Receptor-mediated endocytosis is an integral part of cellular signaling, as it mediates signal attenuation and provides spatial and temporal dimensions to signaling events. In mammalian systems, endocytosis of receptor tyrosine kinases can attenuate the signal outputs, by removing the active receptor pools from the plasma membrane, or it can specify signals at defined sites of action, such as signaling through endosomes (Sigismund et al., 2012). As a sessile organism, plants make use of a large number of receptor-like kinases (RLKs) for cell-cell, shoot-to-root, and inter-kingdom communications (Shiu and Bleecker, 2001). The RLKs with extracellular leucine-rich repeat domain, known as LRR-RLKs, comprise the largest RLK subfamily (Shiu and Bleecker, 2001), and they specify critical aspects of development, environmental response, and immunity by perceiving extrinsic signals (Torii, 2004; Macho and Zipfel, 2014). Increasing evidence shows that the subcellular localization and trafficking routes of LRR-RLKs regulate their function and activity (Ben Khaled et al., 2015). In Arabidopsis, bacterial flagellin peptide flg22 induces the heterodimer formation consisting of the LRR-RLKs FLAGELLIN SENSING2 (FLS2) and BRI1-ASSOCIATED RECEPTOR KINASE (BAK1)/SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE 3 (Chinchilla et al., 2007). This triggers the endocytosis and degradation of the receptor complex to generate transient cellular immune signaling but also to prevent continuous signaling to the same stimulus (Robatzek et al., 2006; Beck et al., 2012). The brassinosteroid (BR) receptor BRASSINOSTEROID INSENSITIVE1 (BRI1) forms a complex with BAK1 (Li et al., 2002; Nam and Li, 2002; Bücherl et al., 2013). BRI1 can undergo constitutive endocytosis independent of BRs, but BRs can elevate BRI1 and BAK1 interaction and reduce the number of available BRI1-BAK1 complexes on the plasma membrane (Geldner et al., 2007; Bücherl et al., 2013; Hutten et al., 2017). CLAVATA1 (CLV1), an LRR-RLK that controls stem cell homeostasis within the shoot meristem (Clark et al., 1997), is downregulated by ligand-dependent internalization upon perception of its ligand CLV3 (Nimchuk et al., 2011). It remains a key question as to where within the cell these LRR-RLKs transduce signals and how different activation states of LRR-RLKs influence their subcellular localization.

Developmental patterning of stomata, adjustable pores on the plant epidermis for gas-exchange and transpiration, relies on intricate cell-cell communication mediated by signaling peptides and their receptors (Lau and Bergmann, 2012; Pillitteri and Torii, 2012). In Arabidopsis, secreted peptides from the EPF family, and their shared receptor LRR-RLKs, ERECTA, ERL1, and ERL2, mediate this process (Rychel et al., 2010). Amongst the plant LRR-RLKs, the ERECTA family offers a unique advantage to study how multiple signals are perceived to achieve cell fate and patterning. EPF2 and EPF1 negatively regulate stomatal development primarily through ERECTA and ERL1, respectively (Hara et al., 2007; Hara et al., 2009; Hunt and Gray, 2009). In contrast, EPF-LIKE9 (EPFL9), also known as Stomagen, promotes stomatal development by competing with EPF2 and, to some extent, with EPF1 for receptor binding (Sugano et al., 2010; Lee et al., 2015; Lin et al., 2017; Qi et al., 2017). Moreover, EPFL4/5/6, a subfamily only expressed in hypocotyls and stems, also act as ligands for the ERECTA family to inhibit stomatal formation when an LRR receptor protein, TOO MANY MOUTHS (TMM), is missing (Abrash and Bergmann, 2010; Abrash et al., 2011). Although the final phenotypic outcomes of these different EPF signaling events are well characterized, the very early step of signal transmission by the receptors remain elusive. While internalization of ERL2 was documented briefly (Ho et al., 2016), it is unknown whether it has any implications in signal transduction or in which subcellular organelle ERL2 was localized.

ERL1 regulates guard cell differentiation in an autocrine manner in addition to enforcing stomatal spacing of neighboring cells in a paracrine manner (Lee et al., 2012; Qi et al., 2017). This dual function of ERL1 can be attributed to its cell-type specific expression patterns as well as its ability to perceive different EPF/EPFL peptide ligands (Shpak et al., 2005; Lin et al., 2017). On the other hand, its sister receptor ERECTA is broadly and ubiquitously expressed in the epidermis (Horst et al., 2015). The specific expression of ERL1 in stomatal meristemoids as well as its dual function as an autocrine and paracrine signaling receptor provide a unique advantage to study how receptor subcellular dynamics translates into the eventual cell fate. Here, we combined genetic, pharmacological, and live imaging approaches to explore the initial events that occurred at ERL1 upon perception of different EPF peptides. Our study shows that EPF1 and EPFL6, the ligands activating the inhibitory stomatal signaling, trigger ERL1 endocytosis into multi-vesicular bodies/late endosomes (MVB/LEs). TMM, which can form a receptor complex with ERL1, is required for the EPF1-induced ERL1 internalization and suppression of stomatal fate but is superfluous for EFPL6-induced ERL1 internalization. Surprisingly, Stomagen interferes with the inhibitory regulation of stomatal differentiation by retaining ERL1 to the endoplasmic reticulum, similar to when endocytosis was pharmacologically blocked by Tyrphostin A23 (Tyr A23) (Santuari et al., 2011) and Endosidin 9–17 (ES9-17) (Dejonghe et al., 2019). Additionally, we extensively examined the effects of Brefeldin A (BFA) on subcellular organelle behaviors in stomatal meristemoids and established optimal conditions to observe BFA body formation in true leaves. Combined, our study reveals a mechanism by which plant cells interpret multiple signals through the subcellular localization and trafficking route of a single receptor.

In this study, we revealed that ERL1 endocytosis accurately reflects EPF signal perception based on three pieces of evidence (Figure 9): first, both EPF1 and EPFL6 peptides trigger ERL1 endocytosis. Second, in the absence of the co-receptor TMM, ERL1 endocytosis is compromised and becomes insensitive to EPF1 application. Third, the cytoplasmic domain of ERL1 is required for ERL1 endocytosis. EPF1 and EPFL6 peptide application increased ERL1 population in endosomes in a dosage-dependent manner (Figures 5 and 6). Our pharmacological application of Wortmannin, BFA, BFA with CHX and Concanamycin A (Figures 2–3, Figure 4—figure supplement 1), indicate that ERL1 is constitutively recycled whereas the receptor activation triggers endocytosis for vacuolar sorting via MVB/LE. In this aspect, the subcellular dynamics of ERL1 resembles that of FLS2, which is also constitutively recycled but rapidly removed from the cell surface upon flg22 perception (Robatzek et al., 2006; Smith et al., 2014). Unlike FLS2, however, a vast majority of ERL1-YFP signal remained at the plasma membrane even after treatment of 5 µM MEPF1 (Figure 5). These differences could be attributed to the roles of FLS2 and ERL1 in immunity vs. development, respectively. FLS2 mediates acute pathogen-induced defense response, whereas ERL1 likely detects endogenous peptides to influence slower processes of cell division and differentiation. A recent study showed, however, that defects in the clathrin-mediated FLS2 endocytosis impair only a subset of FLS2-mediated immune responses (Mbengue et al., 2016). Thus, the precise contributions of endocytosis and cellular response remain open questions.

Figure 9

Download asset Open asset

It has been shown that EPF1, but not EPFL6, requires TMM for the inhibition of stomatal development (Hara et al., 2007; Abrash and Bergmann, 2010). Likewise, structural analyses of the EPF-ERECTA family complexes showed that EPF1, but not EPFL6, requires TMM for binding to the ectodomain of ERECTA family receptors (Lin et al., 2017). Here, we demonstrate that TMM is required for endocytosis triggered by EPF1, but not by EPFL6 (Figures 4, 5, 6 and 9). Thus, at least two populations of ERL1 receptor complexes must be present on the plasma membrane, with and without TMM. The BFA treatment on tmm mutants (Figure 4, Figure 4—figure supplement 1) revealed that in the absence of TMM, ERL1-YFP internalization is compromised but not absent. This is most likely due to the presence of EPF-LIKE peptide (e.g. EPFL4/5/6) family members that do not require TMM. The FRAP analysis indeed detected the different mobility of these two ERL1 compositions on the plasma membrane (Figure 4). Multiple compositions of receptor complexes have also been reported in CLV3 signaling, where CLV1 homomers, CLV2/CORYNE (CRN) heterodimers and CLV1/CLV2/CRN multimers co-exist on the plasma membrane (Somssich et al., 2015). However, only the microdomain-localized CLV1/CLV2/CRN multimers can perceive the sole ligand CLV3. In the case of BRI1 and FLS2, pre-formed BRI1-BAK1 complex was detected regardless of BRs whereas FLS2 forms FLS2-BAK1 complex upon flg22 application (Bücherl et al., 2013; Somssich et al., 2015). These receptor complexes are spatially separated, even though BRI1 and FLS2 share the same co-receptor BAK1 (Bücherl et al., 2013; Somssich et al., 2015; Bücherl et al., 2017; Hutten et al., 2017). On the contrary, both populations of ERL1 complexes are ‘functional’ and ligand-inducible, as they can perceive EPF1 or EPFL6, respectively (Figures 5 and 6). It is possible that the distinct ERL1 receptor complexes reside in different microdomains on the plasma membrane and undergo different trafficking routes upon the correlated ligand perception. EPF1 triggers ERL1 association with BAK1 (Meng et al., 2015). Examining spatiotemporal subcellular dynamics of ERL1 together with TMM and BAK1 at a super resolution scale may reveal the contribution of each receptor complex for specific signal perception and transduction.

ERL1 is retained in the endoplasmic reticulum when treated with exogenous Stomagen. This ERL1 subcellular behavior is associated with its own biological contexts, given that BRI1, an LRR-RK with unrelated function, does not change its plasma membrane localization upon Stomagen application (Figure 7—figure supplement 2). Extensive studies support that the steady state of a protein in its subcellular compartment is interdependent on the anterograde and retrograde trafficking routes (Brandizzi and Barlowe, 2013). For example, a secretory protein is often retained in the endoplasmic reticulum when the downstream secretion pathway is compromised (Zheng et al., 2005). Blocking the endoplasmic reticulum-to-Golgi retrograde trafficking will accelerate protein transport to the cell surface (Fossati et al., 2014). It is thus possible that Stomagen binding prevents the ERL1 endocytosis and the plasma membrane-accumulated ERL1 interferes with the normal transport of incoming ERL1 from the endoplasmic reticulum. Two additional pieces of evidence support this hypothesis. First, when endocytosis is blocked by Tyr A23 (Banbury et al., 2003) or ES7-19, the improved, specific inhibitor of clathrin heavy chain (Dejonghe et al., 2019), strong ERL1 signals become evident in the endoplasmic reticulum (Figure 7E). Second, in leaves of the tmm mutant, EPF1-triggered ERL1 endocytosis is compromised, ERL1 also accumulates in the endoplasmic reticulum (Figure 7C and Figure 7—figure supplement 1).

Stomagen-triggered ERL1 accumulation in endoplasmic reticulum may be highlighting the role of the endoplasmic reticulum-plasma membrane contact sites as a direct communication link between the two compartments (Carrasco and Meyer, 2011). The VAP-RELATED SUPPRESSOR OF TMM (VST) family plasma membrane proteins that interact with integral endoplasmic reticulum proteins, have been reported to facilitate ERECTA family-mediated signaling in stomatal development (Ho et al., 2016). Hence, Stomagen perception by the ERL1-TMM complex on the plasma membrane may directly influence signaling via the contact sites and therefore affect the transport of ERL1 to the cell surface. The endoplasmic reticulum accumulation of ERL1-YFP is notable in the erecta erl1 erl2 mutant background, that is when functional ERECTA and ERL2 genes are missing. Because Stomagen interacts with ERECTA at the early stage of stomatal initiation (Lee et al., 2015), the absence of ERECTA likely confers sensitized, exaggerated response of ERL1-YFP to Stomagen. Alternatively, a dysregulated cell-cell signaling in the erecta erl1 erl2 background may abrogate the buffering of exogenously applied Stomagen peptide.

Posttranslational modifications, such as phosphorylation and ubiquitination, of the receptors have emerged as key regulatory mechanisms of receptor subcellular dynamics in FLS2 and BRI1 (Robatzek et al., 2006; Lu et al., 2011; Martins et al., 2015; Zhou et al., 2018). Specific posttranslational modifications of ERL1 are yet unknown, but perception of different EPF/EPFL peptides may trigger unique phosphorylation codes that influence subsequent trafficking of ERL1. The dominant-negative ERL1 fails to respond to either MEPF1 or Stomagen peptides (Figure 8, Figure 8—figure supplement 1). Clearly, the behaviors of ERL1ΔK-CFP is different from the full-length ERL1-YFP. The accumulation of ERL1 in the endoplasmic reticulum occurs when ERL1-YFP becomes signaling incompetent- (i) by Stomagen peptide perception; (ii) due to the absence of co-receptor and redundant receptors, TMM, ERECTA, and ERL2; or (iii) when ERL1 endocytosis is pharmacologically blocked (Figure 7). Thus, subcellular dynamics of the full-length ERL1 is highly tuned and coordinated. Such coordination is lost by removal of the cytoplasmic domain, which includes juxtamembrane domain, kinase domain, and C-terminal tail region (Figure 8A). It is highly plausible that ERL1ΔK-CFP can no longer be recognized as the original endocytosis cargo by adaptor proteins. For instance, the endocytosis of BRI1 is mediated by direct association with Clathrin adaptor complex AP2 (Di Rubbo et al., 2013). Future studies may reveal the molecular basis of how ERL1 interacts with specific cargo receptors.

