Comparisons of two receptor-MAPK pathways in a single cell-type reveal mechanisms of signalling specificity

Ma, Yan; Flückiger, Isabelle; Nicolet, Jade; Pang, Jia; Dickinson, Joe B.; De Bellis, Damien; Emonet, Aurélia; Fujita, Satoshi; Geldner, Niko

doi:10.1038/s41477-024-01768-y

Download PDF

Article
Open access
Published: 10 September 2024

Comparisons of two receptor-MAPK pathways in a single cell-type reveal mechanisms of signalling specificity

Nature Plants volume 10, pages 1343–1362 (2024)Cite this article

14k Accesses
8 Citations
60 Altmetric
Metrics details

Subjects

Abstract

Cells harbour numerous receptor pathways to respond to diverse stimuli, yet often share common downstream signalling components. Mitogen-activated protein kinase (MPK) cascades are an example of such common hubs in eukaryotes. How such common hubs faithfully transduce distinct signals within the same cell-type is insufficiently understood, yet of fundamental importance for signal integration and processing in plants. We engineered a unique genetic background allowing direct comparisons of a developmental and an immunity pathway in one cell-type, the Arabidopsis root endodermis. We demonstrate that the two pathways maintain distinct functional and transcriptional outputs despite common MPK activity patterns. Nevertheless, activation of different MPK kinases and MPK classes led to distinct functional readouts, matching observed pathway-specific readouts. On the basis of our comprehensive analysis of core MPK signalling elements, we propose that combinatorial activation within the MPK cascade determines the differential regulation of an endodermal master transcription factor, MYB36, that drives pathway-specific gene activation.

Mycoparasitism related targets of Tmk1 indicate stimulating regulatory functions of this MAP kinase in Trichoderma atroviride

Article Open access 15 November 2023

Mechanosensory trichome cells evoke a mechanical stimuli–induced immune response in Arabidopsis thaliana

Article Open access 08 March 2022

Targeting the RAS/RAF/MAPK pathway for cancer therapy: from mechanism to clinical studies

Article Open access 18 December 2023

Main

Multicellular organisms rely on individual cells to detect different stimuli and translate them into various biological responses. Correspondence must be maintained between an extracellular input and its specific cellular response output. Decades of research established chains of molecular events from ligand perception by cell surface receptors to transcriptional outputs in the nucleus. Yet a key question remains: how is pathway-specific information maintained all the way to the nucleus, when multiple receptor pathways are stimulated and converge on a limited number of common signalling intermediates? The mitogen-activated protein kinase (MPK) cascade is a key intermediate module, responding to numerous stimuli in plants, animals and fungi, leading to a diverse array of outputs^1,2,3,4. Understanding the mechanisms underlying specificity in such an hourglass architecture presents a substantial challenge.

In plants, this ‘hourglass problem’ is amplified due to their cell types being more broadly competent, compared with highly specialized mammalian cells. Arabidopsis, for instance, displays over 200 leucine-rich-repeat receptor-like kinases (LRR-RLKs)^5,6, dwarfing the 11 Toll-like receptors found in humans⁷, both representing a subset of their total receptor complement. At the same time, Arabidopsis has much less cell types than any mammal^8,9, enhancing the potential complexity of receptor pathways on a per-cell basis. In plants, a prevailing notion is that common pathway components mediate specific outputs by activating different sets of transcription factors, expressed in different cell types. Yet, this would imply that a single cell-type would accommodate only one receptor/ligand pathway, which is implausible for plant cells. Therefore, it is crucial to understand how different receptors in a single plant cell maintain signalling specificity. However, such studies are scarce due to the challenges of isolating and analysing signalling events at single-cell resolution, as well as the lack of precise cellular readouts.

To address these challenges, we leveraged the ordered, linear differentiation of transparent Arabidopsis roots, allowing direct observation of single cells within an intact organism. Within this context, we established the Arabidopsis root endodermis as a cellular model to compare two well-studied LRR-RLK receptor pathways at single-cell level (Fig. 1a). One pathway uses the FLAGELLIN SENSING2 (FLS2) immune receptor, recognizing the microbial pattern flg22 (refs. ^10,11). The other pathway uses the SCHENGEN3 (SGN3, also called GASSHO1 (GSO1)) receptor and its endogenous Casparian Strip Integrity Factor 1/2 (CIF1/2) ligands^12,13. The endodermis is a specialized cell layer found in all vascular plants, forming a ring-like Casparian strip (CS), a highly localized membrane–cell wall nexus (Fig. 1d) analogous to tight junctions in animal intestinal epithelia¹⁴. This CS is composed of a lignified cell wall, creating an extracellular diffusion barrier in roots^15,16. SGN3 enables the plant to detect developmental or stress-induced discontinuities in its CS, manifested as CIF peptide leakage from inner cell layers^12,17. During endodermal differentiation, the SGN3 pathway drives CS maturation, characterized by fusion of its membrane domains. In situations of chronic or strong stimulation, the SGN3 pathway additionally induces ectopic lignification of cell corners and excess suberization (Fig. 1a)^18,19,20. These precise and distinctive subcellular features of the SGN3 pathway were used as unambiguous and quantifiable signalling readouts.

**Fig. 1: Endodermis-specific FLS2 induces ectopic lignin but cannot replace endogenous SGN3 for Casparian strip domain fusion.**

The SGN3 and FLS2 pathways are regarded to have markedly different functional outputs (a precise developmental programme versus a broad bacterial immune response), yet they display striking similarities in the downstream signalling components. Both receptors use the same small family of co-receptors to bind their respective ligands, flg22 and CIF2 (refs. ^21,22,23). Both require members of the same small family of SERK co-receptors^23,24 and subsequently activate members of the receptor-like cytoplasmic kinase VII family^20,25, which triggers a series of common intracellular signalling events, including reactive oxygen species (ROS) production and activation of MPKs^{20,26,27,28,29}. How are different outputs achieved when pathways share such extensive similarities? Narrowing the broad expression of FLS2 into the endodermal SGN3 domain sharpens the question of signalling specificity in plants, because it finally allows us to address what happens when two pathways are stimulated in a single cell-type.

Using our tailored genetic background, we provide clear evidence for the maintenance of signalling specificity in a single cell despite strong pathway activation. We identify the transcription factor MYB36, a master regulator of endodermal differentiation^19,30, as a key hub driving pathway-specific outputs in the nucleus. Through extensive, cell-type-specific loss and gain-of-function studies of MPKs and MPK kinases (MKKs), we unveil their distinct capacities for transducing both common and specific branches of the two signalling pathways. We demonstrate that so-called ‘common intermediates’, such as MPK cascades, can produce a wide range of outputs through combinatorial activation of common and specific components.

Endodermis-specific FLS2 cannot replace endogenous SGN3 for CS domain fusion

To allow stimulation and direct comparison of two distinct pathways in the differentiating root endodermis (Fig. 1a), we expressed FLS2 under the endodermis-specific SGN3 promoter (pSGN3::FLS2-GFP) in an fls2 mutant background (Fig. 1b). We then knocked out CIF1 and CIF2 genes using CRISPR-Cas9, ensuring that SGN3 signalling only occurs upon external peptide stimulation. We termed the resulting plant line (pSGN3::FLS2-GFP fls2 cif1 cif2) as ‘2-in-Endo’. CIF2 peptide treatments on the cif1 cif2 mutant can efficiently restore the SGN3 pathway function and re-establish a continuous Casparian strip^20,23; flg22 peptide treatments rapidly activate immune signalling in the endodermis from both native³¹ and endodermis-expressed FLS2 in the root^31,32. Using this line, we investigated whether FLS2 could functionally replace SGN3.

SGN3 pathway stimulation causes growth and fusion of the Casparian strip domains (CSD) from aligned islands (called ‘string-of-pearls’ stage) into a continuous belt (Fig. 1d)^12,13,33. Casparian strip domain protein 1 (CASP1) marks this domain and predicts lignin deposition at the CS^26,34. We therefore introduced pCASP1::CASP1-GFP (CASP1-GFP) into 2-in-Endo. We confirmed that CIF2 treatment triggers full connection of CASP1 domains and corresponding CS lignin, leaving no gaps at a stage where the CS is fully mature in wild-type roots (Fig. 1c,d and Supplementary Video 1). In contrast, treatments with flg22 cannot improve CS domain fusion, showing comparable percentage of gap areas to controls (Fig. 1e). This was confirmed by visualizing CASP1-GFP from initiation to maturation (Extended Data Fig. 1a) and by quantifying discontinuity in overviews (Extended Data Fig. 1b,c).

Importantly, this distinct outcome is not due to a lack of FLS2 functionality, as flg22 treatment induces similar ROS production and ectopic lignification as CIF2 treatment (Fig. 1c and Extended Data Fig. 1f). These additional SGN3 pathway outcomes are also observed upon strong CIF2 stimulation^12,20. A third outcome of strong CIF2 stimulation is precocious (nearer to the root tip) suberin deposition, also mimicked by flg22 treatment in 2-in-Endo (Fig. 1a and Extended Data Fig. 1d). Therefore, endodermis-specific FLS2 signalling is able to replicate some outcomes of SGN3, but is unable to mediate the crucial aspect of CASP and CS domain growth and fusion. Even with higher FLS2 receptor levels in an sgn3 background (pCASP1::FLS2-GFP fls2 sgn3, Fig. 1b), FLS2 stimulation cannot induce CASP domain fusion (Extended Data Fig. 1g,h).

This suggests that CS domain fusion is insensitive to FLS2 signalling, clearly indicating the presence of a main SGN3 signalling branch that FLS2 stimulation cannot induce.

Blocking of the propidium iodide (PI) fluorescent tracer provides a sensitive assay to measure CS functionality—it quantifies onset of an operational extracellular diffusion barrier by scoring PI exclusion from the inner vascular tissues (Extended Data Fig. 1e)^15,16. In wild-type roots, blocking of PI uptake coincides with the formation of a continuous CS, assured by a functional SGN3 pathway (Fig. 1f). A delayed PI exclusion can still occur in the absence of a CS through SGN pathway-induced compensatory lignification at cell corners. This is observed in myb36, a CS-lacking mutant, but not in myb36 cif1 cif2 (Fig. 1f). Note that the PI block in myb36 is achieved purely by compensatory lignin, as the mutant lacks any CS lignin (Fig. 3b)¹⁹. This blockage is typically delayed by ~20 cells compared with wild-type plants with a CS (Fig. 1f).

We found that absence of a diffusion barrier in 2-in-Endo (H₂O treatment) is fully rescued by CIF2 treatment, restoring a wild-type diffusion barrier (Fig. 1f and Extended Data Fig. 1e). By contrast, flg22 treatments activate delayed barrier formation in 2-in-Endo, similar to myb36, through ectopic lignin formation (Fig. 1f). In the overexpressing pCASP1::FLS2 line, the diffusion barrier is established earlier by flg22 (Extended Data Fig. 1i), probably due to stronger ectopic lignification (Extended Data Fig. 1g). Nevertheless, in both cases, barriers are still significantly delayed compared with CIF2-stimulated 2-in-Endo or wild-type lines.

Thus, despite being able to induce lignification and suberization in the endodermis, FLS2 signalling cannot functionally replace SGN3 for the growth and fusion of CASP1 and CS domains, indicating that signalling specificity between two similar receptor pathways is maintained within the same cell-type.

Endodermis-expressed FLS2 and SGN3 activate both common and pathway-specific transcriptional responses

To compare transcriptional outputs of FLS2 and SGN3 pathways solely within the endodermis, we conducted comparative RNA-seq analysis of pCASP1::FLS2 fls2 and wild-type seedling roots respectively treated with flg22 and CIF2 at three timepoints (30, 120, 480 min) (Extended Data Fig. 2a, CIF2 responses data from ref. ²⁰). The receptor mutant controls sgn3 and fls2, and displayed minimal gene responses to respective peptides (Fig. 2a). We found a substantial overlap between the SGN3 and the FLS2 pathway, with 469 out of 1,262 differentially expressed (DE) genes in common (Fig. 2a). The most enriched Gene Ontology (GO) categories of both pathways show a high degree of overlap, reflecting general biotic and abiotic stress responses (Fig. 2b). The shared functional outcomes that we observed are reflected in the transcriptional profiles: genes involved in lignin and suberin biosynthesis show matching induction patterns after CIF2 or flg22 treatment in the respective backgrounds (Fig. 2c,d). This commonality cannot be simply attributed to shared cell identity (the endodermis), as a cross analysis of highly flg22-responsive genes in whole wild-type seedlings³⁵ reveals mirrored induction patterns in the endodermis (Extended Data Fig. 2b).

**Fig. 2: Endodermis-expressed FLS2 and SGN3 trigger common and pathway-specific transcriptional responses.**

Despite this overlap, we found 158 and 635 DE genes and GO terms specifically associated with SGN3 and FLS2 pathways, respectively (Fig. 2a and Extended Data Fig. 2d). The greater number of DE genes in the FLS2 pathway could arise from the stronger endodermal CASP1 promoter driving FLS2, compared with the native SGN3 promoter. To validate pathway-specific response genes (Extended Data Fig. 2c), we generated transcriptional fluorescent reporters in 2-in-Endo and found that the majority of reporter lines match the RNA-seq predictions (Extended Data Fig. 2c and Supplementary Table 1). For example, MILDEW RESISTANCE LOCUS O 3 (MLO3) shows a much stronger transcriptional response to CIF2 than to flg22 (Fig. 2f). Other SGN preferential reporter genes include PRK1, LAC3 and MCA2 (Supplementary Table 1 and Video 2). In contrast, the cytochrome P450 family gene CYP715A1 is an FLS2-specific reporter that responds strongly to flg22 but is non-responsive to CIF2 (Fig. 2f), with other examples including WRKY30, LAC1 and PER52 (Supplementary Table 1 and Video 3). Some reporters respond equally well to either flg22 or CIF2 in 2-in-Endo, with many showing non-cell-autonomous responses in surrounding root cell-types (for example, WRKY41, PER71 and ERF105) (Supplementary Table 1).

Intriguingly, our dataset revealed that the top 10% most responsive genes (by fold change) in early timepoints are common to both pathways, whereas genes responding more weakly (but significantly) exhibited more pathway specificity (Extended Data Fig. 2e). We speculate that ‘fast and strong’ responses are often regulated by the common signalling modules of the two pathways.

