared the expression levels of the seven NLRs in response to flg22 (1 hpi and 3 hpi) and Pst AvrRpt2 (5 hpi and 8 hpi) in Col-0 and mkk4mkk5. The results indicate that these genes are indeed specifically up-regulated by the treatments, although we also noticed a mild effect of mock treatment for some genes. In addition, they globally confirm the positive effect of MKK4/5 on NLR up-regulation, even if the differences between the two genotypes were not always statistically significant (Fig. 5). Reasons for this finding could be that MPK3/6 activities are not totally abolished in mkk4mkk5 or that other signaling pathways converge towards NLR up-regulation and can compensate, to some extent, for MPK3/6 impairment.
Expression of NLR genes in response to flg22 and Pst AvrRpt2 is dependent on MPK3/6 activities. Relative expression levels were measured through RT–qPCR experiments in response to flg22, Pst AvrRpt2, and mock treatment at different time points. Mean values and SDs were calculated from four technical repeats. Different letters above the bars indicate statistically significant differences (Wilcoxon–Mann–Whitney test, P<0.05, n=4). n.s., not significant. The experiment was repeated two independent times, with similar results.
To further study the roles of MPK3/6 activities on NLR up-regulation, we performed an expression analysis for AT3G04220 and AT4G11170 (the two NLRs for which the contributions of MKK4/5 are the most obvious) in response to various treatments (infiltrations with Pst WT, AvrRpt2, AvrRpm1, AvrRps4, and hrcC-, and mock treatment) at 5 hpi and 8 hpi, corresponding to the time points at which Pst AvrRpt2 and AvrRpm1, unlike other strains, induce high sustained activation of MPK3/6. In parallel, we also measured the expression levels of the PR1 gene as an indicator of the plant defense responses dependent on the SA pathway. Results revealed that globally the three genes were up-regulated strongly by Pst AvrRpt2 and AvrRpm1, moderately by Pst AvrRps4, and weakly by Pst WT and hrcC-, at a level that was not distinguishable from the mock treatment (Fig. 6A). From this, we inferred that the differences in the induction of AT3G04220, AT4G11170, and PR1 might be mostly due to the differences in MPK3/6 activation, and also that the sustained MPK3/6 activities observed in response to Pst AvrRpt2, AvrRpm1, and AvrRps4 reinforce the transcriptional effects of the sole transient activities caused by Pst WT and hrcC-.
Up-regulation of AT3G04220, AT4G11170, and PR1 is dependent on MPK3/6, NDR1, and EDS1. Relative expression levels were measured at different time points through RT–qPCR experiments in Col-0 in response to various Pst strains (A), in different XVE-AvrRpt2 lines in response to estradiol (B), in Col-0 and various loss-of-function backgrounds in response to Pst AvrRpt2 infiltration (C), and in Col-0 and the same loss-of-function backgrounds in response to flg22 (D). Mean values and SDs were calculated from three technical repeats. Different letters above the bars indicate statistically significant differences (Kruskal–Wallis test followed by nonparametric post-hoc Tukey test, P<0.05, n=3). n.s., not significant. The experiments were repeated two independent times for (A) and at least three independent times for (B–D), with similar results.
To consolidate our interpretation, we investigated the expression levels of AT3G04220, AT4G11170, and PR1 in the XVE-AvrRpt2 lines. As shown in Fig. 6B, the three genes were strongly up-regulated upon estradiol treatment, while the induction of the three genes was compromised fully in the ndr1-1 background and partially in the eds1-2 background. Given the contributions of NDR1 and EDS1 to the activation of MPK3/6 in response to AvrRpt2 (Fig. 2), these results are consistent with the notion that MPK3/6 activities, NDR1, and EDS1 act in the same signaling pathway to promote the expression of AT3G04220, AT4G11170, and PR1 during AvrRpt2-triggered ETI.