Previous work reported that BFA effect (BFA body formation) is different in root vs. shoot tissues (Robinson et al., 2008; Langhans et al., 2011). Here, we extensively documented the effects of BFA on subcellular behaviors using eight endomembrane organelle markers together with endocytic tracer dye (Figure 3, Figure 3—figure supplement 1). Our extensive analyses show that with appropriate concentration and treatment condition, BFA body formation can be recapitulated in leaf cells, just like in root cells. The results could facilitate future studies on recycling and endocytosis of membrane proteins in intact leaves. Taken together, our work revealed the mechanism by which multiple peptide ligands with distinct activities, EPF1, EPFL6, and Stomagen, fine-tune stomatal patterning at the level of the subcellular dynamics of a single receptor, ERL1. Successful development of visible functional peptide ligands and identification of the immediate biochemical events by the ERECTA-family perceiving different EPF peptides will help elucidate the exact roles of receptor trafficking and signaling specifying developmental patterning in plants.

All data generated or analyzed during this study are included in the manuscript and supporting files. Source R codes are provided.

1. 1. Abrash EB
   2. Davies KA
   3. Bergmann DC

   (2011) Generation of signaling specificity in Arabidopsis by spatially restricted buffering of ligand-receptor interactions

   The Plant Cell 23:2864–2879.

   https://doi.org/10.1105/tpc.111.086637

   • PubMed
   • Google Scholar
2. 1. Abrash EB
   2. Bergmann DC

   (2010) Regional specification of stomatal production by the putative ligand CHALLAH

   Development 137:447–455.

   https://doi.org/10.1242/dev.040931

   • PubMed
   • Google Scholar
3. 1. Banbury DN
   2. Oakley JD
   3. Sessions RB
   4. Banting G

   (2003) Tyrphostin A23 inhibits internalization of the transferrin receptor by perturbing the interaction between tyrosine motifs and the medium chain subunit of the AP-2 adaptor complex

   Journal of Biological Chemistry 278:12022–12028.

   https://doi.org/10.1074/jbc.M211966200

   • PubMed
   • Google Scholar
4. 1. Beck M
   2. Zhou J
   3. Faulkner C
   4. MacLean D
   5. Robatzek S

   (2012) Spatio-temporal cellular dynamics of the Arabidopsis flagellin receptor reveal activation status-dependent endosomal sorting

   The Plant Cell 24:4205–4219.

   https://doi.org/10.1105/tpc.112.100263

   • PubMed
   • Google Scholar
5. 1. Ben Khaled S
   2. Postma J
   3. Robatzek S

   (2015) A moving view: subcellular trafficking processes in pattern recognition receptor-triggered plant immunity

   Annual Review of Phytopathology 53:379–402.

   https://doi.org/10.1146/annurev-phyto-080614-120347

   • PubMed
   • Google Scholar
6. 1. Brandizzi F
   2. Barlowe C

   (2013) Organization of the ER-Golgi interface for membrane traffic control

   Nature Reviews Molecular Cell Biology 14:382–392.

   https://doi.org/10.1038/nrm3588

   • PubMed
   • Google Scholar
7. 1. Bücherl CA
   2. van Esse GW
   3. Kruis A
   4. Luchtenberg J
   5. Westphal AH
   6. Aker J
   7. van Hoek A
   8. Albrecht C
   9. Borst JW
   10. de Vries SC

   (2013) Visualization of BRI1 and BAK1(SERK3) membrane receptor heterooligomers during brassinosteroid signaling

   Plant Physiology 162:1911–1925.

   https://doi.org/10.1104/pp.113.220152

   • PubMed
   • Google Scholar
8. 1. Bücherl CA
   2. Jarsch IK
   3. Schudoma C
   4. Segonzac C
   5. Mbengue M
   6. Robatzek S
   7. MacLean D
   8. Ott T
   9. Zipfel C

   (2017) Plant immune and growth receptors share common signalling components but localise to distinct plasma membrane nanodomains

   eLife 6:e25114.

   https://doi.org/10.7554/eLife.25114

   • PubMed
   • Google Scholar
9. 1. Carrasco S
   2. Meyer T

   (2011) STIM proteins and the endoplasmic reticulum-plasma membrane junctions

   Annual Review of Biochemistry 80:973–1000.

   https://doi.org/10.1146/annurev-biochem-061609-165311

   • PubMed
   • Google Scholar
10. 1. Chinchilla D
    2. Zipfel C
    3. Robatzek S
    4. Kemmerling B
    5. Nürnberger T
    6. Jones JD
    7. Felix G
    8. Boller T

    (2007) A flagellin-induced complex of the receptor FLS2 and BAK1 initiates plant defence

    Nature 448:497–500.

    https://doi.org/10.1038/nature05999

    • PubMed
    • Google Scholar
11. 1. Clark SE
    2. Williams RW
    3. Meyerowitz EM

    (1997) The CLAVATA1 gene encodes a putative receptor kinase that controls shoot and floral meristem size in Arabidopsis

    Cell 89:575–585.

    https://doi.org/10.1016/S0092-8674(00)80239-1

    • PubMed
    • Google Scholar
12. 1. Dejonghe W
    2. Kuenen S
    3. Mylle E
    4. Vasileva M
    5. Keech O
    6. Viotti C
    7. Swerts J
    8. Fendrych M
    9. Ortiz-Morea FA
    10. Mishev K
    11. Delang S
    12. Scholl S
    13. Zarza X
    14. Heilmann M
    15. Kourelis J
    16. Kasprowicz J
    17. Nguyen leSL
    18. Drozdzecki A
    19. Van Houtte I
    20. Szatmári AM
    21. Majda M
    22. Baisa G
    23. Bednarek SY
    24. Robert S
    25. Audenaert D
    26. Testerink C
    27. Munnik T
    28. Van Damme D
    29. Heilmann I
    30. Schumacher K
    31. Winne J
    32. Friml J
    33. Verstreken P
    34. Russinova E

    (2016) Mitochondrial uncouplers inhibit clathrin-mediated endocytosis largely through cytoplasmic acidification

    Nature Communications 7:11710.

    https://doi.org/10.1038/ncomms11710

    • PubMed
    • Google Scholar
13. 1. Dejonghe W
    2. Sharma I
    3. Denoo B
    4. De Munck S
    5. Lu Q
    6. Mishev K
    7. Bulut H
    8. Mylle E
    9. De Rycke R
    10. Vasileva M
    11. Savatin DV
    12. Nerinckx W
    13. Staes A
    14. Drozdzecki A
    15. Audenaert D
    16. Yperman K
    17. Madder A
    18. Friml J
    19. Van Damme D
    20. Gevaert K
    21. Haucke V
    22. Savvides SN
    23. Winne J
    24. Russinova E

    (2019) Disruption of endocytosis through chemical inhibition of clathrin heavy chain function

    Nature Chemical Biology 15:641–649.

    https://doi.org/10.1038/s41589-019-0262-1

    • PubMed
    • Google Scholar
14. 1. Di Rubbo S
    2. Irani NG
    3. Kim SY
    4. Xu ZY
    5. Gadeyne A
    6. Dejonghe W
    7. Vanhoutte I
    8. Persiau G
    9. Eeckhout D
    10. Simon S
    11. Song K
    12. Kleine-Vehn J
    13. Friml J
    14. De Jaeger G
    15. Van Damme D
    16. Hwang I
    17. Russinova E

    (2013) The clathrin adaptor complex AP-2 mediates endocytosis of brassinosteroid insensitive1 in Arabidopsis

    The Plant Cell 25:2986–2997.

    https://doi.org/10.1105/tpc.113.114058

    • PubMed
    • Google Scholar
15. 1. Ebine K
    2. Fujimoto M
    3. Okatani Y
    4. Nishiyama T
    5. Goh T
    6. Ito E
    7. Dainobu T
    8. Nishitani A
    9. Uemura T
    10. Sato MH
    11. Thordal-Christensen H
    12. Tsutsumi N
    13. Nakano A
    14. Ueda T

    (2011) A membrane trafficking pathway regulated by the plant-specific RAB GTPase ARA6

    Nature Cell Biology 13:853–859.

    https://doi.org/10.1038/ncb2270

    • PubMed
    • Google Scholar
16. 1. Faraco M
    2. Li Y
    3. Li S
    4. Spelt C
    5. Di Sansebastiano GP
    6. Reale L
    7. Ferranti F
    8. Verweij W
    9. Koes R
    10. Quattrocchio FM

    (2017) A tonoplast P_3B-ATPase Mediates Fusion of Two Types of Vacuoles in Petal Cells

    Cell Reports 19:2413–2422.

    https://doi.org/10.1016/j.celrep.2017.05.076

    • PubMed
    • Google Scholar
17. 1. Foissner I
    2. Sommer A
    3. Hoeftberger M
    4. Hoepflinger MC
    5. Absolonova M

    (2016) Is Wortmannin-Induced reorganization of the trans-Golgi network the key to explain charasome formation?

    Frontiers in Plant Science 7:756.

    https://doi.org/10.3389/fpls.2016.00756

    • PubMed
    • Google Scholar
18. 1. Fossati M
    2. Colombo SF
    3. Borgese N

    (2014) A positive signal prevents secretory membrane cargo from recycling between the golgi and the ER

    The EMBO Journal 33:2080–2097.

    https://doi.org/10.15252/embj.201488367

    • PubMed
    • Google Scholar
19. 1. Friedrichsen DM
    2. Joazeiro CA
    3. Li J
    4. Hunter T
    5. Chory J

    (2000) Brassinosteroid-insensitive-1 is a ubiquitously expressed leucine-rich repeat receptor serine/threonine kinase

    Plant Physiology 123:1247–1256.

    https://doi.org/10.1104/pp.123.4.1247

    • PubMed
    • Google Scholar
20. 1. Geldner N
    2. Anders N
    3. Wolters H
    4. Keicher J
    5. Kornberger W
    6. Muller P
    7. Delbarre A
    8. Ueda T
    9. Nakano A
    10. Jürgens G

    (2003) The Arabidopsis GNOM ARF-GEF mediates endosomal recycling, auxin transport, and auxin-dependent plant growth

    Cell 112:219–230.

    https://doi.org/10.1016/S0092-8674(03)00003-5

    • PubMed
    • Google Scholar
21. 1. Geldner N
    2. Hyman DL
    3. Wang X
    4. Schumacher K
    5. Chory J

    (2007) Endosomal signaling of plant steroid receptor kinase BRI1

    Genes & Development 21:1598–1602.

    https://doi.org/10.1101/gad.1561307

    • PubMed
    • Google Scholar
22. 1. Geldner N
    2. Dénervaud-Tendon V
    3. Hyman DL
    4. Mayer U
    5. Stierhof YD
    6. Chory J

    (2009) Rapid, combinatorial analysis of membrane compartments in intact plants with a multicolor marker set

    The Plant Journal 59:169–178.

    https://doi.org/10.1111/j.1365-313X.2009.03851.x

    • PubMed
    • Google Scholar
23. 1. Hara K
    2. Kajita R
    3. Torii KU
    4. Bergmann DC
    5. Kakimoto T

    (2007) The secretory peptide gene EPF1 enforces the stomatal one-cell-spacing rule

    Genes & Development 21:1720–1725.

    https://doi.org/10.1101/gad.1550707

    • PubMed
    • Google Scholar
24. 1. Hara K
    2. Yokoo T
    3. Kajita R
    4. Onishi T
    5. Yahata S
    6. Peterson KM
    7. Torii KU
    8. Kakimoto T

    (2009) Epidermal cell density is autoregulated via a secretory peptide, EPIDERMAL PATTERNING FACTOR 2 in Arabidopsis leaves

    Plant and Cell Physiology 50:1019–1031.

    https://doi.org/10.1093/pcp/pcp068

    • PubMed
    • Google Scholar
25. 1. Hawes C
    2. Wang P
    3. Kriechbaumer V

    (2018) Labeling the ER for light and fluorescence microscopy

    Methods in Molecular Biology 1691:1–14.

    https://doi.org/10.1007/978-1-4939-7389-7_1

    • PubMed
    • Google Scholar
26. 1. Ho CM
    2. Paciorek T
    3. Abrash E
    4. Bergmann DC

    (2016) Modulators of stomatal lineage signal transduction alter membrane contact sites and reveal specialization among ERECTA kinases

    Developmental Cell 38:345–357.

    https://doi.org/10.1016/j.devcel.2016.07.016

    • PubMed
    • Google Scholar
27. 1. Horst RJ
    2. Fujita H
    3. Lee JS
    4. Rychel AL
    5. Garrick JM
    6. Kawaguchi M
    7. Peterson KM
    8. Torii KU

    (2015) Molecular framework of a regulatory circuit initiating Two-Dimensional spatial patterning of stomatal lineage

    PLOS Genetics 11:e1005374.

    https://doi.org/10.1371/journal.pgen.1005374

    • PubMed
    • Google Scholar
28. 1. Hunt L
    2. Gray JE

    (2009) The signaling peptide EPF2 controls asymmetric cell divisions during stomatal development

    Current Biology 19:864–869.

    https://doi.org/10.1016/j.cub.2009.03.069

    • PubMed
    • Google Scholar
29. 1. Hutten SJ
    2. Hamers DS
    3. Aan den Toorn M
    4. van Esse W
    5. Nolles A
    6. Bücherl CA
    7. de Vries SC
    8. Hohlbein J
    9. Borst JW

