A conserved module regulates receptor kinase signaling in immunity and development

Abstract

Ligand recognition by cell-surface receptors underlies development and immunity in both animals and plants. Modulating receptor signaling is critical for appropriate cellular responses but the mechanisms ensuring this are poorly understood. Here, we show that signaling by plant receptors for pathogen-associated molecular patterns (PAMPs) in immunity and CLAVATA3/EMBRYO SURROUNDING REGION-RELATED peptides (CLEp) in development employs a similar regulatory module. In the absence of ligand, signaling is dampened through association with specific type-2C protein phosphatases (PP2Cs). Upon activation, PAMP and CLEp receptors phosphorylate divergent cytosolic kinases, which, in turn, phosphorylate the phosphatases, thereby promoting receptor signaling. Our work reveals a regulatory circuit shared between immune and developmental receptor signaling, which may have broader important implications for plant receptor kinase-mediated signaling in general.

Receptor kinases (RKs) perceive the immediate cellular environment¹. Leucine-rich repeat (LRR)-RKs represent the largest group of RKs in plants, with over 220 members in the model plant Arabidopsis thaliana (hereafter, Arabidopsis)^2,3. LRR-RKs perceive a wide variety of ligands to regulate diverse aspects of growth, development, reproduction, and stress responses¹. How specificity in LRR-RK signaling is achieved is poorly understood, owing to a lack of known signaling mechanisms and intermediates for many LRR-RK pathways. Several of the best-studied LRR-RKs to date function as cell-surface immune receptors, which perceive pathogen-associated molecular patterns (PAMPs) or endogenous phytocytokines to regulate immunity. In particular, the Arabidopsis LRR-RKs FLAGELLIN SENSING 2 (FLS2) and ELONGATION FACTOR TU RECEPTOR (EFR) perceive the bacterial PAMPs flagellin (or its peptide epitope flg22) and elongation factor-Tu (or its peptide epitope elf18), respectively, to regulate pattern-triggered immunity (PTI)^4–6.

An emerging common theme in RK signaling is the involvement of receptor-like cytoplasmic kinases (RLCKs). RLCKs are homologous to RKs but lack an extracellular domain, and are thought to act in downstream RK signaling⁷. Members of the large RLCK-VII/PBS1-LIKE (PBL) family in particular have emerged as key components of LRR-RK-mediated signaling, such as the close homologs BOTRYTIS-INDUCED KINASE1 (BIK1) and PBL1, which act downstream of several immune-related RKs, including FLS2 and EFR^8–11, and are thus key executors of PTI¹².

Modulation of receptor signaling is critical to prevent inappropriate activation, and previous work has implicated several protein phosphatases in LRR-RK-mediated immune signaling^13–16, including the Arabidopsis PP2Cs POLTERGEIST-LIKE 4 and 5 (PLL4 and 5) and their rice homolog XB15, which were identified as negative regulators of EFR- and XA21-mediated PTI, respectively^14,15. Accordingly, pll4 pll5 mutants exhibited accelerated kinetics of reactive oxygen species (ROS) production in response to the PAMPs elf18 and flg22 as well as to the plant elicitor peptide AtPep1 (Fig. 1a, b), indicating that like BIK1¹⁷, PLL4 and 5 are common components of PTI signaling downstream of multiple LRR-RKs.

Fig 1: A BIK1-PLL regulatory circuit controls PTI activation.

(a-b) PLL4 and PLL5 regulate ROS burst induction by elf18 (100 nM), flg22 (100 nM) or AtPep1 (1 μM) treatments on 4.5-week-old Arabidopsis leaf discs in Col-0 wildtype and pll4 pll5 backgrounds. (a) Values correspond to the mean of n=12 independent leaf discs (± SE) and are expressed in relative light units (RLU). (b) Histograms represent the time to max RLU (T_max RLU) of n=12 independent biological replicates (± SD). p values from two-tailed T-tests are indicated. Box plots show 25^th to 75^th percentile range with a line at the median and whiskers from minimum to maximum values. (c) PLL4 can dephosphorylate EFR in vitro. In vitro phosphatase assay incubating equal amounts of MBP-tagged full length (FL) or C-terminal (C) WT (PLL4) or inactive PLL4 (PLL4*) with autophosphorylated cytosolic domain of MBP-tagged EFR (EFR-CD). The phosphorylation of MBP-EFR-CD WT or kinase dead (EFR*) was detected by anti-phosphothreonine western blot. (d) Schematic representation of PLL4 domains and details of the S-X-X-L domain of PLL4 and PLL5 homologs. Bold amino acids indicate the S-X-X-L motifs targeted for mutagenesis in PLL4^6A or PLL4^6D; numbers correspond to the amino acids positions within the protein. (e) BIK1 phosphorylates the PLL4 N-terminus in a site-specific manner. Autoradiogram of in vitro kinase assay using WT BIK1 or inactive (BIK1*) BIK1 with FL WT or phosphovariant (PLL4^6A) MBP-PLL4*. (f) In planta flg22-triggered PLL4 dissociation from FLS2 is regulated by phosphorylation. CoIP assay of transiently expressed FLS2-GFP and HA-tagged PLL4 or PLL4^6A in N. benthamiana leaves with or without 100 nM flg22 treatment for 10 minutes. CBB: Coomassie brilliant blue. All experiments were performed at least three times with similar results.

When expressed as a maltose-binding protein (MBP)-fusion protein, wild-type (WT) but not catalytically-dead PLL4 (PLL4*, D280N/D573N) directly dephosphorylated the autophosphorylated cytosolic domain of EFR (EFR-CD) in vitro (Fig. 1c). Interestingly, truncation of the non-catalytic N-terminus rendered PLL4 inactive in this assay, in contrast to previous work with the related phosphatase POLTERGEIST (POL)¹⁸, suggesting an important role for this N-terminal region.

It was previously shown that elf18 perception induced dissociation of PLL4 and 5 from EFR in planta¹⁴; however, the mechanisms mediating such dissociation remain unknown. We observed a similar flg22-induced dissociation of PLL4 and 5 from FLS2 in planta using transient expression in Nicotiana benthamiana (Extended Data Fig. 1), in keeping with our observation that these phosphatases also regulate FLS2 signaling (Fig. 1a, b). To understand how PLL4 and 5 are themselves regulated, we interrogated public databases^19,20 for phosphosites within these proteins. Several clustered, conserved sites were identified in the N-terminus of PLL4 and 5 that conformed to a previously identified [S/T]-X-X-L motif (Fig. 1d, Extended Data Fig. 2), which is targeted by BIK1 and PBL1 in other substrates^10,21.