Furthermore, we measured the transcript levels of AT3G04220, AT4G11170, and PR1 in response to Pst AvrRpt2 in the Col-0, mkk4mkk5, ndr1-1, and eds1-2 backgrounds, and could show that proper induction of the three genes requires functional MKK4/5, NDR1, and EDS1 (Fig. 6C). We also noticed differences in the contributions of these components to the induction of NLR and PR1, with that of MKK4/5 being weaker than that of EDS1, which in turn is weaker than that of NDR1. Considering that the impact of MKK4/5 on sustained MPK3/6 activities appears to be stronger than that of EDS1 (Fig. 2B, Supplementary Fig. S4), such differences were not entirely anticipated. An explanation could be, as mentioned previously, that other factors downstream of EDS1 and NDR1, acting additionally to or buffering the MPK3/6 pathway, are involved in the transcriptional regulation of the NLR and PR1 genes in response to Pst AvrRpt2. To confirm the role of sustained MPK3/6 activation, we also compared the induction of AT3G04220, AT4G11170, and PR1 in response to Pst AvrRpm1 in Col-0 and mkk4mkk5, and found that it was compromised in the mkk4mkk5 background, as it was in response to Pst AvrRpt2 (Supplementary Fig. S6).
Finally, we analyzed the expression of AT3G04220, AT4G11170, and PR1 in the Col-0, mkk4mkk5, ndr1-1, and eds1-2 backgrounds in response to the PAMP flg22 at 2 hpi and 8 hpi. Surprisingly, in this condition, we established that loss of function of not only MKK4/5 but also NDR1 and EDS1 impairs the up-regulation of the NLR genes as well as that of PR1 (Fig. 6D). Since EDS1 and NDR1 are not involved in the transient flg22-mediated MPK3/6 activation (Fig. 2F), these findings uncover an unexpected role for the two ETI regulators in some PTI responses independently or downstream of MPK3/6. Interestingly, we also found that in response to flg22, the expression levels of EDS1 and NDR1 are increased but that these increases are compromised in the mkk4mkk5 background compared with Col-0 (Fig. 7). These data further argue in favor of a model in which, during PTI, MPK3/6 activities contribute to the up-regulation of NLR genes upstream of EDS1 and NDR1.
EDS1 and NDR1 are up-regulated in response to flg22 in an MKK4/5-dependent manner. Relative expression levels were measured through RT–qPCR experiments at different time points in Col-0 and mkk4mkk5. Mean values and SDs were calculated from four technical repeats. Different letters above the bars indicate statistically significant differences (Wilcoxon–Mann–Whitney test, P<0.05, n=4). n.s., not significant. The experiments were repeated two independent times, with similar results.
Up-regulation of AT3G04220 is sufficient to activate the SA sector of defense
To determine what the effects of NLR up-regulation can be, we first performed some pathoassays in at3g04220 and at4g11170 lines compared with Col-0 in response to Pst WT. However, we could not detect any significant differences in the pathogen load between the different genotypes, although other studies have succeeded in showing that at4g11170 is more sensitive than WT (Halter et al., 2021).
As an alternative, we created two independent transgenic Arabidopsis lines expressing the coding sequence of AT3G04220 in the at3g04220 background, under the control of a dexamethasone-inducible promoter (DEX-AT3G04220/at3g04220 lines). RT–qPCR revealed that the two lines display leaky expression of the NLR but still specifically respond to dexamethasone treatment, with an induction of more than 10-fold (Fig. 8). Then, we measured the expression levels of the SA markers PR1 and PBS3 in response to dexamethasone or mock treatment in these transgenic two lines and the at3g04220 mutant. We observed that the induction of AT3G04220 is correlated with high induction of the two SA marker genes, in a manner that seemed to be dose- and time-dependent (Fig. 8). As these inductions were not observed in the at3g04220 background, these findings strongly indicate that control of the expression level of AT3G04220 is critical to modulate SA-related defense responses.
Up-regulation of AT3G04220 is sufficient to activate the SA sector of defense. Relative expression levels were measured through RT–qPCR experiments in two independent DEX-AT3G04220/at3g04220 lines and the at3g04220 line at different time points and in response to dexamethasone (+DEX) or mock treatment. Mean values and SDs were calculated from three technical repeats. The experiment was repeated three independent times, with similar results.