    (2017) Visualization of BRI1 and SERK3/BAK1 nanoclusters in Arabidopsis roots

    PLOS ONE 12:e0169905.

    https://doi.org/10.1371/journal.pone.0169905

    • PubMed
    • Google Scholar
30. 1. Jin H
    2. Yan Z
    3. Nam KH
    4. Li J

    (2007) Allele-specific suppression of a defective brassinosteroid receptor reveals a physiological role of UGGT in ER quality control

    Molecular Cell 26:821–830.

    https://doi.org/10.1016/j.molcel.2007.05.015

    • PubMed
    • Google Scholar
31. 1. Kleine-Vehn J
    2. Leitner J
    3. Zwiewka M
    4. Sauer M
    5. Abas L
    6. Luschnig C
    7. Friml J

    (2008) Differential degradation of PIN2 auxin efflux carrier by retromer-dependent vacuolar targeting

    PNAS 105:17812–17817.

    https://doi.org/10.1073/pnas.0808073105

    • PubMed
    • Google Scholar
32. 1. Kohze R
    2. Dieteren CEJ
    3. Koopman WJH
    4. Brock R
    5. Schmidt S

    (2017) Frapbot: an open-source application for FRAP data

    Cytometry Part A 91:810–814.

    https://doi.org/10.1002/cyto.a.23172

    • Google Scholar
33. 1. Kondo T
    2. Kajita R
    3. Miyazaki A
    4. Hokoyama M
    5. Nakamura-Miura T
    6. Mizuno S
    7. Masuda Y
    8. Irie K
    9. Tanaka Y
    10. Takada S
    11. Kakimoto T
    12. Sakagami Y

    (2010) Stomatal density is controlled by a mesophyll-derived signaling molecule

    Plant and Cell Physiology 51:1–8.

    https://doi.org/10.1093/pcp/pcp180

    • PubMed
    • Google Scholar
34. 1. Langhans M
    2. Förster S
    3. Helmchen G
    4. Robinson DG

    (2011) Differential effects of the brefeldin A analogue (6R)-hydroxy-BFA in tobacco and Arabidopsis

    Journal of Experimental Botany 62:2949–2957.

    https://doi.org/10.1093/jxb/err007

    • PubMed
    • Google Scholar
35. 1. Lau OS
    2. Bergmann DC

    (2012) Stomatal development: a plant's perspective on cell polarity, cell fate transitions and intercellular communication

    Development 139:3683–3692.

    https://doi.org/10.1242/dev.080523

    • PubMed
    • Google Scholar
36. 1. Lee JS
    2. Kuroha T
    3. Hnilova M
    4. Khatayevich D
    5. Kanaoka MM
    6. McAbee JM
    7. Sarikaya M
    8. Tamerler C
    9. Torii KU

    (2012) Direct interaction of ligand-receptor pairs specifying stomatal patterning

    Genes & Development 26:126–136.

    https://doi.org/10.1101/gad.179895.111

    • PubMed
    • Google Scholar
37. 1. Lee JS
    2. Hnilova M
    3. Maes M
    4. Lin YC
    5. Putarjunan A
    6. Han SK
    7. Avila J
    8. Torii KU

    (2015) Competitive binding of antagonistic peptides fine-tunes stomatal patterning

    Nature 522:439–443.

    https://doi.org/10.1038/nature14561

    • PubMed
    • Google Scholar
38. 1. Li J
    2. Wen J
    3. Lease KA
    4. Doke JT
    5. Tax FE
    6. Walker JC

    (2002) BAK1, an Arabidopsis LRR receptor-like protein kinase, interacts with BRI1 and modulates brassinosteroid signaling

    Cell 110:213–222.

    https://doi.org/10.1016/S0092-8674(02)00812-7

    • PubMed
    • Google Scholar
39. 1. Lin G
    2. Zhang L
    3. Han Z
    4. Yang X
    5. Liu W
    6. Li E
    7. Chang J
    8. Qi Y
    9. Shpak ED
    10. Chai J

    (2017) A receptor-like protein acts as a specificity switch for the regulation of stomatal development

    Genes & Development 31:927–938.

    https://doi.org/10.1101/gad.297580.117

    • PubMed
    • Google Scholar
40. 1. Lu D
    2. Lin W
    3. Gao X
    4. Wu S
    5. Cheng C
    6. Avila J
    7. Heese A
    8. Devarenne TP
    9. He P
    10. Shan L

    (2011) Direct ubiquitination of pattern recognition receptor FLS2 attenuates plant innate immunity

    Science 332:1439–1442.

    https://doi.org/10.1126/science.1204903

    • PubMed
    • Google Scholar
41. 1. Macho AP
    2. Zipfel C

    (2014) Plant PRRs and the activation of innate immune signaling

    Molecular Cell 54:263–272.

    https://doi.org/10.1016/j.molcel.2014.03.028

    • PubMed
    • Google Scholar
42. 1. Martins S
    2. Dohmann EM
    3. Cayrel A
    4. Johnson A
    5. Fischer W
    6. Pojer F
    7. Satiat-Jeunemaître B
    8. Jaillais Y
    9. Chory J
    10. Geldner N
    11. Vert G

    (2015) Internalization and vacuolar targeting of the brassinosteroid hormone receptor BRI1 are regulated by ubiquitination

    Nature Communications 6:6151.

    https://doi.org/10.1038/ncomms7151

    • PubMed
    • Google Scholar
43. 1. Mbengue M
    2. Bourdais G
    3. Gervasi F
    4. Beck M
    5. Zhou J
    6. Spallek T
    7. Bartels S
    8. Boller T
    9. Ueda T
    10. Kuhn H
    11. Robatzek S

    (2016) Clathrin-dependent endocytosis is required for immunity mediated by pattern recognition receptor kinases

    PNAS 113:11034–11039.

    https://doi.org/10.1073/pnas.1606004113

    • PubMed
    • Google Scholar
44. 1. Meckel T
    2. Hurst AC
    3. Thiel G
    4. Homann U

    (2004) Endocytosis against high turgor: intact guard cells of Vicia faba constitutively endocytose fluorescently labelled plasma membrane and GFP-tagged K-channel KAT1

    The Plant Journal 39:182–193.

    https://doi.org/10.1111/j.1365-313X.2004.02119.x

    • PubMed
    • Google Scholar
45. 1. Meng X
    2. Chen X
    3. Mang H
    4. Liu C
    5. Yu X
    6. Gao X
    7. Torii KU
    8. He P
    9. Shan L

    (2015) Differential function of Arabidopsis SERK family Receptor-like kinases in stomatal patterning

    Current Biology 25:2361–2372.

    https://doi.org/10.1016/j.cub.2015.07.068

    • PubMed
    • Google Scholar
46. 1. Mitsuhashi N
    2. Shimada T
    3. Mano S
    4. Nishimura M
    5. Hara-Nishimura I

    (2000) Characterization of organelles in the vacuolar-sorting pathway by visualization with GFP in tobacco BY-2 cells

    Plant and Cell Physiology 41:993–1001.

    https://doi.org/10.1093/pcp/pcd040

    • PubMed
    • Google Scholar
47. 1. Nakano RT
    2. Matsushima R
    3. Ueda H
    4. Tamura K
    5. Shimada T
    6. Li L
    7. Hayashi Y
    8. Kondo M
    9. Nishimura M
    10. Hara-Nishimura I

    (2009) GNOM-LIKE1/ERMO1 and SEC24a/ERMO2 are required for maintenance of endoplasmic reticulum morphology in Arabidopsis thaliana

    The Plant Cell 21:3672–3685.

    https://doi.org/10.1105/tpc.109.068270

    • PubMed
    • Google Scholar
48. 1. Nam KH
    2. Li J

    (2002) BRI1/BAK1, a receptor kinase pair mediating brassinosteroid signaling

    Cell 110:203–212.

    https://doi.org/10.1016/S0092-8674(02)00814-0

    • PubMed
    • Google Scholar
49. 1. Nekrasov V
    2. Li J
    3. Batoux M
    4. Roux M
    5. Chu ZH
    6. Lacombe S
    7. Rougon A
    8. Bittel P
    9. Kiss-Papp M
    10. Chinchilla D
    11. van Esse HP
    12. Jorda L
    13. Schwessinger B
    14. Nicaise V
    15. Thomma BP
    16. Molina A
    17. Jones JD
    18. Zipfel C

    (2009) Control of the pattern-recognition receptor EFR by an ER protein complex in plant immunity

    The EMBO Journal 28:3428–3438.

    https://doi.org/10.1038/emboj.2009.262

    • PubMed
    • Google Scholar
50. 1. Nimchuk ZL
    2. Tarr PT
    3. Ohno C
    4. Qu X
    5. Meyerowitz EM

    (2011) Plant stem cell signaling involves ligand-dependent trafficking of the CLAVATA1 receptor kinase

    Current Biology 21:345–352.

    https://doi.org/10.1016/j.cub.2011.01.039

    • PubMed
    • Google Scholar
51. 1. Peterson KM
    2. Torii KU

    (2012) Long-term, High-resolution confocal time lapse imaging of Arabidopsis cotyledon epidermis during germination

    Journal of Visualized Experiments 4:4426.

    https://doi.org/10.3791/4426

    • Google Scholar
52. 1. Pillitteri LJ
    2. Torii KU

    (2012) Mechanisms of stomatal development

    Annual Review of Plant Biology 63:591–614.

    https://doi.org/10.1146/annurev-arplant-042811-105451

    • PubMed
    • Google Scholar
53. 1. Postma J
    2. Liebrand TW
    3. Bi G
    4. Evrard A
    5. Bye RR
    6. Mbengue M
    7. Kuhn H
    8. Joosten MH
    9. Robatzek S

    (2016) Avr4 promotes Cf-4 receptor-like protein association with the BAK1/SERK3 receptor-like kinase to initiate receptor endocytosis and plant immunity

    New Phytologist 210:627–642.

    https://doi.org/10.1111/nph.13802

    • PubMed
    • Google Scholar
54. 1. Qi X
    2. Han SK
    3. Dang JH
    4. Garrick JM
    5. Ito M
    6. Hofstetter AK
    7. Torii KU

    (2017) Autocrine regulation of stomatal differentiation potential by EPF1 and ERECTA-LIKE1 ligand-receptor signaling

    eLife 6:e24102.

    https://doi.org/10.7554/eLife.24102

    • PubMed
    • Google Scholar
55. 1. Reyes FC
    2. Buono R
    3. Otegui MS

    (2011) Plant endosomal trafficking pathways

    Current Opinion in Plant Biology 14:666–673.

    https://doi.org/10.1016/j.pbi.2011.07.009

    • PubMed
    • Google Scholar
56. 1. Ritz C
    2. Baty F
    3. Streibig JC
    4. Gerhard D

    (2015) Dose-Response analysis using R

    PLOS ONE 10:e0146021.

    https://doi.org/10.1371/journal.pone.0146021

    • PubMed
    • Google Scholar
57. 1. Robatzek S
    2. Chinchilla D
    3. Boller T

    (2006) Ligand-induced endocytosis of the pattern recognition receptor FLS2 in Arabidopsis

    Genes & Development 20:537–542.

    https://doi.org/10.1101/gad.366506

    • PubMed
    • Google Scholar
58. 1. Robinson DG
    2. Langhans M
    3. Saint-Jore-Dupas C
    4. Hawes C

    (2008) BFA effects are tissue and not just plant specific

    Trends in Plant Science 13:405–408.

    https://doi.org/10.1016/j.tplants.2008.05.010

    • PubMed
    • Google Scholar
59. 1. Rychel AL
    2. Peterson KM
    3. Torii KU

    (2010) Plant twitter: ligands under 140 amino acids enforcing stomatal patterning

    Journal of Plant Research 123:275–280.

    https://doi.org/10.1007/s10265-010-0330-9

    • PubMed
    • Google Scholar
60. 1. Santuari L
    2. Scacchi E
    3. Rodriguez-Villalon A
    4. Salinas P
    5. Dohmann EM
    6. Brunoud G
    7. Vernoux T
    8. Smith RS
    9. Hardtke CS

    (2011) Positional information by differential endocytosis splits auxin response to drive Arabidopsis root meristem growth

    Current Biology 21:1918–1923.

    https://doi.org/10.1016/j.cub.2011.10.002

    • PubMed
    • Google Scholar
61. 1. Shiu S-H
    2. Bleecker AB

    (2001) Plant Receptor-Like kinase gene family: diversity, function, and signaling

    Science Signaling 2001:re22.

    https://doi.org/10.1126/stke.2001.113.re22

    • Google Scholar
62. 1. Shpak ED
    2. Lakeman MB
    3. Torii KU

    (2003) Dominant-negative receptor uncovers redundancy in the Arabidopsis ERECTA Leucine-rich repeat receptor-like kinase signaling pathway that regulates organ shape

    The Plant Cell 15:1095–1110.

    https://doi.org/10.1105/tpc.010413

    • PubMed
    • Google Scholar
63. 1. Shpak ED
    2. McAbee JM
    3. Pillitteri LJ
    4. Torii KU

    (2005) Stomatal patterning and differentiation by synergistic interactions of receptor kinases

    Science 309:290–293.

    https://doi.org/10.1126/science.1109710

    • PubMed
    • Google Scholar
64. 1. Sigismund S
    2. Confalonieri S
    3. Ciliberto A
    4. Polo S
    5. Scita G
    6. Di Fiore PP

