ERECTA family signaling constrains CLAVATA3 and WUSCHEL to the center of the shoot apical meristem

The shoot apical meristem (SAM) generates new organs throughout the life of a plant. As stem cells in the central zone of the dome-shaped SAM slowly divide, some of their progeny are displaced laterally into the peripheral zone and basally into the rib zone. Cells in the peripheral and rib zones rapidly divide, differentiate and are incorporated into forming leaves, flowers and stems. Even though cells are constantly dividing, the SAM size remains stable throughout development owing to a tight balance of proliferation and incorporation of cells into new organs.

The principal regulator of SAM size is a negative feedback loop consisting of WUSCHEL (WUS; AT2G17950) and CLAVATA3 (CLV3; AT2G27250) (Fuchs and Lohmann, 2020). WUS is a homeodomain transcription factor that maintains the pool of stem cells; in its absence stems cells arise but almost immediately differentiate (Laux et al., 1996). WUS is expressed in the organizing center beneath the central zone, and the protein moves up into the central zone through plasmodesmata (Brand et al., 2000; Daum et al., 2014; Mayer et al., 1998; Schoof et al., 2000; Yadav et al., 2011). CLV3 encodes a secreted peptide expressed in the central zone and perceived by multiple plasma membrane localized receptors: CLV1 (AT1G75820), CLV2, BARELY ANY MERISTEM 1 (BAM1; AT5G65700), BAM2 (AT3G49670), CORYNE (CRN), RECEPTOR-LIKE PROTEIN KINASE 2 (RPK2) and CLAVATA3 INSENSITIVE RECEPTOR KINASEs (CIKs) (DeYoung et al., 2006; Fletcher et al., 1999; Hu et al., 2018; Kinoshita et al., 2010; Müller et al., 2008; Shinohara and Matsubayashi, 2015). In the central zone, WUS binds directly to the promoter of CLV3 and activates its expression while CLV3-activated signaling inhibits WUS expression (Brand et al., 2000; Schoof et al., 2000; Yadav et al., 2011), forming a regulated feedback loop.

One of the central questions to understanding the meristem is how the spatial expressions of CLV3 and WUS are established and maintained in the face of the continual turnover of cells. Previous studies have elucidated key mechanisms underlying the apical-basal distributions of CLV3 and WUS. The depth of the WUS expression domain is defined by opposing activity of CLV3 and cytokinin: CLV3 inhibits WUS expression whereas cytokinin signaling promotes it. Both signals are produced in the apical region of the meristem and form a diffusion gradient along the apical basal axis. Cytokinin is perceived in deeper tissue layers than CLV3, which establishes WUS expression at a certain distance from the surface of the SAM (Chickarmane et al., 2012). Confinement of CLV3 expression to the region above the WUS domain is dependent on HAIRY MERISTEM (HAM) transcription factors in the rib zone. Interaction with HAM transcription factors prevents WUS from activating CLV3 transcription, which restricts CLV3 expression to the apical region of the SAM (Zhou et al., 2018). However, it remains unclear why WUS and CLV3 are expressed only around the central vertical axis of the SAM. Previous mathematical models have used implicit or explicit assumptions to define the lateral boundary that confines the expression of WUS and CLV3 (Gruel et al., 2018; Zhou et al., 2018), but little is known about the actual existence of such a lateral signal. A recent model of SAM growth in the clv3 mutant suggested the existence of additional mechanisms sustaining the peripheral zone (Klawe et al., 2020). Here, we present data showing that ERECTA family signaling restricts WUS and CLV3 expression laterally, confining them to the center of the meristem thereby providing a key mechanism for SAM maintenance.

