NopT specifically suppresses NFR1/NFR5-triggered cell death in N. benthamiana.

Agrobacterium strains harboring plasmid DNA encoding NopT or the empty vector (EV) were infiltrated into N. benthamiana leaves. At 12 hpi, a second infiltration was performed with an Agrobacterium strain containing a plasmid with NFR1/NFR5 genes. Dashed red lines indicate leaf discs where different proteins were expressed. (A, B) NopT (A), but not NopTC93S (B), a protease-inactive version of NopT, suppressed NFR1NFR5-induced cell death. (C) Expression of NopM-induced cell death in N. benthamiana. (D) Expression of NopT could not suppress the cell death response triggered by expression of Avr3a/R3a, NopM, BAX, INF1 or AtCERK1. Numbers in (D) represent the number of leaf discs showing cell death and total leaf discs tested. The pictures shown in this figure are representative of at least three independent biological replicates.

NopT interacts with NFR1 and NFR5.

Interactions between NopT and NFR1 or NFR5 were detected using BiFC (Split-YFP) (A, B), Split-LUC complementation (C) and co-IP (D) assays in N. benthamiana leaves. (A, B) For BIFC analysis, nYFP and cYFP tags were fused at the C-terminus of NopT, NFR1, NFR5, and the flagellin receptor FLS2 (negative control). YFP fluorescence signals represent protein-protein interactions. Scale bar=25 μm. (C) For the Split-LUC complementation assay, the nLUC and cLUC tags were fused at the C-terminus of NopT, NFR1, NFR5 and FLS2 (negative control). Luminescence signals represent protein-protein interactions. (D) For the co-IP assay, HA-tagged NFR1 or NFR5 and FLAG-tagged NopT were expressed in N. benthamiana cells followed by immunoprecipitation using an anti-FLAG antibody. Immunoblot analysis was performed using anti-HA and anti-FLAG antibodies (NopTC denotes the truncated version of NopT after autocleavage). The images and immunoblots shown in this figure are representative of three biological replicates.

NopT proteolyzes NFR5 at its JM domain.

Proteins with indicated tags were expressed in N. benthamiana (A), L. japonicus (B) or E. coli (C-G) and detected by immunoblotting. (A) NopT but not NopTC93S cleaves NFR5-GFP protein expressed in N. benthamiana cells by releasing NFR5CD-GFP (NopTC denotes autocleaved NopT). (B) NopT but not NopTC93S cleaves NFR5-Myc expressed in hairy roots of L. japonicus by releasing NFR5CD-Myc (marked by an asterisk). (C) Analysis of the CDs of NFR1 and NFR5 co-expressed with NopT or NopTC93S in E. coli. Cleavage of NFR5CD was observed for NopT but not NopTC93S, while NFR1CD was not proteolyzed by NopT. In the presence of NFR1CD, a slower migrating band was observed, possibly representing phosphorylated NopT (NopTm). (D) A repeat experiment confirmed that NopT is able to cleave NFR5CD (the asterisk indicates the HA-tagged NFR5 cleavage product containing the kinase domain and a C-terminal tail region). (E) Not cleaves His-SUMO-NFR5JM-GFP (His-SUMO and GFP linked by the JM domain of NFR5) in vitro. F.T. indicates proteins in flow through samples after purification with Ni-beads. (F) NopT expressed in E. coli was unable to cleave co-expressed NFR5CD-5m-HA, a modified version of NFR5CD-HA in which five amino acids of the JM were substituted by other residues (S283Y, G294Q, Y303S, A310I, T311Y). (G) NopT expressed in E. coli was unable to cleave co-expressed M. truncatula NFPCD-HA and the NFR5JM-NFPKC-HA fusion protein, while NFPJM-NFR5KC was proteolyzed (KC stands for the kinase domain and a C-terminal tail region).

Phosphorylation of NopT by NFR1 suppresses its proteolytic activity.

