Direct Interaction of Ligand-Receptor Pairs Specifying Stomatal Patterning

Lee_Jin Suk 1 Direct Interaction of Ligand-Receptor Pairs Specifying Stomatal Patterning Jin Suk Lee, Takeshi Kuroha, Marketa Hnilova, Dmitriy Khatayevich, Masahiro M. Kanaoka, Jessica M. McAbee, Mehmet Sarikaya, Candan Tamerler, and Keiko U. Torii Supplemental Materials Supplemental Figures 1-12 Supplemental Tables 1, 2 Supplemental Figure Legends Supplemental Figure 1: Domain structure and amino-acid sequence of EPF1, EPF2, MEPF1, and MEPF2 used for this study (A) Domain structure of EPF1 and EPF2. Nt, N-terminus; Ct, C-terminus; orange shade, signal peptide (SP); gray box, propeptide region (Pro); blue shade, predicted mature EPF (MEPF) domain with eight cysteines (black) in conserved positions. Scale bar for A-C, 10 amino acids. (B) Domain structure of EPF1 and EPF2 used for Co-IP assays. Black, 3xFLAG tag (Tag). (C) Domain structure of recombinant, E. coli-expressed MEPF1 and MEPF2 used for bioassays, QCM, and SPR. Black, single Myc and 6xHis tag (Tag). (D) Amino-acid sequence of EPF1-FLAG and EPF2-FLAG used for Co-IP assays (see B) and MEPF1 and MEPF2 sequence used for bioassays, QCM, and SPR analyses. (Top) Bold, EPF1 and EPF2 coding sequence; plain, linker and 3xFLAG sequence. (Bottom) Underline: signal sequence from pbadg vector; blue, predicted MEPF1 and MEPF2 sequence; plain, linker, cmyc tag, and 6xHis tag.

Lee_Jin Suk 2 Supplemental Figure 2: EPF-FLAG peptides expressed in N. benthamiana are functional Shown are confocal microscopy of liquid cultured 6-day-old abaxial cotyledons from wild type that were incubated with buffer only (mock; A), buffer with immunopurified EPF1-FLAG (B), or buffer with immunopurified EPF2-FLAG (C). Unlike mock treatment, which has no effects on stomatal differentiation (A: dots), application of EPF1-FLAG conferred arrested meristemoids (B: asterisks), a phenotype identical to induced EPF1 overexpression. Application of EPF2-FLAG conferred severe inhibition of asymmetric entry divisions (C) similar to induced EPF2 overexpression. Concentrations of EPF1-FLAG and EPF2-FLAG immunoeluates are approx. 6.2 M and 7.3 M, respectively. Images were taken under the same magnification. Scale bar, 20 m. Supplemental Figure 3: Epitope-tagged receptors expressed in N. benthamiana are predominantly detected in microsomal fractions Shown is an immunoblot of subcellular fractionations of protein extracts from N. benthamiana transiently expressing ERECTA K-GFP, ERL1 K-GFP, and TMM- GFP driven by the CaMV35S promoter. The fractionation was performed as described in the Supplemental Methods. T, total fraction; C, cytosol fraction; M, microsomal (membrane) fraction. Arrowheads indicate the size of each epitopetagged receptor. Experiments were repeated two times. Supplemental Figure 4: LURE2, an unrelated cysteine-rich peptide, does not associate with ERECTA, ERL1, or TMM Co-IP assays of epitope-tagged receptors and LURE2 expressed in N. benthamiana leaves. Top two panels, input (Input). Bottom two panels, immunoprecipitated fraction using anti-gfp antibodies (IP -GFP). Immunoblots using anti-flag antibodies (WB -FLAG) detect LURE2-FLAG signals in input but not in Co-IP fraction, indicating that LURE2 does not associate with ERECTA K,