The transcription factor MYB36 controls SGN3-specific transcriptional responses

Our transcriptional analyses suggested that the transcription factor MYB36, reported to be a key regulator of endodermis differentiation^19,30, also acts as a hub to regulate SGN3-specific responses. We found that genes regulated by MYB36 (downregulated in myb36 compared with wild type³⁰) tend to respond specifically to CIF2 (Fig. 2e). While the scale of induction by CIF2 is relatively weak (log₂ FC < 2), many of these genes are specifically expressed in the endodermis and are known to contribute to CS integrity. For example, expression of CASP genes (CASP1/2/3/4) is crucial for establishing a functional CSD^34,36, while peroxidases such as PER9 and PER64 mediate CS lignification³⁷, and the dirigent protein ESB1 is essential for both CSD formation and lignification³⁸. These genes might therefore carry out SGN3-specific functions and their non-responsiveness to flg22 can straightforwardly explain the inability of flg22 to promote CS domain fusion.

CIF2-induced CASP domain fusion requires de novo protein synthesis²⁰. We hypothesize that MYB36-mediated transcriptional responses upon SGN3 stimulation enable CASP domain fusion, while the FLS2 pathway is incapable of activating MYB36. Consequently, the SGN3 pathway would function independently of MYB36 for ectopic lignification but require MYB36 for CASP and CS domain fusion. MYB36 remains partially functional in sgn3 and cif1 cif2, but only mediates formation of aligned CASP domain islands. The role of SGN3 would therefore be to drive domain fusion by boosting the activity of MYB36.

Predicted MPK phospho-sites in MYB36 are important for CS domain fusion

MYB36 was shown to be phosphorylated by specific MPKs (MPKs 2, 3 and 16) in an in vitro protein microarray³⁹. Three MPK phosphorylation motifs (serine proline) are predicted in MYB36 (Extended Data Fig. 3a). To test the potential necessity and sufficiency of MYB36 phosphorylation at these sites for CS domain fusion, we mutated all three serines to generate a phospho-null variant, MYB36^AAA and a phosphomimic variant, MYB36^DDD.

The untagged MYB36^AAA, introduced in myb36 mutant, can restore CASP1 expression and accumulation in the endodermis, but often failed to complement CASP1 domain fusion (Fig. 3a and Supplementary Table 2). This suggests that MYB36^AAA, while active, is less functional than either the wild-type MYB36^WT or the phosphomimic MYB36^DDD, both of which fully complement CASP1 domain fusion (Fig. 3a and Supplementary Table 2). The fusion defects exhibited by MYB36^AAA can be described as a band of ‘speckles’ roughly aligned with the endodermis central plane (Fig. 3b). These CASP1 speckles consist of many smaller disconnected patches, in comparison with the larger, more connected domains caused by the loss of SGN3 signalling in sgn3 or cif1 cif2 (Fig. 3a–c and Extended Data Fig. 3c). As expected, the ability to fuse CASP1 domains correlates with the formation of a continuous CS lignin (Fig. 3b) and early formation of a diffusion barrier (Fig. 3d).

**Fig. 3: Predicted phospho-sites of MYB36 important for SGN3-dependent CS domain fusion.**

Importantly, we found that the domain fusion defect in myb36 MYB36^AAA is resistant to external CIF2 application (Fig. 3b,c and Extended Data Fig. 3c,d), suggesting that CIF2-induced domain fusion is impaired by the absence of functional phospho-sites in MYB36 (Fig. 3f middle). By contrast, CIF2-induced ectopic lignification is unaffected by MYB36^AAA, since ectopic lignin in myb36 MYB36^AAA is further enhanced by external CIF2 application (Fig. 3b). Accordingly, myb36 MYB36^AAA has a delayed PI block, similar to untransformed myb36 (Fig. 3d).

Our data strongly suggest that the SGN3 pathway operates through multiple branches, one requiring MYB36 activation for CASP domain fusion, probably via MPK phosphorylation, and another triggering compensatory ectopic lignification independent of MYB36 (Fig. 3f left). Specificity could thus arise from the FLS2 pathway’s inability to activate MYB36.

We also found that different C-terminal tags affect the functionality of MYB36^AAA, with large tags such as Turbo-GFP (847 aa) and mCherry (406 aa) apparently enhancing functionality of all MYB36 variants compared with their untagged version (Fig. 3d and Supplementary Table 2), possibly due to enhanced stability. Nevertheless, despite similar accumulation levels, MYB36^AAA-mCherry was still unable to fully complement CASP domain fusion in myb36, unlike the other variants (Extended Data Fig. 3b).

The phosphomimic MYB36^DDD could fully complement CASP domain fusion and CS barrier function in myb36, comparable to MYB36^WT (Fig. 3a–d). Interestingly, the untagged MYB36^DDD shows gain-of-function activity, improving CASP domain fusion in sgn3, whereas MYB36^AAA could not (Fig. 3e,f right). This further supports the idea that the predicted MPK phospho-sites in MYB36 are essential for CASP domain fusion induced by the SGN3 pathway. Furthermore, MYB36 protein accumulation and stability appear to be important determinants of its activity within the SGN3 pathway.

Multiple MPKs contribute to CS integrity in the SGN3 pathway

MPKs are direct, key regulators of transcription factors, and our data indicate their potential in directly regulating MYB36 activity. Yet, several MPKs are reported to be activated by both SGN3 (ref. ²⁰) and FLS2 pathways^27,40. In 2-in-Endo seedling roots, we confirmed that both pathways can activate MPK3 and MPK6 in the endodermis upon flg22 or CIF2 treatments (Fig. 4a). To obtain a more global picture of MPK activities in the endodermis, we generated an expression atlas for all 20 MPKs in Arabidopsis (Supplementary Table 3) by generating and analysing individual transcriptional fluorescent marker lines. Overall, we observed broad expression patterns across different tissues (Supplementary Table 3). We found that at least 12 MPKs are expressed in the endodermis, hence a subset could easily transduce specific signals downstream of SGN3 and FLS2 (Extended Data Fig. 4b,c). Integrating our atlas with RNA-seq³⁰ and translating ribosomal affinity purification (TRAP) data enriched in the endodermal cell file (Extended Data Fig. 4a), we homed in on 8 candidates (MPK2/3/6/9/15/16/17/19). However, none of the tested mpk single mutants show strong defects in CS barrier function or CASP1 domain fusion (Fig. 4b and Extended Data Fig. 5a). We nevertheless observed a slight delay in forming a diffusion barrier in mpk2 (Fig. 4b), fitting with the in vitro phosphorylation data of MYB36 being a substrate for MPK2, 3 and 16 (ref. ³⁹). However, no obvious discontinuities persist at maturity in mpk2 CASP1-GFP (Extended Data Fig. 5a). Moreover, neither T-DNA nor CRISPR mpk2 mpk16 double mutants showed a significant barrier diffusion or CASP1-GFP defect (Fig. 4b and Extended Data Fig. 5a,c), suggesting that higher-order mutants will be required for observing strong and reliable CS formation phenotypes.

Because MPKs are involved in many processes, dissecting their function without pleiotropic effects in higher-order mutants is challenging. To circumvent this problem, several alternative tools were tested. First, we expressed HopAI1, a Pseudomonas syringae effector that specifically deactivates multiple MPKs, including MPK3/4/6 (refs. ^41,42,43), under an endodermis-specific estradiol (EST) inducible promoter (pLTPG15::XVE>>pLexA::HopAI1, hereafter indHopAI1). HopAI1 induction in wild type leads to delayed barrier formation and some ectopic lignin, suggesting that MPK activities are required for CS functionality. No phenotype is observed using an inactive HopAI1^H103A mutant (Fig. 4c and Extended Data Fig. 4d). However, at our confocal resolution, HopAI1 induction does not seem to affect CASP domain fusion (Extended Data Fig. 4d).

Since mpk3 mpk6 double mutants are embryo lethal⁴⁴, we used conditional loss-of-function lines expressing pMPK3::MPK3^TG or pMPK6::MPK6^YG in an mpk3 mpk6 background^45,46. MPK3^TG and MPK6^YG are specifically sensitive to NM-PP1, a bulky analogue of the general kinase inhibitor PP1. Indeed, 1-day inhibitor treatment of these lines delayed the PI block without causing other developmental defects (Fig. 4d).

We also used a gain-of-function approach to investigate whether constitutively active MPK6a and MPK2a can alleviate CS defects in sgn3. We found that expressing MPK6a or MPK2a under an inducible endodermis-specific promoter (ind) in sgn3 reduces CASP1 domain discontinuities at maturity (Fig. 4e). Interestingly, only induction of MPK6a, but not MPK2a, partially restores the diffusion barrier in sgn3 (Fig. 4f and Extended Data Fig. 5b). This correlates with MPK6a activating ectopic ROS and ectopic lignin and suberin formation in sgn3, in contrast to MPK2a (Fig. 4g, and Extended Data Figs. 4e,f, and 7d). MPK6a thus appears to have a broader activity, affecting specific and common branches of the SGN3 pathway, while MPK2a contributes more specifically to CASP domain fusion.

Since MPK6a is sufficient to induce ectopic lignin in sgn3, we investigated whether MPK6 is necessary for SGN3-dependent ectopic lignification by testing whether CIF2-induced ectopic lignin is affected in mpk6. We found that 1-day CIF2 treatment could induce similarly strong ectopic lignin in mpk6 and wild type, as well as in mpk3 (Extended Data Fig. 4g). This suggests that genes in addition to MPK6 must contribute to ectopic lignification downstream of the SGN3 pathway. Despite MPK2a and MPK6a contributing to domain growth and fusion, neither alone is sufficient to complete CS fusion. Therefore, a combination of different MPK activities appears to contribute to the three-branched outputs of the SGN3 pathway, some unique and some overlapping with the FLS2 pathway.

Activation of individual MKKs in the endodermis produces distinct outputs

Arabidopsis contains 80 MAPKKKs and 20 MAPKs, but only 10 MKKs, the direct upstream regulators of MPKs^47,48. MKKs thus represent the narrowest part of the hourglass and the most critical step for information transfer. MKKs tend to selectively activate certain MPK combinations^39,49, suggesting that the activation of different MKKs may translate into distinct combinations of MPK activities, depending on the specific pathway triggered.

Given the relatively small number of MKKs, it is feasible to conduct a comprehensive examination of their activities in the endodermis. Since we did not find even weak phenotypes in any of the single and double mutants tested (Extended Data Fig. 7e,f,g), we decided to use our endodermal-specific induction system and introduced all 10 constitutively active MKKs in the sgn3 background (designated indMKK1a–indMKK10a). The 10 MKKs are classified into phylogenetic groups (A–D)⁴⁷ (Fig. 5a) and interestingly, these groupings match strikingly well with the functional outcomes we observe in the endodermis.

**Fig. 5: Activation of individual MKKs in the endodermis produces distinct outputs.**

For Group A/B members, namely MKK1/2/6/3, we found that induction of each individual MKKa in the sgn3 endodermis can improve CASP1 and CS lignin fusion (Fig. 5b,c), although none is sufficient to rescue the barrier defects of sgn3 in the diffusion assay (Fig. 5e). This can be attributed to the persistence of small gaps in the CS after induction (Fig. 5b), coupled with the fact that these activated variants do not induce compensatory ectopic lignin or ROS (Fig. 5b and Extended Data Fig. 7c,d). Intriguingly, unlike MPK6a, which triggers early excess suberin, MKK1/2/3a induction suppresses suberization in sgn3 roots (Extended Data Fig. 7a). Our findings suggest that MKK1/2/3/6 are involved in transducing the SGN3-specific branch of CS domain fusion, which aligns nicely with in vitro data of MKK1-MPK2 and MKK2-MPK2 phosphorylating MYB36 (ref. ³⁹). This fits our hypothetical model, whereby distinct SGN3 and FLS2 outputs are defined by differential regulation of MYB36 via specific MKK-MPK signalling routes (Fig. 7)⁴⁶.

Group C members MKK4 and MKK5 show broader activities. Induction of MKK4a can moderately improve domain fusion in sgn3 (Fig. 5c) but uniquely promotes ectopic lignin and ROS production, even in the absence of SGN3 signalling (Fig. 5b and Extended Data Fig. 7c,d). This ectopic lignin is sufficient to form a delayed diffusion barrier in contrast to no barrier with non-induced control in indMKK4a sgn3 (Fig. 5e). In addition, indMKK4a triggers early excess suberin accumulation (fluorol yellow staining) and suberin synthesis (GPAT5 expression marker) in sgn3 (Extended Data Fig. 7a,b). These data suggest that MKK4a mediates the compensatory branches of the SGN3 pathway (ectopic lignin and excess suberin), a function akin to MPK6a’s effects. Given that MKK4-MPK6 is activated by FLS2 for immune function^27,50, this cascade could regulate the common downstream outputs of both SGN3 and FLS2 in the endodermis (Fig. 7). MKK5, often considered functionally redundantly with MKK4 in immune signalling²⁷, has little effect on CASP1 domain fusion or ectopic lignification (Fig. 5b,c,e) but does slightly enhance suberin synthesis upon induction in the sgn3 endodermis (Extended Data Fig. 7a,b).

Among Group D members, MKK9 stands out as we found that the others (MKK7/8/10) are either absent or only weakly expressed in the root endodermis (Extended Data Fig. 6a)³⁰. Strikingly, EST induction of MKK7a and MKK9a in sgn3 suppresses both CASP1 and CS lignin accumulation (Fig. 5b,d and Extended Data Fig. 6b,e). Electron microscopy confirms that MKK9a also abolishes ROS accumulation and membrane attachment, typical of a functional CS structure (Extended Data Fig. 7c). This suppression remains prominent in wild-type background with SGN3 present (Fig. 5d and Extended Data Fig. 6c). In wild type, MKK9a suppresses CASP1 accumulation and domain fusion but does not suppress ectopic lignification (Extended Data Fig. 6c), explaining the partial PI block phenotype (Fig. 5f and Extended Data Fig. 6f). MKK9a also has a potent inhibitory effect on suberin in sgn3 (Extended Data Fig. 7b,c), consistent with MKK9’s recognized role in ethylene signalling^51,52 and the known inhibitory effect of ethylene activation on suberin in the endodermis⁵³. The other Group D members, MKK8a and MKK10a, improve CASP1 domain fusion, but not diffusion barrier defects, resembling the action of Group A/B members (Extended Data Fig. 5c,e)^54,55. Western blot analysis confirms the enhanced accumulations of MKK1–10a upon EST induction (Extended Data Fig. 6d).

In summary, our systematic analysis of all 10 MKKs in a single cell-type demonstrates that activation of individual MKKs in the endodermis leads to clearly distinct output patterns. These patterns match both the common and specific outputs upon SGN3 and FLS2 stimulation in the endodermis. The Groups A/B (MKK1/2/3/6) are strong positive regulators specific to domain fusion; Group C MKK4 is the only one capable of promoting ectopic lignin and excess suberin. Finally, among Group D, MKK7 and MKK9 are strong negative regulators capable of suppressing CS development.

