Lipo-chitooligosaccharidic Symbiotic Signals Are Recognized by LysM

Letters pubs.acs.org/acschemicalbiology

Lipo-chitooligosaccharidic Symbiotic Signals Are Recognized by LysM Receptor-Like Kinase LYR3 in the Legume Medicago truncatula Judith Fliegmann,†,‡ Sophie Canova,§,○ Christophe Lachaud,†,‡,○ Sandra Uhlenbroich,∥,⊥,○ Virginie Gasciolli,†,‡ Carole Pichereaux,¶ Michel Rossignol,¶ Charles Rosenberg,†,‡ Marie Cumener,†,‡ Delphine Pitorre,†,‡ Benoit Lefebvre,†,‡ Clare Gough,†,‡ Eric Samain,# Sébastien Fort,# Hugues Driguez,# Boris Vauzeilles,§,□ Jean-Marie Beau,§,□ Alessandra Nurisso,▽ Anne Imberty,# Julie Cullimore,†,‡ and Jean-Jacques Bono*,†,‡ †

INRA, Laboratoire des Interactions Plantes-Microorganismes (LIPM), UMR441, 31326 Castanet-Tolosan, France CNRS, Laboratoire des Interactions Plantes-Microorganismes (LIPM), UMR2594, 31326 Castanet-Tolosan, France § Université Paris-Sud and CNRS, Laboratoire de Synthèse de Biomolécules, Institut de Chimie Moléculaire et des Matériaux d’Orsay, UMR 8182, 91405 Orsay, France ∥ Université de Toulouse, UPS, UMR 5546, Laboratoire de Recherche en Sciences Végétales (LRSV), BP 42617, 31326 Castanet-Tolosan, France ⊥ CNRS, UMR 5546, BP 42617, 31326 Castanet-Tolosan, France ¶ UPS, FR3450 Plateforme de Protéomique, Toulouse, France # Centre de Recherches sur les Macromolécules Végétales (CERMAV, UPR-CNRS 5301), affiliated with the Université Joseph Fourier (UJF) and member of the Institut de Chimie Moléculaire de Grenoble (ICMG, FR-CNRS 2607), BP53, 38041 Grenoble Cedex 9, France □ Centre de Recherche de Gif, Institut de Chimie des Substances Naturelles du CNRS, Avenue de la Terrasse, 91198 Gif-sur-Yvette, France ▽ School of Pharmaceutical Sciences, UNIGE, Quai Ernest Ansermet 30, 1205 Geneva, Switzerland ‡

S Supporting Information *

ABSTRACT: While chitooligosaccharides (COs) derived from fungal chitin are potent elicitors of defense reactions, structurally related signals produced by certain bacteria and fungi, called lipo-chitooligosaccharides (LCOs), play important roles in the establishment of symbioses with plants. Understanding how plants distinguish between friend and foe through the perception of these signals is a major challenge. We report the synthesis of a range of COs and LCOs, including photoactivatable probes, to characterize a membrane protein from the legume Medicago truncatula. By coupling photoaffinity labeling experiments with proteomics and transcriptomics, we identified the likely LCO-binding protein as LYR3, a lysin motif receptor-like kinase (LysM-RLK). LYR3, expressed heterologously, exhibits high-affinity binding to LCOs but not COs. Homology modeling, based on the Arabidopsis CO-binding LysM-RLK AtCERK1, suggests that LYR3 could accommodate the LCO in a conserved binding site. The identification of LYR3 opens up ways for the molecular characterization of LCO/CO discrimination.

L

nonreducing N-acylated glucosamine unit, which may be substituted with various functional groups.3 The major Nod factor of Sinorhizobium meliloti, the rhizobial symbiont of Medicago spp., contains three GlcNAc units and is O6-sulfated on the reducing residue and the glucosamine unit is N-acylated

ipo-chitooligosaccharides (LCOs) regulate two important plant−microbe root endosymbioses: Nod factors, produced by Rhizobia bacteria, are essential signaling molecules for establishing the root nodule nitrogen-fixing symbiosis with legumes, whereas Myc-LCOs, produced by a Glomeromycota fungus, stimulate the arbuscular mycorrhizal (AM) symbiosis, which concerns over 80% of terrestrial plants.1,2 Nod factors and Myc-LCOs are generally composed of three or four β-1,4linked N-acetylglucosamine (GlcNAc) units and a terminal © 2013 American Chemical Society