BIK1 specifically trans-phosphorylated full-length MBP-PLL4* (catalytically-dead PLL4 was used as a substrate to preclude any effect of phosphatase activity) in vitro, confirming that PLL4 is a bona fide BIK1 substrate (Fig. 1e). We then performed liquid chromatography (LC)-mass spectrometry (MS)/MS to identify BIK1-mediated phosphosites on MBP-PLL4*. The majority of identified sites corresponded to the tandem consensus motifs found within this N-terminal S-X-X-L domain (Extended Data Fig. 3a), with L substituted to I in some cases (Fig. 1d), which agrees with previous studies showing that BIK1 targets highly specific motifs within substrate proteins, both in vitro and in vivo^11,21,22. To confirm the specificity of BIK1 phosphorylation, we mutated six putative motif-conforming sites within the PLL4 N-terminus (Fig. 1d) to phospho-dead variants (PLL4^6A). In assays using either full-length PLL4 or the PLL4 N-terminus (PLL4-N), BIK1-mediated trans-phosphorylation of PLL4 was abolished with PLL4^6A (Fig. 1e; Extended Data Fig. 3b), demonstrating that BIK1 specifically phosphorylates these clustered, tandem sites in the PLL4 N-terminus, at least in vitro.

To understand how PLL4 phosphorylation regulates its function, we compared phospho-dead (PLL^6A) or phospho-mimetic (PLL4^6D) mutant variants. Because we were unable to obtain stable transgenic pll4 pll5 lines expressing native promoter-driven PLL4 variants, possibly due to the autoimmunity of pll4 pll5 plants, we examined the in planta function of PLL4 phosphorylation using heterologous expression in N. benthamiana. Treatment with flg22 induced a phosphorylation-dependent shift in the mobility of PLL4 in N. benthamiana, which was not observed with PLL4^6A (Fig. 1f; Extended Data Fig. 3c), suggesting that these N-terminal sites become phosphorylated upon ligand perception. The flg22-induced dissociation of PLL4^6A from FLS2 was partially compromised (Fig. 1f), whereas MBP-PLL4^6D displayed impaired direct interaction with EFR-CD in vitro (Extended Data Fig. 3d). Consistent with a negative role for phosphorylation in PLL4 function, the kinetics of ROS production upon elicitor treatment were dampened by PLL4^6A, while PLL4^6D had no effect (Extended Data Fig. 3e), whereas the plasma membrane association was unaffected in either PLL4 variant (Extended Data Fig. 3f). Together, these data indicate a negative regulatory role for N-terminal phosphorylation on PLL4 function and support a model wherein N-terminal phosphorylation promotes the dissociation of PLL4 from the immune RK complex.

Aside from immunity, many LRR-RKs regulate diverse growth and developmental processes; however, the downstream signaling components of these pathways are often still poorly characterized. LRR-RKs of the CLAVATA 1 (CLV1) and BARELY ANY MERISTEM (BAM1–3) clade perceive endogenous CLE family peptides^23,24. CLE peptides (CLEp) are broadly conserved across land plants²⁵ and regulate important aspects of plant development including stem cell niche maintenance and root development^26,27. Despite the biological importance of CLEp signaling, the molecular components of such pathways are mostly unknown.

In an unbiased forward genetic screen to identify novel components of the CLEp pathway²⁸, two independently isolated mutants were recovered based on their dominant insensitivity to CLE26p²⁹; and both mutants carried a L135F amino acid change in the RLCK-VII subfamily member PBL34. This dominant negative allele (pbl34–2) not only conferred insensitivity to CLE26p, but also to other root-active CLE peptides (Fig. 2a), indicating that PBL34 is required for CLEp signaling. Matching these observations, a PBL34 null allele (pbl34–3) displayed quantitative insensitivity to the same range of root-active CLE peptides (Fig. 2a). The activation of BIK1 by RK-mediated phosphorylation is well-characterized¹², and, consistent with a role downstream of CLEp perception, PBL34 and the cytosolic domain of the CLEp receptor BAM3 (BAM3-CD) were able to directly phosphorylate each other in vitro (Fig. 2b).

Fig 2: The RLCK-VII-5 subfamily is required for CLEp signaling.

(a) The pbl34–2 dominant negative allele is less sensitive to exogenous CLEp than the pbl34–3 loss-of-function allele. 7-day-old seedlings grown on media with 50 nM of indicated CLE peptides. NT: not treated. Letters indicate significant differences within the treatments (ANOVA followed by Tukey test). n=11–50 independent biological replicates. Box plots show 25^th to 75^th percentile range with a line at the median and whiskers from minimum to maximum values. (b) BAM3 and PBL34 transphosphorylate each other in vitro. Autoradiogram of in vitro kinase assay using MBP-tagged WT PBL34 or inactive PBL34 (PBL34*) and GST-tagged WT cytosolic domain (CD) of BAM3 (BAM3-CD) or inactive CD of BAM3 (BAM3*-CD). (c) The L135F mutation disrupts auto- and trans-phosphorylation activity of PBL34. Autoradiogram of in vitro kinase assay incubating equal amounts of GST-tagged BAM3 with MBP-tagged WT PBL34 or mutant forms of PBL34 (PBL34^D275A or PBL34^L135F). Kinase assays in (b) and (c) were performed twice with similar results. (d) CLV3p responses specifically require the RLCK-VII-5 subfamily. 7-day-old seedlings grown on media with CLV3p as indicated. NT: not treated. Letters indicate significant differences within the treatments (ANOVA followed by Tukey test, two-sided). n=26–46. Box plots show 25^th to 75^th percentile range with a line at the median and whiskers from minimum to maximum values. (e) RLCK-VII-5 members are expressed in the root with partially overlapping patterns. Confocal microscopy pictures of 6-day-old seedlings carrying PBL34::3xNLS-VENUS, PBL35::3xNLS-VENUS and PBL36::3xNLS-VENUS constructs, respectively in Col-0 background. yellow channel: 3xNLS-VENUS; cyan: propidium iodide cell wall staining. (f, g) PBL34 is expressed in the root, accumulates in the protophloem and localizes to the cytosol and the plasma membrane. (f) Confocal microscopy images of 5-day-old seedlings expressing PBL34-CIT fusion protein under control of the PBL34 promoter in the rlck-vii-5 triple mutant. PBL34-CIT is detected by anti-GFP antibody immunostaining; yellow channel: CIT; cyan: calcofluor white cell wall staining. (g) Intensity trace of a protophloem cell file showing plasma membrane localization of PBL34-CIT signal.

L135 of PBL34 is highly conserved across the RLCK-VII/PBL family (Extended Data Fig. 4). To understand the dominant negative effect of the pbl34–2 allele, we expressed recombinant PBL34 bearing the causative L135F mutation and tested its kinase activity in vitro. Both auto- and trans-phosphorylation activities of PBL34^L135F were severely reduced compared to WT (Fig. 2c), indicating that kinase activity is essential for PBL34 function in CLEp signaling.

PBL34 belongs to the RLCK-VII-5 subfamily together with PBL35 and PBL36³⁰. Comparison of single and double mutants in these PBL genes revealed that each contributes quantitatively to CLEp sensitivity in root elongation assays (Extended Data Fig. 5a, b). A loss-of-function mutant of all three RLCK-VII-5 subfamily members (rlck-vii-5) led to strongly increased CLEp insensitivity (Fig. 2d, Extended Data Fig. 5a–c), comparable or superior to the pbl34–2 mutant (Extended Data Fig. 5d). A screen of all rlck-vii subfamily polymutants³⁰ revealed that this CLEp insensitivity was unique to rlck-vii-5, indicating that RLCK-VII-5 subfamily PBLs are specifically required for CLEp signaling (Fig. 2d). Notably, the same PBLs were recently similarly implicated in CLEp signaling³¹. In agreement with roles in CLEp signaling, transcriptional reporters indicated that all three RLCK-VII-5 subfamily members are expressed in the root (Fig. 2e).