ERECTA (ER; AT2G26330), ERECTA-LIKE 1 (ERL1; AT5G62230) and ERL2 (AT5G07180), collectively called ERfs, encode plasma membrane-localized leucine-rich repeat receptor-like kinases (Shpak et al., 2004). The activity of ERf receptors is regulated by a group of cysteine-rich peptides belonging to the EPIDERMAL PATTERNING FACTOR/EPF-LIKE (EPF/EPFL) family (Hara et al., 2007, 2009; Lee et al., 2012; Lin et al., 2017). A mitogen-activated protein kinase cascade consisting of YODA, MKK4/5/7/9 and MPK3/6 functions downstream of the receptors (Bergmann et al., 2004; Lampard et al., 2009, 2014; Meng et al., 2012; Wang et al., 2007). ERf signaling controls various developmental processes including stomata formation, above-ground organ elongation, SAM size, leaf initiation and phyllotaxy (Chen et al., 2013; Shpak, 2013; Uchida et al., 2013). In the SAM, three ERfs function redundantly, with single and double mutants having no or extremely weak meristematic phenotypes. Altered meristem development can be observed in the er erl1 erl2 mutant, which has a wider vegetative SAM and forms fewer leaves at almost random divergence angles (Chen et al., 2013; Uchida et al., 2013). Recently, we demonstrated that ERf activity in the SAM is controlled by four ligands – EPFL1 (AT5G10310), EPFL2 (AT4G37810), EPFL4 (AT4G14723) and EPFL6 (AT2G30370) – which are expressed at the periphery of the SAM (Kosentka et al., 2019). In-depth transcriptome profiling of Aquilegia coerulea identified a homolog of EPFL4 as a most connected hub in the flower meristem gene network, further confirming the importance of EPFL4 for meristem maintenance (Min and Kramer, 2020).

Based on altered expression of DR5rev:GFP and PIN1pro:PIN1-GFP markers in er erl1 erl2, we proposed that the decrease in leaf initiation might be a result of altered auxin distribution (Chen et al., 2013); the cause of the increased meristem size, however, has remained unknown. A decrease in organ initiation does not automatically lead to an increase of SAM size. For example, the inflorescence meristem of the pin1 mutant does not show alteration of size or WUS expression, although it fails to produce flower organs (Vernoux et al., 2000). As regulation of SAM size depends on the CLV3/WUS feedback loop, we investigated whether ERfs genetically interact with these two genes and alter their expression. Our experiments indicate that ERfs are important modulators of CLV3 and WUS expression. We propose that ERf and EPFLs are a part of a new regulatory circuit that enables communication between the peripheral and the central zones and specifies the location and size of the stem cell population in the SAM.

In a previous study, 10-day old clv3 er erl1 erl2 seedlings had a bigger SAM compared with both er erl1 erl2 and clv3 (Kimura et al., 2018). Here, we investigate how the ERf and CLV3 signaling pathways control meristem maintenance over time and their effect on leaf initiation and internode formation. Comparison of clv3 with er erl1 erl2 mutants suggests that, although both ERfs and CLV3 control SAM size, they play dominant roles during different developmental stages. At 1 day post germination (DPG) the SAM of er erl1 erl2 is considerably wider (105.1±3.3; mean±s.e.m.) than in the wild type (56.1±1.7; P<1.2×10⁻¹³ in unpaired two-tailed Student's t-test) or in clv3 (83.0±1.4; P<4.2×10⁻⁷) (Fig. 1A,C), suggesting a key role for ERfs in restricting SAM size during embryogenesis. For the first 5 days after germination, the wild type and er erl1 erl2 SAMs do not substantially further increase in width whereas clv3 SAM size continues to increase, indicating that post embryogenesis CLV3 signaling plays the primary role in SAM size maintenance (Fig. 1A). The two pathways also contribute differently to leaf initiation, with ERfs promoting leaf initiation and CLV3 slightly inhibiting it (Fig. 1B).