Received: February 19, 2013 Accepted: June 28, 2013 Published: June 28, 2013 1900

dx.doi.org/10.1021/cb400369u | ACS Chem. Biol. 2013, 8, 1900−1906

ACS Chemical Biology

Letters

Figure 1. Synthesis of photoactivatable lipo-chitooligosaccharide probes. Compounds 1, 2, and 3 are lipo-chitooligosaccharides corresponding to the major Nod factors produced by Sinorhizobium meliloti (1, 2) and the major Myc-LCOs produced by Rhizophagus irregularis (3). Aryl-azido (6, 7) or benzophenone (8, 9) photoaffinity probes were prepared from chitooligomers 4 and 5. The corresponding photoactivatable lipid analogues (in the form of their activated succinimidyl esters 10 and 11) were obtained either from 4-amino-3-hydroxybenzoic acid and 1-iodoundec-4Z-ene (for 10) or from methyl 3-formylbenzoate, 3-methoxyphenylmagnesium bromide, and 1-bromoheptane (for 11).

generated from roots of a dmi3 mutant (which does not establish endosymbioses) compared to a wild-type line.12 As MtNFBS2 could represent a new player in the perception of LCOs and notably Myc-LCOs, we decided to further characterize this protein in dmi3 cell cultures and to identify it. To this aim, MtNFBS2 was thoroughly characterized in the membrane fraction from cell cultures of the dmi3 mutant. A saturation binding experiment performed with a fixed concentration (0.9 nM) of the NodSm factor, LCO-IV( 35S,C16:2Δ2,9), and increasing amounts of the same unlabeled compound indicated a high affinity for the binding site with a Kd of 15 nM (Supplementary Table 1). The affinities of different LCOs and COs were then determined in competition experiments with the labeled ligand and expressed as Ki values. The affinity of the nonsulfated homologue was similar to the sulfated ligand, whereas an LCO with only two GlcNAc residues (LCO-II(C16:1Δ9)) or CO-IV (tetra-Nacetyl-chitotetraose) showed low affinities (Ki > 5 μM) (Supplementary Table 1). Different Myc-LCOs (LCO-IV(S,C16:0), LCO-IV(S,C18:1Δ9), and its nonsulfated homologue) all bind MtNFBS2 with high affinities (Ki from 6 to 11 nM) (Supplementary Table 1, Supplementary Figure 1), indicating that MtNFBS2 specifically recognizes the symbiotic LCO structure of both Nod- and Myc-LCOs. To identify the binding protein in MtNFBS2 we designed photoreactive probes based on the procedure used to synthesize aromatic LCO analogues that tightly bound to MvNFBS2.13 Starting with nonsulfated and sulfated precursors 4 and 5, two benzamide glycolipids were synthesized, equipped with either an aryl azide motif (6, 7) or a benzophenone moiety (8, 9), both substituted by an alkoxy-alkyl chain (Figure 1).

with a C16:2 fatty acid and partially O6-acetylated (NodSmIV(Ac,S,C16:2Δ2,9), compounds 1 and 2, Figure 1), whereas the AM fungus Rhizophagus irregularis produces a mixture of simpler sulfated and nonsulfated LCOs (3, Figure 1).1 Structurally related GlcNAc-compounds, such as bacterial peptidoglycan (PGN) or chitooligosaccharides (COs) derived from fungal chitin, play an important role in parasite−host interactions by triggering plant immune responses.4 Hence, plants discriminate symbionts from pathogens via the perception of closely related signaling molecules. Although other proteins have been shown to bind LCOs5 and COs,6 recent work has highlighted the role of lysin motif (LysM)containing proteins, such as extracellular receptor-like proteins (LYMs) and transmembrane receptor-like kinases (LysMRLKs, encoded by the LYR and LYK genes)7 as receptors for GlcNAc-containing signals.8 A LysM-RLK from Arabidopsis thaliana (AtCERK1) and LYMs have been shown to bind chitin, COs, and PGN.8,9 Nod factor responses rely on the presence of two LysM-RLKs,3,8,10 which in the model legume Medicago truncatula are called NFP (a LYR) and LYK3.3 In this species Myc-LCO responses partially depend on NFP, but as neither NFP nor LYK3 are essential for mycorrhization, it has been hypothesized that other Myc factor receptors are present.1,3 Biochemical studies, using radiolabeled Nod factors, identified a high affinity (Kd = 4 nM) Nod factor-binding site (NFBS2) in the plasmamembrane of M. varia cell cultures.11 This site discriminates LCOs from COs and recognizes both sulfated and nonsulfated Nod factors. In M. truncatula, a similar site (MtNFBS2) was found to be independent of NFP and to be more abundant (at least 5-times) in a cell culture line 1901