Expression of a PBL34-CITRINE (PBL34-CIT) fusion protein under control of the PBL34 promoter complemented CLEp responsiveness in the loss-of-function pbl34–3 single mutant (Extended Data Fig. 5e, f) as well as the rlck-vii-5 polymutant (Extended Data Fig. 5g, h), corroborating the predominant role of PBL34 within the clade. PBL34-CIT displayed plasma membrane association (Fig. 2f, g), and was broadly expressed in both the root and shoot apical meristems (Fig. 2f; Extended Data Fig. 5i), consistent with a role in CLEp signaling immediately downstream of CLEp receptors. The CLEp-insensitive pbl34–2 mutant had no effect on immune-related ROS production (Extended Data Fig. 5k), while the rlck-vii-8 polymutant (which includes loss-of-function mutations in both BIK1 and PBL1) responded normally to CLEp treatment in root length assays (Fig. 2d and Extended Data Fig. 6a), indicating pathway-specific RLCK-VII dependency.

We next examined the function of the RLCK-VII-5 clade in specific CLEp-dependent processes controlled by the primary CLV1 and BAM clade LRR-RKs. Consistent with a role in primary CLEp receptor function, CLE45p inhibited protophloem differentiation in a BAM3³² and RLCK-VII-5 dependent manner (Fig. 3a). CLE40p regulates root quiescent center (QC) stem cell maintenance³³, where BAM1, BAM2³⁴, and RLCK-VII-5s are expressed (Fig. 2e, f). CLE40p promoted loss of quiescence and induced QC cell divisions in WT, but not in rlck-vii-5 or bam1 bam2 mutants (Fig. 3b, c). In shoot and floral meristems, CLV3p and highly redundant CLE peptides signal through CLV1 and BAM receptors to limit stem cell proliferation^24,35, and disruption of CLV3p/CLEp signaling thereby results in increased floral organ numbers. Consistent with a general role of RLCK-VII-5s in CLEp perception, rlck-vii-5 mutants also displayed a mildly increased carpel number, which was never observed in WT plants; this rlck-vii-5 phenotype was dramatically enhanced in a sensitized clv1 mutant background (Fig. 3d–e). Collectively, these data demonstrate that RLCK-VII-5s are critical for diverse CLEp-CLV1/BAM developmental outputs.

Fig. 3: The RLCK-VII-5 subfamily functions downstream of multiple primary CLEp receptors.

(a) RLCK-VII-5 kinases control phloem differentiation through CLE45p/BAM3 signaling pathway. Representative confocal images of 5-day-old-seedling meristems (n≥15 independent biological replicates). In pbl34 mutants differentiated phloem files could be observed frequently, which was never observed in Col-0 wildtype. bam3 and rlck-vii-5 are CLE45p-resistant. White asterisks: phloem cell files. (b-c) RLCKV-VII kinases control CLE40p/BAM1/BAM2/POL mediated QC division. CLE40p triggers QC division dependent on RLCK-VII-5 kinases and BAM1 BAM2; while untreated pol displays ectopic QC divisions mimicking CLE40p treatment (b) Representative pictures of confocal microscopy of 5-day-old seedlings treated or not with 100 nM CLE40p. black: calcofluor white cell wall staining. (c) Corresponding quantification of QC division phenotypes in control or 100 nM CLE40p treated plants. n=15–32 independent biological replicates. (d-e) rlck-vii-5 mutants display a mild extra carpel phenotype by itself and enhances the clv1 single mutant. (d) Representative images of carpels per flower with white asterisk indicating a carpel and (e) table representing the average number of carpels per flower (±SD) for every flower on the primary inflorescence of 30 individual 6-week-old plants per genotype and the percent distribution of carpel number per flower.

The roles of specific RLCK-VII family members in immune or CLEp signaling suggests that these LRR-RK pathways utilize similar but genetically distinct signaling components. Interestingly, while PLL4 and 5 regulate immunity, pol and pll1 mutants were first identified as partial suppressors of clv1^36,37, suggesting that they negatively regulate CLEp signaling. Ectopic QC divisions were observed in pol single mutants in the absence of CLE40p (Fig. 3b, c), phenocopying CLEp treatment in WT plants. We observed no such CLE-related phenotypes in pll4 pll5 mutants (Extended Data Fig. 6a–c), whereas pol-6 and pll1–1 displayed WT ROS production upon flg22 or elf18 treatment (Extended Data Fig. 6d), together indicating that, as with RLCK-VII/PBLs, immune and CLEp signaling pathways utilize specific PLL family members.

The clustered S-X-X-L domain phosphosites identified in PLL4 are highly conserved across the PLL family (Extended Data Fig. 2), and database interrogation revealed phosphorylation at seven tandem sites within this region of POL (Fig. 4a). Given the similar roles of POL, PLL1 and PBL34 in CLEp signaling to those of PLL4, 5 and BIK1, PBL1 in PTI, we reasoned that the regulatory mechanisms could be conserved. To test this, we performed in vitro kinase assays with recombinant MBP-tagged POL fragments and confirmed that PBL34 directly phosphorylates the N-termini of POL and PLL1 (Fig. 4b; Extended Data Fig. 7a, b). As with PLL4 and BIK1, we performed phosphosite identification via MS on MBP-POL following kinase assays with PBL34, and confirmed that PBL34 primarily phosphorylates POL within its S-X-X-L domain (Extended Data Fig. 7c). While the PBL34-targeted POL S-X-X-L sites represent bona fide phosphosites (given their presence in publicly available Arabidopsis datasets, Extended Data Fig. 2d), due to expression in limited tissues and endogenous CLEp signaling we were however unable to test whether CLEp perception triggers POL or PLL1 phosphorylation in planta. Nevertheless, PBL34-mediated in vitro transphosphorylation of the POL N-terminus was lost when 7 S-X-X-L motif sites were mutated to phospho-dead variants (POL-N^7A) (Fig. 4b), further confirming that PBL34 specifically phosphorylates the N-terminal S-X-X-L domain of POL. Furthermore, as with PLL4 in immunity, we observed that a phosphomimetic variant of POL (POL^7D) was compromised in association with both BAM3 (Fig. 4c) and CLV1 (Extended Data Fig. 7d) in planta, as well as showed reduced direct interaction with BAM3-CD in vitro (Extended Data Fig. 7e). POL is a plasma membrane associated protein owing to its N-myristoylation, which can be blocked by a G2A mutation³⁸. As with PLL4, the S-X-X-L phosphosites did not affect the plasma membrane association of POL (Extended Fig. 7f).

Fig. 4: Conservation of the regulatory mechanism in PAMP and CLEp receptor complexes.