Constitutively active MKK9 inhibits CS formation via suppression of MYB36

The specific impact of MKK9 activation on CASP1 accumulation and fusion suggests that it could be a powerful suppressor of MYB36 activity. We tested this by introducing MYB36^WT, MYB36^AAA or MYB36^DDD tagged with GFP-TurboID (here labelled -GFP) into the sgn3 indMKK9a background (Fig. 3d and Supplementary Table 2). With MYB36^WT-GFP or MYB36^AAA-GFP, indMKK9a still suppresses CASP1 accumulation in sgn3 (Fig. 6a,b). The phosphomimic MYB36^DDD-GFP, by contrast, abolishes the inhibitory effect of indMKK9a on CASP1 accumulation (Fig. 6a,b). This indicates that MKK9 suppression is conditioned by the phosphorylation status of MYB36.

**Fig. 6: Constitutively active MKK9 inhibits CS formation via suppressing MYB36.**

We found that MKK9a induction reduces accumulation of MYB36^WT-GFP and MYB36^AAA-GFP in endodermal nuclei (Fig. 6c,d), and MYB36 levels correspond closely to the extent of CS suppression (Fig. 6a,c and Extended Data Fig. 8a). Conversely, MYB36^DDD-GFP levels were unaffected by MKK9a induction, and MYB36^DDD-GFP levels positively correlated with the rescue activity (Fig. 6 and Extended Data Fig. 8a,b). MKK9 signalling thus appears to destabilize wild type and the phospho-null variant of MYB36 to suppress its function. Intriguingly, in sgn3 indMKK9a lines expressing high levels of MYB36^DDD-GFP, we could observe ectopic CASP1 patches outside of the CSD (Extended Data Fig. 8a), mimicking hyperstimulation of the SGN3 pathway¹². We conclude that the phosphomimic variant of MYB36 not only has an ability to complement the sgn3 phenotype, but also reiterates SGN3 hyperstimulation, firmly placing phosphorylated MYB36 as an important downstream player of the SGN3 pathway.

The FLS2 pathway intertwines closely with ethylene signalling^54,55,56 known to activate MKK1/3/9^51,52,57, and MKK7 was found to contribute to flg22-induced ROS burst⁵⁸. Hypothetically, an activation of MKK9 (and/or MKK7) by the FLS2 pathway could account for the inability of flg22 stimulation to induce closure of CS domains by negatively regulating MYB36. This negative regulation could influence the activation from positive regulators of MYB36 activity and thus exacerbate the difference between SGN3 and FLS2 signalling outputs. If this were the case, flg22-FLS2 signalling should be able to interfere with CIF2-SGN3 signalling when simultaneously activated in the same cell. Indeed, we found that pre/co-treatment of flg22 with CIF2 leads to suppression of CIF2-induced CASP1 domain fusion in 2-in-Endo (Fig. 6e), lending support to such a model (Fig. 7b).

**Fig. 7: Hypothetical model of two receptor pathways in a single cell-type.**

Discussion

Signalling specificity is maintained at single-cell level

The complexity and intricacy of cellular signalling are increasingly recognized in plants^59,60. Understanding the signalling dynamics at single-cell level is fundamental for delineating overall organismal responses. To enable both stimulation and observation of pathway activities at single-cell resolution, we utilized the unique characteristics of the SGN3 pathway, allowing for precise and quantifiable readouts in a non-dividing, easily observable cell.

Our study underlines the importance of examining signalling specificity beyond the constraints of tissue specificity and compartmentalization. We demonstrated that the FLS2 pathway in the endodermis could not substitute the SGN3 pathway in establishing a continuous Casparian strip (Fig. 1). This pertains to endogenous signalling, since the endodermis has the inherent capacity for FLS2 immune signalling³¹.

SGN3 as a neo-functionalized immune receptor?

A simplistic approach to achieve specificity at single-cell level would be to have dedicated sets of receptors, signalling components and transcription factors that are insulated between one pathway and another in the same cell. However, evolution appears to have favoured a more elegant solution—reusing key signalling components from diverse processes, a common theme among multicellular organisms^61,62.

In plants, despite the expansion of the receptor repertoire for detecting a growing number of specific inputs, intracellular signalling pathways share striking similarities in their structure and components. It is thus reasonable to believe that outputs of many pathways share a common core. This is supported by the extensive interconnections among plant hormonal pathways⁶³ and the substantial overlaps observed in transcriptomic responses to biotic and abiotic stresses^64,65,66. Our data expand this idea and show that stimulation of a developmental and an immunity receptor kinase pathway can induce common ROS, lignin and suberin production (Fig. 1). This is corroborated by transcriptional profiles, showing matching induction patterns for genes involved in lignin and suberin biosynthesis following either CIF2 or flg22 treatments (Fig. 2). These commonalities raise an intriguing question: Could both pathways have originated from a shared signalling pathway module, having diverged over time to perceive distinct inputs while maintaining partially overlapping outputs? This fits the prevalent view of plant evolution being driven by cycles of gene duplication and neo-functionalization⁶⁷. Once plants developed a mechanism for ROS production and lignification during defence, they may have adapted these responses for more precise developmental purposes, leading to the emergence of the primary function of the present-day SGN3 pathway, possibly through initially subtle divergences in the substrate specificity of the receptor kinase domains.

Intriguingly, our comparative RNA-seq analysis revealed substantial overlaps among early, highly responsive genes between the two pathways. Yet, despite this remarkable overlap, the FLS2 immune receptor kinase cannot replicate the central developmental function of the SGN3 pathway.

Combinatorial activation of MPK cascades can encode specific signalling outcomes

There is a remarkable potential for the combinatorial activation of kinases to encode signalling specificity in plants. Beyond the members of the MPK cascade, other extended kinase families, including several families of calcium-dependent kinases (CDKs)⁶⁸, also transduce signals from LRR-RLKs. Our study focuses on the MPK cascade’s potential to confer a basic level of specificity, yet we acknowledge the role of additional signalling elements such as CDKs and phospholipids in contributing to the overall output specificity. Intriguingly, plant MPK cascades by themselves already exhibit a complex web-like structure, contrasting with the simpler, more linear pathways found in yeast and mammals^62,69,70. Their expansion in plants has enhanced their capacity to form numerous combinations of MPKKK-MKK-MPK^49,62. While our understanding of plant MPK network interactions has been advanced through global analyses and in vitro methods^39,71,72,73, the mechanisms controlling specificity within these cascades remain largely unknown. MPK3 and MPK6, for example, are reported to be activated by numerous signalling pathways associated with diverse cellular functions^4,74. Similarly, MPK1/2/7/14 are activated by various stress pathways^75,76. This makes it challenging to understand how MPKs would mediate specific cellular responses. Our reporter expression atlas (Supplementary Table 3) and previous tissue-wide expression analyses indicate that MPKs are broadly expressed across tissues^77,78, providing potential for combinatorial action. Our 2-in-Endo line ensures that downstream MPKs activated by both FLS2 and SGN3 pathways encounter identical sets of target proteins. This allowed us to assign the specific activity of SGN3 to its differential activation of MYB36, shown to be an MPK target in vitro³⁹ (Fig. 3). Other endodermis-expressed MYB transcription factors, such as MYB15 and MYB41 respectively involved in lignin and suberin production^79,80, are also reported to be phosphorylated by MPKs^81,82 and might be mediators of the pathway responses common to FLS2 and SGN3, that is, ectopic lignin and excess suberin production (Fig. 7).

In plants, MKK–MPK interactions are known to be promiscuous, which might be highly relevant for generating signalling specificity^39,49,71. Each MKK can generate a potentially unique pattern of MPK activation; each MPK could receive inputs from different MKKs depending on the pathway involved; and individual MPKs could activate both common and unique substrates⁸³, suggesting an extensive interaction network. We propose that such combinatorial MPK activation patterns generate distinct ‘flavours’ that can drive specific outputs.

Elegant tissue-specific manipulations of MKKs in the stomatal lineage have demonstrated overlapping roles for MKKs 4/5 and 7/9 during early stages, and unique activities of MKK7/9 in late stages of stomatal development^84,85,86. In addition, MKK4/5, but not MKK1/2/3/8/9, contribute to root meristem development⁸⁷. These analyses revealed that MKKs have divergent functions over time along a complex developmental trajectory. Our simpler endodermal system uses a differentiating, non-dividing cell-type that allows distinct receptor stimulation and readouts at the same developmental stage. In our system, analysis of all ten MKKs provides compelling evidence that each MKK can activate distinct outputs within the same cell-type (Fig. 5). These functional groupings align well with MKK phylogeny and can account for both the common and unique outputs of the SGN3 and FLS2 pathways (Fig. 7).

On the basis of the matching readouts, we propose that the SGN3 pathway activates MKK1/2/3/6 (Groups A/B) to ensure domain fusion during CS development while suppressing excess suberin. Only upon strong or prolonged stimulation would the SGN3 pathway boost MKK4/5 (Group C) activity to promote ectopic lignin and excess suberin. By avoiding activation of negative regulators of CS development such as MKK9, the SGN3 pathway might further promote effective MYB36 activation. Among all MKKs, only MKK4 could enhance ROS production and lignification, consistent with a reported role of MKK4 in promoting ROS⁸⁸. ROS homoeostasis could be achieved by balancing the activities of different MKKs downstream of SGN3, favouring CS domain closure upon low and transient stimulation, and preventing pleiotropic ectopic lignin and suberin accumulation until stronger, persistent stimulations are reached.

The FLS2 pathway in the endodermis, by contrast, might have a weaker ability to activate MKK1/2/3/6 (Groups A/B) but more potently activates MKK4/5 (Group C), leading to enhanced ectopic lignin and suberin. In addition, the inability of FLS2 to activate MYB36 may be reinforced by a stronger activation of MKK7/9, which would actively suppress MYB36 activity, fitting with the observed inhibitory effect of flg22 on CIF2 action (Fig. 6e). Previous analyses of the FLS2 pathway have painted much more complex and confusing pictures, with at least two independent MPK cascades, MKKK3/5-MKK4/5-MPK3/6 and MEKK1-MKK1/2-MPK4^27,29,89,90. Parts of this were found to be due to perturbations of guarding resistance proteins^43,91,92,93.

We conclude that single cell-type analyses performed in a whole-tissue context can offer much clearer insights into signalling pathways. The complexities observed in previous studies may arise from readouts relying on protoplast experiments and/or use of total plant tissues. It is also intriguing to speculate that cells might switch between signalling states, with strong stimulation of one pathway actively suppressing another. Such dynamic modulation would allow a cell to respond to diverse stimuli, although not necessarily simultaneously.

Limitations of the study

This study provides new and conclusive evidence of divergent functional outputs of ten individual MKKs and two MPKs in a single cell-type, using a gain-of-function approach. However, future efforts require complementary loss-of-function analyses to genetically elucidate the MKK-MPK routes downstream of each pathway that lead to these distinct outputs. The major barrier is the redundancies and pleiotropic effects that arise from broad perturbation of multiple MPKs and/or MKKs. Potential workarounds, such as cell-type-specific MPK inhibition by HopAI1 phosphothreonine lyase activity, appears to be restricted to a subset of MPKs^41,42,43. Future studies would benefit from testing additional approaches for precise tissue-specific MPK knockouts.

Our data support a compelling model of possible differential phosphorylation of MYB36 downstream of FLS2 and SGN3 pathways, driving output specificity. However, we have been unable to obtain sufficient amounts of MYB36 from 2-in-Endo to determine their phosphorylation states after peptide treatment, due to the small number of responsive endodermal cells in total root cells. In planta biochemical validation of MYB36 phosphorylation would require a cell-type-specific enrichment and phosphoproteomics approach. This approach could also help confirm and further delineate the MKK-MPK network coding output specificity in the endodermis.

Methods

Plant materials

Arabidopsis thaliana ecotype Columbia (Col-0) was used as wild-type controls and the background ecotype for all experiments. The mutants fls2 (SALK_062054C)¹⁰, sgn3-3 (SALK_043282)¹⁷, cif1 cif2²⁰(generated by CRISPR-Cas9), myb36-2 (GK- 543B11)¹⁹ and pSGN3::SGN3-mVenus in wild type¹⁷ were previously described. The 2-in-Endo line was generated in two steps: (1) pSGN3::FLS2-3myc-GFP was introduced into fls2 (SALK_062054C) background and (2) cif1 cif2 null alleles were obtained by CRISPR-Cas9 using a single sgRNA (5′-gctttggtttaggactggag-3′) that targets both genes^20,37. The pCASP1::FLS2-3myc-GFP fls2 (SAIL691_C04) described previously³² was crossed with sgn3-3 (SALK_043282) to obtain pCASP1::FLS2-GFP fls2 sgn3.

T-DNA mutant seeds mpk2 (SALK_047422C)⁹⁴, mpk3 (SALK_151594)⁹⁵, mpk6 (SALK_073907)⁹⁵, mpk9 (SALK_064439C), mpk16 (SALK_059737C), mpk17 (SALK_020801C)⁹⁶, mpk19 (SALK_075D213C), mkk4 (SALK_0188040C)⁹⁷, mkk5 (SALK_047797C)⁹⁸ and mkk9 (SAIL_60_H06)⁵² were provided by the Nottingham Arabidopsis Stock Centre (NASC). Most T-DNA lines in our study were previously analysed and reported to show elimination of transcript/protein levels in the cited publications, with mkk4 (SALK_0188040C) showing reduced transcript level⁹⁷. All lines were verified for homozygosity. See Supplementary Table 4 for corresponding T-DNA screening primers. Homozygous double mutants mpk2 mpk16 (T-DNA) and mkk4 mkk5 (T-DNA) were generated by crossing the above indicated T-DNA mutants. mpk2 mpk16 double mutant lines #54-9 and #55-13 were generated by CRISPR-Cas9 using MPK2-targeting sgRNA1 (5′-ccgataggccgaggcgcgta-3′) sgRNA2 (5′-tcttttgcatagttgcttcg-3′) and MPK16-targeting sgRNA1 (5′-agagtttcgagacatttacg-3′) sgRNA2 (5′- atacaccgactgcgatattc-3′). mkk4 mkk5 (TILLING)^46,99, mkk1 (SALK_027645)¹⁰⁰ mkk2 (SAIL 511_H01)¹⁰¹ and mkk3 (SALK_051970)¹⁰² were obtained from Jean Colcombet, while mpk3 mpk6 pMPK6::MPK6^YG and mpk3 mpk6 pMPK3::MPK3^TG were obtained from refs. ^45,46.