Letters

The activated N-hydroxysuccinimide esters of these lipid chain analogues (10, 11) were synthesized from appropriate simple aryl building blocks and alkyl halides and selectively coupled to the oligosaccharidic backbones (Figure 1, Supplementary Figure 2; Supporting Information). In this way the photoreactive groups were positioned adjacent to the nonreducing unit of the oligosaccharide, thus very close to a potential receptor recognition domain. Competition experiments with the radiolabeled NodSm ligand showed that the sulfated aryl-azide analogue 7 and the benzophenone analogue 9 exhibited high affinities for MtNFBS2 (Ki of 9 and 13 nM, respectively), close to that of LCO-IV(S,C16:2Δ2,9). Therefore, the nonsulfated homologues 6 and 8 were radiolabeled with 35S-sulfate14 and used for photoaffinity labeling experiments. The membrane fraction from dmi3 and wild-type cell cultures were incubated with 3 nM radiolabeled azido derivative 7, irradiated at 254 nm to promote cross-linking, and the proteins were then analyzed by SDS-PAGE. Autoradiography revealed the presence of a photolabeled polypeptide with an apparent molecular mass close to 100 kDa for the dmi3 sample but not for wild-type, thus reflecting the expected difference in abundance of the binding protein (Figure 2a). Labeling of this polypeptide was completely abolished when the incubation was performed in the presence of an excess (2 μM) of either Nod factor, 1, or sulfated or nonsulfated Myc-LCOs, 3, but not with an excess of CO-IV. A competition experiment revealed that the photolabeled polypeptide has an affinity of 22 nM for the Nod factor, 1 (Supplementary Figure 3a,b), close to that of MtNFBS2. The use of the 35S benzophenone derivative 9 resulted in the same specific labeling of a 100 kDa polypeptide (Supplementary Figure 3c). In order to identify the binding protein, we performed in parallel a proteomic and a transcriptomic analysis, based on the premise that the protein and transcript of the binding protein would be more abundant in dmi3 than in the wild-type cell cultures. The quantitative analysis of the polypeptides in the 100 kDa zone, resolved by SDS-PAGE, identified 1973 and 1889 proteins in the wild-type and the dmi3 samples, respectively. Of the 1431 proteins common to both cell lines, 186 were more abundant in dmi3 as estimated by peptide intensity quantification and spectral counting methods (Supplementary Table 2). Up-regulated candidate proteins with carbohydrate-binding domains included the LysM-RLKs LYR3 (8.6-fold up) and LYK9 (2.3-fold up). Comparative transcriptomic analysis of the two cell lines revealed a high number of differentially expressed genes, including 324 genes with at least 10-fold higher expression in dmi3 (p < 0.001) (Supplementary Table 3). Of these genes, LYR3 was 11.9-fold up-regulated, whereas LYK9 and LYR4 were also up-regulated but to a lesser extent (2.5- and 2.4-fold, Supplementary Table 4). NFP, LYK3, and two LYM genes were either down or not regulated in dmi3, suggesting that they do not encode the binding protein. LYR3 and LYK9 identified by both transcriptomics and proteomics, as well as LYR4, a close homologue of LYR3,7 were selected for LCO-binding tests. These proteins are predicted to contain three LysMs in their ectodomain (ECD), followed by a single trans-membrane spanning domain (TM) and either an inactive (LYR3, LYR4) or an active (LYK9) kinase domain. They were expressed in Nicotiana benthamiana leaves as fusion proteins with the yellow fluorescent protein (YFP). Since the expression of LYK9 induced a cell death reaction in N.