(a) Schematic representation of POL domains and details of the S-X-X-L domain of POL. Bold amino acids indicate the S-X-X-L residues targeted for mutagenesis in POL^7A or POL^7D; numbers correspond to the amino acid positions within the protein. (b) PBL34 phosphorylates the POL N-terminus in a site-specific manner. Autoradiogram of in vitro kinase assay of MBP-tagged C-terminal (C), N-terminal fragments of WT POL (POL-N) or phospho-dead variant (POL-N^7A) with WT PBL34 or inactive (PBL34*). Kinase assays were performed twice with similar results. (c) BAM3 and POL interact in planta. CoIP assay of transiently expressed BAM3-GFP and HA-tagged WT POL, POL^7A or POL^7D variants in N. benthamiana leaves. CBB: Coomassie brilliant blue. Experiments were performed three times with similar results. (d) Quantification of the complementation of pol pll phenotype in the shoot by POL^7A and POL^7D variants. POL-HA and POL^7A-HA fusion proteins fully complement the seedling lethality of the pol pll1 double mutant. POL^7D-HA protein fusion only partially rescues the seedling lethality phenotype. n≥117 independent biological replicates (seedlings) per genotype. (e) S-X-X-L phosphosites negatively regulate POL function in the sensitized clv1–15 pol-6 background. Table shows the average number of carpels per flower (±SD) for every flower on the primary inflorescence of individual 6-week-old plants for each genotype and the percent distribution of carpel number per flower (n=90 flowers per genotype). (f) Schematic representation of the conserved signaling mechanism between PTI and CLEp signaling pathways exemplified by FLS2 and BAM3 signaling pathways. In the absence of the ligand, PLL family phosphatases damper signaling by inhibiting RK phosphorylation (e.g. PLL4,5 or POL,PLL1). Perception of the apoplastic peptide ligands (e.g. PAMPs or CLEps) by their cognate RKs leads to co-receptor recruitment and activation of specific RLCK-VII members. These RLCK-VII/PBLs phosphorylate PLLs at conserved N-terminal sites, negatively regulating PLLs and promoting their dissociation from the RK complex and appropriate activation of signaling.

To probe the role of POL phosphorylation in CLEp-mediated development, we transformed pol pll1^−/+ null mutants with WT or mutant POL variants under the native POL promoter and isolated viable pol pll1 double mutants to determine ability to complement their seedling lethal phenotype³⁹. Consistent with a negative regulatory role for S-X-X-L sites, POL and POL^7A but not POL^7D complemented pol pll1, whereas POL^7D lines displayed gain-of-function CLV3-phenotypes across plant tissues and developmental stages, and in multiple independent transgenic lines (Fig. 4d; Extended Data Fig. 8). The increased carpel number of clv mutants can be reverted in combination with pol or pll1 single mutants³⁹, in which BAM RKs are still present. We observed that WT POL fully complemented this reversion in a clv1–15 pol-6 background, whereas POL^7D had no effect on carpel number, as expected (Fig. 4e). Complementation with POL^7A, by contrast, caused a quantitative increase in carpel number (Fig. 4e), further confirming that N-terminal phosphorylation negatively regulates POL function in CLV1/BAM signaling.

Together, our data reveal a conserved regulatory circuit that controls LRR-RK signaling in both immunity and development (Fig. 4f), in which specific PLL family phosphatases and RLCK-VII/PBL family kinases control the activation of ligand-binding receptors in distinct pathways. Members of multiple RLCK families function downstream of RK complexes in immunity¹², and the RLCK-VIII family member MAZZA was recently reported to contribute to signaling downstream of CLEp perception⁴⁰. Distinct RLCK-VIIs have also been identified downstream of LRR-RKs in other developmental processes, such as SCHENGEN 1 (SGN1)/PBL15, which functions downstream of the receptor GASSHO1(GSO1)/SGN3 to regulate Casparian strip formation⁴¹, further indicating that LRR-RKs share conserved signaling modules downstream of receptor activation. In addition to their roles reported here, RLCK-VII-5s were also reported to function downstream of LIPOOLIGOSACCHARIDE-SPECIFIC REDUCED ELICITATION (LORE)⁴², a G-lectin type RK that triggers immunity in response to bacterial 3-OH-FAs⁴³. It will be of interest to see if these additional RK pathways may also be regulated by a similar molecular circuitry.

Methods

Plant growth and materials

All the mutants investigated in this study are in the Arabidopsis thaliana Col-0 wildtype background. Col-0 was used as the control for the phenotypic analyses. The allele named pbl34–2 in this manuscript carries a C to T point mutation in PBL34 (At5g15080), which leads to the substitution of leucine 135 by phenylalanine. pbl34–3 (SALK_126209), pol-6 (SALK_009469.29.99.f), pll1–1 (SAIL_319_C08), clv1–15 (WiscDsLox489–492B1), bam1–4 (SALK_107290), bam2–4 (SAIL_1053_E09), pll4 (SALK_203257C), pll5(SALK_044162C), pll4–1 pll5–1 double mutants were obtained from stock centers or described before¹⁴. Higher order rlck-vii seed stocks were previously described³⁰. Unless otherwise detailed, Arabidopsis plants were grown in a controlled environment growth chamber at 150 μmol light intensity, 60% relative humidity, and 20°C in a 10-h light cycle. N. benthamiana plants were grown in a controlled environment chamber at 120 μmol light intensity, 45–60% relative humidity, and 19–21°C in a 12-h light cycle.

CLEp-induced root growth inhibition assays

Seeds were sterilized, sown on half-strength Murashige and Skoog media supplemented with 0.3% sucrose and 1% agar, stocked at 4°C for 48h and grown vertically under continuous light of 120 μmol intensity at 22°C. Synthetic CLE peptides were obtained from a commercial supplier (Genscript) at > 80% purity, diluted in sterile water and used at the indicated concentration. Root lengths were measured on 600 dpi scans of the plates with Fiji software⁴⁴ using the Simple Neurite Tracer plug-in⁴⁵.

CLV3 peptide root elongation assay

Seeds were sterilized for 10 minutes in 70% ethanol with 0.1% Triton X-100, rinsed in 70% ethanol three times, plated onto ½ Murashige and Skoog (MS-Research Products International), pH 5.7 with 8 g of Phytoagar (RPI) per liter. Seeds were stratified for 48 hours at 4°C. After stratification, seeds were germinated horizontally, under continuous light in a Percival growth chamber set to 22°C for four days. Seedlings at 4 days after germination (DAG) were transferred to vertical ½ MS plates with or without CLV3 peptide (>95% purity, Biomatik) for mock and peptide treatment respectively. Seedlings on vertical plates were allowed to grow for 4 days after transfer, then were scanned and measured using ImageJ software.

Phenotypic analysis of QC response to peptide

Sterilized seeds were plated onto ½ Murashige and Skoog (MS-Research Products International), pH 5.7 with 8 g of Phytoagar (RPI) per liter with or without CLE40p peptide (>95% purity, Biomatik) for peptide and mock treatment respectively. They were stratified for 48 hours at 4°C and grown until 5 DAG before imaging. 5 DAG seedlings were stained for 5 minutes in ClearSee solution⁴⁶ containing 0.1% Calcofluor white (ChemCruz), and stained roots were dissected and mounted in ClearSee solution. Roots were observed on a Zeiss 710 inverted confocal microscope and scored into categories. Calcofluor white was excited with 405nm, and emission was collected from 410 to 551 nm.