The published constructs pCASP1::CASP1-GFP with FastRed selection cassette³⁶ and pGPAT5::mCitrine-SYP122 (ref. ⁵³) were introduced into respective backgrounds by floral dipping. For pathway-specific reporters, promoter fusion constructs in Supplementary Table 2 were introduced into the 2-in-Endo background. For the MPK expression atlas (Supplementary Table 3), promoter fusion constructs pMPK1-20::NLS-3xmVenus were introduced into wild-type background.

Plant growth conditions

Plant seeds were surface sterilized in 70% ethanol + 0.05% Tween 20 for 7–10 min, then washed in 96% ethanol and dried in sterile conditions. Seeds were stratified for 2 days at 4 °C in the dark on half-strength Murashige and Skoog (½MS) + 0.8% agar (Roth) plates containing 500 mg l⁻¹ MES buffer (Duchefa). Seedlings were grown vertically for 5–6 days at 23 °C under continuous light before analysis.

Plasmid construction and plant transformation

The In-Fusion Advantage PCR Cloning kit (Clontech) and Multisite Gateway Cloning Technology (Invitrogen) were used for generation of constructs.

All inducible constructs (ind) were assembled by triple Gateway recombination reaction (LR) between entry clones P4-pELTP (LTPG15)-XVE-P1r (promoter)¹⁰³, L1-HopAI1-L2, L1-MPKa-L2 or L1-MKKa-L2 (gene of interest), and R2-3xFlag-L3 (C-terminal tag) into the destination vector pF7m34GW, containing a FastRed plant selection cassette with a 35S terminator. Complementary (c)DNA entry clones of HopAI1 and HopAI1^H103A were synthesized by Genscript⁴¹; MPK6a (D218G E222A)¹⁰⁴ was obtained from Jean Colcombet; MPK2a (Q188G E192A), MKK1a (T218E S224D), MKK2a (T220D T226E), MKK3a (S235E T241D), MKK4a (T224D S230E), MKK5a (T215E S221E), MKK6a (S221D T227E), MKK7a (S193E S199D), MKK8a (S195D S201E), MKK9a (S195E S201E) and MKK10a (S197E) were cloned from Col-0 cDNA and mutations obtained through site-directed mutagenesis.

All MYB36 constructs were assembled by triple Gateway recombination reaction between entry clones P4-pMYB36(4kb)-R1 (native promoter), L1-MYB36^WT-L2, L1-MYB36^AAA-L2 or L1-MYB36^DDD-L2 (gene of interest), and R2-MYB36-native terminator(250bp)-L3, R2-mCherry-4G-L3 or R2-SuperfolderGFP-TurboID(BirA*)-6His-TEV-3FLAG-L3 (terminator or C-terminal tag) into destination vectors pF7m34GW,0 containing a FastRed plant selection cassette or pFG34GW, containing a FastGreen plant selection cassette. Genomic (g)DNA entry clones of wild-type MYB36 were cloned from Col-0 gDNA, and mutants MYB36^AAA (S18A, S146A, S169A) and MYB36^DDD (S18D, S146D, S169D) were synthesized by Genscript.

The pathway-specific reporter constructs (Supplementary Table 1) and MPK reporter constructs (Supplementary Table 3) were made, first by generating promoter entry clones using In-Fusion: the endogenous promoter sequences (~2 kb before the start codon) were cloned from Col-0 gDNA or JAtY clones. The reporter constructs were then assembled by double Gateway recombination reaction between a promoter entry clone and L1-nuclear localization signal (NLS) fused to a triple mVenus fluorescent tag (3xmVenus)-L2 or L1-NLS-3xmScarlet-L2 into the destination vector pFG24GW (FastGreen), pFR7m24GW (FastRed) or pB7m24GW (BASTA selection). See Supplementary Tables 4 and 5 for plasmid and primer details.

All plasmids were transformed by heat shock into Agrobacterium tumefaciens GV3101 strain with pSoup plasmid (pMP90) and then transformed into the corresponding plant lines using the floral dip method^105,106. On the basis of the plant selection marker, we typically obtained 10–30 transgenic lines for initial characterization or analyses at T1 generation. From these, 2–6 independent T2 transgenic lines were isolated for further characterization. For all T2 analysis in this paper, all plants were pre-selected using fluorescent seedling selection marker. Independent homozygous T3 lines were confirmed by non-segregation of the selection marker and used for western blot experiments or as background for further introduction of constructs. For certain experiments, T1 populations were used to fully capture the inherent expression variability due to random transformation events. This T1 analysis ensures that statistically significant phenotypic differences are indeed reflective of the constructs’ effects themselves and not transgene expression variability. The use of fluorescent seed selection markers (FastRed or FastGreen)¹⁰⁷ substantially reduces the occurrence of false positives during our T1 and T2 selection process and facilitates our analysis without the stress of antibiotic selection.

Seedling growth and treatments

The flg22 peptide from Pseudomonas aeruginosa (QRLSTGSRINSAKDDAAGLQIA) was ordered from EZBioLab, and CIF2 peptide (DY*GHSSPKPKLVRPPFKLIPN, * indicates sulfated tyrosine) from Peptide Specialty Laboratories (https://www.peptid.de/). Peptides were dissolved in deionized MilliQ sterile water at the stock concentration of 1 mM and diluted in melted ½ MS agar medium to indicated concentrations. Peptide treatments: seedlings were grown vertically for 4–5 days, then transferred with care onto plates containing peptides or H₂O control to grow vertically for 1 day before analysis.

For time-lapse imaging and videos with peptide treatments, 3-day-old seedlings grown vertically on ½ MS plates were transferred to a hydroponic system (a 12-well Multiwell plate (CytoOne) with 6.5 ml liquid ½ MS in each well) to grow for 2 days. These 5-day-old seedlings were then transferred inside a chambered coverglass (Thermo Scientific) and then covered with a ½ MS agar block containing 1 µM peptide. A volume of 50 μl of liquid ½ MS containing the same peptide concentration was added through a channel in the middle of the agar block, directly reaching the roots.

EST (10 mM stock, Sigma, solvent dimethylsulfoxide (DMSO)) and NM-PP1 (Adipogen, solvent DMSO) were diluted to 5 µM in melted ½ MS agar medium, and 0.05% v/v DMSO was used as control treatments. For estradiol (EST) treatments, seedlings were grown vertically on EST or DMSO plates for 5 days or transferred to EST or DMSO to grow for 1 day. For NM-PP1 treatments, seedlings were grown vertically on ½ MS plates for 4 days and transferred to NM-PP1 or DMSO plates to grow vertically for 1 day.

Fluorescence microscopy

Imaging was performed using Leica SP8, Leica Stellaris or Leica Stellaris 5 WLL IR confocal laser scanning microscopes and Leica Thunder light microscope (DM6BZ). Images were taken with a ×63 water immersion objective (Leica SP8, Leica Stellaris) at zoom factor 1 for overviews and zoom factor 5 for surface views. Time-lapse images of Fig. 2f and Supplementary Videos 2 and 3 were taken with a ×63 oil immersion objective (Leica Stellaris 5 WLL IR). Time-lapse images of Extended Data Fig. 3d and Supplementary Video 1 were taken with a ×40 oil immersion objective (Leica Stellaris). Root scans for fluorol yellow (suberin staining) and pGPAT5:mCitrine-SYP122 (suberin synthesis) were acquired using a Leica Thunder light microscope with ×10 dry objective and tile-scan (10% overlap). Unless indicated otherwise, confocal images comparing genotypes/treatments were taken following the ‘four identical criteria’: same position in the roots, the same laser detection intensity, the same laser scanning area (zoom) and the same interval of slices for Z-stack projection. For overview images, maximum projection of z-stacks (step size = 1 µm) are shown. For surface-view images, maximum projection of z-stacks (step size = 0.36 µm) are shown.

The excitation and detection windows were set as follows: for live imaging, GFP (488 nm, 495–550 nm), PI (488 nm, 600–690 nm), mVenus (514 nm, 520–565nm), mCitrine (514 nm, 555–570 nm), mScarlet (561 nm, 580–640 nm) and mCherry (561 nm, 585–680 nm). For time-lapse imaging with Leica Stellaris 5 WLL IR: mVenus (507 nm, 520–560 nm) and mScarlet (555 nm, 580–600 nm). For fixed samples, sequential scanning was used to avoid interference between the following fluorescence channels: calcofluor white (405 nm, 415–435 nm), GFP (488 nm, 495–525 nm) and basic fuchsin (561 nm, 600–650 nm). For suberin analysis on Thunder: fluorol yellow (488 nm, 500–550 nm) and mCitrine (488 nm, 500–550 nm).

Propidium iodide penetration assay

PI penetration assays assess the stage of the formation of a functional CS diffusion barrier^15,16. Seedlings were incubated in 10 µg ml⁻¹ PI for 10 min and rinsed in H₂O before imaging. PI block was identified as the endodermal cell achieved the exclusion of PI signal from the inner side. The number of endodermal cells was counted from the onset of elongation until PI block. For roots with no functional barrier, PI was not blocked near the root–hypocotyl junction and roots are thus shown as individual data points in the category ‘near hypocotyl’ and excluded from numerical statistical tests.

Suberin quantification assay

Suberin was stained using the methanol-based fluorol yellow protocol^20,32. Seedlings were fixed and cleared in methanol for at least 3 days at 4 °C, changing the methanol at least once. The cleared seedlings were transferred to a freshly prepared solution of Fluorol Yellow 088 (0.01% in methanol) and incubated for 1 h at room temperature or overnight at 4 °C in the dark. The stained seedlings were rinsed in methanol and transferred to a freshly prepared solution of aniline blue (0.5% in methanol) for 1 h counterstaining in the dark. Finally, the seedlings were washed briefly in water before mounting and imaging with the Leica Thunder light microscope.

Fixation and lignin staining

Fixation and staining were performed using an adapted Clearsee protocol^108,109. Briefly, seedlings were fixed in 4% paraformaldehyde PBS solution at 4 °C overnight using 6-well or 12-well plates (CytoOne), then washed twice for 1 min with PBS. Once fixed, seedlings were cleared in Clearsee solution for at least 24 h under gentle shaking. Fixed and cleared samples were then incubated overnight in a Clearsee solution supplemented with 0.2% basic fuchsin and 0.1% calcofluor white for combined cell wall and lignin staining. Once the dye solution was removed, samples were rinsed overnight in Clearsee twice before mounting and observation using confocal microscopes.

Electron microscopy for ROS and lignin detection

Visualization of H₂O₂ (ROS) around the CS was performed with the cerium chloride assay described previously^20,37. This detection method is based on Ce³⁺ ions reacting with H₂O₂, forming electron-dense cerium perhydroxide precipitates, which are detected by transmission electron microscopy. Briefly, seedlings were incubated in 50 mM MOPS buffer (pH 7.2) containing freshly prepared 10 mM cerium chloride (CeCl₃) (Sigma) for 45 min. Seedlings were then washed twice in MOPS buffer for 5 min and fixed in 2.5% glutaraldehyde solution (EMS) in 100 mM phosphate buffer (pH 7.4) for 1 h at room temperature. Then, they were postfixed in 1% osmium tetroxide (EMS) with 1.5% potassium ferrocyanide (Sigma) in phosphate buffer for 1 h at room temperature. The plants were rinsed twice in distilled water and then dehydrated in ethanol solution (Sigma) at gradient concentrations (30%, 40 min; 50%, 40 min; 70%, 40 min; 100%, 1 h ×2). This was followed by infiltration in Spurr resin (EMS) at gradient concentrations (Spurr 33% in ethanol, 4 h; Spurr 66% in ethanol, 4 h; Spurr 100%, 8 h ×2) and finally polymerized for 48 h at 60 °C in an oven. Ultrathin sections (50 nm) were cut transversally at 1.5, 2 or 3 mm from the root tip on a Leica Ultracut ultramicrotome (Leica Mikrosysteme) and picked up on a 2 × 1 mm nickel slot grid (EMS) coated with a polystyrene film (Sigma).

Visualization of lignin deposition around the CS and cell corner was done using the permanganate potassium (KMnO₄, Sigma) staining protocol^110,111. The sections were post-stained with 1% KMnO₄ in H₂O for 45 min and rinsed several times with H₂O.

Micrographs were taken with a CM100 transmission electron microscope (FEI) at an 80 kV acceleration voltage with a TemCamF416 digital camera (TVIPS) using the software EM-MENU 4.0 (TVIPS). Panoramic views were aligned using the software IMOD¹¹².

Ectopic ROS quantification from EM images

For Extended Data Fig. 7d EM ROS quantification, all samples from the selected lines—indMKK2a, indMKK4a, indMPK6a and indMPK2a in sgn3 CASP1-GFP with and without estradiol induction were collected, prepared and imaged in a single experiment. We assessed 5 roots per sample and quantified the ROS staining patterns categorically cross an average of 42 endodermal cells per sample: from the TEM images, each endodermal cell was carefully examined for ROS staining patterns (electron-dense dots and patches) occurring outside of the Casparian strip region, typically between the endodermis–cortex boundary. These ectopic ROS patterns were categorized as no, yes low or yes strong on the basis of their intensity.

RNA-seq experiments and analysis

Sample preparation for RNA-seq experiments was as previously described²⁰. Briefly, wild-type, sgn3, fls2 and pCASP1::FLS2-GFP fls2 seeds were grown vertically on a sterile mesh on solid ½ MS plates for 5 days and transferred onto fresh ½ MS solid medium containing 1 μM peptides and mock plates (mock and peptide treatments always transferred in parallel). After 30-, 120- and 480-min incubation on plates, whole roots from 2 square plates (50 ml) were cut and collected for each sample, then immediately frozen in liquid nitrogen. Three biological replicates for each sample were conducted on three different days. Total RNA was extracted with a TRIzol-adapted ReliaPrep RNA extraction kit (Promega).

Libraries were prepared by the Lausanne Genome Technology Facility following the same protocol²⁰. RNA quality control was performed on a Fragment Analyzer (Advanced Analytical Technologies). Total RNA (1,000 ng) was used to prepare RNA-seq libraries with the Illumina TruSeq Stranded mRNA reagents (Illumina) on a Sciclone liquid handling robot (PerkinElmer) using a PerkinElmer-developed automated script. The resulting library was used for cluster generation with the Illumina TruSeq SR Cluster Kit v4 reagents and sequenced on the Illumina HiSeq 2500 system using TruSeq SBS Kit v4 reagents. The Illumina Pipeline Software v.2.20 was used to process sequencing data.