(A) In vivo phosphorylation assay with proteins expressed in L. japonicus roots (wild-type and nfr1-1 mutant plants) using Zn2+-Phos-tag SDS-PAGE. Phosphorylation of NopTC93S was induced by inoculation with rhizobia (Mesorhizobium loti MAFF303099) and was largely dependent on NFR1. (B) The relative protein amount of each lane, as shown in (A), was quantified with ImageJ software (three biological replicates). The value of the control band in each gel was set to 1 for comparison. Values are means ± SEM. (C) NFR1CD but not NFR1CD-K351E phosphorylates NopT in E. coli. The phosphorylated full-length form of NopT could be dephosphorylated by CIAP (gel band shift). Abbreviations: NFR1CD-P, autophosphorylated NFR1; NopTP, phosphorylated NopT; NopT, non-phosphorylated NopT; NopTC, autocleaved NopT. (D) The phoshorylation sites of NopT identified by LC-MS were either substituted to alanine (A) or aspartate (D). The indicated NopT variants were subsequently tested for autocleavage and NFR5CD proteolysis in E. coli. Wild-type NopT (WT) and protease-dead NopTC93S were included into the analysis.

NopT regulates rhizobial infection in L. japonicus.

(A) Analysis of rhizobial infection in L. japonicus roots inoculated with GUS-labelled S. fredii NGR234 (wild-type; WT) or a nopT knockout mutant (NGR234ΔnopT; abbreviated as ΔnopT in other panels). Scale bar=100 µm. (B) GUS staining pictures showing roots of L. japonicus plants expressing GUS expression under control of the NIN promoter (pNIN:GUS). Plants were inoculated with NGR234 or NGR234ΔnopT and analyzed at 7 dpi. Scale bar=1 mm. (C) Infection data (ITs, infection threads) for roots shown in (A) at 7 dpi (n=10, Student’s t-test; * indicates P < 0.01). (D) Quantification of GUS staining sites for roots shown in (B) (n=10, Student’s t-test, P < 0.01). (E) Nodule primordia formation in L. japonicus wild-type roots inoculated with NGR234 (WT), NGR234ΔnopT or NGR234 over-expressing nopT (pT7:NopT). Roots were analyzed at 14 dpi (n=16, Student’s t-test: P < 0.01). (F) Infection data for wild-type roots inoculated with NGR234 (WT), NGR234ΔnopT and NGR234ΔnopT expressing indicated NopT variants at 7 dpi (n=5, Student’s t-test: P < 0.01). (G) Analysis of rhizobial infection in hairy roots of L. japonicus (wild-type) expressing GFP (EV, empty vector control) or NopT. Plants were inoculated with DsRed-labeled M. loti MAFF303099 and analyzed at 5 dpi (n=8, Student’s t–test: P < 0.01). (H) Expression of NFR5 and NFR5-NFPKC in hairy roots of nfr5-3 mutant plants. Plants were inoculated with GFP-labelled NGR234 (WT) or NGR234ΔnopT and analyzed at 8 dpi. Roots expressing NFR5-NFPKC showed high numbers of infection foci for both strains whereas significant differences were observed for NFR5 expressing roots (n=10, Student’s t-test: P < 0.01).

S. fredii NopT proteins cleave NFR5 and working model for NopT of NGR234.

(A) NopT of S. fredii NGR234 and homologs from other rhizobial strains were co-expressed with NFR5CD in E. coli (NopT234, NopT of NGR234; NopT1110 and NopT2110, NopT proteins of B. diazoefficiens USDA110; NopT257, NopT of S. fredii USDA257; NopT103, NopT of S. fredii HH103). Immunoblot analysis indicated NFR5 cleavage by NopT proteins from S. fredii strains. (B) Expression of NopT257 in N. benthamiana could not inhibit the cell death triggered by co-expressed NFR1 and NFR5. (C) L. japonicus Gifu pNIN:GUS plants were inoculated with S. fredii NGR234, NGR234ΔnopT and NGR234ΔnopT expressing NopT257 (ΔnopT+NopT257). Roots were subjected to GUS staining at 7 dpi. Scale bar=2.5 mm. (D) Quantitative analysis of GUS-stained roots shown in panel C (n=19, Student’s t–test: P < 0.01). (E) A proposed model for NopT of NGR234 interacting with NFRs. NopT and NopTC (autocleaved NopT) proteolytically cleave NFR5 at the JM domain to release the intracellular domain of NFR5 (cleaved NFR5). NFR1 phosphorylates full-length NopT to block its proteinase activity. NopTC cannot be phosphorylated by NFR1.