Cloning

PCR products were amplified from plant DNA or plasmid templates (ABRC) using primers listed in Supplementary Table 1. Mutations were generated using DpnI-mediated site-directed mutagenesis using primers listed in Supplementary Table 1. For Gateway cloning, PCR products were successively transferred to pDONR vectors by BP reaction (Invitrogen) and pDEST by LR reaction (Invitrogen) according to the manufacturer’s protocols. pPBL34::gPBL34-CIT was cloned by Gibson strategy (NEB) into a modified pCAMBIA1305,1 plasmid carrying a FASTRED seed selection marker. POL, PLL1, PLL4, and PBL34 fragments were cloned into pOPINM using InFusion (Takara). The PBL34 promoter was cloned by restriction enzyme cloning into a modified pCAMBIA1305,1 carrying a 3xNLS-VENUS cassette. The PBL35 and PBL36 promoters were introduced into a pCAMBIA1305,1 3xNLS-VENUS by Gibson cloning (NEB). The POL native promoter (3.6 kb-5’, 1.0 kb-3’) was integrated into the pMOA34 binary gateway destination vector via PCR and standard restriction cloning. For transient localization experiments in N. benthamiana, POL or PLL4 variants were cloned behind the 35S promoter binary vector and N-myristoylatable mTurquoise2 was cloned into behind the Ubiquitin10 promoter binary vector.

Plant transformation

Binary vectors were introduced into Arabidopsis via Agrobacterium tumefaciens-mediated (strain GV3101 pMP90) transformation by standard floral dipping. Transgenic lines were selected on hygromycin selection media (35 mg/l) or FASTRED seed expression. Single insertion lines were studied.

pbl34–2 mutant isolation

pbl34–2 mutants were isolated as described²⁸. CLE26p insensitivity was confirmed in the M3 generation and resistant plants were backcrossed to Col-0. The F1 was uniformly insensitive to CLE26p treatment, suggesting the dominance of this allele. The causative mutation was mapped by whole-genome sequencing of a bulk of 100 seedlings resistant to CLE26p versus 100 sensitive ones as described⁴⁷.

Genotyping

The pbl34–2 mutation was genotyped with a CAPS strategy. A 730bp PCR product was amplified with the Phire kit (Thermo Fisher). The subsequent PCR product was digested with AflII restriction enzyme, which cuts the wild-type product into 340bp + 390bp fragments but not the pbl34–2 product. Primers for genotyping are listed in Supplementary Table 2.

Protein alignment and phylogenic tree

Protein alignments were performed using CLUSTALW (https://www.ebi.ac.uk/Tools/msa/clustalo/). The output file was uploaded into MEGA X software (https://www.megasoftware.net/) to generate the corresponding phylogenic trees.

Statistical analyses

Statistical analyses were performed on RStudio software (www.rstudio.com/) or on Prism software (https://www.graphpad.com/scientific-software/prism/). ANOVA analyses followed by Tukey tests were performed with a confidence level of 95%. Specific tests used are indicated in figure captions.

Recombinant protein expression and purification

All proteins were expressed in Escherichia coli strain BL21(DE3) Rosetta pLysS unless otherwise noted. BIK1 or BIK1* (kinase dead, K105A/K106A), PLL4, POL, PLL1, and PBL34 variants expressed as 6xHis-MBP fusion proteins in the pOPINM vector. The EFR cytosolic domain was expressed using pMAL-c4E (MBP-EFR-CD or kinase-dead MBP-EFR*-CD, D849N) in BL21(DE3) Rosetta pLysS or pET-28a(+) (6xHis-EFR-CD) in BL21(DE3)-VR2-pACYC-LamP E. coli, respectively. The cytosolic domain of BAM3 WT or kinase dead (BAM3*, D836N) was cloned into a modified pET-28a(+) backbone and expressed as a 6xHis-GST fusion protein (GST-BAM3-CD). All proteins were purified using Amylose Resin (NEB) or HisPur Cobalt Resin (Thermo) for MBP or 6xHis, 6xHis-MBP, and 6xHis-GST fusions, respectively.

In vitro kinase assays

Approximately 1 μg of kinase was incubated with approximately 1 μg of substrate protein in kinase buffer (25 mM Tris-Cl pH 7.4, 5 mM MnCl₂, 5 mM MgCl₂, 1 mM DTT). Reactions were initiated by addition of 5 μM ATP plus 0.5 μCi ³²P-γ-ATP in a final reaction volume of 30 μl. Reactions were carried out at 25 °C for 30 min and stopped by addition of SDS-loading dye and heating at 70 °C for 10 min. Proteins were resolved by SDS-PAGE, transferred to PVDF membrane, and stained with Coomassie brilliant blue G-250. Autoradiographs were imaged using an Amersham Typhoon phosphorimager (GE Healthcare). For non-radioactive autophosphorylation of EFR, approximately 10 μg of MBP-EFR-CD was incubated in kinase buffer (as above) with 10 μM ATP in a final volume of 100 μl for 1 hour at 25 °C. Free ATP was removed by equilibration into storage buffer (25 mM Tris-Cl pH 7.4, 100 mM NaCl, 10% glycerol, 1 mM DTT) using a 10,000 MWCO centrifugal filtration device (Millipore). For analysis by MS, in vitro kinase assays were performed as above without the addition of ³²P-γ-ATP.

In vitro phosphatase assays

Approximately 1 μg of autophosphorylated MBP-EFR-CD was mixed with approximately 1 μg of phosphatase in buffer (HEPES pH 6.8, 5 mM MgCl₂, 5 mM MnCl₂, 200 mM NaCl, 5% glycerol, 1 mM DTT). Reactions were carried out for 90 min at 25 °C. Phosphorylation was monitored by blotting with anti-pThr (anti-phosphothreonine, Cell Signaling Technology 9381, diluted 1:1000 in TBST-5% gelatin from cold water fish skin).

ROS production assays

ROS burst assays were conducted as previously described^10,48. For assays with N. benthamiana, leaf discs were harvested two days after infiltration and equilibrated overnight in sterile water and used for assays at three days after infiltration. For kinetic analyses , RLUs were collected in 30-s intervals for 60 min. T_max RLU was defined as the interval with the highest total value.

Transient expression

Leaves of 4-week-old N. benthamiana leaves were infiltrated with A. tumefaciens carrying constructs as indicated in figure captions. In all cases cultures were co-infiltrated with A. tumefaciens carrying a P19 suppressor of gene silencing construct.

Protein extraction and co-immunoprecipitation

For co-IP, N. benthamiana leaves were detached and bisected 2 days post-infiltration. Leaf halves were equilibrated in liquid MS 1% sucrose (1–2 hours) and subsequently vacuum infiltrated with MS or MS+PAMP as indicated in figure captions. Tissue was frozen and ground in liquid nitrogen. Protein extraction and immunoprecipitation were performed as described previously¹⁰ using GFP-trap (Chromotek) or GFP-clamp⁴⁹ resin, as indicated. Proteins were separated by SDS-PAGE and blotted onto PVDF membrane. Membranes were blocked and probed in TBST-5% non-fat milk using anti-GFP (HRP-conjugated B-2, sc-9996 HRP, Santa Cruz, 1:5000 dilution) or anti-HA (HRP-conjugated, 12013819001, Roche, 1:3000 dilution).