The Lausanne Genomic Technologies Facility performed the data processing using their in-house RNA-seq pipeline, as described in ref. ²⁰. Briefly, purity-filtered read trimming for adapters and low-quality sequences was done with Cutadapt (v.1.8)¹¹³ and removal of reads matching ribosomal RNA sequences with fastq_screen (v.0.11.1). Low complexity reads were filtered with reaper (v.15-065)¹¹⁴. Cleaned reads were aligned against A. thaliana TAIR10 genome using STAR (v.2.5.3a)¹¹⁵ and read counts per gene locus were obtained with htseqcount (v.0.9.1)¹¹⁶ using A. thaliana TAIR10 Ensembl 39 gene annotation. RSeQC (v.2.3.7)¹¹⁷ was used to evaluate the quality of the data alignment.

Statistical analysis for genes was performed in R (3.5.3). Genes with low counts were filtered out according to the rule of one count per million (cpm) in at least one sample. Library sizes were scaled using trimmed mean of M values normalization and log transformed into cpm (EdgeR package v.3.24.3)¹¹⁸. Data were corrected for the experimental batch effect using the removeBatchEffect function (limma). Statistical quality controls were performed through pairwise sample correlations, clustering and sample Principal component analysis using batch corrected normalized data. Differential expression was computed using the limma-trend approach by fitting all samples into one linear model¹¹⁹. The experimental batch factor was added to the model. Two-sided moderated t-test was used for each pairwise comparison (peptide-treated vs H₂O-treated) per timepoint. The adjusted P value was computed using the Benjamini–Hochberg method, controlling for false discovery rate. To determine genes significantly regulated by CIF2 or flg22 (*s in Fig. 2 and Extended Data Fig. 2), the adjusted P value was set to be below 0.05 (P_adj < 0.05).

Heat maps of gene lists in Fig. 2c,d were selected on the basis of published genes involved in lignin and suberin biosynthesis^20,120. Figure 2e genes were based on the intersection of RNA-seq datasets in ref. ³⁰ and microarray datasets in ref. ¹⁹ that report significant downregulation in myb36 mutants compared with wild type. myb36 vs wild type regulation heat map (left) was generated using the data in ref. ³⁰. Figure 2b genes were based on genes significantly upregulated by flg22 in whole seedlings⁴¹. The heat map of wild type treated with flg22 (left) in Extended Data Fig. 2b was generated using the data in ref. ⁴¹.

GO-enrichment analyses were performed using the clusterProfiler R package (3.18.1) with ‘Biological process’, and P value was adjusted using the Benjamini–Hochberg method with a cut-off of 0.05. GO annotations used the org.At.tair.db R package (3.12.0). Gene ratio represents the gene counts in a GO category that are differentially regulated by a pathway divided by the total number of genes in that GO category.

For Fig. 4a and Supplementary Fig. 7, seedlings were grown vertically for 5 days on sterilized mesh on solid ½ MS media. The seedlings were treated by adding 20 ml liquid ½ MS containing respective peptides or H₂O control to submerge the roots. At 5, 10 and 15 min, liquid was drained and roots were cut (Fig. 4a), or whole seedlings were collected (Source Data Supplementary Fig. 7b). For Extended Data Fig. 5b and 6d, seedlings were grown for 5–6 days on sterilized mesh on solid ½ MS media supplemented with EST or DMSO.

Root tissues collected were immediately frozen in liquid nitrogen and ground using Tissue Lyser II (Qiagen). Extraction buffer (50 mM Tris HCl pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 1% Triton X-100, 50 mM β-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, 100 μM Na orthovanadate, 5 mM Na fluoride, 2.5 mM MgCl₂, 1x complete EDTA-free protease inhibitor cocktail) was added to the frozen samples. The samples were briefly mixed by vortexing and centrifuged at 21,000 × g for 15 min at 4 °C. The supernatant was transferred to new tubes and the total protein concentration was measured using the Bradford method (Thermo Fisher). After adjusting total protein concentration, the proteins were denatured by addition of 3x (v/w) NuPAGE LDS sample buffer + 10 mM dithiothreitol (Invitrogen, NP0007) and heated for 5 min at 95 °C.

Equal amounts of protein were loaded (~20–40 μg per lane) and separated on 10% or gradient (4–12%) pre-cast acrylamide gels (Eurogentec). After electrophoresis, the separated proteins were transferred onto PVDF membranes (Immobilon-E, Merck-Millipore) using Pierce FastBlotter G2 semi-dry blotting (Thermo Fisher). The blot was blocked in 1% blocking buffer in TBS (western blocking reagent, Roche) for 1 h at room temperature before incubation with primary antibodies: anti-p44/42 (1:1,000, Cell Signaling Technologies, 4370), anti-MPK6 (1:10,000, Sigma, A7104), anti-FLAG conjugated M2 peroxidase (1:500, Sigma, A8592) and anti-actin (1:2,000, Sigma, A0605) in 0.5% blocking buffer + TBS-0.05% Tween overnight at 4 °C. After washing the blot two to three times with 1x TBS + 0.1% Tween (TBS-T), the blot was treated with secondary antibodies (anti-rabbit, 1:10,000, Agrisera for anti-p44/42; anti-rabbit HRP ECL, 1:30,000 for anti-MPK6; anti-mouse HRP, 1:5,000 for actin, Promega) in 0.5% blocking buffer + TBS-0.05% Tween for 2 h and then washed 5–7 times with TBS-T. For loading control detection of MPK6 and actin, the original membrane was stripped (Thermo Scientific). Signal was detected with SuperSignal West Pico Plus and Femto kits (Thermo Scientific).

Quantification and statistical analysis

All fluorescence microscopy images were processed using Fiji (ImageJ2 v.2.9.0/1.53t) with Fiji macros to semi-automate the processing and quantification¹²¹. Macros will be made available on request.

Percentages (%) of gaps in CASP1-GFP (Fig. 1e and Extended Data Fig. 1h) were based on maximum projections of surface-view images (GFP channel) after thresholding to distinguish background from signal. Quantification represents the percentage of uncovered area (background) of a fused CASP1-GFP belt reconstructed from connecting individual domains using a selection Bush tool. Fully connected CASP1-GFP yields 0% of gaps.

Numbers of holes per 100 µm of CASP1-GFP (Figs. 3e, 4e, 5c, 6e and Extended Data Fig. 1c) were based on maximum projections of overviews (GFP channel) after thresholding to distinguish background from signal. Steps for quantification are detailed in Extended Data Fig. 1b. Each data point represents the average number of holes per 100 µm calculated from a single overview image.

Quantifications of total grey intensity of nuclei signals (Fig. 2f) were calculated from ‘sum slice projection’ of overview images at each hour. This method sums the intensity values at each pixel location throughout the stack and projects the sum onto a two-dimensional image. This is useful for visualizing the total intensity contribution from all slices in the stack, especially in the cases of overlapping nuclei. Thresholding and region of interest selection were then applied to isolate nuclei signals from noise and background. Similar procedures were used for Fig. 6d to calculate total grey intensity of MYB36 nuclei signals from maximum projection of overview images.

For quantification of CASP1 domain properties: number of fragments per 100 µm (Fig. 3c) and the average size of fragments (Extended Data Fig. 3c) were based on the same data points from maximum projections of surface-view images. Thresholding followed by particle analysis (no minimum size) was used to quantify the number of fragments and the average size of fragments in a single image. The length of CASP1 in µm for a single image was measured and used to calculate the number of fragments per 100 µm.

Quantification of CS coverage (Figs. 5d, 6b and Extended Data Fig. 6e) is represented by the percentage of CASP1 length coverage normalized to wild type. Total CASP1 length (measured after applying skeletonize (Fiji)) of each overview image (maximum projection) was measured and then normalized to the average CASP1-GFP length in wild-type roots from the same experiments to calculate the percentage of CS coverage for each data point.

All graphs and statistical analyses were performed using R 4.0.4 (http://www.R-project.org/). For multiple comparisons within the same dataset, analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test was applied when linear model assumptions were met. Unbalanced Tukey’s test was used for unequal replication. For comparing the mean of one sample with a control or treatment versus no treatment, Welch’s t-test was used: for multiple comparisons using t-tests, P values were adjusted by Bonferroni correction. In boxplots, boxes span the first to third quartiles, with the bold line representing the median. Whiskers indicate maximum and minimum values, except outliers (defined as >1.5× interquartile range above the third quartile or below the first quartile, respectively). Outliers are always shown as dots. Bold lines in dot plots represent the mean. Each point in the boxplot or dot plot represents a data point from an individual root (n). Statistical parameters, error bars and n are reported in the Figure legends.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The full RNA-seq dataset is deposited at the GEO depository, accession: GSE270950. All other data are available in the article and extended data. Plasmids and transgenic plant seeds generated in this study will be made available on request, but we may require a payment and/or a completed Materials Transfer Agreement if there is potential for commercial application. Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, N.G. (niko.geldner@unil.ch). Source data are provided with this paper.

Code availability

This study did not generate new code. R scripts used to generate graphs or Fiji macros used for image processing and quantifications will be made available on request.