Figure 2. Identification of MtNFBS2. (a) Photoaffinity labeling of a 100 kDa polypeptide (arrowhead). Membrane proteins extracted from M. truncatula wild-type or dmi3 cell cultures were treated with radiolabeled aryl-azido photoaffinity probe 6, with or without competing LCOs or COs. (b) LCO-binding to M. truncatula LysMRLKs. YFP fusion proteins were expressed in N. benthamiana leaves and detected by Western blotting using anti-GFP antibodies (upper panel). Binding experiments were performed using 0.9 nM LCOIV(35S,C16:2Δ2,9) (lower panel). 1, LYR3/YFP; 2, LYR3-NFP/YFP; 3, LYR4-NFP/YFP; 4, LYK9-NFP/YFP; c, non-transformed control leaves.

benthamiana leaves (Supplementary Figure 4a), as observed for AtCERK1,15 chimeric constructs were designed that contained the TM and kinase domain of NFP (which does not induce cell death), fused to the different ECDs. Western blotting and fluorescent microscopy showed that all of the proteins were well expressed and located at the plasma membrane (Figure 2b, Supplementary Figure 4b). Equilibrium binding assays showed specific binding of the NodSm factor 1 only to the membrane extracts containing LYR3/YFP or the LYR3-NFP/YFP chimera (Figure 2b), indicating that binding is specific to the ECD of LYR3. Physical interaction between LYR3 and LCOs was confirmed by photoaffinity labeling (Supplementary Figure 5). Scatchard analysis of a saturation experiment (Figure 3a) showed the presence of a single class of binding sites for LYR3/ YFP, exhibiting a high affinity (Kd = 25 nM) for the NodSm factor 1, similar to that of MtNFBS2 (Kd = 15 nM, Supplementary Table 1). The protein showed similar high 1902



Letters

length of the acyl chain from C16:2 to C18:2 led to an approximate 3-fold increase in affinity (Supplementary Table 1). However, LYR3 does not discriminate the structure of the acyl chain as similar affinities were found for LCOs with C16 chains containing 0, 1, or 2 unsaturations (Figure 3b, Supplementary Table 1). In summary, the similarities in LCO/CO binding selectivity and affinities of MtNFBS2, the photolabeled 100 kDa polypeptide, and LYR3 suggest that LYR3 is the binding protein of MtNFBS2. However, we cannot exclude that another LCO-binding protein with characteristics close to those of LYR3 also exists in the dmi3 culture. The crystal structure of the ECD of the Arabidopsis LysMRLK AtCERK1 in complex with chitopentaose (CO-V) has recently been solved.18 Alignment of the ECDs of LYR3 and LYR4 with AtCERK1 shows conservation of the three LysM domains, and although sequence identity is weak (∼20%, Figure 4a), conservation of interdomain cysteine residues and of hydrophobic amino acids, involved respectively in disulfide bridges and secondary structures, is sufficient for homology modeling. The model of LYR3 (Figure 4b, Supplementary Figure 6) presents three disulfide bridges and five putative Nglycosylation sites accessible on the surface of the protein. Each LysM domain is characterized by two canonical β-strands and α-helices. The LysM2 domain of LYR3 is particularly structurally close to that of AtCERK1 with the presence of an extra helix, α3, limited by two proline residues (Pro159 and Pro164). This helix, together with the loop between β1 and α1 (Asn131 to Phe136) forms a groove similar to the CO-binding site present in LysM2 of AtCERK1 (Figure 4c, Supplementary Figure 7). The robustness of ligand docking was first validated by redocking CO-IV in the LysM2 domain of AtCERK1. As the prediction was excellent (Supplementary Figure 8a), CO-IV and sulfated CO-IV were docked in the LysM2 of LYR3 (Figure 4d, Supplementary Figure 8b,c). The best pose, according to the GoldScore, fits in the predicted groove. The entire oligosaccharide establishes contacts with the protein involving mainly backbone atoms. Phe129, Thr132, Asp133, and Tyr163 form a pocket in which the N-acetyl group of the second GlcNAc (numbered from the nonreducing end) is deeply buried and fixed via hydrogen bonds to Tyr163 (Figure 4d). This residue, as well as Asp133, also interacts with the third GlcNAc. Further contacts are established at both ends via Gly160, Thr134, and Gly165. In the model, the sulfate group does not interact with the protein surface, which is in accordance with the binding characteristics of LYR3. Docking of Nod factor 1 to the CO-binding groove suggests that the proximal part of the acyl chain makes contact with the LysM2 domain in the region delimited by Pro159, Val162, and Pro164 (Figure 4e). This is consistent with the photoaffinity labeling of LYR3 by both aryl-azido and benzophenone groups localized in this part of the LCO. However, the distal part of the acyl chain can adopt a variety of orientations, either pointing toward the solvent or establishing contacts with the protein. As the length of the acyl chain is important for high-affinity binding (Supplementary Table 1),11 this may suggest that it either binds to unknown residues in the LysM2 domain or extends outward to bind to another LysM domain in a homodimer. Such a homodimerization mechanism has been proposed for the CO-VIII/AtCERK1 interaction.18 Alternatively the LCO could bind to another binding site in LYR3, which remains to be discovered.