    (2012) Endocytosis and signaling: cell logistics shape the eukaryotic cell plan

    Physiological Reviews 92:273–366.

    https://doi.org/10.1152/physrev.00005.2011

    • PubMed
    • Google Scholar
65. 1. Smith JM
    2. Salamango DJ
    3. Leslie ME
    4. Collins CA
    5. Heese A

    (2014) Sensitivity to Flg22 is modulated by ligand-induced degradation and de novo synthesis of the endogenous flagellin-receptor FLAGELLIN-SENSING2

    Plant Physiology 164:440–454.

    https://doi.org/10.1104/pp.113.229179

    • PubMed
    • Google Scholar
66. 1. Somssich M
    2. Ma Q
    3. Weidtkamp-Peters S
    4. Stahl Y
    5. Felekyan S
    6. Bleckmann A
    7. Seidel CA
    8. Simon R

    (2015) Real-time dynamics of peptide ligand-dependent receptor complex formation in planta

    Science Signaling 8:ra76.

    https://doi.org/10.1126/scisignal.aab0598

    • PubMed
    • Google Scholar
67. 1. Sugano SS
    2. Shimada T
    3. Imai Y
    4. Okawa K
    5. Tamai A
    6. Mori M
    7. Hara-Nishimura I

    (2010) Stomagen positively regulates stomatal density in Arabidopsis

    Nature 463:241–244.

    https://doi.org/10.1038/nature08682

    • PubMed
    • Google Scholar
68. 1. Takagi J
    2. Renna L
    3. Takahashi H
    4. Koumoto Y
    5. Tamura K
    6. Stefano G
    7. Fukao Y
    8. Kondo M
    9. Nishimura M
    10. Shimada T
    11. Brandizzi F
    12. Hara-Nishimura I

    (2013) MAIGO5 functions in protein export from Golgi-associated endoplasmic reticulum exit sites in Arabidopsis

    The Plant Cell 25:4658–4675.

    https://doi.org/10.1105/tpc.113.118158

    • PubMed
    • Google Scholar
69. 1. Torii KU

    (2004) Leucine-rich repeat receptor kinases in plants: structure, function, and signal transduction pathways

    International Review of Cytology 234:1–46.

    https://doi.org/10.1016/S0074-7696(04)34001-5

    • PubMed
    • Google Scholar
70. 1. Uchida N
    2. Lee JS
    3. Horst RJ
    4. Lai HH
    5. Kajita R
    6. Kakimoto T
    7. Tasaka M
    8. Torii KU

    (2012) Regulation of inflorescence architecture by intertissue layer ligand-receptor communication between endodermis and phloem

    PNAS 109:6337–6342.

    https://doi.org/10.1073/pnas.1117537109

    • PubMed
    • Google Scholar
71. 1. Uchida N
    2. Takahashi K
    3. Iwasaki R
    4. Yamada R
    5. Yoshimura M
    6. Endo TA
    7. Kimura S
    8. Zhang H
    9. Nomoto M
    10. Tada Y
    11. Kinoshita T
    12. Itami K
    13. Hagihara S
    14. Torii KU

    (2018) Chemical hijacking of auxin signaling with an engineered auxin-TIR1 pair

    Nature Chemical Biology 14:299–305.

    https://doi.org/10.1038/nchembio.2555

    • PubMed
    • Google Scholar
72. 1. Uemura T
    2. Kim H
    3. Saito C
    4. Ebine K
    5. Ueda T
    6. Schulze-Lefert P
    7. Nakano A

    (2012) Qa-SNAREs localized to the trans-Golgi network regulate multiple transport pathways and extracellular disease resistance in plants

    PNAS 109:1784–1789.

    https://doi.org/10.1073/pnas.1115146109

    • PubMed
    • Google Scholar
73. 1. Zheng H
    2. Camacho L
    3. Wee E
    4. Batoko H
    5. Legen J
    6. Leaver CJ
    7. Malhó R
    8. Hussey PJ
    9. Moore I

    (2005) A Rab-E GTPase mutant acts downstream of the Rab-D subclass in biosynthetic membrane traffic to the plasma membrane in tobacco leaf epidermis

    The Plant Cell 17:2020–2036.

    https://doi.org/10.1105/tpc.105.031112

    • PubMed
    • Google Scholar
74. 1. Zhou J
    2. Liu D
    3. Wang P
    4. Ma X
    5. Lin W
    6. Chen S
    7. Mishev K
    8. Lu D
    9. Kumar R
    10. Vanhoutte I
    11. Meng X
    12. He P
    13. Russinova E
    14. Shan L

    (2018) Regulation of Arabidopsis brassinosteroid receptor BRI1 endocytosis and degradation by plant U-box PUB12/PUB13-mediated ubiquitination

    PNAS 115:E1906–E1915.

    https://doi.org/10.1073/pnas.1712251115

    • PubMed
    • Google Scholar

Acceptance summary:

This work provides a better understanding of ligand-dependent control of intracellular receptor distribution. We are very pleased to publish this paper at eLife because it outlines how vesicle trafficking contributes to signal integration during stomata differentiation in Arabidopsis.

Decision letter after peer review:

[Editors’ note: the authors submitted for reconsideration following the decision after peer review. What follows is the decision letter after the first round of review.]

Thank you for submitting your work entitled "The manifold actions of signaling peptides on subcellular dynamics of a receptor specify stomatal cell fate" for consideration by eLife. Your article has been reviewed by a Senior Editor, a Reviewing Editor, and three reviewers. The following individual involved in review of your submission has agreed to reveal their identity: Grégory Vert (Reviewer #2).

While the reviewers showed interest in your cell biological analysis, they also discussed several shortcomings, which precludes its publication at eLife in its present form. On one hand, they ask you to improve the quantification and image quality throughout your manuscript. On the other hand, they request additional experiments, including the generation of new lines (e.g. kinase dead ERL1), to further support your conclusion. It is policy of eLife to avoid lengthy re-review processes with uncertain outcomes. Therefore, we decided to reject the paper at this stage. Please find the specific comments below.

Reviewer #1:

The authors use genetics, microscopy and pharmacological treatments to determine sub cellular dynamics of the receptor ERL1 upon perceiving distinct ligands in the stomatal lineage of Arabidopsis. Through pharmacological experiments they show that the receptor ligand pair goes to the MVB and the vacuole upon endocytosis and can distinguish activated receptor routes towards degradation from general receptor endocytosis. Furthermore, the authors propose an attractive hypothesis where the coreceptor TMM modulates internalization of ERL1 in a ligand-specific manner. The experiments are well controlled, described in detail both in terms of methodology and sample size and the paper is written clearly and with an easy to follow structure. I have several concerns that I would like to see addressed and a couple of suggestions that would help clarity and visualization purposes.

I would like to stress that I am not a cell biologist and not 100% familiar with the limitations of the pharmacological assays.

A) Major concerns:

1) The data clearly shows that upon ligand perception the ERL1 gets internalized, send to MVB and likely degraded in the vacuole. While it is extremely interesting that these dynamics get modulated by co-receptor and the specific ligand, to me the mechanistic proof that internalization is relevant is somewhat lacking. Particularly when considering that very little of the PM-localized receptor gets internalized (unlike for FSL2, which the authors also discuss). If endocytosis and MVB mediated degradation is relevant then I would expect a patterning and clustering phenotype upon treatment with an endocytosis inhibitor (ES9-17 or Tyr A23) or even with Wm that blocks MVBs. Even though these experiments are tricky and probably must be done in a pulsed set-up, I think the manuscript would profit from phenotypic analysis of the leaf epidermis upon blocking the aforementioned processes (particularly endocytosis).

2) I am wondering why the authors focused on mEPF1 only since they are having mEPF2 available and Lee et al., 2012 showed that mEPF2 also binds to ERL1. Since the two peptides have not quite the same genetic role (one seems to inhibit the stomatal lineage at earlier stages than the other) it would be interesting to see if mEPF2 treatments induce the same internalization and sub-cellular dynamics and whether EPF2-ERL1 internalization is TMM dependent, too.

3) I feel somewhat uneasy about the fact that the authors use BFA treatment to infer anything about "general recycling" of the receptor. In my understanding BFA bodies are stalled TGN vesicles, which can go either to the MVB-vacuole pathway or get recycled to the PM. Therefore the BFA treatments and resulting BFA bodies rather indicate general internalization/endocytosis rather than recycling. To show recycling the authors would need to perform BFA washout experiments, but this is likely not possible since so much PM signal persists. I would like to ask the authors to change recycling for endocytosis/internalization throughout the manuscript.

Along the same lines, Figure 7 shows the receptor recycling as something isolated on the upper right corner. In my opinion this also goes through the TGN and should be drawn accordingly. Everything gets internalized through endocytosis and form TGN/EE. Then they either cycle back to the PM (recycling) or go to LE/MVB to be targeted to the vacuole. Please change the model accordingly and indicate which steps are inhibited by BFA, Wm, and ES9-17 and Tyr A23 in Figure 7.

B) Concerns in Figures:

4) General comments regarding figures:

- I would like to suggest that for all the quantifications of cells with bodies (eg. Figure 1C,E; Figure 2F, Figure 3E, F etc.) to label the y-axis with "% (or n) of YFP-positive bodies/endosomes/etc.".

- In addition, I would appreciate that for all plots the jitter of the data is indicated much like for the violin plots. It is currently missing in Figure 1C, 1E, 1G, Figure 2G.

- Finally, all bar charts should be box plots or violin plots (Figure 3G; Figure 4 – Figure supplement 1 B and C)

5) Figure 1:

I don't quite get what the difference between ERL1 vs. SYP22 is compared to SYP22 vs. ERL1. Please elaborate in the legend.

6) Figure 2:

In panel A it is not clear that the leftmost π stainings are different images than the ones showing YFP channels. Please split phenotype and YFP pictures into panel A and panel B, respectively.

For the FRAP experiments, I wonder why there seems to be more recovery albeit slower recovery in tmm. Is this something that was observed in all 9 replicates? If yes, please quantify and discuss this data, too. If not, I would suggest to use a different representative fluorescence recovery curve for tmm. And please also elaborate on why WT vs. tmm was so unbalanced in terms of replicates (2G, 3 vs. 9 respectively).

7) Figure 3.

In Figure 3E, it is really hard to see the FM4-64 internalization. Maybe use a more focussed and enlarged field of view?

I think that Figure 3F and Figure 3G should be combined or at the very least should be both transformed into violin plots with jitter so that they can be compared. It would also be much easier to read if it indicated in the figure and not just the legend that one is ERL1-YFP bodies and the other is ERL1deltaK-YFP bodies.

Please indicate the n for all treatments in 3G in the figure legend and also show as jitter on a violin or box plot. Otherwise it is really hard to judge the quality of this data.

8) Figure 5.

Please specify exact p values in Figure 5E and 5F as you did in Figure 4E and 4F.

9) Figure 6.

In 6B and 6C it is really hard to see the lines indicating where the fluorescence intensity is measured.

In 6D I would suggest to choose other colors than black dark gray and light gray. Or bigger difference between dark grey and light grey.

C) Concerns in text:

10) subsection “ERL1 is internalized through multivesicular bodies to vacuolar pathway in stomatal meristemoids”.

What do you mean with differentiating meristemoids? Are not all meristemoids differentiating?

11) Subsection “TMM is required for the process of ERL1 endocytosis in true leaves”.

I would appreciate if these quantifications would be included as box plots with jitter in the Figure 2—figure supplement 2 so that we can see the n, the variation in the data etc.

12) Subsection “An antagonistic EPFL peptide, Stomagen, elicits retention of ERL1-YFP in the endoplasmic reticulum”.

For the Stomagen experiments the authors use the er erl1 erl2 triple mutant to remove potential redundancy. I am a little worried that for some of the EPF1 and EPF6l experiments the results would look different in the triple mutant as well. I would appreciate the authors discussing this issue here or in the Discussion section.

13) Subsection “An antagonistic EPFL peptide, Stomagen, elicits retention of ERL1-YFP in the endoplasmic reticulum”.

I think that the data does not warrant such a strong conclusion. The ER accumulation is seen independent of TMM but might be stronger in tmm background. But for this statement to be valid the authors have to compare fluorescence intensity in wt vs. tmm background.

14) Subsection “Pharmacological treatment”.

I am a bit worried by the immensely different incubation times for MEPF1 / MEPFL6 compared to Stomagen (10minutes vs. 1hour). Would we also see ER retention if MEPF1 was incubated for 1h, too? Please show that this is not the case.

Reviewer #2:

The manuscript by Qi et al., reports on the characterization of the dynamics of the ERECTA-LIKE1 receptor and its dependence on the building of a functional receptor complex. It is a very interesting manuscript that provides a detailed analysis of ERL1 trafficking upon ligand binding or activation. Below are listed the most important points that need to be addressed by authors to improve the quality and readability of their contribution.

1) The confocal microscopy images require some more work to fully convince the reader :

- Figure 2A, rather than quantifying the ratio of cells showing ERL1 positive endosomes in WT and tmm, it would be more informative to quantify the number of ERL1-positive endosomes per cells in WT and tmm. To do so, z-stacks across the whole cell and max projection must be performed to avoid bias from uneven distribution of endosomes between, cell surface and middle section.

This comment applies to other figures. For example, the authors quantified the number of endosomes in a given confocal section in subsequent figures (Figure 5A) but must provide quantification per cell using z-stacks and max projection.