We then evaluated the SA sector of defense in Col-0, at3g04220, and at4g11170 lines in response to infiltration with Pst AvrRpt2 and flg22, through the expression levels of SID2, PBS3, PR1, and CBP60g. CBP60g is a transcription factor that acts as a master regulator of SA synthesis and signaling (Huang et al., 2020). However, we could not detect any significant differences between the three genotypes in these conditions (Supplementary Fig. S7). This result is not totally surprising. Indeed, as we showed in this study, several NLRs are up-regulated during PTI and ETI and, therefore, when one is missing, others might take over and secure an appropriate implementation of the SA sector of defense. Incidentally, the absence of a difference between Col-0, at3g04220, and at4g11170 is consistent with the fact that we did not observe a higher susceptibility to Pst WT in the single mutants compared with Col-0. Finally, a recent study revealed that ectopic expression of both AT3G04220 and AT4G11170 in tobacco leaves led to an important accumulation of SA and cell death, while simultaneous depletion of multiple NLR levels through overexpression of the E3 ligase SNIPER1 compromises SA-dependent PTI responses (Tian et al., 2021). Overall, these data further support our own results and conclusions.
Discussion
Regulation of sustained and transient MPK3/6 activations
One of the first results we obtained by comparing the effects of different Pst strains expressing different effectors is that sustained activations of MPK3/6 are characteristic of ETI mediated by the CNLs RPS2 and RPM1 (Fig. 1A, B). Indeed, although sustained activation could also be detected in response to Pst AvrRps4, this activation was notably weaker and cannot yet be clearly associated with an NLR receptor (Fig. 1A, Supplementary Fig. S1). These findings are consistent with previous reports (Tsuda et al., 2013; Cui et al., 2017; Su et al., 2018; Ngou et al., 2020). An explanation for this characteristic could be that sustained MPK3/6 activities are mostly mediated by CNLs. Another possibility could be that sustained MPK3/6 activities are a consequence of the recognition of effectors acting at the level of the cell membrane, as is the case for AvrRpt2 and AvrRpm1, which both target RIN4. In this model, the mechanisms of sustained MPK3/6 activations would be reminiscent of those allowing transient activations (Lang and Colcombet, 2020).
We also demonstrated that sustained MPK3/6 activations depend on NDR1 and EDS1 (Fig. 2). Because NDR1 is an integrin-like protein involved in the association between the plasma membrane and the cell wall, as well as a master regulator of the AvrRpt2- and AvrRpm1-triggered ETI (Knepper et al., 2011), it is not really surprising to find it upstream of sustained MPK3/6 activations. In AvrRpt2-triggered ETI, EDS1 is known to buffer the SA sector of defense (Venugopal et al., 2009; Bhandari et al., 2019), yet the underpinning molecular mechanisms remain enigmatic. Here, our results suggest that EDS1 might act upstream of sustained MPK3/6 activation to fulfill this function, which is actually in line with the fact that sustained MPK3/6 activities can also buffer the SA sector of defense (Tsuda et al., 2013).
Three additional points are worth mentioning with regard to the regulation of MPK3/6 activations. First, EDS1 and NDR1 are not required for transient MPK3/6 activations (Fig. 2F), indicating that if both transient and sustained MPK3/6 activations could originate at the cell membrane, there are some decisive discrepancies in the molecular mechanisms of these two phenomena. Second, the ndr1-1 mutation partially rescues the K3CA-2 phenotype (Fig. 4), strongly suggesting that the NLR signaling component NDR1 acts both upstream and downstream of sustained MPK3/6 activities. The possible functions of NDR1 downstream of MPK3/6 have yet to be uncovered. Finally, the eds1-2 mutation compromises the abundance and activity of K3CA-2 (Fig. 4C), hinting that EDS1 might be involved in a positive feedback regulation downstream of MPK3 activation allowing the high and sustainable accumulation of the active kinase. This could be in line with the model where sustained MPK3/6 activations are achieved through a regulatory loop dependent on systemic acquired resistance (Wang et al., 2018).