Figure 3. Affinity and selectivity of LYR3 for LCOs vs COs. Membrane extracts of LYR3/YFP expressed in N. benthamiana were incubated with 0.9 nM LCO-IV(35S,C16:2Δ2,9). (a) Scatchard plot of a saturation experiment using LCO-IV(S,C16:2Δ2,9). Competitive inhibition by (b) Myc LCOs, (c) LCOs of different backbone length, and (d) COs; see insets for legends.

affinities for sulfated and nonsulfated Myc-LCOs (Figure 3b). The affinity for LCOs increased as the number of GlcNAc units increased from LCO-II to LCO-V (Figure 3c). COs were poorly recognized, since the affinity for the best CO (CO-VI) was at least 200-fold lower than for the best LCO (Figure 3d). PGN, which is perceived by some LysM proteins,16,17 did not compete for LCO binding (data not shown). Varying the 1903



Letters

Figure 4. Modeling of the ectodomain of LYR3. (a) Sequence alignment with shading according to conservation of amino acids (identical or conserved residues are shaded in black if they are present in all three or in gray if they are present in two sequences). LysMs are denoted by doubleheaded arrows. Structural elements of AtCERK1-ECD (PDB code 4EBZ) are indicated as β (β-strands) and α (α-helices). Cysteine residues connected by disulfide bridges are marked by identically colored dots. Predicted N-glycosylation sites of LYR3 are indicated by blue triangles. Red squares mark the regions of AtCERK1 that compose the CO-binding groove. (b) Model of LYR3-ECD with disulfide bridges linking the three LysM domains. (c) Superimposition of the model of LysM2 of LYR3 (green) with the LysM2 of AtCERK1-ECD (beige) in complex with CO-IV. (d) Best docking pose for CO-IV on LysM2 of LYR3 with hydrogen bonds as blue dotted lines; stick models of the amino acid residues that bind the ligand directly are sketched on top of the space-filling model of the binding groove. (e) Five docking poses for Nod factor LCO-IV(S,C16:2Δ2,9), carbohydrate moiety in yellow, lipid moiety in red, interacting with LysM2 of LYR3. Accessible surface is colored from brown for lipophilic to blue for hydrophilic potential. Relative to panel d, the model is turned at 90° along the vertical axis.

Sequence analysis and homology modeling performed on LysM2 of LYR4 showed that the putative CO-binding groove lacks the two proline residues at the extremities of the α3 helix (Figure 4a, Supplementary Figure 6) and is possibly blocked by the bulky Arg156 side chain (Supplementary Figure 9c). Attempts to dock CO-IV in the LysM2 did not result in satisfactory binding modes. In conclusion, the LysM-RLK of M. truncatula, LYR3, is a high-affinity LCO-binding protein, which discriminates LCOs from COs. High-affinity binding occurs to the LysM-containing ectodomain of this protein and requires an oligosaccharide backbone of at least 3 units and an acyl chain with at least 16 carbon atoms (Figure 3), similar to the basic structure of LCOs produced by symbiotic microorganisms. LYR3 recognizes fungal Myc-LCOs with affinities similar to those of rhizobial

Nod factors, suggesting that it could play a role in either or both of the legume-rhizobia and AM root endosymbioses. However, the ligand specificity of LYR3 is more in accordance with the recognition of Myc-LCOs than the specific Nod factors produced by S. meliloti.3 Further studies using molecular genetics will be necessary to elucidate the biological function of LYR3. However, since LYR3 is predicted to contain an inactive kinase-like domain, it might assemble with an active kinase in a multicomponent receptor complex, as suggested for the Arabidopsis chitin-binding receptors AtLYK4 and AtCERK1;19 interestingly AtLYK4 (At2g23770) is the closest Arabidopsis homologue to LYR3.7 LYR4, the close homologue of LYR3 in M. truncatula, does not bind LCOs (Figure 2), which together with the chitinbinding ability of AtLYK4 suggests that minor changes in the 1904



■

LysM domains may determine LCO binding. This view is supported by homology modeling using the structure of the ectodomain of Arabidopsis AtCERK1, which suggests conservation of the binding site for the CO-moiety in LysM2 of LYR3, but not LYR4. The question of how LYR3 is able to accommodate the lipid moiety remains to be answered. The biochemical identification of LYR3 opens up new avenues for uncovering the mechanisms of LCO/CO perception at the molecular level.