- Quantification of % of cells showing BFA or Wm bodies is not appropriate. Please also quantify BFA body number and size per cell. This is a better readout of endocytic defects.

- Some of the images would deserve higher magnification or enlarged panels to better visualize endosomes, lack of endosomes, or BFA/Wm bodies.

2) BFA has been widely characterized and used in Arabidopsis roots where it yields bodies containing aggregated TGN surrounded by Golgi. In aerial parts, BFA is known to trigger the collapsing of Golgi into the ER. Whether it also yields endosomal defects in stomatal meristemoids must be documented or tested if undocumented. The images shown the effect of BFA being very zoomed out, it is hard to evaluate the effect of BFA.

3) BFA leads, according to the authors, to the aggregation of TGN indicating that proteins trafficking through the TGN will end up (recycled or not) in BFA bodies. The fact that ERL1 is observed in BFA bodies in tmm upon CHX is therefore puzzling. One would anticipate a decrease (or even absence) of ERL1 if internalization is compromised. A better quantification of BFA bodies (BFA body number/cell and size; see 2)) is required to better characterize possible defects in ERL1 internalization. If the result stands, the authors will have to provide some explanations in the Discussion section. Regardless, a more direct investigation of ERL1 internalization kinetics in WT or tmm background by VAEM imaging would GREATLY strengthen the authors conclusions. I am not sure the authors have access to VAEM, but if so, I strongly encourage them to try.

4) The FRAP experiment raises some questions. First, the recovery of ERL1-YFP in erl1 is almost absent in my opinion. There is hardly any recovery compared to T0 (slope is flat). As such, it is hard to interpret and one may come up with a very different conclusion where ERL1 is largely immobile at the cell surface, and that it is more mobile in tmm. The authors should (i) show the prebleach intensity on their plot and (ii) image faster after photobleaching to catch the initial recovery phase better.

5) The internalization of ERL1 and targeting to MVB suggests that ERL1 is routed to the vacuole for degradation, which the authors have not shown. The authors must therefore assess ERL1 degradation by western blot if ERL1 is detectable in leaf extracts and/or by confocal imaging inhibiting the lytic activity of the vacuole (dark conditions of concanamycin A).

6) The ER retention of ERL1 upon stomagen application is puzzling yet very interesting. The authors must back up their observations by monitoring the endogenous ERL1 protein by immunofluorescence (or alternatively an ERL1-HA or FLAG if the authors do not have an ERL1 antibody) to ascertain that the YFP fusion protein, although active, represents the behavior of genuine ERL1 for ER retention. Second, the authors must address whether this is specific or ERL1 by testing the influence of stomagen on other secreted proteins (PM proteins) that woumd be expressed (e,dogenously or artificially) in stomata.

Reviewer #3:

This manuscript by Qi et al. investigated how the ERL1 receptor interprets different signal inputs and how this affects its subcellular dynamics. Overall, the manuscript is rather descriptive and fails to provide convincing mechanistic insight, leaving many loose ends. In addition, the text needs to be streamlined with respect to terminology (MVBs, endosomes, BFA-bodies, etc) and what is actually shown on the figures to avoid mis- or over-interpretation. Furthermore, the conclusions need to be more accurately phrased.

1) Regarding the statement that "TMM is required for ERL1 endocytosis", I would phrase this differently. TMM is required to form an active signaling complex, and in the absence of such an active signaling complex, no endocytosis occurs (as is also the case upon Stomagen treatment or when the kinase domain is removed). How do the authors explain the Stomagen-mediated accumulation of ERL1 in the endoplasmic reticulum? This must be a different pool of ERL1 distinct from the Stomagen-bound ERL1. Do the authors assume a regulatory mechanism whereby Stomagen-bound ERL1 signals to control the production and ER retention of new ERL1? In this context, I would conclude that it is not inefficient ERL1 endocytosis that leads to retention in the ER, but an inactive receptor that stays on the membrane which leads to no new secretion. Does the ERL1Δkinase also accumulate in the ER?

2) With the current data, it cannot be concluded that kinase activity is required for ERL1 internalization. There could be internalization motifs in the kinase domain. To conclusively demonstrate this, an ERL1 kinase dead variant needs to be analyzed with respect to internalization. In the absence of TMM, un-activated ERL1 receptors are not readily targeted for endocytosis and remain stable at the membrane. What happens with ERL1Δkinase endocytosis in the presence of ligand?

3) The authors use different markers to reveal the localization of ERL1-YFP. However, a typical experiment to study endocytosis is the co-localization with FM4-64. Figure 2—figure supplement 2A and Figure 3E provide some idea, but the individual endosomes should be tracked. BFA bodies and MVBs are a proxy to study this, but then the phrasing in the text should be accordingly.

4) TMM is required for endocytosis of ERL1, and not recycling. BFA blocks the secretion of ERL1. One would expect that if there is less endocytosis, there is also less ERL1 in BFA bodies. However, based on the data in Figure 2 this is not the case. Also, Figure 2B shows an increased ERL1-YFP intensity upon BFA treatment in tmm, so both the number of cells with BFA bodies and the signal intensity would be meaningful to report. I would rephrase the conclusion in subsection “TMM is required for the process of ERL1 endocytosis in true leaves”: the absence of tmm does not impact the ERL1 secretory pathway, as reflected in the presence of ERL1-YFP positive BFA bodies.

5) The authors invoke that distinct EPF/EPFL peptide ligands activate a sub-population of ERL1 receptor complex to internalize through a TMM-based discriminatory system. This seems independent of distinct expression domains of these peptides as purified EPFL6 does not have the same effect as EPF1. The authors suggest distinct localization in nanodomains, but do not investigate this. It would be valuable to see FRAP analyses on ERL1-YFP +/1 EPF1 or EPFL6 in erl1 and erl1 tmm.

6) Throughout the manuscript, MVBs should be used when describing the "large dots" on the figures. I am not sure if the resolution allows to see endosomes. Are Figure 4E-F and Figure 5E-F really showing number of endosomes or should this be MVBs?

7) The authors write "TMM is required for endocytic sorting of ERL1 to MVBs, which is a hallmark for receptor degradation". I guess that given the fact that in tmm ERL1-YFP signal increases in the ER, it is rather difficult to investigate this statement using e.g. Western Blot.

8) Given the recent knowledge on Tyr A23, I would only include the more specific ES9-17. Is this effect on ES9-17 specific for ERL1-YFP, or are other receptor proteins also accumulating in the ER?

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "The manifold actions of signaling peptides on subcellular dynamics of a receptor specify stomatal cell fate" for consideration by eLife. Your article has been reviewed by Christian Hardtke as the Senior Editor, Jürgen Kleine-Vehn as the Reviewing Editor, and two reviewers. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

As the editors have judged that your manuscript is of interest, but as described below that additional experiments are required before it is published, we would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). First, because many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts. We are also offering, if you choose, to post the manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is "in revision at eLife". Please let us know if you would like to pursue this option. (If your work is more suitable for medRxiv, you will need to post the preprint yourself, as the mechanisms for us to do so are still in development.)

Your manuscript reveals remarkable insights into peptide ligands and how they modulate the intracellular distribution of their receptors. We invite you to resubmit your paper for final acceptance, considering the following main suggestions of the reviewers:

I) The reviewers ask you to clarify several issues, such as statistical evaluation and quantifications, which may not require novel experiments.

II) During their discussion, the reviewers defined a better characterization and integration of the ERL1Δkinase line, as an essential part of the revision, and one that unfortunately requires additional experimentation.

III) The reviewers judge that some experiments aimed at a better integration of your data into developmental context would substantially strengthen the impact of your manuscript. On the other hand, they do not believe that such an insight is essential for acceptance of the manuscript.

Please, see the detailed comments of the reviewers below.

Reviewer #1:

This manuscript by Qi et al., investigated how the ERL1 receptor interprets different signal inputs and how this affects its subcellular dynamics. There are several novel and interesting insights, but still some loose ends remain. The authors have addressed several of the comments raised previously and improved the manuscript.

Reviewer #2:

The authors undertook a remarkable effort to improve the cell biological aspects of this manuscript and I do appreciate the new experiments and careful quantification of the BFA experiments and analysis of new marker lines with expert Dr. Akira Yoshinari. In addition, the ConcavA treatments and the clear visualization of ERL1 in vacuoles adds significantly to this manuscript. However, I would like to stress two of my previous concerns again.

1) Previous point 1:

From a mechanistic/genetic perspective, the functional relevance of ERL1 endocytosis and targeting to MVB/LE and vacuole is still missing. I do completely understand that a phenotypic assay is not doable due to long-time cytotoxic effects of BFA or Wm. I also agree that strong genetic support can only come through analysis of how ERL1 is modified (phosphorylation etc) and how these modifications affect endocytosis, but this is clearly out of scope of this article. Nevertheless, in Qi et al., 2017, the authors show that the EPF1-ERL1 module directly represses MUTE in an autocrine manner. When EPF1 expression is induced, a clear decrease in MUTE-YFP fluorescence can be observed already 15hours post induction. I wonder whether induction of EPF1 together with mock or toxin treatment (Wm?) in the first couple of hours could reveal differences in how efficiently MUTE-YFP is downregulated. This could give at least somewhat functional support to the hypothesis that the endocytosis and targeting to the vacuole is required for efficient downregulation of key targets by the EFP1-ERL1 module.

2) Previous point 12:

Even though the authors stated in their response that they elaborated on ER ERL1 and ERL2 redundancy in the discussion, I cannot find new information or statements regarding the fact that the Stomagen-induced ER-retainment of ERL1 is much stronger in the triple erecta-erl1-erl2 background. I am aware that this is a subtle difference but as the authors confirm an important one. So please elaborate on this in the results (line 298ff) and clearly mention that Stomagen has been shown to interact initially with ER before acting through ERL1 and ERL2 at later stages, which is why a triple mutant is needed here and ER retainment in the single erl1 is much weaker.

We took the constructive criticisms from the three Peer Reviewers very seriously, and performed the following experiments to fully address their concerns.

i) Concanamycin A treatment to seedlings expressing ERL1-YFP: The results demonstrate that ERL1-YFP is destined to a vacuole (see Revised Figure 2).

ii) Stomagen application to BRI1pro::BRI1-GFP seedlings: BRI1-GFP is endongenously expressed in leaf epidermis, includingstomatal meristemoids. The results show that Stomagen peptide does not influence subcellular localization of BRI1, thereby supporting the notion that the effects of Stomagen on ERL1 is specific (see Revised Figure 8–figure supplement 2).

iii) Extensive documentation of BFA body formation in stomatal meristemoids: We used eight different endomembrane organelle markers with an endocytic tracer dye FM4-64 to fully document the behavior of endomembrane system upon BFA treatment in meristemoids. Our results clearly demonstrate that, at low concentration (30 microM), BFA body formation occurs just like in root cells (See Revised Figure 3 and Figure 3–figure supplement 1).

Furthermore, we re-analyzed BFA and Wm treatment data as suggested by the Reviewers, and provided enlarged, high-resolution confocal images as much as possible. By analyzing the number of endosomes/BFA/Wm bodies per cell as well as the volume ratio, we were able to support our original conclusion that TMM is critical for the internalization of EPF1-perceived ERL1. Please see our point-by-point response for specifics.

Reviewer #1:

[…] A) Major concerns:

1) The data clearly shows that upon ligand perception the ERL1 gets internalized, send to MVB and likely degraded in the vacuole. While it is extremely interesting that these dynamics get modulated by co-receptor and the specific ligand, to me the mechanistic proof that internalization is relevant is somewhat lacking. Particularly when considering that very little of the PM-localized receptor gets internalized (unlike for FSL2, which the authors also discuss). If endocytosis and MVB mediated degradation is relevant then I would expect a patterning and clustering phenotype upon treatment with an endocytosis inhibitor (ES9-17 or Tyr A23) or even with Wm that blocks MVBs. Even though these experiments are tricky and probably must be done in a pulsed set-up, I think the manuscript would profit from phenotypic analysis of the leaf epidermis upon blocking the aforementioned processes (particularly endocytosis).

We thank reviewer 1 for finding our work extremely interesting. We agree with reviewer 1 that it would be ideal if we could use endocytosis inhibitors to block ERL1 endocytosis and then characterize the resulting stomatal pattering. Unfortunately, extended treatment of these inhibitors for several days to observe stomatal development is not possible due to high cytotoxicity of these inhibitors. All of our results clearly show the subcellular dynamics of ERL1 upon perceiving its ligands with different functions (EPF1, Stomagen, and EPFL6). Furthermore, the dominant-negative receptor that cannot signal is compromised in internalization. In the future, identification of receptor modification which influence internalization (e.g. phosphorylation, ubiquitination) may fully answer reviewer 1's question, which are out-of-scope of this manuscript.

2) I am wondering why the authors focused on mEPF1 only since they are having mEPF2 available and Lee et al., 2012 showed that mEPF2 also binds to ERL1. Since the two peptides have not quite the same genetic role (one seems to inhibit the stomatal lineage at earlier stages than the other) it would be interesting to see if mEPF2 treatments induce the same internalization and sub-cellular dynamics and whether EPF2-ERL1 internalization is TMM dependent, too.

We have previously reported that mEPF1 and ERL1 are both expressed in stomatal meristemoids and participate in autocrine signaling to restrict stomatal differentiation potential (Qi et al., 2017), in addition to their well-established role in enforcing the stomatal one-cell spacing rule (Hara et al., 2007). The clear expression pattern of ERL1 in the meristemoids and the known consequence of autocrine signaling make mEPF1-ERL1 a unique and powerful system to address how signal perception (activation vs inactivation) triggers receptor subcellular dynamics. While mEPF2 can also be perceived by ERL1, its major receptor, ERECTA is ubiquitously expressed at very high level (Lee et al., 2012), and we predict that this will interfere with the proper interpretation. In the revised manuscript, we emphasized our rationale in the Introduction.