The most dramatic phenotype is observed when both signaling pathways are deactivated. The clv3 er erl1 erl2 mutant has a considerably larger SAM immediately after germination (Fig. 1A). Post embryonically, the SAM increases dramatically in size but does not form lateral organs and internodes (Fig. 1D,E). In rare occasions the mutant will form one or two leaves or produce structures resembling stigmas, but it never forms a stem even after more than 40 days of growth (Fig. S1A,B). The meristematic nature of the dome-like structure in clv3 er erl1 erl2 is consistent with the presence of cells with dense cytoplasm and without chlorophyll in the outer cell layers (Fig. 2A; Fig. S1C). Moreover, the epidermal layer is composed of very small cells and the guard cells are absent, indicating absence of differentiation (Fig. S1D). The synergistic function of CLV3 and ERfs in the SAM is also evident in the clv3 er erl2 mutant: although the er erl2 mutant has a meristem indistinguishable from the wild type, these two mutations enhance the width of the clv3 SAM, and er erl2 reduces leaf initiation in the clv3 background (Fig. S1E,F). Finally, CLV3 regulates SAM size and leaf initiation in concert with the meristematic ERf ligands EPFL1, EPFL2, EPFL4 and EPFL6. The size of the SAM is dramatically increased in clv3 epfl1 epfl2 epfl4 epfl6 plants, comparable with the increase we observed in clv3 er erl1 erl2 mutants (Fig. 1F). Similar to our previous findings (Kosentka et al., 2019), the four EPFLs function redundantly in regulation of SAM size, as we observed a drastic increase only in the pentuple clv3 epfl1 epfl2 epfl4 epfl6 mutant.

Taken together, these findings indicate that ERf and CLV3 signaling pathways synergistically restrict SAM size. The extent of their individual contributions varies at different developmental stages. Before germination both contribute to SAM width maintenance with ERfs playing the primary role. After germination their roles switch, with CLV3 playing the dominant role and ERf signaling becoming auxiliary. In addition, synergistic function of ERf and CLV3 is essential for differentiation of organs at the periphery of the meristem and for growth of internodes.

CLV3 regulates meristem size by inhibiting expression of WUS (Brand et al., 2000; Müller et al., 2006; Schoof et al., 2000). Although expression of WUS is increased in an er erl1 erl2 background, the increase is relatively moderate: four- to six-fold at 5 DPG (Chen et al., 2013; Uchida et al., 2013) (Fig. 1G). The significance of ERfs for regulation of WUS expression becomes more evident in the absence of CLV3 signaling. In 5 DPG clv3 er erl1 erl2 seedlings we observed up to ∼30-fold increase in WUS expression compared with the single clv3 mutation and a ∼1000-fold increase over the wild type (Fig. 1G). The massive increase of WUS expression in clv3 er erl1 erl2 is unlikely to be due simply to a bigger meristem size, as the other meristematic marker STM increases only ∼5.5-fold compared with clv3 and 11-fold compared with the wild type. The large increase of WUS expression in clv3 er erl1 erl2 suggests that its expression is synergistically regulated by CLV3 and ERfs.

To study genetic interactions between ERfs and WUS and to compare them with CLV3 and WUS genetic interactions, we measured SAM size in wus, wus er erl1 erl2, wus clv3 and wus clv3 er erl1 erl2 mutants at 3 and 5 DPG. Although SAM width varied in individual seedlings (Fig. S2A), the four mutants were statistically indistinguishable (Fig. 2B) suggesting that during early seedling growth wus is epistatic to both clv3 and er erl1 erl2. This conclusion is supported by histological analysis: the shoot apices of wus, wus er erl1 erl2 and wus clv3 er erl1 erl2 mutants did not have the classic dome-like SAM structure consisting of multiple layers of small, evenly shaped and tightly packed stem cells (Fig. 2A). In the mutants, the shoot apices were composed of only two layers of small cells with some of those cells dividing periclinally, a sign of premature differentiation (Fig. S2B). A previous analysis of wus er erl1 erl2 using 10 DPG seedlings indicated that its SAM is bigger than that of wus, suggesting additive effects of ERf and WUS (Kimura et al., 2018). This conclusion was supported by the ability of er erl1 erl2 mutations to partially rescue initiation of stamens and carpels in the wus background (Kimura et al., 2018). However, our data and analysis of wus er erl1 erl2 does not support the hypothesis of additive ERf and WUS interactions. At 10 DPG in many wus er erl1 erl2 seedlings we observed a narrow region between forming leaf primordia (Fig. S2C). Although in some seedlings the area between forming leaves was indeed large, it did not contain stem cells with the characteristic dense cytoplasm (Fig. S2C). Based on morphology, cells in that region are differentiated: they are highly vacuolated, and some L2 layer cells divide in orientations other than anticlinal. We did occasionally observe meristem-like aggregations of small cells with dense cytoplasm; however, those structures were always small in diameter and asymmetrically localized, often at the axil of a leaf. These structures may be either axillary meristems or leaf primordia arising from a few erratically localized stem cells. Thus, although initiation of new meristematic regions or leaf primordia might be altered in wus er erl1 erl2 compared with wus, there is no rescue of the central zone maintenance. Our analysis of wus er erl1 erl2 flower structure in 2-month-old plants indicated that ERf family mutations were unable to rescue carpel or stamen initiation in the wus background (Table S1). While analyzing flower development we observed formation of stigma-like structures at the tips of sepals and the formation of stigma-like tissue in the area of the shoot apex in older wus er erl1 erl2 plants, but in flowers that emerge soon after bolting we never observed the formation of carpels. Although we used the same alleles of WUS and ERfs, we cannot reproduce the wus er erl1 erl2 flower structure data described by Kimura and colleagues (Kimura et al., 2018). In sum, our data indicate that wus is epistatic to er erl1 erl2 in regulation of the SAM central zone width and in the flower meristem.