Functions of sustained and transient MPK3/6 activations
It is unclear whether sustained MPK3/6 activations can give rise to new functions compared with transient activations (Lang and Colcombet, 2020). Here, we provide evidence that NLR up-regulation controlled by MPK3/6 is not attributable to a specific pattern of activation, but that both transient and sustained MPK3/6 activations are able to up-regulate these genes (Figs 5, 6). Moreover, our results suggest that sustained activation caused by specific pathogen strains could reinforce the effects of transient activation on the expression of NLRs and PR1 (Fig. 6A). However, a full understanding of how the transition between transient MPK3/6 activation, elicited by PAMP perception, and sustained activation, elicited by effector recognition, converges towards NLR up-regulation, and thereby impacts the strength of the defense responses, is still missing. This is notably a reason why we could not draw conclusions about the resistance phenotypes of lines affected at the same time in the transient and sustained patterns of MPK3/6 activities (Fig. 3, Supplementary Figs S4, S5).
If MPK3/6 activities positively control the up-regulation of some NLR genes, the mechanisms underlying these processes are currently unknown. Interestingly, the induction of AT3G04220 and AT4G11170 in response to flg22 has been shown to be regulated by promoter DNA methylation and the actions of WRKY transcription factors (Yu et al., 2013; Halter et al., 2021). Since the functions of WRKY transcription factors are known to be modulated by MPK3/6, either as direct substrates or through the actions of VQ-domain-containing proteins (Weyhe et al., 2014), further investigations into the links between these different actors would seem promising. An alternative could be that MPK3/6 activities inhibit the nonsense-mediated mRNA decay pathway, which plays an important role in the regulation of NLR expression levels (Jung et al., 2020).
Our results (Fig. 6C, D) also leave room for the notion that factors other than MPK3/6 are instrumental in the up-regulation of the NLRs. For instance, Tian et al. (2021) showed that TNL inductions in PTI are compromised by inhibitors of calcium signaling, suggesting that CDPKs might also be involved in these processes. The specificities of this pathway compared with the MAPK pathway remain to be elucidated.
NLR up-regulation: crosstalk between PTI and ETI
Up-regulation of NLR genes in a PTI context has previously been documented (Yu et al., 2013). Nevertheless, a comprehensive analysis of the regulation and consequences of this phenomenon is still lacking. Here, we demonstrated that MPK3/6, as well as the NLR signaling components EDS1 and NDR1, contribute to the up-regulation of two NLR genes, AT3G04220 and AT4G11170, upon both PAMP and effector treatments (Fig. 6). We further showed that up-regulation of AT3G04220 is sufficient to activate the SA sector of defense (Fig. 8). Altogether, these results support the model presented in Fig. 9. In this model, we propose that the transcriptional up-regulation of some NLR genes would result in higher protein levels, which, once above a certain threshold, would be sufficient to trigger auto-activation, irrespective of pathogen effectors. Thereby, MPK3/6 activities would bridge PTI and ETI by regulating NLR expression levels, and allowing modulation or ‘priming’ of NLR activation in response to not only effectors but also PAMPs. As a consequence, these NLRs should play a critical role during PTI. Remarkably, a concomitant and independent study obtained similar results and came to similar conclusions (Tian et al., 2021). By revealing that TNL accumulation as well as several TNL signaling components, including EDS1, are required to mount an efficient SA-dependent PTI response, the authors of this study pinpointed the same crosstalk between PTI and ETI as we did. However, they interpreted the need for TNL signaling components as an event downstream of TNL activation, whereas our own data indicate that the NLR signaling components NDR1 and EDS1 might contribute to NLR activation through NLR up-regulation.