Lee_Jin Suk 3 ERL1 K, or TMM. Asterisks, non-specific bands. All experiments were repeated two times. Supplemental Figure 5: Expression and purification of bioactive recombinant MEPF peptides Shown are SDS-PAGE gels stained with Coomassie Brilliant Blue showing expression and purification of bacterially expressed MEPF1-His (A) and MEPF2- His (B). Lanes 1, 2, 5, and 6: Bacterial lysate carrying MEPF1 (lanes 1, 2) and MEPF2 (lanes 5, 6) in the absence (lanes 1, 5) or presence (lanes 2, 6) of L- arabinose for induction. Lanes 3, 7: Eluates from Ni 2+ -NTA column containing purified MEPF1-His (lane 3) and MEPF2-His (lane 7) proteins. Lanes 4, 8: Purified, dialyzed, and refolded peptide solution (lane 4, MEPF1-His; lane 8, MEPF2-His) used for bioassays in Figs. 2, 5. The positions of molecular mass markers in kilodaltons are indicated on the left. Asterisks indicate the size of each His-tagged peptide. Supplemental Figure 6: Separation of properly folded MEPF1 by reverse phase chromatography followed by mass spectrometry and bioassays (A) HPLC chromatogram of purified MEPF1-His. Refolded MEPF1 solution was fractionated by HPLC using a C18 reverse-phase column. Each peak in the UV chromatogram (215 nm) was analyzed by MALDI-TOF mass spectrometry. Based on molecular mass data, peak 1, 2, and 3 contain MEPF1 isoforms. The peptide in each peak was lyophilized, re-dissolved in MS medium, and subsequently analyzed using bioassays. (B) MALDI-TOF spectrum of fraction 1 from (A). A single-charged peptide corresponding to expressed MEPF1 lacking the signal peptide from pbadg vector was observed at m/z = 9,386.9 ([M+H] + ) and a double-charged at m/z = 4,695.1 ([M+2H] 2+ ). (C) Confocal images of cotyledon epidermis from wild-type seedling grown 5 days in a solution with fraction 1 from (A). The epidermis is devoid of stomata, with arrested meristemoids, indicating that fraction 1 contains bioactive MEPF1 isoform. (D)

Lee_Jin Suk 4 MALDI-TOF spectrum of fraction 2 from (A). A single-charged peptide corresponding to expressed MEPF1 including the signal peptide from pbadg vector was observed at m/z = 11,457.7 ([M+H] + ) and a double-charged at m/z = 5,732.1 ([M+2H] 2+ ). (E) Confocal images of cotyledon epidermis from wild-type seedling grown 5 days in a solution with fraction 2 from (A). Again, the epidermis is devoid of stomata with arrested meristemoids, indicating that fraction 2 contains bioactive MEPF1 isoform. (F) MALDI-TOF spectrum of fraction 3 from (A). A single-charged peptide corresponding to expressed MEPF1 including the signal peptide from pbadg vector was observed at m/z = 11,461.2 ([M+H] + ) and a double-charged at m/z = 5,734.2 ([M+2H] 2+ ). (G) Confocal images of cotyledon epidermis from wild-type seedling grown 5 days in a solution with fraction 3 from (A). Stomatal differentiation (dots) is visible, indicating that fraction 3 contains only inactive, misfolded MEPF1. Supplemental Figure 7: Separation of properly-folded, bioactive MEPF2 by reverse-phase chromatography followed by mass-spectrometry and bioassays (A) HPLC chromatogram of purified, refolded MEPF2-His. Each peak in the UV chromatogram was collected and analyzed by MALDI-TOF mass spectrometry as described in fig. S3. The peaks 1 and 2 were subsequently analyzed for bioassays. (B) MALDI-TOF spectrum of fraction 1 from (A). A single-charged peptide corresponding to expressed MEPF2 lacking the signal peptide from pbadg vector was observed at m/z = 9,515.9 ([M+H] + ) and a double-charged at m/z = 4,759.3 ([M+2H] 2+ ). (C) Confocal images of cotyledon epidermis from wildtype seedling grown 5 days in a solution with fraction 1. Severe inhibition of asymmetric division is observed, indicating that this fraction contains bioactive MEPF2. (D) MALDI-TOF spectrum of fraction 2 from (A). Low molecular weight contaminant was observed at m/z = 1,721.2 ([M+H] + ). (E) Confocal images of cotyledon epidermis from wild-type seedling grown 5 days in a solution with fraction 2 from (A), with visible stomatal differentiation (dots).