In situ RNA hybridization and GUS reporter analysis show an increased expression of CLV3 and WUS in the vegetative SAM of er erl1 erl2 mutants (Kimura et al., 2018; Uchida et al., 2013). To quantify the spatial expression of these two genes we generated transgenic plants expressing nuclear localized EGFP (H2B-EGFP) under the CLV3 and WUS cis-regulatory sequences. Fluorescent reporter analysis suggests that the CLV3 expression domains in the wild type and in er erl2 mutants are very similar (Fig. 3A), consistent with the indistinguishable SAM size in the wild type and the mutant during seedling development (Chen et al., 2013). In the 1-day old seedlings, the CLV3 reporter was expressed in the L1 epidermal layer, the L2 subepidermal layer and in top two cell layers of the L3 zone. When seen from overhead, the CLV3 reporter expression region appears oval; it is slightly broader along the axis connecting the boundaries of cotyledons (width #1 in Fig. 3B,C). In the er erl1 erl2 mutant, the expression of the CLV3 reporter is expanded more than twice in both directions, along the shorter and longer axes, leading to a dramatic increase in the lateral area of expression. In addition, we often observed extended expression of the CLV3 reporter in the L1 layer in the periphery of the SAM (Fig. 3A) as has been reported previously (Kimura et al., 2018). The er erl1erl2 mutations had only minor (if any) impact on the depth of CLV3 expression.

Next, we analyzed the expression of the WUS reporter. In the er erl2 mutant the reporter was expressed in the top two cell layers of the L3 zone (Fig. 4A). When seen from above, the area of expression was oval, similar to the CLV3 expression, but smaller in size. The expression of our WUS reporter just below the L2 layer in the SAM is consistent with previously published data for the wild type (Daum et al., 2014; Mayer et al., 1998). As the expression of the WUS reporter was inconsistent and difficult to detect in the wild-type seedlings, we crossed transgenic er erl2 with the wild type and analyzed expression of the reporter in the F1 generation. As both er and erl2 are recessive mutations and F1 has a wild-type phenotype we called these seedlings WT on Fig. 4. The pattern of WUS expression in the wild type and er erl2 was indistinguishable (Fig. 4B). However, in the er erl1 erl2 mutant there was a dramatic increase in the reporter expression in the lateral direction (Fig. 4). We did not observe significant changes in the reporter expression along the apical-basal axis. Taken together, our analysis of WUS and CLV3 expression in the er erl1 erl2 mutant suggests that ERf signaling laterally restricts the expression of these two genes.