3) I feel somewhat uneasy about the fact that the authors use BFA treatment to infer anything about "general recycling" of the receptor. In my understanding BFA bodies are stalled TGN vesicles, which can go either to the MVB-vacuole pathway or get recycled to the PM. Therefore the BFA treatments and resulting BFA bodies rather indicate general internalization/endocytosis rather than recycling. To show recycling the authors would need to perform BFA washout experiments, but this is likely not possible since so much PM signal persists. I would like to ask the authors to change recycling for endocytosis/internalization throughout the manuscript.

Along the same lines, Figure 7 shows the receptor recycling as something isolated on the upper right corner. In my opinion this also goes through the TGN and should be drawn accordingly. Everything gets internalized through endocytosis and form TGN/EE. Then they either cycle back to the PM (recycling) or go to LE/MVB to be targeted to the vacuole. Please change the model accordingly and indicate which steps are inhibited by BFA, Wm, and ES9-17 and Tyr A23 in Figure 7.

We truly thank reviewer 1 for raising an important point. We agree that BFA treatment affects not only recycling but also internalization/endocytosis. As reviewer 1 mentions, "BFA washout" time course experiments is generally performed to measure the rate of recycling from BFA body to Plasma membrane. Given that our results indicate that the main effects of tmm mutation is reduced internalization/endocytosis, we do not believe that the additional BFA washout experment is necessary. In response to reviewer 1, we have changed recycling to recycling and endocytosis/internalization throughout the manuscript. Likewise, we have revised the model figure to correctly reflect the literature and our inhibitor results (now Figure 9).

During the revision process, we extensively discussed how to address the questions raised by the reviewers regarding our BFA treatment experiments. We invited an expert, Dr. Akira Yoshinari, who studies subcellular dynamics of polarly-localized membrane transporters. Through collaboration, we performed a series of BFA-treatment experiments in the context of stomatal meristemoids using a variety of subcellular membrane markers (revised Figure 3 and Figure 3—figure supplement 1; see our specific response to reviewer 2). Furthermore, we performed a pharmacological treatment using Concanamycin A, a specific inhibitor of vacuolar H⁺-ATPase that is known to reduce protein degradation in the lytic vacuole (e.g. Kleine-Vehn et al., 2008). We now demonstrate that ERL1 indeed accumulates in a vacuole (see revised Figure 2). In summary, all the additional experiments we performed support our original findings that ERL1, upon activated by EPF peptides, is subjected to endocytosis/internalization destined to vacuole and the process gets compromised in the absence of TMM.

B) Concerns in Figures:

4) General comments regarding figures:

- I would like to suggest that for all the quantifications of cells with bodies (eg. Figure 1C,E; Figure 2F, Figure 3E, F etc.) to label the y-axis with "% (or n) of YFP-positive bodies/endosomes/etc.".

We have modified the y-axis labels and re-analyzed our data to plot the number of

YFP positive bodies/endosomes/etc per cell (see revised Figure 4F, Figure 6—figure supplement 1B, C). For a quantitative analysis of BFA bodies, as also requested by reviewer 2, we have re-analyzed the voxel YFP intensity within the cell (representing endosomes and BFA bodies) per voxel YFP intensity of cell periphery (plasma membrane) and plotted the data to reveal subtle reduction in tmm mutant background (see revised Figure 4D). This analysis indeed revealed a reduction in the BFA body signals by the loss-of-function mutation in TMM. We additionally reanalyzed Wm treatment and plotted boxplots as number of Wm bodies per cell (see revised Figure 4F, Figure 4 —figure supplement 2D).

- In addition, I would appreciate that for all plots the jitter of the data is indicated much like for the violin plots. It is currently missing in Figure 1C, 1E, 1G, Figure 2G.

Provided for all box plot graphs (see revised Figure 1C, Figure 4D, 4F, Figure 6—figure supplement 1B, C).

- Finally, all bar charts should be box plots or violin plots (Figure 3G; Figure 4 – Figure supplement 1 B and C)

Done. The original Figure 3G is a percentage of all cell counted. We removed the chart and just indicated the percentage in the text.

5) Figure 1:

I don't quite get what the difference between ERL1 vs. SYP22 is compared to SYP22 vs. ERL1. Please elaborate in the legend.

We thank reviewer 1 for asking for the clarification.

ERL1 vs. SYP43 indicates how many ERL1-YFP endosomes co-localize with SYP43-RFP. Co-localized endosomes are in white arrowheads.

SYP43 vs. ERL1 indicates how many SYP43-RFP endosomes co-localize with ERL1-YFP endosomes. Co-localized endosomes are in white arrowheads.

The % colocalization are not identical because some endosomes only accumulate ERL1-YFP whereas some other endosomes only accumulate SYP43-RFP. This disparity is most evident when comparing SYP22-RFP and ERL1-YFP, because almost all endosomes accumulates YFP signals alone, whereas nearly all of the RFP endosomes co-localize with YFP signals. To be clear to readers, we now describe this in the figure legend. Thank you.

6) Figure 2:

In panel A it is not clear that the leftmost π stainings are different images than the ones showing YFP channels. Please split phenotype and YFP pictures into panel A and panel B, respectively.

Reviewer 1 is correct. Our intention in the original Figure was to show the stomatal cluster phenotype of tmm. To be unambiguous, we revised the figure to split the figure panels (now Figure 4A and B).

For the FRAP experiments, I wonder why there seems to be more recovery albeit slower recovery in tmm. Is this something that was observed in all 9 replicates? If yes, please quantify and discuss this data, too. If not, I would suggest to use a different representative fluorescence recovery curve for tmm. And please also elaborate on why WT vs. tmm was so unbalanced in terms of replicates (2G, 3 vs. 9 respectively).

We have used FRAP Bot program (www.frapbot.kohze.com) to normalize and fit our fluorescence recovery curve to a single-parameter exponential model. The fitted curve, but not the original fluorescence intensity changes, are provided in the graphs (revised Figure 4G). As far as our understanding, recovery curve is a standard way to determine the half time, and similar curve fitting has been done by many others in the field (e.g. Martinière et al., 2012; Zhang et al., 2016). We performed 10 independent FRAP experiments, and the FRAP Bot program chose the experiment that passed their criteria (we thought it would be better for the algorithm to perform unbiased curve fitting, rather than we choose the 'representative' data). To be clear, we indicated this in our revised Method section. Furthermore, in response to a similar critique from reviewer 2, we now provide a graph of observed florescence intensity change in the representative FRAP experiment (Figure 4—figure supplement 3 in the revision) as well as confocal images throughout a representative FRAP experiment (Video 2 and Video 3).

7) Figure 3.

In Figure 3E, it is really hard to see the FM4-64 internalization. Maybe use a more focussed and enlarged field of view?

Thank you for the suggestion. As also requested by reviewer 2, we now provide enlarged field of view throughout our confocal microscopy image figures.

I think that Figure 3F and and Figure 3G should be combined or at the very least should be both transformed into violin plots with jitter so that they can be compared. It would also be much easier to read if it indicated in the figure and not just the legend that one is ERL1-YFP bodies and the other is ERL1deltaK-YFP bodies.

Please indicate the n for all treatments in 3G in the figure legend and also show as jitter on a violin or box plot. Otherwise it is really hard to judge the quality of this data.

As suggested, we removed the Figure 3G, and instead, indicated the percentage of cells with endosome bodies within the manuscript.

8) Figure 5.

Please specify exact p values in 5E and 5F as you did in 4E and 4F.

We have performed Welch’s Two-Sample T-Test and the exact p values for pairwise comparison are specified (now revised Figures 6E, 6F, Figure 7E, and 7F).

9) Figure 6.

In 6B and 6C it is really hard to see the lines indicating where the fluorescence intensity is measured.

As suggested, we made the lines thicker (Figure 8 in the revision). Thank you.

In 6D I would suggest to choose other colors than black dark gray and light gray. Or bigger difference between dark grey and light grey.

Done (gray arrows blue and cyan now).

C) Concerns in text:

10) subsection “ERL1 is internalized through multivesicular bodies to vacuolar pathway in stomatal meristemoids”.

What do you mean with differentiating meristemoids? Are not all meristemoids differentiating?

Text changed to late meristemoids.

11) Subsection “TMM is required for the process of ERL1 endocytosis in true leaves”.

I would appreciate if these quantifications would be included as box plots with jitter in the Figure 2—figure supplement 2 so that we can see the n, the variation in the data etc.

Thanks for the suggestion. As also requested by other two Reviewers, we have plotted the number of YFP+ BFA/Wm bodies per cell (per ERL1-YFP expressing meristmoid) and replotted as box plots with jitter (now Figure 4—figure supplement 2 in the revision).

12) Subsection “An antagonistic EPFL peptide, Stomagen, elicits retention of ERL1-YFP in the endoplasmic reticulum”.

For the Stomagen experiments the authors use the er erl1 erl2 triple mutant to remove potential redundancy. I am a little worried that for some of the EPF1 and EPF6l experiments the results would look different in the triple mutant as well. I would appreciate the authors discussing this issue here or in the Discussion section.

We appreciate reviewer 1 for commenting on this, as we agree this is a valid point. Our group has extensively studied a redundancy among three ERECTA family genes and their specific interaction with TMM as well as different EPF peptides (e.g. Shpak, 2005; Hara, 2009; Lee, 2012; Lee, 2015); often these subtle yet important differences are ignored by the scientific community. Through these studies, we have found that whereas EPF1 has rather effects to activate ERL1, Stomagen acts stepwise throughout stomatal lineages, earlier with ERECTA, and ERL1 and ERL2 at a later stage). This complicates the interpretation. We elaborated this in the Discussion section.

13) Subsection “An antagonistic EPFL peptide, Stomagen, elicits retention of ERL1-YFP in the endoplasmic reticulum”.

I think that the data does not warrant such a strong conclusion. The ER accumulation is seen independent of TMM but might be stronger in tmm background. But for this statement to be valid the authors have to compare fluorescence intensity in wt vs. tmm background.

We have changed the sentence to "suggesting that the absence of TMM intensifies the accumulation of ERL1 in the endoplasmic reticulum".

14) Subsection “Pharmacological treatment”.

I am a bit worried by the immensely different incubation times for MEPF1 / MEPFL6 compared to Stomagen (10minutes vs. 1hour). Would we also see ER retention if MEPF1 was incubated for 1hour, too? Please show that this is not the case.

Due to very low amount of bioactive MEPF1 peptide we currently have, we did not perform additional time course experiments. Having said that, using very robust cheically inducible EPF1 overexpression system (which we used previously to robustly trigger EPF1 overexpression in time-controlled manner; Lee et al., 2012, Qi et al., 2017). We performed time course overexpression (which occurs with in less than 1 hour) up to 8 hours, and we did not observe any ERL1-YFP signal in the ER. Thus, it is highly unlikely that the different subcellular localization pattern of ERL1-YFP by MEPF1 or Stomagen application is owing to 50 min difference. If reviewer 1 request, we would be happy to provide the data as Supplemental Figure.

Reviewer #2:

The manuscript by Qi et al., reports on the characterization of the dynamics of the ERECTA-LIKE1 receptor and its dependence on the building of a functional receptor complex. It is a very interesting manuscript that provides a detailed analysis of ERL1 trafficking upon ligand binding or activation. Below are listed the most important points that need to be addressed by authors to improve the quality and readability of their contribution.

We thank reviewer 2 for finding our manuscript very interesting.

1) The confocal microscopy images require some more work to fully convince the reader :

- Figure 2A, rather than quantifying the ratio of cells showing ERL1 positive endosomes in WT and tmm, it would be more informative to quantify the number of ERL1-positive endosomes per cells in WT and tmm. To do so, z-stacks across the whole cell and max projection must be performed to avoid bias from uneven distribution of endosomes between, cell surface and middle section.

This comment applies to other figures. For example, the authors quantified the number of endosomes in a given confocal section in subsequent figures (Figure 5A) but must provide quantification per cell using z-stacks and max projection.

We sincerely appreciate reviewer 2's expert suggestions. We would like to emphasize that all of our quantitative analyses of ERL1-YFP endocytosis were performed by using maximum intensity projection of Z-stack images covering the entire meristmoids, not using single-plane images. We clarified this in the Materials and methods section.

- Quantification of % of cells showing BFA or Wm bodies is not appropriate. Please also quantify BFA body number and size per cell. This is a better readout of endocytic defects.

As suggested by reviewer 2 and all other reviewers, we re-analyzed our images and quantified BFA or Wm body number per cell (revised Figure 4D, F, Figure 4 —figure supplement1D, E, Figure 6—figure supplement 1B, C) and plotted as boxplots with individual data points as jitter.

For comparing the BFA-body formation between WT vs. tmm, in order to see subtle differences, we additionally performed quantification of the voxel signal intensity of inside of the cell (corresponding to endosomes/BFA bodies) per plasma membrane. This analysis was indeed very informative, as we were able to show a subtle yet statistically different voxel signal intensity ratio between WT and tmm (revised Figure 4D). It is well known that BFA strongly inhibits recycling, but also interferes with endocytosis/internalization. The subtle yet significant reduction in BFA bodies in tmm supports our original conclusion that ERL1 internalization requires its co-receptor TMM. Thank you for the great suggestion!