PTI and ETI responses are intricately interrelated around the induction of NLR expression in a process involving MPK3/6 activation and the ETI signaling components EDS1 and NDR1. In the PTI context, upon PAMP perception by pattern-recognition receptors (PRR) and signal transduction by receptor-like cytoplasmic kinases (RLCK), NLR up-regulation is modulated by transient MPK3/6 activities, NDR1, and EDS1, with the latter two acting downstream or independently of the former. In the ETI context, the effectors AvrRpt2 and AvrRpm1, injected by the bacteria through a type III secretory system (T3SS) and targeting RIN4, are recognized by the CNLs RPS2 and RPM1, thereby triggering convergent signaling pathways involving NDR1 and EDS1, as well as nucleus-accumulating and sustainably active MPK3/6, to further promote NLR expression. Contrary to what happens in PTI, in ETI, NDR1 and EDS1 act upstream of sustained MPK3/6 activities. The recognition of other effectors, such as AvrRps4, can also induce NLR expression but to a lesser extent, through the activation of only the EDS1 signaling branch. A positive feedback loop downstream of MPK3/6 and requiring EDS1 contributes to the sustained activation of MPK3/6. Finally, NLR transcriptional up-regulation leads to the implementation of the SA sector of defense, through an NLR activation process that remains to be determined.
Overall, our findings bring new perspectives to the emerging model of the plant immune system in which defense responses are extensively and dynamically modulated by diverse interactions between PTI and ETI (Lu and Tsuda, 2021; Yuan et al., 2021b). They also raise important questions that await answers. One such question that seems to be of utmost interest is related to the way components of NLR signaling promote the up-regulation of NLR genes during PTI. As elements of answers, our results indicate that EDS1 and NDR1 do not affect transient MPK3/6 activations during PTI but that their expression levels are increased in an MPK3/6-dependent manner (Figs 2F, 7), suggesting that in this context, and contrary to ETI, MPK3/6 contributes to NLR up-regulation upstream of EDS1 and NDR1 (Fig. 9). However, we cannot totally exclude the possibility that EDS1 and NDR1 might act on NLR up-regulation independently of MPK3/6 activities through a process that would be reinforced by the promotion of EDS1 and NDR1 expression. Interestingly, a recent report demonstrated that EDS1 could interact with some LRR-RPs as well as some RLCKs, and contribute to PAMP-triggered ethylene accumulation, thereby offering an explanatory framework, based on physical mechanisms, for the link between ETI components and PTI responses (Pruitt et al., 2021). In the same order of idea, it was shown that NDR1 can interact with RIN4 (Day et al., 2006), which can interact with RPS2 (Mackey et al., 2003), which in turn can interact with FLS2, the LRR-RK responsible for perceiving flg22 (Qi et al., 2011). Taken together, these data outline a complex membrane-associated signalosome machinery that involves components of both PTI and ETI (Lu and Tsuda, 2021; Dongus and Parker, 2021). Unravelling the complexity of this machinery would constitute a great step forward for our understanding of plant immune responses.
Supplementary data
The following supplementary material are available at JXB online.
Table S1. Sequences of primers used for qPCR experiments.
Fig. S1. Sustained MAPK activation is characteristic of RPS2/RPM1-mediated ETI responses.
Fig. S2. AvrRpt2 induction in the XVE-AvrRpt2 lines.
Fig. S3. flg22-mediated versus mock-mediated MPK3/6 activation.
Fig. S4. Misregulation of MPK3/6 activities in K3CA-2, mpk3-1, mpk6-4, and mkk4mkk5 lines.
Fig. S5. Resistance phenotypes against the Pst AvrRps4 strain.
Fig. S6. Up-regulation of AT3G04220, AT4G11170, and PR1 in response to Pst AvrRpm1 infiltration in Col-0 and mkk4mkk5.
Fig. S7: Expression analysis of the SA sector of defense in Col-0, at3g04220, and at4g11170.
Acknowledgements
The authors would like to thank Dr Kenishi Tsuda and Dr Wei Zhang for sharing plant and bacterial materials.
Author contributions
JL, with the assistance of JC, designed the experimental setups and processed the data. JL, BG, and JB performed the experiments. JL wrote the manuscript with contributions from JB and JC.
Conflict of interest
The authors declare no conflict of interest.
Funding
This study was supported by the French Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE).
Data availability
The data concerning the biological replicates are available from the corresponding author, Julien Lang, upon request. All other data supporting the findings of this study are available within the paper and within its supplementary materials published online.
Comments