The very broad expression of CLV3 in the L1 layer of the er erl1 erl2 mutant is difficult to explain simply by expansion of the central zone. Although epfl1 epfl2 epfl4 and epfl1 epfl2 epfl6 mutants exhibit only a ∼1.25-fold increase in SAM width compared with the wild type (Kosentka et al., 2019), they express four to six times more CLV3 (Fig. 5A). These facts coupled with the epistatic nature of the wus mutation motivated us to investigate whether WUS and CLV3 are the direct targets of the ERf signaling pathway. We treated epfl1 epfl2 eplf4 seedlings exogenously with either the EPFL4 peptide or the EPFL6 peptide for 6 h. RT-qPCR analysis revealed a significantly decreased expression of WUS and CLV3 in response to both peptides (Fig. 5B). Several other genes, that have altered expression in er erl1 erl2 such as STM (Fig. 1G), MONOPTEROS (MP) (Chen et al., 2013) and CLV1 (Fig. S3), did not change expression after the peptide treatment (Fig. 5B), suggesting specificity in the downregulation of CLV3 and WUS. The decrease in WUS and CLV3 mRNA levels was dependent on the presence of functional ERf receptors, as it was not observed when er erl1 erl2 seedlings were treated with EPFL4 (Fig. 5C).

Next, we analyzed whether the ability of EPFLs to suppress expression of WUS and CLV3 is dependent on CLV3 function. The clv3-9 allele carries a point mutation in the coding region of CLV3 that disrupts its function but not expression. The clv3 epfl1 epfl2 epfl4 epfl6 mutant was treated with EPFL6 for 1, 3 and 6 h. The 1 h treatment produced a statistically significant decrease in CLV3 levels, and the 3 h treatment produced a statistically significant decrease in the WUS levels. Interestingly, the 6 h treatment did not lead to a further reduction of CLV3 or WUS expression. At the same time, a treatment of seedlings with EPFL6 in the presence of the translational inhibitor cycloheximide had a very strong impact on the steady-state mRNA levels of WUS and CLV3. WUS and CLV3 levels decreased approximately eleven and six times, respectively, after 3 h of treatment (Fig. 5D). Expression of two other analyzed genes, MP and CLV1, did not change (Fig. 5D). Collectively, our data imply that WUS and CLV3 are downstream targets of the ERf signaling pathway, and the ability of ERfs to inhibit WUS and CLV3 expression is independent of protein biosynthesis.

The er erl1 erl2 mutant has a reduced sensitivity to CLV3 peptide (Kimura et al., 2018) suggesting that ERfs might have additional roles in regulation of the CLV3 signaling pathway. The reduced sensitivity of the mutant to CLV3 might be related to reduced expression of several CLV3 receptors: CLV1, BAM1 and BAM2 (Fig. S3). However, the role of ERfs in CLV1 expression is likely to be indirect and complex. Although in the epfl1 epfl2 epfl4 background we did not observe any effects of EPFLs on the CLV1 expression, in clv3 epfl1 epfl2 epfl4 epfl6 after 6 h of treatment with EPFL6 we detected, instead of an increase, a very small decrease in the CLV1 expression (Fig. 5D).

Our analysis of genetic interactions uncovered a synergy between ERf/EPFL and CLV3 in the maintenance of SAM size and the regulation of organogenesis. Although clv3 and er erl1 erl2 mutants have bigger SAMs, they form stems, leaves and flowers. In the quadruple mutant the growth of the meristem is unrestricted, one or two leaves form only sporadically, and cells at the periphery of the meristem are unable to differentiate into internode tissues. Synergistic phenotypes most often result from redundancy between paralogs or when pathways converge on a specific node (Pérez-Pérez et al., 2009). As WUS is the core regulator of SAM size and is the primary target of CLV3, we investigated whether the two signaling pathways converge on that transcription factor. Our genetic analysis determined that wus is epistatic to er erl1 erl2 in the regulation of SAM size, suggesting that WUS could be a downstream target of the ERf/EPFL pathway. Treatment of seedlings with exogenous EPFLs for 3 or 6 h reduced steady-state levels of WUS mRNA only in the presence of functional ERf receptors. Considering that the average length of the cell cycle in the SAM is over 30 h (R. Jones et al., 2017), the decrease of WUS accumulation cannot be attributed to a decrease in the size of the SAM. Moreover, when cycloheximide was included in the treatment, EPFLs were still able to change steady-state levels of WUS, suggesting that EPFLs control WUS independently of protein biosynthesis. Consistent with previously published data (Gordon et al., 2009), we noticed an increased accumulation of WUS in the presence of cycloheximide (Fig. S3A). Our experiments do not distinguish whether ERf/EPFL signaling controls WUS transcription or its mRNA stability. We also do not know why EPFL6 has a stronger impact on WUS expression in the presence of cycloheximide. Perhaps there is a negative unstable regulator of WUS, and during cycloheximide treatment its concentration drops which enhances the impact of EPFLs on WUS transcription. Alternatively, an EPFL-induced increase in WUS mRNA degradation might be more evident if an inhibition of translation elongation alters stability of WUS mRNA.