References

Lewis, T. S., Shapiro, P. S. & Ahn, N. G. Signal transduction through MAP kinase cascades. Adv. Cancer Res. 74, 49–139 (1998).
Article CAS PubMed Google Scholar
Colcombet, J. & Hirt, H. Arabidopsis MAPKs: a complex signalling network involved in multiple biological processes. Biochem. J. 413, 217–226 (2008).
Article CAS PubMed Google Scholar
Arthur, J. S. C. & Ley, S. C. Mitogen-activated protein kinases in innate immunity. Nat. Rev. Immunol. 13, 679–692 (2013).
Article CAS PubMed Google Scholar
Zhang, M. & Zhang, S. Mitogen‐activated protein kinase cascades in plant signaling. J. Integr. Plant Biol. 64, 301–341 (2022).
Article PubMed Google Scholar
Lehti-Shiu, M. D., Zou, C., Hanada, K. & Shiu, S.-H. Evolutionary history and stress regulation of plant receptor-like kinase/pelle genes. Plant Physiol. 150, 12–26 (2009).
Article CAS PubMed PubMed Central Google Scholar
Soltabayeva, A. et al. Receptor-like kinases (LRR-RLKs) in response of plants to biotic and abiotic stresses. Plants 11, 2660 (2022).
Article CAS PubMed PubMed Central Google Scholar
Liu, M. & Zen, K. Toll-Like receptors regulate the development and progression of renal diseases. Kidney Dis. 7, 14–23 (2021).
Article CAS Google Scholar
Osumi-Sutherland, D. et al. Cell type ontologies of the Human Cell Atlas. Nat. Cell Biol. 23, 1129–1135 (2021).
Article CAS PubMed Google Scholar
Apelt, F. et al. Shoot and root single cell sequencing reveals tissue- and daytime-specific transcriptome profiles. Plant Physiol. 188, 861–878 (2021).
Article PubMed Central Google Scholar
Zipfel, C. et al. Bacterial disease resistance in Arabidopsis through flagellin perception. Nature 428, 764–767 (2004).
Article CAS PubMed Google Scholar
Chinchilla, D., Bauer, Z., Regenass, M., Boller, T. & Felix, G. The Arabidopsis receptor kinase FLS2 binds flg22 and determines the specificity of flagellin perception. Plant Cell 18, 465–476 (2005).
Article PubMed Google Scholar
Doblas, V. G. et al. Root diffusion barrier control by a vasculature-derived peptide binding to the SGN3 receptor. Science 355, 280–284 (2017).
Article CAS PubMed Google Scholar
Nakayama, T. et al. A peptide hormone required for Casparian strip diffusion barrier formation in Arabidopsis roots. Science 355, 284–286 (2017).
Article CAS PubMed Google Scholar
Barberon, M. & Geldner, N. Radial transport of nutrients: the plant root as a polarized epithelium. Plant Physiol. 166, 528–537 (2014).
Article PubMed PubMed Central Google Scholar
Alassimone, J., Naseer, S. & Geldner, N. A developmental framework for endodermal differentiation and polarity. Proc. Natl Acad. Sci. USA 107, 5214–5219 (2010).
Article CAS PubMed PubMed Central Google Scholar
Naseer, S. et al. Casparian strip diffusion barrier in Arabidopsis is made of a lignin polymer without suberin. Proc. Natl Acad. Sci. USA 109, 10101–10106 (2012).
Article CAS PubMed PubMed Central Google Scholar
Pfister, A. et al. A receptor-like kinase mutant with absent endodermal diffusion barrier displays selective nutrient homeostasis defects. ELife 3, e03115 (2014).
Article PubMed PubMed Central Google Scholar
Hosmani, P. S. et al. Dirigent domain-containing protein is part of the machinery required for formation of the lignin-based Casparian strip in the root. Proc. Natl Acad. Sci. USA 110, 14498–14503 (2013).
Article CAS PubMed PubMed Central Google Scholar
Kamiya, T. et al. The MYB36 transcription factor orchestrates Casparian strip formation. Proc. Natl Acad. Sci. USA 112, 10533–10538 (2015).
Article CAS PubMed PubMed Central Google Scholar
Fujita, S. et al. SCHENGEN receptor module drives localized ROS production and lignification in plant roots. EMBO J. 39, e103894 (2020).
Article CAS PubMed PubMed Central Google Scholar
Schulze, B. et al. Rapid heteromerization and phosphorylation of ligand-activated plant transmembrane receptors and their associated kinase BAK1. J. Biol. Chem. 285, 9444–9451 (2010).
Article CAS PubMed PubMed Central Google Scholar
Sun, Y. et al. Structural basis for flg22-induced activation of the Arabidopsis FLS2-BAK1 immune complex. Science 342, 624–628 (2013).
Article CAS PubMed Google Scholar
Okuda, S. et al. Molecular mechanism for the recognition of sequence-divergent CIF peptides by the plant receptor kinases GSO1/SGN3 and GSO2. Proc. Natl Acad. Sci. USA 117, 2693–2703 (2020).
Article CAS PubMed PubMed Central Google Scholar
Chinchilla, D. et al. A flagellin-induced complex of the receptor FLS2 and BAK1 initiates plant defence. Nature 448, 497–500 (2007).
Article CAS PubMed Google Scholar
Alassimone, J. et al. Polarly localized kinase SGN1 is required for Casparian strip integrity and positioning. Nat. Plants 2, 16113 (2016).
Article CAS PubMed Google Scholar
Lee, Y., Rubio, M. C., Alassimone, J. & Geldner, N. A mechanism for localized lignin deposition in the endodermis. Cell 153, 402–412 (2013).
Article CAS PubMed Google Scholar
Asai, T. et al. MAP kinase signalling cascade in Arabidopsis innate immunity. Nature 415, 977–983 (2002).
Article CAS PubMed Google Scholar
Smith, J. M. & Heese, A. Rapid bioassay to measure early reactive oxygen species production in Arabidopsis leave tissue in response to living Pseudomonas syringae. Plant Methods 10, 6 (2014).
Article PubMed PubMed Central Google Scholar
Suarez-Rodriguez, M. C. et al. MEKK1 is required for flg22-induced MPK4 activation in Arabidopsis plants. Plant Physiol. 143, 661–669 (2007).
Article CAS PubMed PubMed Central Google Scholar
Liberman, L. M., Sparks, E. E., Moreno-Risueno, M. A., Petricka, J. J. & Benfey, P. N. MYB36 regulates the transition from proliferation to differentiation in the Arabidopsis root. Proc. Natl Acad. Sci. USA 112, 12099–12104 (2015).
Article CAS PubMed PubMed Central Google Scholar
Zhou, F. et al. Co-incidence of damage and microbial patterns controls localized immune responses in roots. Cell 180, 440–453.e18 (2020).
Article CAS PubMed PubMed Central Google Scholar
Emonet, A. et al. Spatially restricted immune responses are required for maintaining root meristematic activity upon detection of bacteria. Curr. Biol. 31, 1012–1028.e7 (2021).
Article CAS PubMed Google Scholar
Roppolo, D. et al. Functional and evolutionary analysis of the Casparian strip membrane domain protein family. Plant Physiol. 165, 1709–1722 (2014).
Article CAS PubMed PubMed Central Google Scholar
Roppolo, D. et al. A novel protein family mediates Casparian strip formation in the endodermis. Nature 473, 380–383 (2011).
Article PubMed Google Scholar
Li, B. et al. Phosphorylation of trihelix transcriptional repressor ASR3 by MAP KINASE4 negatively regulates Arabidopsis immunity. Plant Cell 27, 839–856 (2015).
Article CAS PubMed PubMed Central Google Scholar
Barbosa, I. C. R. et al. Directed growth and fusion of membrane-wall microdomains requires CASP-mediated inhibition and displacement of secretory foci. Nat. Commun. 14, 1626 (2023).
Article CAS PubMed PubMed Central Google Scholar
Rojas-Murcia, N. et al. High-order mutants reveal an essential requirement for peroxidases but not laccases in Casparian strip lignification. Proc. Natl Acad. Sci. USA 117, 29166–29177 (2020).
Article CAS PubMed PubMed Central Google Scholar
Wang, P. et al. Surveillance of cell wall diffusion barrier integrity modulates water and solute transport in plants. Sci. Rep. 9, 4227 (2019).
Article PubMed PubMed Central Google Scholar
Popescu, S. C. et al. MAPK target networks in Arabidopsis thaliana revealed using functional protein microarrays. Genes Dev. 23, 80–92 (2009).
Article CAS PubMed PubMed Central Google Scholar
Bi, G. et al. Receptor-like cytoplasmic kinases directly link diverse pattern recognition receptors to the activation of mitogen-activated protein kinase cascades in Arabidopsis. Plant Cell https://doi.org/10.1105/tpc.17.00981 (2018).
Li, X. et al. Flagellin induces innate immunity in non-host interactions that is suppressed by Pseudomonas syringae effectors. Proc. Natl Acad. Sci. USA 102, 12990–12995 (2005).
Article CAS PubMed PubMed Central Google Scholar
Zhang, J. et al. A Pseudomonas syringae effector inactivates MAPKs to suppress PAMP-induced immunity in plants. Cell Host Microbe 1, 175–185 (2007).
Article CAS PubMed Google Scholar
Zhang, Z. et al. Disruption of PAMP-induced MAP kinase cascade by a Pseudomonas syringae effector activates plant immunity mediated by the NB-LRR protein SUMM2. Cell Host Microbe 11, 253–263 (2012).
Article CAS PubMed Google Scholar
Wang, H., Ngwenyama, N., Liu, Y., Walker, J. C. & Zhang, S. Stomatal development and patterning are regulated by environmentally responsive mitogen-activated protein kinases in Arabidopsis. Plant Cell 19, 63–73 (2007).
Article PubMed PubMed Central Google Scholar
Xu, J. et al. A chemical genetic approach demonstrates that MPK3/MPK6 activation and NADPH oxidase‐mediated oxidative burst are two independent signaling events in plant immunity. Plant J. 77, 222–234 (2014).
Article CAS PubMed Google Scholar
Su, J. et al. Regulation of stomatal immunity by interdependent functions of a pathogen-responsive MPK3/MPK6 cascade and abscisic acid. Plant Cell 29, 526–542 (2017).
Article CAS PubMed PubMed Central Google Scholar
Ichimura, K. et al. Mitogen-activated protein kinase cascades in plants: a new nomenclature. Trends Plant Sci. 7, 301–308 (2002).
Article CAS Google Scholar
Jiang, M. et al. Mitogen-activated protein kinase and substrate identification in plant growth and development. Int. J. Mol. Sci. 23, 2744 (2022).
Article CAS PubMed PubMed Central Google Scholar
Andreasson, E. & Ellis, B. Convergence and specificity in the Arabidopsis MAPK nexus. Trends Plant Sci. 15, 106–113 (2010).
Article CAS PubMed Google Scholar
Bi, G. & Zhou, J.-M. MAP kinase signaling pathways: a hub of plant–microbe interactions. Cell Host Microbe 21, 270–273 (2017).
Article CAS PubMed Google Scholar
Xu, J. et al. Activation of MAPK kinase 9 induces ethylene and camalexin biosynthesis and enhances sensitivity to salt stress in Arabidopsis. J. Biol. Chem. 283, 26996–27006 (2008).
Article CAS PubMed Google Scholar
Yoo, S.-D., Cho, Y.-H., Tena, G., Xiong, Y. & Sheen, J. Dual control of nuclear EIN3 by bifurcate MAPK cascades in C₂H₄ signalling. Nature 451, 789–795 (2008).
Article CAS PubMed PubMed Central Google Scholar
Barberon, M. et al. Adaptation of root function by nutrient-induced plasticity of endodermal differentiation. Cell 164, 447–459 (2016).
Article CAS PubMed Google Scholar
Tena, G., Boudsocq, M. & Sheen, J. Protein kinase signaling networks in plant innate immunity. Curr. Opin. Plant Biol. 14, 519–529 (2011).
Article CAS PubMed PubMed Central Google Scholar
Mersmann, S., Bourdais, G., Rietz, S. & Robatzek, S. Ethylene signaling regulates accumulation of the FLS2 receptor and is required for the oxidative burst contributing to plant immunity. Plant Physiol. 154, 391–400 (2010).
Article CAS PubMed PubMed Central Google Scholar
Tintor, N. et al. Layered pattern receptor signaling via ethylene and endogenous elicitor peptides during Arabidopsis immunity to bacterial infection. Proc. Natl Acad. Sci. USA 110, 6211–6216 (2013).
Article PubMed PubMed Central Google Scholar
Zhang, T. et al. Ethylene‐induced stomatal closure is mediated via MKK1/3–MPK3/6 cascade to EIN2 and EIN3. J. Integr. Plant Biol. 63, 1324–1340 (2021).
Article CAS PubMed Google Scholar
Rufián, J. S. et al. The bacterial effector HopZ1a acetylates MKK7 to suppress plant immunity. New Phytol. 231, 1138–1156 (2021).
Article PubMed Google Scholar
Baluška, F. & Mancuso, S. Deep evolutionary origins of neurobiology: turning the essence of ‘neural’ upside-down. Commun. Integr. Biol. 2, 60–65 (2009).
Article PubMed PubMed Central Google Scholar
Krysan, P. J. & Colcombet, J. Cellular complexity in MAPK signaling in plants: questions and emerging tools to answer them. Front. Plant Sci. 9, 1674 (2018).
Article PubMed PubMed Central Google Scholar
Bardwell, L., Zou, X., Nie, Q. & Komarova, N. L. Mathematical models of specificity in cell signaling. Biophys. J. 92, 3425–3441 (2007).
Article CAS PubMed PubMed Central Google Scholar
Ma, Y. & Nicolet, J. Specificity models in MAPK cascade signalling. FEBS Open Bio https://doi.org/10.1002/2211-5463.13619 (2023).
Article PubMed PubMed Central Google Scholar
Shigenaga, A. M. & Argueso, C. T. No hormone to rule them all: interactions of plant hormones during the responses of plants to pathogens. Semin. Cell Dev. Biol. 56, 174–189 (2016).
Article CAS PubMed Google Scholar
Li, J.-R., Liu, C.-C., Sun, C.-H. & Chen, Y.-T. Plant stress RNA-seq nexus: a stress-specific transcriptome database in plant cells. BMC Genomics 19, 966 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kang, W.-H. et al. Transcriptome profiling of abiotic responses to heat, cold, salt, and osmotic stress of Capsicum annuum L. Sci. Data 7, 17 (2020).
Article CAS PubMed PubMed Central Google Scholar
Cohen, S. P. & Leach, J. E. Abiotic and biotic stresses induce a core transcriptome response in rice. Sci. Rep. 9, 6273 (2019).
Article PubMed PubMed Central Google Scholar
Flagel, L. E. & Wendel, J. F. Gene duplication and evolutionary novelty in plants. New Phytol. 183, 557–564 (2009).
Article PubMed Google Scholar
Valmonte, G. R., Arthur, K., Higgins, C. M. & MacDiarmid, R. M. Calcium-dependent protein kinases in plants: evolution, expression and function. Plant Cell Physiol. 55, 551–569 (2014).
Article CAS PubMed Google Scholar
Bardwell, L. Mechanisms of MAPK signalling specificity. Biochem. Soc. Trans. 34, 837–841 (2006).
Article CAS PubMed PubMed Central Google Scholar
Raman, M., Chen, W. & Cobb, M. H. Differential regulation and properties of MAPKs. Oncogene 26, 3100–3112 (2007).
Article CAS PubMed Google Scholar
Lee, J. S., Huh, K. W., Bhargava, A. & Ellis, B. E. Comprehensive analysis of protein–protein interactions between Arabidopsis MAPKs and MAPK kinases helps define potential MAPK signalling modules. Plant Signal. Behav. 3, 1037–1041 (2008).
Article PubMed PubMed Central Google Scholar
Feilner, T. et al. High throughput identification of potential Arabidopsis mitogen-activated protein kinases substrates. Mol. Cell. Proteom. 4, 1558–1568 (2005).
Article CAS Google Scholar
Hoehenwarter, W. et al. Identification of novel in vivo MAP kinase substrates in Arabidopsis thaliana through use of tandem metal oxide affinity chromatography. Mol. Cell. Proteom. 12, 369–380 (2013).
Article CAS Google Scholar
Yue, J. & López, J. M. Understanding MAPK signaling pathways in apoptosis. Int. J. Mol. Sci. 21, 2346 (2020).
Article CAS PubMed PubMed Central Google Scholar
Danquah, A. et al. Identification and characterization of an ABA‐activated MAP kinase cascade in Arabidopsis thaliana. Plant J. 82, 232–244 (2015).
Article CAS PubMed Google Scholar
Sözen, C. et al. Wounding and insect feeding trigger two independent MAPK pathways with distinct regulation and kinetics. Plant Cell 32, 1988–2003 (2020).
Article PubMed PubMed Central Google Scholar
Menges, M. et al. Comprehensive gene expression atlas for the Arabidopsis MAP kinase signalling pathways. New Phytol. 179, 643–662 (2008).
Article CAS PubMed Google Scholar
Schmid, M. et al. A gene expression map of Arabidopsis thaliana development. Nat. Genet. 37, 501–506 (2005).
Article CAS PubMed Google Scholar
Chezem, W. R., Memon, A., Li, F.-S., Weng, J.-K. & Clay, N. K. SG2-type R2R3-MYB transcription factor MYB15 controls defense-induced lignification and basal immunity in Arabidopsis. Plant Cell 29, 1907–1926 (2017).
Article CAS PubMed PubMed Central Google Scholar
Shukla, V. et al. Suberin plasticity to developmental and exogenous cues is regulated by a set of MYB transcription factors. Proc. Natl Acad. Sci. USA 118, e2101730118 (2021).
Article CAS PubMed PubMed Central Google Scholar
Hoang, M. H. T. et al. Phosphorylation by AtMPK6 is required for the biological function of AtMYB41 in Arabidopsis. Biochem. Biophys. Res. Commun. 422, 181–186 (2012).
Article CAS PubMed Google Scholar
Kim, S. H. et al. Phosphorylation of the transcriptional repressor MYB15 by mitogen-activated protein kinase 6 is required for freezing tolerance in Arabidopsis. Nucleic Acids Res. 45, gkx417 (2017).
Article Google Scholar
Rayapuram, N. et al. Quantitative phosphoproteomic analysis reveals shared and specific targets of Arabidopsis mitogen-activated protein kinases (MAPKs) MPK3, MPK4, and MPK6. Mol. Cell. Proteom. 17, 61–80 (2018).
Article CAS Google Scholar
Lampard, G. R., Lukowitz, W., Ellis, B. E. & Bergmann, D. C. Novel and expanded roles for MAPK signaling in Arabidopsis stomatal cell fate revealed by cell type-specific manipulations. Plant Cell 21, 3506–3517 (2009).
Article CAS PubMed PubMed Central Google Scholar
Wengier, D. L., Lampard, G. R. & Bergmann, D. C. Dissection of MAPK signaling specificity through protein engineering in a developmental context. BMC Plant Biol. 18, 60 (2018).
Article PubMed PubMed Central Google Scholar
Lampard, G. R., Wengier, D. L. & Bergmann, D. C. Manipulation of mitogen-activated protein kinase kinase signaling in the Arabidopsis stomatal lineage reveals motifs that contribute to protein localization and signaling specificity. Plant Cell Online 26, 3358–3371 (2014).
Article CAS Google Scholar
Lu, X. et al. RGF1-RGI1, a peptide-receptor complex, regulates Arabidopsis root meristem development via a MAPK signaling cascade. Mol. Plant 13, 1594–1607 (2020).
Article CAS PubMed Google Scholar
Ren, D., Yang, H. & Zhang, S. Cell death mediated by MAPK is associated with hydrogen peroxide production in Arabidopsis. J. Biol. Chem. 277, 559–565 (2002).
Article CAS PubMed Google Scholar
Ichimura, K., Casais, C., Peck, S. C., Shinozaki, K. & Shirasu, K. MEKK1 is required for MPK4 activation and regulates tissue-specific and temperature-dependent cell death in Arabidopsis. J. Biol. Chem. 281, 36969–36976 (2006).
Article CAS PubMed Google Scholar
Sun, T. et al. Antagonistic interactions between two MAP kinase cascades in plant development and immune signaling. EMBO Rep. 19, e45324 (2018).
Article PubMed PubMed Central Google Scholar
Kong, Q. et al. The MEKK1-MKK1/MKK2-MPK4 kinase cascade negatively regulates immunity mediated by a mitogen-activated protein kinase kinase kinase in Arabidopsis. Plant Cell 24, 2225–2236 (2012).
Article CAS PubMed PubMed Central Google Scholar
Takagi, M. et al. Disruption of the MAMP-induced MEKK1-MKK1/MKK2-MPK4 pathway activates the TNL immune receptor SMN1/RPS6. Plant Cell Physiol. 60, 778–787 (2018).
Article Google Scholar
Takagi, M. et al. Simultaneous mutations in SMN1 and SUMM2 fully suppress the dwarf and autoimmune phenotypes of Arabidopsis mpk4 mutant. Plant Signal. Behav. 17, 2046412 (2022).
Article PubMed PubMed Central Google Scholar
Zhang, J. et al. MKK4/MKK5-MPK1/MPK2 cascade mediates SA-activated leaf senescence via phosphorylation of NPR1 in Arabidopsis. Plant Mol. Biol. 102, 463–475 (2020).
Article CAS PubMed Google Scholar
Zhu, Q. et al. A MAPK cascade downstream of IDA–HAE/HSL2 ligand–receptor pair in lateral root emergence. Nat. Plants 5, 414–423 (2019).
Article CAS PubMed Google Scholar
Goyal, R. K. et al. Analysis of MAPK and MAPKK gene families in wheat and related Triticeae species. BMC Genomics 19, 178 (2018).
Article PubMed PubMed Central Google Scholar
Romero-Hernandez, G. & Martinez, M. Opposite roles of MAPKKK17 and MAPKKK21 against Tetranychus urticae in Arabidopsis. Front. Plant Sci. 13, 1038866 (2022).
Article PubMed PubMed Central Google Scholar
Li, K. et al. AIK1, a mitogen-activated protein kinase, modulates abscisic acid responses through the MKK5-MPK6 kinase cascade. Plant Physiol. 173, 1391–1408 (2016).
Article PubMed PubMed Central Google Scholar
Zhao, C. et al. EDR1 physically interacts with MKK4/MKK5 and negatively regulates a MAP kinase cascade to modulate plant innate immunity. PLoS Genet. 10, e1004389 (2014).
Article PubMed PubMed Central Google Scholar
Gao, M. et al. MEKK1, MKK1/MKK2 and MPK4 function together in a mitogen-activated protein kinase cascade to regulate innate immunity in plants. Cell Res. 18, 1190–1198 (2008).
Article CAS PubMed Google Scholar
Teige, M. et al. The MKK2 pathway mediates cold and salt stress signaling in Arabidopsis. Mol. Cell 15, 141–152 (2004).
Article CAS PubMed Google Scholar
Dóczi, R. et al. The Arabidopsis mitogen-activated protein kinase kinase MKK3 is upstream of group C mitogen-activated protein kinases and participates in pathogen signaling. Plant Cell 19, 3266–3279 (2007).
Article PubMed PubMed Central Google Scholar
Siligato, R. et al. MultiSite gateway-compatible cell type-specific gene-inducible system for plants. Plant Physiol. 170, 627–641 (2015).
Article PubMed PubMed Central Google Scholar
Berriri, S. et al. Constitutively active mitogen-activated protein kinase versions reveal functions of Arabidopsis MPK4 in pathogen defense signaling. Plant Cell 24, 4281–4293 (2012).
Article CAS PubMed PubMed Central Google Scholar
Clough, S. J. & Bent, A. F. Floral dip: a simplified method for Agrobacterium‐mediated transformation of Arabidopsis thaliana. Plant J. 16, 735–743 (1998).
Article CAS PubMed Google Scholar
Zhang, X., Henriques, R., Lin, S.-S., Niu, Q.-W. & Chua, N.-H. Agrobacterium-mediated transformation of Arabidopsis thaliana using the floral dip method. Nat. Protoc. 1, 641–646 (2006).
Article CAS PubMed Google Scholar
Ursache, R., Fujita, S., Tendon, V. D. & Geldner, N. Combined fluorescent seed selection and multiplex CRISPR/Cas9 assembly for fast generation of multiple Arabidopsis mutants. Plant Methods 17, 111 (2021).
Article CAS PubMed PubMed Central Google Scholar
Kurihara, D., Mizuta, Y., Sato, Y. & Higashiyama, T. ClearSee: a rapid optical clearing reagent for whole-plant fluorescence imaging. Development 142, 4168–4179 (2015).
CAS PubMed PubMed Central Google Scholar
Ursache, R., Andersen, T. G., Marhavý, P. & Geldner, N. A protocol for combining fluorescent proteins with histological stains for diverse cell wall components. Plant J. 93, 399–412 (2018).
Article CAS PubMed Google Scholar
Hepler, P. K., Fosket, D. E. & Newcomb, E. H. Lignification during secondary wall formation in Coleus: an electron microscopic study. Am. J. Bot. 57, 85–96 (1970).
Article Google Scholar
Kolbeck, A. et al. CASP microdomain formation requires cross cell wall stabilization of domains and non-cell autonomous action of LOTR1. ELife 11, e69602 (2022).
Article CAS PubMed PubMed Central Google Scholar
Kremer, J. R., Mastronarde, D. N. & McIntosh, J. R. Computer visualization of three-dimensional image data using IMOD. J. Struct. Biol. 116, 71–76 (1996).
Article CAS PubMed Google Scholar
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 17, 10–12 (2011).
Article Google Scholar
Davis, M. P. A., van Dongen, S., Abreu-Goodger, C., Bartonicek, N. & Enright, A. J. Kraken: a set of tools for quality control and analysis of high-throughput sequence data. Methods 63, 41–49 (2013).
Article CAS PubMed PubMed Central Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Anders, S., Pyl, P. T. & Huber, W. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015).
Article CAS PubMed Google Scholar
Wang, L., Wang, S. & Li, W. RSeQC: quality control of RNA-seq experiments. Bioinformatics 28, 2184–2185 (2012).
Article CAS PubMed Google Scholar
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 (2010).
Article CAS PubMed Google Scholar
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Article PubMed PubMed Central Google Scholar
Ursache, R. et al. GDSL-domain proteins have key roles in suberin polymerization and degradation. Nat. Plants 7, 353–364 (2021).
Article CAS PubMed PubMed Central Google Scholar
Schneider, C. A., Rasband, W. S. & Eliceiri, K. W. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675 (2012).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank the Cellular Imaging Facility (CIF), Genomic Technologies Facility (GTF) and Electron Microscopy Facility (EMF) of the University of Lausanne and N. Pullen for technical support; G. Ingram for critically reading the paper; S. Zhang and J. Colcombet for kindly sharing seed materials; and all Geldner lab members for sharing materials. Y.M. was supported by an EMBO non-stipendiary long-term fellowship (ALTF 711-2018), a FEBS Long-term Fellowship and by 20-MSCA-IF-2018 (grant number 846050). The work was supported by SNSF grant 310030 to N.G.