- Some of the images would deserve higher magnification or enlarged panels to better visualize endosomes, lack of endosomes, or BFA/Wm bodies.

As suggested, we now provide enlarged panels throughout (e.g. revised Figure 1A, 1D, Figure 4E, Figure 5D).

2) BFA has been widely characterized and used in Arabidopsis roots where it yields bodies containing aggregated TGN surrounded by Golgi. In aerial parts, BFA is known to trigger the collapsing of Golgi into the ER. Whether it also yields endosomal defects in stomatal meristemoids must be documented or tested if undocumented.

We thank reviewer 2 for bringing up the previous knowledge to our attention.

Indeed, Robinson et al., (2006) and Langhans et al., (2011) reported that BFA triggers re-absorption of Golgi membrane to the ER in leaves. These authors used rather high concentrations of BFA (50 µM for the former report and 90, 180, and 360 µM for the latter report) when observing the collapse of Golgi to ER. By contrast, we used low BFA concentration (30 µM), a condition generally used to block recycling/endocytosis in roots.

Since we did not find previous literatures that study membrane organelle dynamics in BFA-treated stomatal meristemoids, we took this revision process as an opportunity for the extensive documentations. For this purpose, we collaborated with Dr. Akira Yoshinari, an expert in subcellular dynamics of membrane transporters in plants (Dr. Yoshinari is the second author of our revised manuscript).

We collected eight different membrane organelle markers: GFP-HDEL (ER); ST-

YFP (Golgi); YFP-SYP32 (cis-Golgi); GFP-SYP43 (TGN/EE); YFP-RabA1e (TGN, PM); YFP-RabA5d (uncharacterized endosomes); ARA6-GFP (MVB/late endosomes, PM); YFP-ARA7 (MVB/late endosomes), and co-stained with the endocytosis tracer FM4-64, which initially stains plasma membrane.

First, we treated GFP-HDEL/FM4-64 and ST-YFP/FM4-64 with BFA at different concentrations (0, 30, 90, 180 µM). Treatment of low BFA concentration at 30 µM (which we used throughout our study) does not confer any discernable effects on ER (GFP-HEDL, revised Figure 3A, B). Under this condition, BFA bodies are surrounded by Golgi marker ST-YFP (Figure 3C), just as expected for the BFA treatment in roots (Figure 3D) (Langhans et al., 2011). In contrast, treatment of BFA at very high concentration, 180 µM, conferred aberrant spherical structure of ER (Figure 3A, B), which was also previously reported (Nakano et al., 2009). Moreover, we observed Golgi absorbed ER, which confers ring-like structure of membrane accumulating ST-YFP (Figure 3C). In short, we recapitulated the previous observations of collapsing of Golgi into the ER only at extremely high BFA concentration at 180 µM.

Next, we investigated the intactness of different endomenbrane organelles in stomatal meristemoids when treated in 30 µM BFA. For this purpose, seven additional endomembrane markers were carefully examined (YFP-SYP32, N-STYFP, GFP-SYP43, YFP-RabA1e, YFP-RabA5d, ARA6-GFP and YFP-ARA7 costained with FM4-64. All marker lines show the formation of characteristic BFA bodies in the meristemoids, without any collapse of Golgi into ER (Figure 3—figure supplement 1).

Taken together, we conclude that BFA treatment at 30 µM, in our hands, confers the formation of characteristic BFA bodies in the stomatal meristemoids in leaf, and thereby supporting the validity of our BFA treatment to study ERL1 subcellular dynamics. We strongly believe that our extensive documentation will serve as a guide to the field, and we thank reviewer 2 again for the excellent advice.

The images shown the effect of BFA being very zoomed out, it is hard to evaluate the effect of BFA.

As also requested by reviewers 1 and 3, we now provide enlarged field of view throughout our confocal microscopy image figures whenever possible.

3) BFA leads, according to the authors, to the aggregation of TGN indicating that proteins trafficking through the TGN will end up (recycled or not) in BFA bodies. The fact that ERL1 is observed in BFA bodies in tmm upon CHX is therefore puzzling. One would anticipate a decrease (or even absence) of ERL1 if internalization is compromised. A better quantification of BFA bodies (BFA body number/cell and size; see 2)) is required to better characterize possible defects in ERL1 internalization. If the result stands, the authors will have to provide some explanations in the Discussion section. Regardless, a more direct investigation of ERL1 internalization kinetics in WT or tmm background by VAEM imaging would GREATLY strengthen the authors conclusions. I am not sure the authors have access to VAEM, but if so, I strongly encourage them to try.

We have re-analyzed the effects of BFA on ERL1-YFP in WT vs. tmm backgrounds by quantifying the ratio of YFP+ signals within the cell (voxel) per YFP+ signals at plasma membrane using our maximal intensity projection of Z-stack images (formerly Figure 2C, now Figure 4D). The analysis revealed a statistically significant reduction of ERL1-YFP in BFA bodies in tmm mutant, but it is not completely absent. These results suggest that in the absence of TMM, ERL1-YFP internalization is compromised but not absent. This is most likely due to the presence of EPF-LIKE peptide family members that do not require TMM (e.g. EPFL4 and EPFL6). We discussed this point in the Discussion. We agree that VAEM (or TIRF) microscopy would be an interesting next step to study receptor internalization dynamics in the future.

4) The FRAP experiment raises some questions. First, the recovery of ERL1-YFP in erl1 is almost absent in my opinion. There is hardly any recovery compared to T0 (slope is flat). As such, it is hard to interpret and one may come up with a very different conclusion where ERL1 is largely immobile at the cell surface, and that it is more mobile in tmm. The authors should (i) show the prebleach intensity on their plot and (ii) image faster after photobleaching to catch the initial recovery phase better.

We have used FRAP Bot program (www.frapbot.kohze.com) to normalize and fit our fluorescence recovery curve to a single-parameter exponential model. The fitted curve, but not the original fluorescence intensity changes, are provided in the original graphs. As far as our understanding, recovery curve is a standard way to determine the half time, and similar curve fitting has been done by many others in the field (e.g. Martinière et al., 2012; Zhang et al., 2016).

As requested by reviewer 2, we are providing both a graph of observed prebleach florescence intensity and intensity recovery in representative FRAP experiments (Figure 4—figure supplement 3 in the revision) as well as confocal images throughout a representative FRAP experiment (Video 2 and Video 3).

5) The internalization of ERL1 and targeting to MVB suggests that ERL1 is routed to the vacuole for degradation, which the authors have not shown. The authors must therefore assess ERL1 degradation by western blot if ERL1 is detectable in leaf extracts and/or by confocal imaging inhibiting the lytic activity of the vacuole (dark conditions of concanamycin A).

We thank reviewer 2 for the excellent suggestion. With Dr. Yoshinari, we performed Concancmycin A treatment of ERL1-YFP seedlings that were pretreated with FM4-64 following the published protocols (e.g. Kleine-Vehn et al., 2008). Indeed, we observed ERL1-YFP accumulation in the vacuole of stomatal meristemoids (see revised Figure 2). The data support our original conclusion that ERL1, upon activated by EPF peptides, is subjected to internalization and targeted to MVP and destined to the vacuole.

6) The ER retention of ERL1 upon stomagen application is puzzling yet very interesting. The authors must back up their observations by monitoring the endogenous ERL1 protein by immunofluorescence (or alternatively an ERL1-HA or FLAG if the authors do not have an ERL1 antibody) to ascertain that the YFP fusion protein, although active, represents the behavior of genuine ERL1 for ER retention. Second, the authors must address whether this is specific or ERL1 by testing the influence of stomagen on other secreted proteins (PM proteins) that woumd be expressed (e,dogenously or artificially) in stomata.

We do not have endogenous antibody that specifically detect ERL1, and immunostain of epidermal cells have been unsuccessful in our hands. Our Stomagen treatment experiments have been performed multiple times, not just by Confocal microscopy observations but also using biochemical Endo-H treatment.

As suggested by reviewer 2, we treated Stomagen peptide on Arabidopsis seedlings expressing BRI1-GFP driven by its endogenous promoter (BRI1pro::BRI1-GFP). As reported previously (Friedrichsen et al., 2000), BRI1 is ubiquitously expressed in the epidermis, including meristemoids and stomata (see Figure 8—figure supplement 3). The application of Stomagen did not change plasma-membrane localization of BRI1. We did not detect BRI1-GFP in the ER, indicating that the influence of Stomagen on the ER retention is not universal to LRR-RLK.

Reviewer #3:

This manuscript by Qi et al. investigated how the ERL1 receptor interprets different signal inputs and how this affects its subcellular dynamics. Overall, the manuscript is rather descriptive and fails to provide convincing mechanistic insight, leaving many loose ends. In addition, the text needs to be streamlined with respect to terminology (MVBs, endosomes, BFA-bodies, etc) and what is actually shown on the figures to avoid mis- or over-interpretation. Furthermore, the conclusions need to be more accurately phrased.

We sincerely appreciate reviewer 3's expert comments. Our work reports extensive documentation of receptor subcellular localization upon perceiving signaling ligands with different activities as well as when the receptor cannot signal, using the system in which one cell type (stomatal meristemoids) respond to multiple different, and conflicting signals. Thus, we believe our finding add new insight and useful knowledge (together with extensive study of BFA treatment in the meristemoids now presented in Figure 3) to the field. We agree with reviewer 3, and we took our best effort to accurately phrase the observations and results in the revision.

1) Regarding the statement that "TMM is required for ERL1 endocytosis", I would phrase this differently. TMM is required to form an active signaling complex, and in the absence of such an active signaling complex, no endocytosis occurs (as is also the case upon Stomagen treatment or when the kinase domain is removed). How do the authors explain the Stomagen-mediated accumulation of ERL1 in the endoplasmic reticulum? This must be a different pool of ERL1 distinct from the Stomagen-bound ERL1. Do the authors assume a regulatory mechanism whereby Stomagen-bound ERL1 signals to control the production and ER retention of new ERL1? In this context, I would conclude that it is not inefficient ERL1 endocytosis that leads to retention in the ER, but an inactive receptor that stays on the membrane which leads to no new secretion. Does the ERL1Δkinase also accumulate in the ER?

We agree with reviewer 3 that TMM is required to form an active signaling complex, and in the absence of the active signaling complex endocytosis of the receptor gets compromised. We have revised our text throughout the manuscript to be clear. As for the effects of Stomagen on ERL1 accumulation in the ER, together with the pharmacological studies using TyrA23 and ES9-17, suggests that reduced signal activation by Stomagen will result in stalling of ERL1 in the ER. We wish to study the exact subcellular mechanism of the observed stalling in the future.

2) With the current data, it cannot be concluded that kinase activity is required for ERL1 internalization. There could be internalization motifs in the kinase domain. To conclusively demonstrate this, an ERL1 kinase dead variant needs to be analyzed with respect to internalization. In the absence of TMM, un-activated ERL1 receptors are not readily targeted for endocytosis and remain stable at the membrane. What happens with ERL1Δkinase endocytosis in the presence of ligand?

We fully concur with reviewer 3. We have no intention to claim that the kinase activity is required for the internalization. We are fully aware of the fact that removal of the entire cytoplasmic domain will remove potential binding interface with endocytosis adaptor proteins, including those in AP complex or TPLATE complex. Our intention here is to show that a dominant-negative receptor, which is known to severely interfere with it signal transduction pathway (Shpak et al., 2003; Lee et al., 2012), is largely compromised in internalization. We agree that to examine whether the kinase activity affect internalization, kinase-dead variant should be used. We revised our text throughout to clarify this point. We have previously shown that dominant negative ERL1DKinase override EPF1 peptide application (i.e. EPF1 peptide has no effects on epidermis of ERL1DKinase; Lee et al., 2012). Thus, we predict that ERL1DKinase would not be strongly influenced by the peptide application.

3) The authors use different markers to reveal the localization of ERL1-YFP. However, a typical experiment to study endocytosis is the co-localization with FM4-64. Figure 2—figure supplement 2A and Figure 3E provide some idea, but the individual endosomes should be tracked. BFA bodies and MVBs are a proxy to study this, but then the phrasing in the text should be accordingly.

This is a great suggestion. We observed co-localized FM4-64 endosomes and ERL1-YFP signals (now provided in revised Figure 1A), which indeed serves as a starting point to investigate the subcellular dynamics of ERL1. It is evident from our Concanamycin A treatment (revised Figure 2) that FM4-64 signals are eventually detected to a tonoplast and ERL1-YFP targeted into a vacuole. Having said that, in our hands it was not trivial to track the exact snapshots of ERL1 subcellular loalizations in a time-resolved fashion by simply following FM4-64 signals in different endomembrane organelles (e.g. early endosomes, late endosomes, etc). We strongly believe that the use of well-studied endomembrane organelle markers combined with pharmacological analysis provide robust and reliable results. These markers and inhibitors are widely used in the field.

4) TMM is required for endocytosis of ERL1, and not recycling. BFA blocks the secretion of ERL1. One would expect that if there is less endocytosis, there is also less ERL1 in BFA bodies. However, based on the data in Figure 2 this is not the case. Also, Figure 2B shows an increased ERL1-YFP intensity upon BFA treatment in tmm, so both the number of cells with BFA bodies and the signal intensity would be meaningful to report. I would rephrase the conclusion in subsection “TMM is required for the process of ERL1 endocytosis in true leaves”: the absence of tmm does not impact the ERL1 secretory pathway, as reflected in the presence of ERL1-YFP positive BFA bodies.