■

Letters

ASSOCIATED CONTENT

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Address ○

(S.C.) EDELRIS, 115, Avenue Lacassagne, 69003 Lyon, France. (C.L.) MRC Protein Phosphorylation Unit, College of Life Sciences, Sir James Black Centre, University of Dundee, Dundee DD1 5EH, Scotland. (S.U.) F-star Biotechnology Ltd., Moneta B280, Babraham Research Campus, Cambridge CB22 3AT, U.K.

METHODS

Synthesis of LCOs and Photoaffinity Probes. LCOs, harboring different oligosaccharide backbones and lipid moieties were synthesized by combining biotechnological and chemical approaches.1,11,13,20 An aryl-azido lipid analogue was prepared from 4-amino-3hydroxybenzoic acid and 1-iodoundec-4Z-ene, while a benzophenone-containing lipid analogue was prepared from methyl 3formylbenzoate, 3-methoxyphenylmagnesium bromide, and 1-bromoheptane (see also Supporting Information). Radiolabeling with 35S was done by enzymatic sulfation of nonsulfated homologues using the sulfotransferase NodH and radioactive PAPS as sulfate donor. 14 Fractionation of Plant Material and Binding Assays. Membrane fractions were prepared from M. truncatula cell suspension cultures or N. benthamiana leaves, as described11 (see Supporting Information). For equilibrium binding assays, 10−50 μg of membrane protein was incubated in the presence of 0.4−2 nM 35S-NodSm factor in a total volume of 200 μL of binding buffer. Competition assays were performed as described.11 For photoaffinity labeling, 300 μg of membrane protein was incubated with 3 nM radioactive photoactivatable ligand (7 or 9) in binding buffer. Samples were irradiated after 1 h incubation either for 5 min at 254 nm (azido derivative) or for 10 min at 365 nm (benzophenone derivative). The labeled membrane fraction was recovered and further analyzed by SDS-PAGE and autoradiography. Proteomic Analysis. Gel pieces of the 100 kDa polypeptide zone, resolved by SDS-PAGE, were digested with trypsin at 37 °C overnight. The peptide mixtures were analyzed by nano-LC−MS/MS using an Ultimate3000 system coupled to an LTQ-Orbitrap spectrometer. Data were obtained from the Medicago 3.5.1 protein database using the Mascot Daemon software (2.3.2). Mascot results were parsed with the in-house developed software Mascot File Parsing and Quantification (MFPaQ, 4.021), and protein hits were automatically validated. Output files from Mascot searches were uploaded with MFPaQ for spectral counting and peptide intensity measurement (see Supporting Information). Transcriptomic Analysis. RNA was prepared from the dmi3 and the wild-type cell cultures and hybridized to 4x44 K Medicago gene expression microarrays (Agilent) as described in Supporting Information (four biological replicates, each). Data were analyzed using GenespringGX9, and 863 probes, which showed a 10-fold difference in expression (p < 0.001), were retrieved. Binary Vector Construction. Gateway and Golden Gate technology was used to generate constructs harboring protein coding sequences, fused 3′ with the coding sequence for YFP (see Supporting Information). Binary plasmids were transformed by electroporation into Agrobacterium tumefaciens strain LBA4404 carrying the ternary plasmid pBBRvirGN54D.22 Plant Transformation. Overnight cultures of A. tumefaciens strains harboring the binary plasmids were used for syringe-infiltration of N. benthamiana leaves (see Supporting Information). After 3 days, leaves were observed for expression of YFP-fusion proteins, harvested, and stored at −80 °C before extraction. Molecular Modeling. The ectodomain of AtCERK1 (AtCERK1ECD) co-crystallized with CO-V (PDB code 4EBY) was used as template for homology modeling of LYR3 and LYR4 using MOE (2010.10). Surface calculations were performed with Sybylx (1.3). Docking of CO and Nod factor LCO-IV(S,C16:2Δ2,9) was performed with GOLD v.523 (see Supporting Information).