For flg22-induced PLL4-HA bandshift, extractions were performed as above. Extracts were then diluted into 1x phosphatase buffer (HEPES pH 6.8, 5 mM MgCl₂, 5 mM MnCl₂, 200 mM NaCl, 5% glycerol) with or without 200 units λPPase (NEB) and incubated for 40 minutes at RT. Samples were then analyzed by western blot as described above.

For detection of POL-HA from POL::POL-HA lines, tissue from 20 8-day-old seedlings grown on ½ MS was flash frozen in liquid nitrogen and protein extraction was performed as above. Proteins were separated by SDS-PAGE (4–12% Bolt Gel run in 1x Bolt MOPS SDS running buffer from Thermo) and transferred to PVDF membrane. Membranes were blocked and probed in TBST-4% non-fat milk using anti-HA antibody as primary (Clone 3F10, Roche, 1:1000 dilution) and anti-rat HRP-conjugated antibody as secondary (Polyclonal antibody, Santa Cruz Biotech, 1:5000 dilution).

Sample preparation for mass spectrometry

Affinity-purified protein samples were separated on a 4–12% NuPAGE gel (Invitrogen). The gel was stained with InstantBlue (abcam) and the band corresponding to MBP-PLL4* or MBP-POL* was excised, cut into smaller pieces and washed three times with 50% acetonitrile, 50 mM ammonium bicarbonate (50 % AcN/ABC), 30 min each, followed by dehydration in acetonitrile, 10 min. Gel pieces were then reduced with 10 mM DTT for 30 min at 45 °C followed by alkylation with 55 mM iodoacetamide for 20 min at room temperature, and a further three washes with 50% AcN/ABC, 30 min each. Gel pieces were dehydrated again with acetonitrile before rehydration with 40 μl trypsin (Pierce Trypsin Protease, MS-Grade, catalog no. 90058) working solution (100 ng trypsin in 50 mM ammonium bicarbonate, 5% (v/v) acetonitrile). Where required, gel pieces were covered with 50 mM ammonium bicarbonate to a final volume before incubation at 37 °C overnight. Tryptic peptides were extracted from the gel pieces three times in an equal volume of 50% acetonitrile, 5% formic acid (Pierce LC-MS-Grade, catalog no. 85178), 30 min each. Extracted peptides were dried in a speed-vac and resuspended in 2% acetonitrile/0.2% trifluoroacetic acid (Merck, catalog no. 302031). A total of four biological replicates for each sample type was submitted.

LC-MS/MS analysis

Approximately 35% of each sample was analysed using an Orbitrap Fusion™ Tribrid™ Mass Spectrometer (Thermo Fisher Scientific) coupled to a U3000 nano-UPLC (Thermo Fisher Scientific). The dissolved peptides were injected onto a reverse phase trap column NanoEase m/z Symmetry C18, beads diameter 5 μm, inner diameter 180 μm × 20 mm length (Waters). Trap column flowrate was 20 μl/min in 2% acetonitrile, 0.05% TFA. Peptides were eluted from trap column onto the analytical column NanoEase m/z HSS C18 T3 Column, beads diameter 1.8 μm, inner diameter 75 μm × 250 mm length (Waters). The column was equilibrated with 3% B (B: 80% acetonitrile in 0.05% formic acid (FA), A: 0.1% FA) before subsequent elution with the following steps of a linear gradient: 2.5 min 3% B, 5 min 6.3% B, 13 min 12.5% B, 50 min 42.5% B, 58 min 50% B, 61 min 65% B, 63 min 99% B, 66 min 99% B, 67 min 3% B, 90 min 3% B. The flow rate was set to 200 nl/min. The mass spectrometer was operated in positive ion mode with nano-electrospray ion source. Molecular ions were generated by applying voltage +2.2kV to a conductive union coupling the column outlet with fused silica PicoTip emitter, ID 10 μm (New Objective, Inc.) and the ion transfer capillary temperature was set to 275°C. The mass spectrometer was operated in data-dependent mode using a full scan, m/z range 300–1,800, nominal resolution of 120,000, target value 1 × 10⁶, followed by MS/MS scans of the 40 most abundant ions. MS/MS spectra were acquired using normalized collision energy of 30%, isolation width of 1.6 m/z, resolution of 120,000, and a target value set to 1 × 10⁵. Precursor ions with charge states 2–7 were selected for fragmentation and put on a dynamic exclusion list for 30 seconds. Multistage activation was applied for detection of −98, −49, or −32.7 from the precursor (corresponding to the neutral loss of phosphoric acid from 1+, 2+, and 3+ charge states, respectively) during any of the MS/MS scans. The minimum automatic gain control target was set to 5 × 10³ and intensity threshold was calculated to be 4.8 × 10⁴. The peptide match feature was set to the preferred mode and the feature to exclude isotopes was enabled.

Data processing and peptide identification

Peak lists in the form of Mascot generic files were prepared from raw data files using MS Convert (Proteowizard) and sent to a peptide search on Mascot server v2.7 using Mascot Daemon (Matrix Science, Ltd.) against an in-house constructs and contaminants database and the E. coli K12 protein database. Tryptic peptides with up to 1 possible mis-cleavage and charge states +2, +3 were allowed in the search. The following peptide modifications were included in the search: carbamidomethylated Cysteine (fixed), oxidized Methionine (variable) and phosphorylated Serine, Threonine and Tyrosine (variable). Data were searched with a monoisotopic precursor and fragment ion mass tolerance 10 ppm and 0.8 Da respectively. Decoy database was used to validate peptide sequence matches. Mascot results were combined in Scaffold v4.4.0 (Proteome Software Inc.) and filtered to show only phospho-peptides. Peptide and protein identifications were accepted if peptide probability and protein threshold was ≥ 80.0% and 99% respectively. Data was then exported to Excel (Microsoft) for further processing. For each phospho-peptide, spectral counts from all biological replicates were summed. Spectral counts from miss-cleaved peptides identifying the same phosphorylation site were then summed to give final spectral counts for each site.

In vitro pulldowns

Approximately 6 μg each of bait and prey proteins were mixed to 100 μl final volume in buffer (25 mM Tris-Cl pH 7.4, 100 mM NaCl, 0.2% Triton-X, 1 mM DTT). 30 μl was removed (“input”) and the remaining sample was mixed with 50 μl of Amylose Resin (NEB) in a final volume of 500 μl. Samples were mixed at RT for 30 min. The resin was washed three times with buffer and enriched proteins were eluted with 50 μl SDS-loading dye (“pulldown”). Samples were separated by SDS-PAGE, transferred to PVDF, and imaged by blotting with anti-Polyhistidine (Sigma H1029), anti-GST (Upstate 06332), or anti-MBP (NEB E8032) antibodies (all diluted 1:10,000 in TBST-5% non-fat milk powder).