Author information

Yan Ma
Present address: Gregor Mendel Institute of Molecular Plant Biology (GMI), Austrian Academy of Sciences, Vienna BioCenter, Vienna, Austria
Joe B. Dickinson
Present address: Department of Fundamental Microbiology, Biophore, UNIL-Sorge, University of Lausanne, Lausanne, Switzerland
Aurélia Emonet
Present address: Max Planck Institute for Plant Breeding Research, Cologne, North Rhine-Westphalia, Germany
Satoshi Fujita
Present address: Plant Science Research Laboratory (LRSV), UMR5546 CNRS/Université Toulouse 3, Auzeville Tolosane, France
These authors jointly supervised this work: Yan Ma, Niko Geldner.

Authors and Affiliations

Department of Plant Molecular Biology, Biophore, UNIL-Sorge, University of Lausanne, Lausanne, Switzerland
Yan Ma, Isabelle Flückiger, Jade Nicolet, Jia Pang, Joe B. Dickinson, Damien De Bellis, Aurélia Emonet, Satoshi Fujita & Niko Geldner

Authors

Yan Ma
View author publications
Search author on:PubMed Google Scholar
Isabelle Flückiger
View author publications
Search author on:PubMed Google Scholar
Jade Nicolet
View author publications
Search author on:PubMed Google Scholar
Jia Pang
View author publications
Search author on:PubMed Google Scholar
Joe B. Dickinson
View author publications
Search author on:PubMed Google Scholar
Damien De Bellis
View author publications
Search author on:PubMed Google Scholar
Aurélia Emonet
View author publications
Search author on:PubMed Google Scholar
Satoshi Fujita
View author publications
Search author on:PubMed Google Scholar
Niko Geldner
View author publications
Search author on:PubMed Google Scholar

Contributions

Y.M. and N.G. conceived, designed and coordinated the project. Y.M., I.F., J.N., J.P., J.B.D., D.D.B. and A.E. performed all experimental work and data analysis. S.F. and A.E. provided materials and were involved in the discussion of the work. Y.M. and N.G. wrote the paper with feedback from all co-authors.

Corresponding authors

Correspondence to Yan Ma or Niko Geldner.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Plants thanks the anonymous reviewers for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Endodermis-specific FLS2 induces ectopic lignin but cannot replace endogenous SGN3 for Casparian strip (CS) domain fusion.

(a) CASP1-GFP overviews of 2-in-Endo after 1d treatments of H₂O, 100 nM CIF2 or flg22 from initiation to maturation (7^th, 11^th, 16-17^th endodermal cells after onset of elongation). (b) Schematics showing steps to quantify CS discontinuity in overviews to incorporate many more endodermal cells at once: 1. CASP1-GFP overviews: maximum projection of z-stacks capturing CS network in endodermal cell layer; 2. Particle Analysis (Fiji): count fragments and calculate corresponding number of holes; 3. Skeletonize (Fiji): convert area to lines and measure CS length in µm; 4. Calculate ‘Number of holes per 100 µm’ as in (C). (c) Quantification of CS discontinuity in 2-in-Endo CASP1-GFP overviews at mature CS stage after constitutive (5d) treatments with H₂O, 100 nM CIF2 or 100 nM flg22 (n ≥ 8). (d) Suberization quantification of 2-in-Endo after constitutive (5d) treatments with H₂O, 100 nM CIF2 or flg22. Suberin was stained with Fluorol Yellow and quantified along the root axis divided by three zones: unsuberized (near the root-tip); patchy; and continuous (near hypocotyl). Data presented as percentage of total root length (Mean ± SE; n = 10). Zones with the same letter are not significantly different among treatments (p < 0.05, one-way ANOVA and Tukey test). (e) Diffusion barrier of 2-in-Endo visualized by Propidium Iodide (PI, red) penetration at 25^th endodermal cell after onset of elongation. Arrowheads mark PI signal exclusion from the inner side of an endodermal cell (PI block). (f) Transmission Electron Microscopy (TEM) showing in situ H₂O₂ (ROS) detection with CeCl₃ (left) and lignin detection with KMnO₄ (right) at the CS of 2-in-Endo after 1d H₂O, 100 nM CIF2 or flg22 treatments. Micrographs obtained from sections ~1.5 mm away from the root tip. Brackets indicate CS: uniformly light grey cell wall areas with membrane attachment (left) and dense black staining (right). Arrowheads indicate ectopic accumulations of ROS (left) and lignin (right). (g) CS & CASP1 domain fusion phenotypes of pCASP1::FLS2-GFP fls2 sgn3 CASP1-GFP after 1d treatment with H₂O, 100 nM flg22 or 1 µM flg22. Endodermal cell surface views at mature CS stage (16-17^th cell) show CASP1-GFP (green), lignin stained with Basic Fuchsin (magenta) and overlaps of the two channels (Merged). Arrowheads highlight ectopic lignin. ‘outer’, cortex-facing, ‘inner’, stele-facing endodermal surface. (h) Quantifications for (G) showing the percentage of gaps in CASP1-GFP pCASP1::FLS2-GFP fls2 sgn3 after 1d treatments (n ≥ 9). (i) PI penetration assay quantifies the CS barrier function of pCASP1::FLS2-GFP fls2 sgn3 with 1d treatments of H₂O and flg22 at 100 nM and 1 µM concentrations compared to controls. Roots with no barrier show PI penetration near root-hypocotyl junction, shown in the ‘near hypocotyl’ category and are excluded from numerical statistical test (n ≥ 10). For (C)(H)(I), groups with the same letter are not significantly different (P < 0.05, one-way ANOVA and Tukey test). Unbalanced Tukey test used for unequal replication.

Extended Data Fig. 2 Endodermis-expressed FLS2 and SGN3 activate both common and pathway-specific transcriptional responses.

(a) Experimental setups for RNAseq analysis comparing CIF2 and flg22 responses in root endodermis. Transcriptional data for CIF2 responses are from Fujita et al.²⁰. WT.CIF2, sgn3.CIF2: wild-type with functional SGN3 and sgn3 treated with 1 µM CIF2 vs H₂O control. FLS2.flg22, fls2.flg22: pCASP1::FLS2-GFP fls2 and fls2 treated with 1 µM flg22 vs H₂O control. Fold change (log₂FC) were calculated by comparing peptide (1 µM flg22 or 1 µM CIF2) to H₂O treatment at each time point (30, 120, 480 mins). (b) Heatmaps of genes regulated by FLS2 in whole seedlings – significantly upregulated genes in flg22-treated wild-type (left) (*, P < 0.001)³⁵ – are also strongly induced by CIF2 (WT.CIF2) and flg22 (FLS2.flg22) in the endodermis (right) (*P < 0.05). Colour code: fold change degree (log₂FC). (c) Heatmaps of fold change for reporter genes. Candidates are predicted to be preferentially CIF2 (green) or flg22 (yellow) responsive, or equally responsive (blue) in the endodermis (*P < 0.05). Reporter line analyses for selected genes are shown on the right (circles), with colour indicating observed pathway specificity (details see Supplementary Table 1). Corresponding heatmap of fold-change in myb36 compared to Wild-type (*P < 0.001, NA in grey)³⁰ shown on the left. Colour code: fold change degree (log₂FC). (d) Gene ontology (GO) enrichment analysis showing enriched GO-terms unique to FLS2 and SGN3 pathways upon stimulation in the endodermis. Colour: GO-enrichment significance. Dot size: gene ratio (the gene counts in a GO category differentially regulated by a pathway divided by the total number of genes in that GO category). (e) Workflow to determine the pathway specificity of top 10% most responsive genes and least responsive (P < 0.05, ranked by Log₂FC) at 30 min. Genes that are significantly upregulated by one pathway but not the other are considered specific, genes that are significantly upregulated by both are considered common (Cut-off: Log₂FC > 1 and P < 0.05).