It has previously been noticed that ERECTA and CLV3 function along different spatial axes: CLV3 preferentially regulates meristem height and ERECTA regulates meristem width (Mandel et al., 2016). Four ligands that regulate the activity of ERfs in the SAM are mostly expressed at the periphery of the meristem and are excluded from the central zone and organizing center (Kosentka et al., 2019). EPFL1 expression in the peripheral zone under the KANADI promoter fully rescues meristematic defects of epfl1 epfl2 epfl4 epfl6 (Kosentka et al., 2019). In contrast, although ERfs are endogenously expressed throughout the SAM, their function in the center of the meristem is crucial for SAM maintenance. ERECTA expressed under the CLV3 promoter rescues meristematic defects significantly better compared with its expression under the KANADI promoter (Kosentka et al., 2019). The distinct expression of ERfs and EPFLs in the SAM is similar to their distinct expression during leaf tooth initiation, in which ERfs are expressed more strongly at the tip of the tooth, whereas EPFL2 is excluded from the tip and is expressed in the surrounding sinus tissues (Tameshige et al., 2016). Expression of ERECTA under the DR5 promoter exclusively at the tip of the leaf tooth rescues its growth (Tameshige et al., 2016), just as expression of ERECTA under the CLV3 promoter in the center of the meristem rescues meristematic phenotypes (Kosentka et al., 2019). Previously, we have proposed that the signaling occurs in the peripheral region of the SAM where expression of EPFL and ERf overlaps, a pattern similar to that observed in the leaf tooth. The present quantitative analysis of CLV3 and WUS expression in the er erl1 erl2 mutant confirms that ERf functions in restricting their expression laterally.

We propose that ERfs restrict the size of the central zone by inhibiting expression of CLV3 and WUS in the periphery of the SAM (Fig. 6). This control is especially important during establishment of the SAM during embryogenesis. After germination, CLV3 signaling can partially substitute for ERfs in the lateral inhibition of WUS expression. However, when both signaling pathways are disrupted, as in the clv3 er erl1 erl2 mutant, WUS expression becomes rampant, and the SAM expands without restraint. Increased expression of CLV3 in the L1 layer of er erl1erl2 and wus er erl1 erl2 (Kimura et al., 2018) suggests that an as-yet unidentified signal induces CLV3 expression in the epidermis, consistent with a previously proposed model (Gruel et al., 2016).

Our recent mathematical model describing regulatory networks in the SAM identified nonlinear cooperativity of EPFL and CLV3 signals in the control of the lateral boundary of WUS expression (Liu et al., 2020). The model demonstrates the paradoxical role of EPFLs in regulation of WUS expression: they directly downregulate WUS and, at the same time, they upregulate it by inhibiting expression of CLV3. The downregulation plays the dominant role, as blocking it in the model leads to expression patterns very similar to the er erl1 erl2 mutant. The model also suggested that the inhibition of CLV3 expression by EPFLs is necessary to shift CLV3 expression vertically from the organizing center to the top of the SAM.