Lee_Jin Suk 5 Supplemental Figure 8: Real-time binding curves detecting direct ligandreceptor interactions by QCM Shown are representative QCM real-time binding curves for the attachment of bioactive recombinant MEPF1 (A-C) and MEPF2 (D-F) to surface-immobilized ERECTA K-GFP (A,D: ERECTA, red), ERL1 K-GFP (B, E: ERL1, green), TMM- GFP (C, F: TMM, blue), and GFP (A-F: GFP, black). Arrows indicate the time point of injection of 300 nm recombinant peptide solutions (A-C: MEPF1, D-F: MEPF2). Reduction in frequency (Hz) of quartz crystals indicates total mass shifts due to ligand-receptor interaction. See Supplemental Methods for experimental details. Supplemental Figure 9: Heat denaturation severely compromises binding of MEPF1 and MEPF2 to receptors Shown are observed SPR kinetic sensograms fitted by least square regression with Langmuir adsorption model from representative SPR binding assay upon injection (t=0) of: (A) heat-denatured MEPF1 peptide solution (300 nm); and (B) heat-denatured MEPF2 peptide solution (300 nm) to biosensor chips immobilized with ERECTA K-GFP (ERECTA, red), ERL1 K-GFP (ERL1, green), and TMM- GFP (TMM, blue). The peptide solution was heat denatured at 70 C for 30 min. Experiments were repeated at least three times. Supplemental Figure 10: Functional epitope-tagged receptors expressed in Arabidopsis are predominantly detected in microsomal fractions (A-F) Immunoblots of subcellular fractionations of protein extracts from 2-weekold seedlings of Arabidopsis wild type (WT) and transgenic lines expressing each functional epitope-tagged receptor in its respective null mutant background: proerecta::erecta-yfp in erecta (A); proerl1::erl1-yfp in erl1 (B);

Lee_Jin Suk 6 proerecta::erecta-flag in erecta (C); proerl1::erl1-flag in erl1 (D); protmm::tmm-yfp in tmm (E); protmm::tmm-ha in tmm (F). The fractionation was performed as described in the Supplemental Methods. T, total fraction; C, cytosol fraction; M, microsomal (membrane) fraction. Arrowheads indicate the size of each epitope-tagged receptor. Non-specific bands are detected in immunoblots using anti-ha antibody (TMM-HA) and anti-gfp antibody (TMM- YFP and ERL1-YFP). 6% (for panels A-E) or 8% (for panel F) acrylamide gels were used for SDS-PAGE. Asterisks, non-specific bands. Molecular mass markers are indicated in kilodaltons (kd). Supplemental Figure 11: Dominant-negative receptors link ERECTA and ERL1 predominantly with EPF2- and EPF1-mediated signaling, respectively (A-F) Confocal microscopy of 12-day-old abaxial cotyledons from wild type (WT; A), erecta (B), ERECTA K in erecta (C), epf2 (D), erecta erl1 erl2 (E), erl1 (F), ERL1 K in erl1 (G), and epf1 (H). Both dominant-negative ERECTA and epf2 mutation confer excessive entry asymmetric divisions (C,D; brackets). This phenotype is less notable in erecta (B; brackets). erecta erl1 erl2 exhibits massive stomatal clusters (E). erl1 exhibits no discernible phenotype (F). Both dominantnegative ERL1 and epf1 mutation confer stomatal pairing (G, H; asterisks). Scale bar, 20 m. (I-K) Quantitative analysis of 12-day-old abaxial cotyledon epidermis. (I) Stomatal density (p=0.02, one-way ANOVA). (J) Stomatal pairing density (p<0.0001, one-way ANOVA). (K) Non-stomatal epidermal cell density (p<0.0001, one-way ANOVA). Asterisk and double asterisk represent groups significantly different (p<0.01) from others by Tukey's HSD test. WT, epf1, ERL1 K erl1: n=8; epf2 and ERECTA K erecta: n=5. Total numbers of stomata and non-stomatal epidermal cells counted: WT (301, 841); epf1 (390, 940); ERL1 K erl1 (378, 708); epf2 (211, 1657); ERECTA K erecta (230, 2147). Bars, means. Error bars, s.e.m.