Extended Data Fig. 3 Predicted phospho-sites of MYB36 are crucial for SGN-dependent CASP and CS domain fusion.

(a) MYB36 protein diagram, with locations of three putative MPK phosphorylation sites (Serine Proline - SP) and two MYB-type HTH DNA binding domains (MYB). (b) CASP1-GFP phenotypes and MYB36 accumulations in myb36 CASP1-GFP transformed with MYB36^WT-mCherry, MYB36^AAA-mCherry, MYB36^DDD-mCherry. CASP1-GFP overviews (upper-panels) at 16-17^th cell and MYB36-mCherry nuclei signals (lower-panels) from the same lines at 7-8^th cells are shown for each genotype. GFP and mCherry in gradient intensity. (c) Quantification of CASP1 domain property: Average size of CASP1-GFP fragments in surface views, same data as Fig. 3c. myb36 CASP1-GFP transformed with untagged MYB36^WT, MYB36^AAA, MYB36^DDD (WT, AAA, DDD) are treated with 1d H2O (grey) and 1 µM CIF2 (pink) (n ≥ 10). (d) CASP1-GFP time lapse images of myb36 CASP1-GFP untagged MYB36^AAA (top) and cif1 cfi2 CASP1-GFP (bottom) treated with 1 µM CIF2 from 0-4 hours. Maximum projections of roots at mature CS (16-17^th cell) are shown. Scale bars = 20 µm.

Extended Data Fig. 4 Multiple MPKs contribute to CS integrity downstream of the SGN3 pathway.

(a) MPK expression analysis in the endodermis. MPKs 1-20 from Arabidopsis thaliana are classified into 4 phylogenetic groups using full length sequences⁴⁷. Endodermis-enriched RNA-seq data³⁰ (Endo) and Translating Ribosomal Affinity Purification (TRAP) analysis (Andersen & Vermeer, personal communication) that detect mRNA transcripts in distinct compartments of endodermis and xylem poles using different promoters: SCR (Meristematic endodermis); LOG3 (Xylem pole protoxylem); LOG4 (Xylem pole endodermis); XPP (Xylem pole); CASP1 (Early unsuberized endodermis); ELTP or LTPG15 (The entire differentiated endodermis); GPAT5 (Suberized endodermis). CASP1 and MYB36 are used as references. (b) MPKs with endodermal expression around CS initiation (7-8^th cell). Analyses based on individual pMPK::NLS-3xmVenus lines (1 to 20) in wild-type. See Supplementary Table 3 for full atlas analysis. Expression in all cells (green +), expression in some but not all cells (yellow +_p), no expression in all cells (red –). ‘!’ marks the selected MPKs for further analyses. (c) Examples of pMPK::NLS-3xmVenus reporter with endodermal signals during CS development. White arrows indicate endodermal nuclei. Cell walls are stained with Calcofluor white (magenta – upper) or with PI (red – lower). Scale bars = 100 µm. (d) CASP1-GFP phenotype with endodermal suppression of multiple MPK activities by inducing HopAI1 vs. inactive HopAI^H103A in wild-type at 16-17^th cell. (e) Suberization quantification with or without MPK6a or MPK2a induction in sgn3 CASP1-GFP. Suberin stained with Fluorol Yellow and quantified along the root axis divided by three zones: unsuberized; patchy; and continuous. Data presented as percentage of total root length (Mean ± SE; n ≥ 10). Zones with the same letter are not significantly different (p < 0.001, Student’s t-test). (f) TEM showing in situ H₂O₂ (ROS) detection with CeCl₃ at the CS of 2-in-Endo after 5d DMSO control or EST induction of MPK6a and MPK2a in sgn3 CASP1-GFP. Micrographs obtained from sections ~3 mm from the root tip. Brackets indicate CS: uniformly light grey cell wall areas with membrane attachment. Arrowheads indicate accumulations of ROS at endodermal cell corners. (g) CS lignin & CASP1 phenotypes of mpk3 and mpk6 compared to wild-type, after 1d treatment of 1 µM CIF2. Surface views at 16-17^th cell show CASP1-GFP (green), Lignin (magenta) and channel overlaps (Merged). Arrowheads highlight ectopic lignin on the cortex-facing side.

Extended Data Fig. 5 Multiple MPKs contribute to CS integrity downstream of the SGN3 pathway (continued).

(a) CASP1 phenotypes of T-DNA mutants mpk2, mpk16, mpk3, mpk6, mpk2 mpk16 and two independent lines of mpk2 mpk16 CRIPSR mutants. CASP1-GFP surface views at 16-17^th cell are shown. (b) Immunoblot (IB) against anti-FLAG shows accumulations of the C-terminally 3xFLAG tagged MPK2a in sgn3 CASP1-GFP background. IB against MPK6 shows accumulations of MPK6a in sgn3 CASP1-GFP background. T3 seedling roots grown on DMSO or 5 µM EST were harvested. Loading controls: IBs with anti-Actin antibody or Ponceau S staining. See Supplementary Fig. 7 for uncropped original blots. This experiment was repeated three times. (c) Allelic information of two independent mpk2 mpk16 CRIPSR mutant lines. Pink arrowheads indicate positions for homozygous deletions or insertion. Amino acid numbers in brackets indicate the size of protein fragment due to early stop codon. gRNAs used to target MPK2 and MPK16 see Supplementary Table 4.

Extended Data Fig. 6 Activation of individual MKKs in the endodermis produces distinct outputs.

(a) MKK expression analysis in the endodermis. MKKs 1-10 from Arabidopsis thaliana are classified into 4 phylogenetic groups using full length sequence⁴⁷. Endodermis-enriched RNA-seq data³⁰ (Endo) and TRAP analysis (Andersen & Vermeer, personal communication) that detect mRNA transcripts in distinct compartments of endodermis and xylem poles using different promoters: SCR (Meristematic endodermis); LOG3 (Xylem pole protoxylem); LOG4 (Xylem pole endodermis); XPP (Xylem pole); CASP1 (Early unsuberized endodermis); ELTP or LTPG15 (The entire differentiated endodermis); GPAT5 (Suberized endodermis). (b) CASP1 & lignin phenotypes upon EST induction of MKK9a in sgn3 CASP1-GFP at 16-17^th cell. Left panels show overviews and right panels show surface views with partial suppression for CASP1-GFP (green), Lignin (magenta) and channel overlaps (Merged). (c) CASP1 phenotypes after one-day EST induction of MKK9a in wild-type CASP1-GFP at new growth (15^th and 30^th cell -EST induction), and older growth (40^th cell - no induction). Right panels show zoomed surface views at 30^th cell for CASP1-GFP (green), lignin (magenta) and channel overlaps (Merged). Arrowheads highlight ectopic lignin on the cortex-facing side. (d) Immunoblots (IB) against anti-FLAG showing accumulations of the C-terminally 3xFLAG tagged indMKK1-10a in sgn3 CASP1-GFP background upon EST induction. T3 seedlings roots grown on DMSO or 5 µM EST were harvested. N.S., non-specific bands. Loading controls: IBs with anti-Actin antibody or Ponceau S staining. See Supplementary Fig. 7 for uncropped original blots. This experiment was repeated two to three times. (e) Quantification of CS coverage upon induction of MKK7a and MKK9a in sgn3 CASP1-GFP, MKK9a in WT CASP1-GFP (T1 roots, n ≥ 8). At mature CS stage, total CASP1 length of each overview image measured normalized to the average CASP1-GFP length in wild-type to calculate percentage of CS coverage for each data point. (f) PI penetration assay upon endodermal induction of MKK9a in WT CASP1-GFP (n ≥ 10). Data from three independent T2 lines and independent T1 seedlings are included (T2-3 and T2-4 data are shown in Fig. 5f). For (E)(F), Welch’s T-test compares DMSO with EST for each genotype, p-values were adjusted for multiple comparisons by Bonferroni correction (***p < 0.001, **p < 0.01, *p < 0.05). For EST induction, all seedlings were grown for 5d on 5 µM EST or DMSO.

Extended Data Fig. 7 Activation of individual MKKs in the endodermis produces distinct outputs (continued).

(a) Suberization quantification upon induction of MKK1/2/3/4/5/6/9a in sgn3 CASP1-GFP. Suberin stained with Fluorol Yellow and quantified along the root axis divided by three zones: unsuberized; patchy; and continuous. Data presented as percentage of total root length (Mean ± SE; n ≥ 10). (b) Suberin synthesis quantification upon induction of MKK4a, MKK5a and MKK9a in sgn3 pGPAT5::mCitrine-SYP122 (SYP122 is a membrane marker). Suberin synthesis is marked by GPAT5 transcription. Data presented as percentage of total root length (Mean ± SE; n ≥ 10). Zones in each genotype with the same letter are not significantly different upon treatment (p < 0.05, Student’s t-test for equal replicates, Welch’s t-test for unequal replicates). Zones with c/c’ are increased compared to a/a’, b/b’ decreased compared to a/a’. (c) TEM showing in situ H₂O₂ (ROS) detection with CeCl₃ at the CS of 2-in-Endo after 5d DMSO or EST induction of MKK1/2/3/4/5/9a in sgn3 CASP1-GFP. Micrographs obtained from sections ~1.5-2 mm from the root tip. Brackets indicate CS: uniformly light grey cell wall areas with membrane attachment. Dashed bracket indicates the absence of an CS. Arrowheads indicate accumulations of ROS. (d) TEM quantification of ectopic ROS at endodermal cell corner after 5d DMSO or EST induction of induction MPK2a, MPK6a, MKK2a and MKK4a in sgn3 CASP1-GFP. All samples were collected, prepared, and imaged from a single experiment. Each horizontal line represents a single root. Each dot represents an endodermal cell. Different coloured dots illustrate the categorization of ectopic ROS phenotypes (no, yes low, yes strong) which representative images on the right. (e) CASP1 phenotypes of mkk2, mkk3, mkk4, mkk5, mkk9, mkk4 mkk5 at 16-17^th cell. (f) CASP1 & lignin phenotypes of one-day 1 µM CIF2 treated wild-type and mkk4 mkk5 CASP1-GFP at 11-12th Cell. Arrowheads highlight ectopic lignin on cortex-facing side. (g) PI penetration assay quantifies the CS barrier function of T-DNA mutants mkk1, mkk2, mkk3, mkk4, mkk5, mkk9, mkk4 mkk5 and mkk4 mkk5 (TILLING) (n ≥ 10). Groups with the same letter are not significantly different (P < 0.05, one-way ANOVA and Tukey HSD). For EST induction, all seedlings were grown for 5 d on 5 µM EST or DMSO control before analysis unless indicated otherwise.

Extended Data Fig. 8 Constitutively active MKK9 inhibits CS formation via suppressing MYB36.

(a) CASP1-GFP and corresponding MYB36 accumulations in sgn3 CASP1-GFP indMKK9a transformed with MYB36^WT-GFP, MYB36^AAA-GFP, MYB36^DDD-GFP with or without EST induction. CASP1-GFP overviews from different plant lines at mature CS stage (16-17^th cell) are shown in upper panels. Nuclei signals of MYB36-GFP from the same lines are taken around the onset of CASP1-GFP (7-8^th cells) and shown in corresponding panels below. GFP signals are shown in gradient intensity. Scale bars = 20 µm. (b) Immunoblots (IB) against anti-FLAG show accumulations of the C-terminally 3xFLAG tagged indMKK9a in sgn3 CASP1-GFP (T3 homozygous for MKK9a) transformed with MYB36^WT-GFP, MYB36 ^AAA-GFP, MYB36^DDD-GFP with and without EST induction. Seedling roots of the same T2 lines used in Fig. 6a–d were grown on DMSO or 5 µM EST. Loading controls: IBs with anti-Actin antibody. Numbers above each lane show the FLAG (MKK9a) intensity normalized to corresponding Actin loading control. ‘>’ indicate MKK9a, ‘n.s.’ indicate non-specific bands. See Supplementary Fig. 7 for uncropped original blots. Due to limited seed material, this experiment could only be performed once.

Supplementary information

Supplementary Information

Supplementary Tables 1–5 and Supplementary Fig. 7d–l.

Reporting Summary

Supplementary Video 1

Video 1. CASP1-GFP phenotypes of 2-in-Endo with CIF2 or flg22 treatments. Related to Fig. 1. Total imaging time 5.5 h, each frame taken at 30 min intervals. Scale bars, 25 µm.

Supplementary Video 2

Video 2. SGN preferential reporter pPRK1::NLS-3xmVenus in 2-in-Endo. Related to Fig. 2 and Supplementary Table 1. Total imaging time 8 h, each frame taken at 1 h intervals. Scale bars, 25 µm.

Supplementary Video 3

Video 3. FLS2 preferential reporter pWRKY30::NLS-3xmVenus in 2-in-Endo. Related to Fig. 2 and Supplementary Table 1. Total imaging time 8 h, each frame taken at 1 h intervals. Scale bars, 25 µm.

Source data

Source Data Supplementary Fig. 7

Additional western blot repeats and uncropped original blots. a, Original blots for Fig. 4a. Note the original arrangement of 1 μM flg22 was loaded 15 min, 10 min and 5 min from left to right. b, Immunoblots on whole seedlings of 2-in-Endo roots treated with H₂O, 1 µM CIF2 or flg22 for 5, 10, 15 min. Immunoblots (IB) against p42/44 show phosphorylated MAPKs. Loading controls: Ponceau S staining. c, IB on seedling roots of 2-in-Endo treated with H₂O, 1 µM CIF2 or flg22 for 5, 10, 15 min, and mpk6 treated with H₂O for 5, 10, 15 min. IB against anti-MPK6. Loading controls: Ponceau S staining.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Ma, Y., Flückiger, I., Nicolet, J. et al. Comparisons of two receptor-MAPK pathways in a single cell-type reveal mechanisms of signalling specificity. Nat. Plants 10, 1343–1362 (2024). https://doi.org/10.1038/s41477-024-01768-y

Download citation

Received: 01 September 2023
Accepted: 19 July 2024
Published: 10 September 2024
Version of record: 10 September 2024
Issue date: September 2024
DOI: https://doi.org/10.1038/s41477-024-01768-y

This article is cited by

Two residues reprogram immunity receptors for nitrogen-fixing symbiosis
- Magdalini Tsitsikli
- Bine Simonsen
- Simona Radutoiu
Nature (2025)
Exploring the function of plant root diffusion barriers in sealing and shielding for environmental adaptation
- Yi-Qun Gao
- Yu Su
- Dai-Yin Chao
Nature Plants (2024)