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was funded in parts by the French Agence Nationale de la Recherche (contracts ANR-08-BLAN-0208-01 “Sympasignal”, ANR-09-BLAN-0241 “MycSignalling”, NT05-4_42720 “NodBindsLysM”, ANR-12-BSV7-0001-01 “SYMNALING”, ANR-2010-JCJC-1602-01 LCOinNONLEGUMES), the Marie Curie Actions of the European Community (contract MRTNCT-2006-035546 “NODPERCEPTION”), Région Midi-Pyrénées and the CNRS (PhD grant to S.U.). This work was supported by funds from the “Laboratoire d’Excellence (LABEX)” entitled TULIP [ANR-10-LABX-41]. We thank C. Dunand (LRSV, UMR CNRS-UPS 5546, Toulouse, France) and V. Le Berre and L. Trouilh (Biochips Platform, Genotoul, Toulouse, France) for microarray hybridization; F. Debellé, P. Gamas and J. Gouzy for access to genomic data before publication; and L. Deslandes for gateway plasmids. J.-J.B. acknowledges F. Gressent and B. Hogg for their primary work on NFBS2 and J. Dénarié and R. Ranjeva for their long-lasting support.

■

REFERENCES

(1) Maillet, F., Poinsot, V., Andre, O., Puech-Pages, V., Haouy, A., Gueunier, M., Cromer, L., Giraudet, D., Formey, D., Niebel, A., Martinez, E. A., Driguez, H., Becard, G., and Dénarié, J. (2011) Fungal lipochitooligosaccharide symbiotic signals in arbuscular mycorrhiza. Nature 469, 58−63. (2) Oldroyd, G. E. D., Murray, J. D., Poole, P. S., and Downie, J. A. (2011) The rules of engagement in the legume-Rhizobial symbiosis. Annu. Rev. Genet. 45, 119−144. (3) Gough, C., and Cullimore, J. (2011) Lipo-chitooligosaccharide signalling in endosymbiotic plant-microbe interactions. Mol. PlantMicrob. Interact. 24, 867−878. (4) Boller, T., and Felix, G. (2009) A renaissance of elicitors: Perception of microbe-associated molecular patterns and danger signals by pattern-recognition receptors. Annu. Rev. Plant Biol. 60, 379−406. (5) Etzler, M. E., Kalsi, G., Ewing, N. N., Roberts, N. J., Day, R. B., and Murphy, J. B. (1999) A Nod factor-binding lectin with apyrase activity from legume roots. Proc. Natl. Acad. Sci. U.S.A. 96, 5856−5861. (6) Asensio, J. L., Cañada, F. J., Siebert, H.-C., Laynez, J., Poveda, A., Nieto, P. M., Soedjanaamadja, U. M., Gabius, H.-J., and JiménezBarbero, J. (2000) Structural basis for chitin recognition by defense proteins: GlcNAc residues are bound in a multivalent fashion by extended binding sites in hevein domains. Chem. Biol. 7, 529−543. (7) Arrighi, J.-F., Barre, A., Ben Amor, B., Bersoult, A., Soriano, L. C., Mirabella, R., de Carvalho-Niebel, F., Journet, E.-P., Ghérardi, M.,

1905



Letters

compatible plasmid dramatically increases Agrobacterium-mediated plant transformation. Plant Mol. Biol. 43, 495−502. (23) Verdonk, M. L., Cole, J. C., Hartshorn, M. J., Murray, C. W., and Taylor, R. D. (2003) Improved protein-ligand docking using GOLD. Proteins: Struct. Funct. Bioinf. 52, 609−623.