Confocal microscopy

5 to 6-day-old seedlings were imaged using an SP8 (Leica) inverted confocal microscope. Samples were prepared in a drop of 0.04 mg/ml propidium iodide solution. CITRINE (CIT) and VENUS fluorophores were exited at 514 nm and emitted light recorded between 520 and 555 nm. Propidium iodide was excited at 488 and 514 nm and fluorescent light recorded between 600 and 700 nm. CIT/VENUS and propidium iodide channels were sequentially acquired. For localization in N. benthamiana, mYPet (POL) or eYFP (PLL4) and mTurquoise2 controls were excited at 514 and 458 nm, respectively, and emitted light emitted light recorded between 516–593 nm and 460–520 nm, respectively, on a Zeiss 710 confocal microscope. Localization controls and single transformed cells were used to exclude channel bleed through. Figures were prepared using Fiji software or Zen (Zeiss). PBL34-CIT shoot apical meristem images were acquired on a Zeiss 710 with an InverterScope attachment as described⁵⁰. CIT and chlorophyll were excited at 514 nm and emitted light emitted light recorded for CIT (519–548 nm) and chlorophyll (602–728 nm).

Immunolocalization

5-day-old-seedlings of transgenic line expressing pPBL34::gPBL34-CIT in the rlckv-vii-5 background were used for whole mount immunolocalization of the PBL34-CIT protein fusion as described previously⁵¹ and combined with calcofluor white cell wall staining. Primary anti-GFP rabbit (Abcam) antibody was used at 1:600 and secondary Alexa Fluor 546 anti-rabbit antibody was used at 1:500 (Molecular Probes).

Carpel counts

Genotype confirmed seeds were sterilized, stratified, and germinated as above. Four DAG, seedlings were transplanted to soil (7 parts top soil to 1 part sand with pesticide) and kept at high humidity for 3–5 days under continuous 24-h light at 23 °C. Seedlings were removed from high humidity and allowed to grow to full maturity, with gentle staking to prevent tangling at ~3 weeks after transplant. After 5 weeks of growth, the entire number of flowers produced on the primary inflorescence were quantified for carpel number under a dissecting microscope. Data was analyzed in PRISM as above.

Isolation of POL complementation lines

POL constructs were introduced into pol-6 pll1–1/+ line and transgene fixed lines were isolated in the pol-6 pll1–1 background (or pol-6 pll1–1/+ if necessary). Fixed lines were used for complementation analysis and expression analysis. Initial complementation was determined by identifying the ratio of viable to seedling lethal phenotypes for each POL variant. The WT and 7A lines displayed 100% viable phenotypes, while the 7D variants displayed a mixed set of phenotypic ratios. Viable 7D lines were transplanted and assayed for post-seedling stem cell defects.

qPCR

Bulk 8-day-old pol pll1 POL::POL-HA seedlings (12 for WT phenotype, 20 for seedling lethal phenotypes, and an approximate weight mix for mixed phenotype) were ground in liquid nitrogen using SPEX Genogrinder and glass beads. The resulting powder was used as the starting material for a standard RNA extraction using a Qiagen Plant Mini Kit (Qiagen). ~1 ug of RNA from resulting extraction was used as template in standard cDNA synthesis reaction (BioRad iScript cDNA Synthesis Kit). 200 ng of resulting cDNA was used in qRT-PCR reactions to quantify POL expression levels (PowerUp SYBR Green 5x Master Mix-Thermo). CDKA1 was used as an equalization housekeeping gene. Data was analyzed using the ΔΔCT method for three biological replicates with three technical replicates each.

POL complementation analysis

For seedling analyses, seeds of POL substitution lines were sterilized and plated on ½ MS plates (described above). 8 day old seedlings were assayed for seedling lethal pol pll double mutant or WT phenotype. Representative individual plants were genotyped to confirm correct genotype. For adult plant analyses, 5 day-old-seedlings were transplanted to soil (as above). Plants were grown for 4–6 weeks and phenotypes were assayed accordingly.

Data availability

Identifiers for published or publicly available lines are provided in the Methods. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE⁵² partner repository with the dataset identifier PXD031441 and 10.6019/PXD031441. All other relevant data are available from the corresponding authors upon request.

Extended Data

Extended Data Fig. 1. PLL4 and PLL5 dynamically associate with EFR and FLS2 in ligand-dependent manner.

(a) elf18 triggers EFR/PLL4,5 dissociation in planta. CoIP assay of transiently expressed EFR-GFP and HA-tagged PLL4 or PLL5 in N. benthamiana leaves with or without treatment with 1 μM of elf18 for 10 minutes. (b) flg22 triggers FLS2/PLL4,5 dissociation in planta. CoIP assay of transiently expressed FLS2-GFP and HA tagged PLL4 or PLL5 in N. benthamiana leaves with or without treatment with 1 μM of flg22 for 10 minutes. CBB: Coomassie brilliant blue. Experiments were performed twice with similar results.

Extended Data Fig. 2. The S-X-X-L domain is conserved among the PLL family.

(a) Schematic representation of PLLs protein domains. light blue: S-X-X-L domain, green: catalytic domains. Numbers indicates amino acid residues. (b) Protein alignment of the S-X-X-L domain of POL family members. Pink square: S-X-X-L residues; bold: residues targeted for mutagenesis in current study in POL (black arrow head) and PLL4 (white arrow head). (c) Sequence logo of the S-X-X-L domain of the POL/PLL family created from the alignment in panel B using the WebLogo3 online application (http://weblogo.threeplusone.com). (d) Table resuming the in silico predicted phosphorylation in the S-X-X-L domain identified in PhosPhAt4¹⁹ and/or Athena databases²⁰. (e) Phylogenetic tree of the POL family based on protein alignment. PP2C38 is used as an outgroup.

Extended Data Fig. 3. PLL4 regulates PTI in a phosphorylation dependent manner.

(a) BIK1-dependent phosphosites identified on MBP-PLL4* by LC-MS/MS following in vitro kinase assays. Numbers indicate sum of total spectra identified across 4 replicates corresponding to phosphorylation at the specified residue(s) in PLL4. Some spectra were ambiguous for two phosphorylation positions. Sites within the S-X-X-L domain of PLL4 are outlined and spectra corresponding to sites targeted for mutagenesis are highlighted. (b) BIK1 phosphorylates S-X-X-L motif sites within the PLL4 N-terminus. Autoradiogram of in vitro kinase assay PLL4-N (WT) or PLL4^6A-N. Kinase assays were performed three times with similar results. (c) Treatment with 100 nM flg22 induces a phosphosite-dependent mobility shift in PLL4-HA. N. benthamiana leaves transiently expressing FLS2-GFP and PLL4-HA or PLL4^6A-HA were treated with or without 100 nM flg22 for 10 min prior to protein extraction, λPPase treatment, and blotting. Assays were performed three times with similar results. (d) PLL4 phosphomimetic (PLL4^6D) mutation disrupts direct interaction between PLL4 and EFR-CD in vitro. Amylose pulldown assay of 6xHis-tagged cytosolic domain (CD) of EFR with MBP-tagged WT version (PLL4) or phosphovariant (PLL^6D) of PLL4. Assays were performed three times with similar results. (e) Expression of PLL4^6A dampens PTI responses in N. benthamiana. ROS burst induction by elf18 (100 nM) on leaf discs of N. benthamiana transiently expressing FLS2-GFP and PLL4-HA variants (curves show mean values ± SE). Bottom panel displays the time to maximum response; letters indicate statistically different values (one-way ANOVA with Dunnett’s test); n=24 leaf discs from 3 biologically-independent leaves. Box plots show 25^th to 75^th percentile range with a line at the median and whiskers from minimum to maximum values. Assays were performed three times with similar results. (f) S-X-X-L phosphorylation does not alter accumulation of PLL4-YFP at the plasma membrane. Confocal images showing PLL4-YFP (WT or mutant variants; yellow channel) and Myr-mTQ2 (myristoylatable-mTurquoise2, PM marker; blue channel) following co-expression in N. benthamiana leaves.