The Arabidopsis thaliana ecotype Columbia was used as the wild type. The following mutants used in the study have been described previously: er-105 erl1-2 erl2-1 (Shpak et al., 2004), epfl1 epfl2 epfl4, epfl1 epfl2 epfl6 and epfl1 epfl2 epfl4 epfl6 (Kosentka et al., 2019), clv3-9 (Nimchuk et al., 2015) and wus null allele (SAIL_150_G06; Sonoda et al., 2007). They are all in the Columbia background.

To create clv3 er erl2, clv3 er erl1 erl2 and wus er erl1 erl2 plants, clv3-9 and wus/+ were crossed with er erl1/+ erl2. To create wus clv3 plants, clv3-9 was crossed with wus/+. To create wus clv3 er erl1 erl2 plants clv3 er erl1/+ erl2 was crossed with wus/+ er erl1/+ erl2. The higher-order mutants were identified in subsequent generations based on the phenotype. The homozygous status of wus and erl1 were confirmed when necessary by genotyping as described previously (Kosentka et al., 2017; Sonoda et al., 2007). To create clv3 epfl1 epfl2 epfl4 epfl6 plants, the clv3-9 mutant was crossed with epfl1/+ epfl2 epfl4 epfl6. The epfl mutations were genotyped as described previously (Kosentka et al., 2019).

The WUSp:H2B-GFP:35St construct (pESH 746) was created by fusing the 4.5 kb sequence of the WUS promoter (Yadav et al., 2009) with H2B-EGFP and the 35S terminator. The CLV3p:H2B-GFP:CLV3t construct (pESH 747) was created by fusing the 1.5 kb sequence of the CLV3 promoter with H2B-EGFP and the 1.2 kb CLV3 terminator. The inserts were generated by overlap extension PCR and inserted into the binary vector pPZP222 between BamHI and SalI restiction sites. The template for amplifying the H2B-EGFP sequence was a plasmid from the Z. Nimchuk lab (UNC Chapel Hill, USA). Both constructs were examined by sequencing of amplified regions. The pESH746 and pESH747 were transformed into an Agrobacterium tumefaciens strain GV3101/pMP90 by electroporation and introduced into the wild type (Columbia ecotype) and er erl1/+ erl2 plants by the floral dip method. T1 transgenic plants were selected using gentamycin resistance. Heterozygous erl1/+ plants in the T1 generation were identified by genotyping.

Plants were grown as previously described (Kosentka et al., 2017) under an 18 h light/6 h dark cycle (long days) at 21°C. For analysis of SAM size and leaf initiation seedlings were grown on modified Murashige and Skoog medium plates supplemented with 1% (w/v) sucrose. For all experiments, seeds were stratified for 2 days at 4°C before germination.

To analyze expression of genes after EPFL4 and EPFL6 treatment, epfl1 epfl2 epfl4 mutants were grown on modified Murashige and Skoog medium plates for 5 days (3 DPG). Then 60 seedlings (EPFL4 treatment) or 10 seedlings (EPFL6 treatment) per biological replicate were transferred to 1 ml of liquid Murashige and Skoog medium containing 10 µM of EPFL4 or EPFL6. The purification of EPFL4 and EPFL6 peptides has been described previously (Lin et al., 2017). EPFL peptides were dissolved in 10 mM Bis-Tris, 100 mM NaCl (pH 6.0). For mock treatment, a buffer solution of equal volume was added to the medium (92.6 µl in the EPFL4 experiment and 8.7 µl in the EPFL6 experiment). To analyze the effect of EPFL6 in the presence of cycloheximide, seedlings were incubated for 10 min in Murashige and Skoog medium containing 10 µM cycloheximide and then 10 µM of EPFL6 was added to the ‘treatment’ group. For each treatment there were three biological replicas.