Lee_Jin Suk 7 Supplemental Figure 12: Model of ligand-receptor actions during stomatal development (A) Cartoon of stomatal development. A subset of protodermal cells (P) adopt stomatal-lineage fate and become meristemoid mother cells (MMC), undergoing asymmetric entry division. Daughter cells of an MMC are a meristemoid (M) and a stomatal-lineage ground cell (SLGC). A meristemoid reiterates asymmetric amplifying divisions, but eventually differentiates into a guard mother cell (GMC), which terminally differentiates into paired guard cells (GCs) constituting a stoma. EPF2 (purple) is secreted from MMCs and early meristemoids (Hara et al. 2009; Hunt and Gray 2009). EPF1 (cyan) is secreted from late meristemoids and GMCs (Hara et al. 2007). (B) Hypothetical cell-cell interaction between EPF2-expressing MMC early M (purple) and its neighbor cell (white). EPF2-ERECTA (red) ligandreceptor pair inhibits entry division initiating stomatal cell lineage, while TMM (black) may prevent ERECTA's inhibition in the MMC early M (purple). In addition, in nonstomatal neighbor cells TMM may fine-tune EPF2-ERECTA signaling. (C) EPF1- ERL1 (orange) ligand-receptor pair inhibits stomatal differentiation and orients asymmetric spacing divisions. In stomatal precursor cells (late Ms and GMCs) TMM associates with ERL1 to prevent ERL1 from inhibiting stomatal differentiation. In contrast, TMM cooperatively acts in SLGC to inhibit secondary asymmetric division from occurring adjacent to the pre-existing stomatal precursor (i.e. enforcing onecell-spacing rule). ERECTA is expressed high in protodermal tissues, while ERL1 and TMM are expressed in stomatal-lineage cells, in MMCs, meristemoids, and SLGCs (Nadeau and Sack 2002; Shpak et al. 2005). In (B) and (C), the presence of ERECTA-ERL1 heterodimers may contribute to signaling redundancy. (D) Excessive EPF2, either by overexpression or recombinant MEPF2 peptide application, confers ERECTA-mediated inhibition of asymmetric entry division, resulting in an epidermis solely composed of pavement cells (confocal image, cell periphery highlighted in purple). (E) Excessive EPF1, either by overexpression or recombinant MEPF1 peptide application, confers ERL1-mediated inhibition of stomatal differentiation, resulting in an epidermis devoid of stomata. While asymmetric entry and amplifying divisions (sequence of cell divisions numbered) still occur, stomatal precursors

Lee_Jin Suk 8 trans-differentiate into SLGC-like cells (asterisk) instead of guard cells (confocal image, cell periphery highlighted in cyan). Stoichiometry and localization of each receptor species remain uncertain. References for Supplemental Figure Legends Hara K, Kajita R, Torii KU, Bergmann DC, Kakimoto T. 2007. The secretory peptide gene EPF1 enforces the stomatal one-cell-spacing rule. Genes Dev 21: 1720-1725. Hara K, Yokoo T, Kajita R, Onishi T, Yahata S, Peterson KM, Torii KU, Kakimoto T. 2009. Epidermal cell density is auto-regulated via a secretory peptide, EPIDERMAL PATTERNING FACTOR2 in Arabidopsis leaves. Plant Cell Physiol 50: 1019-1031. Hunt L, Gray JE. 2009. The signaling peptide EPF2 controls asymmetric cell divisions during stomatal development. Curr Biol 19: 864-869. Nadeau JA, Sack FD. 2002. Control of stomatal distribution on the Arabidopsis leaf surface. Science 296: 1697-1700. Shpak ED, McAbee JM, Pillitteri LJ, Torii KU. 2005. Stomatal patterning and differentiation by synergistic interactions of receptor kinases. Science 309: 290-293.