Both reviewer 1 and reviewer 2 also commented on this. As our response to other reviewers, we have re-analyzed the effects of BFA on ERL1-YFP in WT vs. tmm backgrounds by quantifying the ratio of YFP+ signals within the cell (voxel) per YFP+ signals at plasma membrane using our maximal intensity projection of Z-stack images (formerly Figure 2C, now Figure 4D). The analysis revealed a statistically significant reduction of ERL1-YFP in BFA bodies in tmm mutant. As reviewer 3 thoughtful points out, we revised our Results section based on the new analysis. Thank you.

5) The authors invoke that distinct EPF/EPFL peptide ligands activate a sub-population of ERL1 receptor complex to internalize through a TMM-based discriminatory system. This seems independent of distinct expression domains of these peptides as purified EPFL6 does not have the same effect as EPF1. The authors suggest distinct localization in nanodomains, but do not investigate this. It would be valuable to see FRAP analyses on ERL1-YFP +/1 EPF1 or EPFL6 in erl1 and erl1 tmm.

This is a very interesting point, and certainly it is our future direction to pursue specific nanodomain formation, using FRAP and elaborated microscopy, such as VAEM (as encouraged by reviewer 2).

6) Throughout the manuscript, MVBs should be used when describing the "large dots" on the figures. I am not sure if the resolution allows to see endosomes. Are Figure 4E-F and Figure 5E-F really showing number of endosomes or should this be MVBs?

As suggested, we have modified to MVP/LE, which is also part of late endosome.

7) The authors write "TMM is required for endocytic sorting of ERL1 to MVBs, which is a hallmark for receptor degradation". I guess that given the fact that in tmm ERL1-YFP signal increases in the ER, it is rather difficult to investigate this statement using e.g. Western Blot.

We agree that our original statement might have been too strong. The sentence was removed.

As requested by reviewer 2, we have done pharmacological treatment of ERL1-

YFP seedlings with Concanamycin A, which specifically inhibits vacuolar H⁺ATPase and thus reduce protein degradation in the lytic vacuole. We followed the published protocols (e.g. Kleine-Vehn et al., 2008) and pre-treated with FM464 to track a tonoplast. Our result show that ERL1 indeed accumulates in a vacuole in the presence of Concanamycin A (see revised Figure 2), indicating that ERL1 is targeted to a vacuole. We discussed receptor degradation in the context of our new results.

8) Given the recent knowledge on Tyr A23, I would only include the more specific ES9-17. Is this effect on ES9-17 specific for ERL1-YFP, or are other receptor proteins also accumulating in the ER?

ES9-17 is very new, and as far as we believe we are one of the very first to use this inhibitor in the context of shoot stomatal meristemoids. Although TyrA23 inhibits endocytosis is due to acidification, thereby non-specific manner (Dejonghe, 2016), it has been widely used to block internalization.

Indeed, since the publication of Dejonghe paper, it is still used to study endocytosis of LRR-RLKs (e.g. Song et al., 2019). While we agree with reviewer 3 that ES9-17's effects on endocytosis is more specific and thus clean, we strongly believe that presenting our data that both TyrA23 and ES9-17 triggers the ER accumulation of ERL1-YFP would strengthen our conclusion. To address the specificity to ERL1-YFP, we think that testing Stomage application to other LRR-RLK expressed endogenously in the leaf epidermis would be the better control. To this end, as also suggested by reviewer 2, we examined the effects of Stomagen peptide on the subcellular localization of BRI1 using a published

BRI1pro::BRI1-GFP reporter line (Friedrichsen et al., 2000). The Stomagen application did not change plasma-membrane localization of BRI1, thus the influence of Stomagen is likely specific to ERL1-YFP. Please see Figure 8—figure supplement 3 or our revision.

[Editors’ note: what follows is the authors’ response to the second round of review.]

Your manuscript reveals remarkable insights into peptide ligands and how they modulate the intracellular distribution of their receptors. We invite you to resubmit your paper for final acceptance, considering the following main suggestions of the reviewers:

We truly thank for the positive and encouraging comments.

I) The reviewers ask you to clarify several issues, such as statistical evaluation and quantifications, which may not require novel experiments.

As requested by reviewer 1, we have in addition performed two-way ANOVA for analysis for those data with two parameters, treatment and genotype (Figure 4D, 4F, Figure 4—figure supplement 1, Figure 4—figure supplement 2D, Figure 5-Figure supplement 1B, 1C in the revision). Accordingly, the additional R scripts for the analysis are provided as Source code.

II) During their discussion, the reviewers defined a better characterization and integration of the ERL1Δkinase line, as an essential part of the revision, and one that unfortunately requires additional experimentation.

We agree that testing how ERL1ΔK responds to different ligands is very important to complete our story. As you might hopefully recognize, however, this has been extremely challenging for us. The Torii laboratory at the University of Texas at Austin has been under lock down due to COVID-19 surge in the US, notably terrible in the Southern states including Texas. Even worse, we ran out of bioactive recombinant MEPF1 peptide.

After UT-Austin gave us permission to conduct time-sensitive experiments in July, my postdoc, Dr. Pengfei Bai, who is now a co-author, worked day and night to express, purify, refold, and test bioactivities of their recombinant MEPF1 peptide. The bioactive peptide was then shipped to Japan, where Dr. Akira Yoshinari performed the requested experiments.

As you can see in revised Figure 8, neither MEPF1 nor Stomagen peptide application triggered subcellular localization changes of the dominant-negative ERL1ΔK. ERL1ΔK remained predominantly at the plasma membrane upon MEPF1 or Stomagen application. These new findings suggest that ERL1ΔK is dysfunctional in ligand-induced subcellular trafficking. Because ERL1ΔK lacks the entire cytoplasmic kinase domain, including juxtamembrane domain, kinase domain, and C-terminal tail region, we predict that proper interaction site with endocytosis cargo adaptors may have been lost in ERL1ΔK. While identifying specific cargo adaptors for ERL1 is a future study that is out-of-scope of this manuscript, our additional experiments highlights a highly-tuned, complex actions of antagonistic peptides on ERL1 subcellular dynamics is a highly tuned, coordinated process. We have revised our main text to include these new findings (subsection “Cytoplasmic domain of ERL1 is required for proper subcellular trafficking behavior”, Discussion section).

III) The reviewers judge that some experiments aimed at a better integration of your data into developmental context would substantially strengthen the impact of your manuscript. On the other hand, they do not believe that such an insight is essential for acceptance of the manuscript.

We appreciate the editors' advice. Please see our detailed response to reviewer 2 regarding this point.

Please, see the detailed comments of the reviewers below.

Reviewer #2:

The authors undertook a remarkable effort to improve the cell biological aspects of this manuscript and I do appreciate the new experiments and careful quantification of the BFA experiments and analysis of new marker lines with expert Dr. Akira Yoshinari. In addition, the ConcavA treatments and the clear visualization of ERL1 in vacuoles adds significantly to this manuscript. However, I would like to stress two of my previous concerns again.

We truly thank reviewer 2 for commending our "remarkable effort to improve the cell biological aspects of this manuscript".

1) Previous point 1:

From a mechanistic/genetic perspective, the functional relevance of ERL1 endocytosis and targeting to MVB/LE and vacuole is still missing. I do completely understand that a phenotypic assay is not doable due to long-time cytotoxic effects of BFA or Wm. I also agree that strong genetic support can only come through analysis of how ERL1 is modified (phosphorylation etc) and how these modifications affect endocytosis, but this is clearly out of scope of this article. Nevertheless, in Qi et al., 2017, the authors show that the EPF1-ERL1 module directly represses MUTE in an autocrine manner. When EPF1 expression is induced, a clear decrease in MUTE-YFP fluorescence can be observed already 15hours post induction. I wonder whether induction of EPF1 together with mock or toxin treatment (Wm?) in the first couple of hours could reveal differences in how efficiently MUTE-YFP is downregulated. This could give at least somewhat functional support to the hypothesis that the endocytosis and targeting to the vacuole is required for efficient downregulation of key targets by the EFP1-ERL1 module.

We appreciate reviewer 2's comments. As we all know, chemical inhibitors of membrane trafficking, such as BFA, TyrA23, and Concanamycin A, are a very powerful tool to delineate subcellular dynamics of membrane proteins of interest, whether receptor kinases or transporters. However, there are major problems with the use of these chemical inhibitors for extended period of time to study developmental outcomes.

First and foremost, these inhibitors are not specific to one plasma membrane protein of interest, but rather affect many proteins. It is well known that BFA, for instance, affects recycling of PIN auxin efflux carriers (e.g. Kleine-Vehn et al., 2008), which also plays a role in stomatal patterning (e.g. Le et al., 2014). ES9-17 blocks endocytosis of not only ERL1, but also PIN2 (Iwatate et al., 2020), BRI1 (Dejonghe et al., 2019), and likely many other membrane proteins. Wm could somehow block endocytosis (Emans et al., 2002; Dettmer et al., 2006) in addition to vacuolar sorting. It is therefore impossible to conclude that a developmental phenotype caused by these inhibitors are due to trafficking defects of a single membrane protein. Second, as reviewer 2 acknowledges, these inhibitors possess high cellular toxicity. Extended treatment of seedlings with these chemicals will impact seedling growth and cell proliferation, making it difficult, if not impossible, to characterize stomatal development phenotype. We appreciate that reviewer 2 suggested using a reporter MUTE-YFP line as a quicker marker (~15 hours). However, in our experience, addition of these chemicals for only a few hours would impact the health of the seedlings, thus the interpretation of these experiments would be very tricky.

Our manuscript reports unique, subcellular dynamics of a single receptor (ERL1) by different peptide ligands with distinct biological activities, thereby suggesting the intricate, fine-tuning of developmental patterning by multiple peptide signals. Future analysis, such as identifying critical post-translational modifications of ERL1 and characterizing their impacts on receptor subcellular behavior, will provide a full picture of how intricate regulation of receptor dynamics directly specifies stomatal patterning.

2) Previous point 12:

Even though the authors stated in their response that they elaborated on ER ERL1 and ERL2 redundancy in the discussion, I cannot find new information or statements regarding the fact that the Stomagen-induced ER-retainment of ERL1 is much stronger in the triple erecta-erl1-erl2 background. I am aware that this is a subtle difference but as the authors confirm an important one. So please elaborate on this in the results (line 298ff) and clearly mention that Stomagen has been shown to interact initially with ER before acting through ERL1 and ERL2 at later stages, which is why a triple mutant is needed here and ER retainment in the single erl1 is much weaker.

We fully agree with reviewer 2 that the Endoplasmic Reticulum retention of ERL1-YFP is notable in the erecta erl1 erl2 mutant background, i.e. when functional ERECTA and ERL2 genes are missing. As reviewer 2 astutely points out, we also think that the lack of ERECTA, the major receptor for the early stage of stomatal initiation, confers sensitized, exaggerated response of ERL1-YFP to Stomagen. This is because ERL1 is the only receptor of the family to perceive Stomagen. In addition, it is known that the lack of ERECTA triggers excessive entry divisions of stomatal cell lineages and clustering of meristemoids, where ERL1 is expressed (e.g. Shpak et al., 2005, Lee et al., 2012, and Qi et al., 2017). The dysregulated cell-cell signaling in the erecta erl1 erl2 background likely abrogate the 'buffering' of exogenously applied Stomagen peptide.

In response to reviewer 2, we have elaborated on the main text, stating the following sentences. We sincerely thank reviewer 2 for the keen, thoughtful suggestion.

"The endoplasmic reticulum retention of ERL1-YFP is notable in the erecta erl1 erl2 mutant background, i.e. when functional ERECTA and ERL2 genes are missing. The signals in erl1 ("WT") are much weaker. Because Stomagen interacts with ERECTA at the early stage of stomatal initiation, the absence of ERECTA likely confers sensitized, exaggerated response of ERL1-YFP to Stomagen. Alternatively, a dysregulated cell-cell signaling in the erecta erl1 erl2 background may abrogate the buffering of exogenously applied Stomagen peptide." (Discussion section.)

Author details

1. Xingyun Qi

   Howard Hughes Medical Institute and Department of Biology, University of Washington, Seattle, United States

   Present address

   Department of Biology, Rutgers University, Camden, United States

   Contribution

   Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1261-1362

2. Akira Yoshinari

   Institute of Transformative Biomolecules (WPI-ITbM), Nagoya University, Aichi, Japan

   Contribution

   Resources, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7083-5674

3. Pengfei Bai

   Howard Hughes Medical Institute and Department of Molecular Biosciences, The University of Texas at Austin, Austin, United States

   Contribution

   Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2281-2881

4. Michal Maes

   Howard Hughes Medical Institute and Department of Biology, University of Washington, Seattle, United States

   Present address

   Institute of Systems Biology, Seattle, United States

   Contribution

   Resources, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7406-3982

5. Scott M Zeng

   1. Howard Hughes Medical Institute and Department of Molecular Biosciences, The University of Texas at Austin, Austin, United States
   2. Department of Physics, University of Washington, Seattle, United States

   Contribution

   Software, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3146-8207

6. Keiko U Torii

   1. Howard Hughes Medical Institute and Department of Biology, University of Washington, Seattle, United States
   2. Institute of Transformative Biomolecules (WPI-ITbM), Nagoya University, Aichi, Japan
   3. Howard Hughes Medical Institute and Department of Molecular Biosciences, The University of Texas at Austin, Austin, United States

   Contribution

   Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing

   For correspondence

   ktorii@utexas.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6168-427X

• 2,206

  Page views

• 418

  Downloads

• 10

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.