Extended Data Fig. 4. L135F is conserved among the PBL clade.

Protein alignment of the PBL family. Magenta square: RLCK-VII-5 clade; cyan square: RLCK-VII-8 clade; yellow square: the highly conserved L residue.

Extended Data Fig. 5. RLCK-VII-5 clade kinases are specifically required for CLEp perception.

(a-c) RLCK-VII-5 members are semi-redundant in the CLEp signaling. (a) Root length of 8-day-old seedlings grown on media with or without 100 nM CLV3p. NT: not treated. Kruskal Wallis non-parametric ANOVA test, **** indicates p value <0.0001, *** indicates p value <0.001, * indicates p value <0.001, ns: not significant, n=30 independent biological replicates. (b) Corresponding growth ratio inhibition observed in (a) calculated as following: $\left(\frac{rootlengt{h}_{NT}-rootlengt{h}_{CLV3}}{rootlengt{h}_{NT}}\right)\times 100$. All of the rlck-vii-5 mutant combinations were less sensitive to CLV3p than the Col-0 background. (c) 7-day-old-seedlings grown on media with 100 nM of indicated CLE peptides. NT: not treated. Letters indicate significant differences within the treatments (ANOVA followed by Tukey test, two-sided). n=26–46 independent biological replicates. (d) rlck-vii-5 is less sensitive to CLV3p and CLE45p treatments than the dominant negative pbl34–2 mutant. 7-day-old-seedlings grown on media complemented with 100 nM of indicated peptides. NT: not treated. n=19–45 independent biological replicates. (e-f) Complementation of pbl34–3 mutants expressing PBL34::gPBL34-CIT construct. 7-day-old seedlings grown on media complemented with 50 nM CLEp. NT: not treated. Letters indicate significant differences within the treatments (ANOVA followed by Tukey test, two-sided). (e) Complementation assay on CLE45p media, n=37–58 independent biological replicates. (f) Complementation assay on CLE26p media. n=17–28 independent biological replicates. (g-h) Complementation of rlck-vii-5 mutants expressing PBL34::gPBL34-CIT construct. (g) Eight-day-old seedlings grown on media supplemented with 100 nM CLV3p (g; n=25–31 independent biological replicates) or (h) seven-day-old seedlings grown on media supplemented with 50 nM CLE45p (h; n=22–47 independent biological replicates). NT: not treated. Letters indicate significant differences within the treatments (ANOVA followed by Tukey test, two-sided). (i) Expression of PBL34::gPBL34-CIT in the shoot apical meristem. Yellow, PBL34-CIT; blue, chlorophyll. Image represents cross section through L4 of SAM capturing the organizing center. Low levels of expression were detected in all cell layers and zones of the SAM and developing floral primordia and meristems. (k) ROS production in response to 100 nM flg22 or elf18. Curves show mean values ± SE, n=12 independent leaf discs. All box plots show 25^th to 75^th percentile range with a line at the median and whiskers from minimum to maximum values. ROS assays were performed three times with similar results.

Extended Data Fig. 6. Pathway-specific roles of PLL family members.

(a) Root length of 7-day-old seedlings in response to 15 nM CLE45p. NT=not treated (n=15–32 independent biological replicates). Different letters indicate statistically significant differences (p<0.01, ANOVA followed by Tukey test, two-sided). Box plots show 25^th to 75^th percentile range with a line at the median and whiskers from minimum to maximum values. (b) QC divisions in response to 100 nM CLE40p. (c) Quantification of response shown in (b); n=10–18 independent biological replicates. (d) PAMP-induced ROS in pll mutants. Curves show mean values ± SE, n=12 independent leaf discs. ROS assays were performed three times with similar results.

Extended Data Fig. 7. Conservation of RK-PBL-PLL circuitry in CLEp signaling.

(a) POL is a substrate of active PBL34. Autoradiogram of in vitro kinase assay incubating equal amounts of MBP-tagged POL with MBP-tagged WT PBL34 or mutant forms of PBL34 (PBL34^D275A or PBL34^L135F). (b) PBL34 phosphorylates PLL1 in vitro. In vitro kinase assay incubating equal amounts of MBP-tagged WT version (PBL34) or inactive (PBL34*) of PBL34 recombinant protein with MBP-tagged N-terminus (PLL1-N), catalytically-dead full length (PLL1*-FL), or catalytically-dead C-terminus (PLL1*-C). CBB: Coomassie brilliant blue. Kinase assays in (a) and (b) were performed twice with similar results. (c) PBL34-dependent MBP-POL phosphosites identified by LC-MS/MS following in vitro kinase assays. Numbers indicate sum of total spectra identified across 3 replicates corresponding to phosphorylation at the specified residue(s) in POL. Some spectra were ambiguous for two phosphorylation positions. Sites within the S-X-X-L domain of POL are outlined and spectra corresponding to sites targeted for mutagenesis are highlighted. (d) POL phosphorylation status determines its interaction with CLV1 in planta. CoIP assay of GFP-tagged CLV1 with HA-tagged WT POL or phosphovariants (POL^7A or POL^7D). (e) POL phosphosites control direct interaction with BAM3 in vitro. Amylose pulldown assay using equal amounts of GST-tagged cytosolic domain (CD) of BAM3 with MBP-tagged WT (POL) or phosphomimetic (POL^7D) variants of POL. Experiments in (d) and (e) were performed three times with similar results. (f) S-X-X-L phosphorylation does not regulate POL-YFP accumulation at the plasma membrane. Confocal images showing POL-YFP (WT or mutant variants; yellow channel) and Myr-mTQ2 (myristoylatable-mTurquoise2, PM marker; blue channel) following co-expression in N. benthamiana leaves.

Extended Data Fig. 8. PBL-phosphorylation sites negatively regulate POL function.

(a-c) POL phosphovariants complement pol pll1 to varying degrees. (a) Representative pictures of 4-week-old plants expressing different phosphovariants of POL-HA protein fusion (WT, POL^7D or POL^7A) under control of native POL promoter. (b) Representative pictures of 6-week-old stems displaying post-seedling stem cell defect-terminated silique phenotype. white asterisk: terminated silique; cyan arrow head: one successful silique formation for entire line of POL^7D-HA #1. (c) Corresponding quantification of the shoot complementation based on terminated silique phenotype. n≥942 siliques from ≥30 independent biological replicates (plants) per genotype. (d) Detailed pictures of terminated flower compared to WT non-terminated one. cyan arrow head: presence of the pistil; white arrow head: absence of pistil. Scale bar: 3 mm. (e) Western blot showing expression of POL-HA in POL::POL-HA complementation lines. (f) Relative fold change (ΔΔC_T) expression analysis of POL in complementation lines. n=3 independent biological replicates. Error bars indicate ±SD.