Huguet, T., Geurts, R., Dénarié, J., Rougé, P., and Gough, C. (2006) The Medicago truncatula lysin motif-receptor-like kinase gene family includes NFP and new nodule-expressed genes. Plant Physiol. 142, 265−279. (8) Gust, A. A., Willmann, R., Desaki, Y., Grabherr, H. M., and Nürnberger, T. (2012) Plant LysM proteins: modules mediating symbiosis and immunity. Trends Plant. Sci. 17, 495−502. (9) Shinya, T., Motoyama, N., Ikeda, A., Wada, M., Kamiya, K., Hayafune, M., Kaku, H., and Shibuya, N. (2012) Functional characterization of CEBiP and CERK1 homologs in Arabidopsis and rice reveals the presence of different chitin receptor systems in plants. Plant Cell Physiol. 53, 1696−1706. (10) Broghammer, A., Krusell, L., Blaise, M., Sauer, J., Sullivan, J. T., Maolanon, N., Vinther, M., Lorentzen, A., Madsen, E. B., Jensen, K. J., Roepstorff, P., Thirup, S., Ronson, C. W., Thygesen, M. B., and Stougaard, J. (2012) Legume receptors perceive the rhizobial lipochitin oligosaccharide signal molecules by direct binding. Proc. Natl. Acad. Sci. U.S.A. 109, 13859−13864. (11) Gressent, F., Drouillard, S., Mantegazza, N., Samain, E., Geremia, R. A., Canut, H., Niebel, A., Driguez, H., Ranjeva, R., Cullimore, J., and Bono, J.-J. (1999) Ligand specificity of a high-affinity binding site for lipo-chitooligosaccharidic Nod factors in Medicago cell suspension cultures. Proc. Natl. Acad. Sci. U.S.A. 96, 4704−4709. (12) Hogg, B. V., Cullimore, J. V., Ranjeva, R., and Bono, J.-J. (2006) The DMI1 and DMI2 early symbiotic genes of Medicago truncatula are required for a high-affinity nodulation factor-binding site associated to a particulate fraction of roots. Plant Physiol. 140, 365−373. (13) Grenouillat, N., Vauzeilles, B., Bono, J.-J., Samain, E., and Beau, J.-M. (2004) Simple synthesis of nodulation-factor analogues exhibiting high affinity towards a specific binding protein. Angew. Chem., Int. Ed. 43, 4644−4646. (14) Gressent, F., Cullimore, J. V., Ranjeva, R., and Bono, J.-J. (2004) Radiolabeling of lipo-chitooligosaccharides using the NodH sulfotransferase: a two-step enzymatic procedure. BMC Biochem. 5, 4. (15) Pietraszewska-Bogiel, A., Lefebvre, B., Koini, M. A., KlausHeisen, D., Takken, F. L. W., Cullimore, J. V., and Gadella, T. W. J. (2013) Interaction of Medicago truncatula Lysin motif receptor-like kinases, NFP and LYK3, produced in Nicotiana benthamiana leaf induces defence-like responses. PLoS One 8, e65055. (16) Willmann, R., Lajunen, H. M., Erbs, G., Newman, M.-A., Kolb, D., Tsuda, K., Katagiri, F., Fliegmann, J., Bono, J.-J., Cullimore, J. V., Jehle, A. K., Götz, F., Kulik, A., Molinaro, A., Lipka, V., Gust, A. A., and Nürnberger, T. (2011) Arabidopsis lysin-motif proteins LYM1 LYM3 CERK1 mediate bacterial peptidoglycan sensing and immunity to bacterial infection. Proc. Natl. Acad. Sci. U.S.A. 108, 19824−19829. (17) Liu, B., Li, J.-F., Ao, Y., Qu, J., Li, Z., Su, J., Zhang, Y., Liu, J., Feng, D., Qi, K., He, Y., Wang, J., and Wang, H.-B. (2012) Lysin motifcontaining proteins LYP4 and LYP6 play dual roles in peptidoglycan and chitin perception in rice innate immunity. Plant Cell 24, 3406− 3419. (18) Liu, T., Liu, Z., Song, C., Hu, Y., Han, Z., She, J., Fan, F., Wang, J., Jin, C., Chang, J., Zhou, J.-M., and Chai, J. (2012) Chitin-induced dimerization activates a plant immune receptor. Science 336, 1160− 1164. (19) Wan, J., Tanaka, K., Zhang, X.-C., Son, G. H., Brechenmacher, L., Nguyen, T. H. N., and Stacey, G. (2012) LYK4, a Lysin motif receptor-like kinase, is important for chitin signaling and plant innate immunity in Arabidopsis. Plant Physiol. 160, 396−406. (20) Rasmussen, M. O., Hogg, B., Bono, J.-J., Samain, E., and Driguez, H. (2004) New access to lipo-chitooligosaccharide nodulation factors. Org. Biomol. Chem. 2, 1908−1910. (21) Bouyssié, D., de Peredo, A. G., Mouton, E., Albigot, R., Roussel, L., Ortega, N., Cayrol, C., Burlet-Schiltz, O., Girard, J.-P., and Monsarrat, B. (2007) Mascot File Parsing and Quantification (MFPaQ), a new software to parse, validate, and quantify proteomics data generated by ICAT and SILAC mass spectrometric analyses. Mol. Cell. Proteomics 6, 1621−1637. (22) van der Fits, L., Deakin, E. A., Hoge, J. H. C., and Memelink, J. (2000) The ternary transformation system: constitutive virG on a 1906


Lipo-chitooligosaccharidic Symbiotic Signals Are Recognized by LysM

Recommend Documents