To measure leaf initiation and SAM size, one, three and five DPG seedlings were fixed overnight with ethanol: acetic acid [9:1 (v/v)]. After fixation, samples were rehydrated with an ethanol series to 30% (v/v) ethanol and cleared in chloral hydrate solution. Chloral hydrate:water:glycerol 8:1:1 (w/v/v) solution contained KOH at 10 mM concentration to prevent degradation of tissues due to high acidity of chloral hydrate. In our experience the acidity of chloral hydrate (Sigma-Aldrich) varies from batch to batch, and the necessity to add KOH should be tested experimentally. Microscopic observations of meristematic regions by DIC microscopy were performed as described previously (Chen et al., 2013). Tissue samples were fixed overnight in acetic acid:ethanol (1:9) at room temperature, dehydrated with a graded series of ethanol, and infiltrated with polymethacryl resin Technovit 7100 (Heraeus Kulzer) followed by embedding and polymerization in Technovit 7100. Then, 7 μm sections were prepared using a Leica RM-6145 microtome. The tissue sections were stained with 0.02% Toluidine Blue O and observed under bright-field illumination. Pictures of older seedlings and the analysis of flower structure was carried out using a Leica MZ16 FA stereomicroscope.

Imaging of the SAMs in T2 plants expressing WUSp:H2B-GFP:35St or CLV3p:H2B-GFP:CLV3t was carried out using a Leica SP8 White Light Laser Confocal microscope. EGFP was excited using a 488-nm White Light Supercontinuum Laser (Leica Microsystems). SCRI Renaissance 2200 (SR2200) was excited with a Diode 405 nm ‘UV’ laser (Leica Microsystems). EGFP and SR2200 fluorescence emission was collected with HyD ‘Hybrid’ Super Sensitivity SP Detector (Leica Microsystems ) and PMT SP Detector (Leica Microsystems). Sequential line scanning was used to generate z-stacks. For better observation of fluorescence in the SAM of live 1-day old seedlings, one cotyledon was removed before imaging. SR2200 (Renaissance Chemicals) was prepared as in Musielak et al. (2015) except for the omission of paraformaldehyde. Samples were incubated in ∼350-500 μl of staining solution under vacuum for ∼1-2 min at room temperature, washed with 1× PBS buffer to remove excess dye, and imaged.

The Fiji image processing package was used for all quantitative image measurements. Two-dimensional slices were used to measure the height of reporter expression in WUSp:H2B-GFP:35St plants. Three dimensional images were used for all other measurements. Apical area measurements were acquired by drawing a circle around the expression domain using the freehand measurement tool and data were analyzed using one-way ANOVA with Tukey's post-hoc test. The er erl1 erl2 mutants in the T2 generation were identified based on stomata clustering.

Total RNA was isolated from the above-ground tissues of seedlings using the Spectrum Plant RNA Isolation Kit (Sigma-Aldrich). The RNA was treated with RNase-free RQ1 DNase (Promega). First-strand complementary cDNA was synthesized with LunaScript™ RT SuperMix Kits (New England Biolabs). Quantitative PCR was performed with a CFX96 Touch Real-Time PCR Detection System (Bio-Rad) using Sso Fast EvaGreen Supermix (Bio-Rad). Each experiment contained three technical replicates of three biological replicates and was performed in a total volume of 10 µl with 4 µl of 10× or 50× diluted cDNA. Cycling conditions were as follows: 3 min at 95°C; then 40 repeats of 10 s at 95°C, 10 s at 52°C for ACTIN2 and STM; 10 s at 55°C for WUS; 10 s at 57°C for MP; 10 s at 50°C for CLV1, BAM1, BAM2 and BAM3, and 10 s at 68°C, followed by the melt-curve analysis. Cycling conditions for CLV3 were 3 min at 95°C; then 40 cycles of 95°C for 10 s and 60°C for 10 s, followed by the melt-curve analysis. Primers for ACTIN2, STM, WUS and MP (Chen et al., 2013) as well as for CLV3 (Chiu et al., 2007) have been described previously. Primers for CLV1, BAM1, BAM2 and BAM3 were as in Nimchuk et al. (2015). The fold difference in gene expression was calculated using relative quantification by the 2^−ΔΔCT algorithm.

We thank Zachary Nimchuk for sharing with us seeds of clv3-9 and wus (SAIL_150_G06) and a construct containing the H2B-EGFP sequence. We thank Richard Maradiaga and Timothy Herman for technical assistance.

The peer review history is available online at https://dev.biologists.org/lookup/doi/10.1242/dev.189753.reviewer-comments.pdf