Analyses of the functions of all 15 effectors from S. fredii NGR234 in preventing NFR1- and NFR5-induced cell death in N. benthamiana leaves.

15 genes from S. fredii NGR234 encoding effector proteins were amplified and cloned into binary vector for expression of Strep-tagged proteins in N. benthamiana. NFR1 and NFR5 from L. japonicus were C-terminally tagged with HA and Myc, respectively, and cloned into binary vectors for expression in N. benthamiana. All plasmids were electroporated into Agrobacterium tumefaciens EH105.

(A) Agrobacterium strain harboring each effector gene or control empty vector (EV) were infiltrated into N. benthamiana. 12 hour post infiltration (hpi), agrobacterium strains harboring NFR1 and NFR5 were mixed and hand-infiltrated in the same leaf discs as shown in (B). 48 hpi, cell death phenotypes were evaluated (C). At least three replicates in the same leaf were performed to test the function of each effector in suppressing cell death-triggered by NFR1 and LjNFR5. (D) Results of all 15 effector proteins in suppressing programmed cell death by overexpression of NFR1 and NFR5. (E) Abundance of NFR1, NFR5, and NopT proteins, as measured by immunoblot with specific antibodies. Leaf discs expressing NFR1, NFR5, and NopT or NFR1 and NFR5 and with EV were used for immunoblot analyses. Actin was used as loading control. (F) Another example of suppression of programmed cell death by NopT in the leaf discs coexpressing NFR1 and NFR5. Rep1, Rep2, and Rep3 represent three technical repeats in the same leave.

Split–LUC assay testing the interactions between different NopT mutants and NFRs.

(A)The interaction between NopT with mutated acylation site and NFR1/NFR5. NopTAAA (G50A, C51A, and C52A). (B) The protein expressing of indicating genes in the leave of (A). (C) NopC93S interact with NFR1 and NFR5. (D) The interaction between NopTΔN50 and NFR5. NopTΔN50 represents a truncated NopT with deletion of about 50 a.a. from its N–terminus.

NopT cleaves NFR5 not not NFR1.

Proteins with indicated tags were expressed in either N. benthamiana or E. coli and detected by immunoblotting. (A) NopT did not cleave NFR1-GFP expressed in N. benthamiana leaves. (B) NopT expressed in E. coli did not cleave co-expressed SUMO-NFR5KC-HA. KC: Kinase domain and C-terminal domain

NopT interacts with the JM domain of NFR5.

NopT interacts with NFR5JM in in vitro pull down assay. IB: immunoblotting, IP: immunoprecipitation. His-SUMO-NFR5JM-GFP: a recombinant protein containing His tag, SUMO tag, the JM domain of NFR5, and GFP.

Conserved domains and residues of NFR5 and related proteins.

(A) Domain and motif annotation of NFR5. LysM: Lysin-Motif; TM: transmembrane domain; JM: juxtamembrane domain; KD: kinase domain; CT: C-terminal domain. (B) Alignment of amino acid residues of AtLYK5, LjLYS11, NFR5 and MtNFP close to the RRKK motif in the JM domain of NFR5. Five highly conserved residues marked by yellow coloration (S283, G294, Y304, A310, T311 in NFR5) and the brown box indicates the start region of kinase domain. The green frame delineates the RRKK motif in NFR5 and the red-framed residues are similar to the NopT autocleavage region.

NopT cleaves NFR5 at the juxtamembrane domain.

(A) Cleavage assay using two mutant versions of NFR5CD in the presence of NopT or AvrPphB. NFR5CD 288-294A represents a mutant NFR5CD with residues 288 to 294 replaced with seven alanines; NFR5CD 288-294PphB represents a mutant NFR5CD with residues 288 to 294 replaced with AvrPphB recognition sites. (B) Autocleavage assay, using different NopT variants with single amino acid mutations. H205A and D220A indicate His-205 and Asp-220 replaced with alanine, respectively. (C-F) NopT cleavage assay using 19 variants of NFR5CD with three adjacent amino acids replaced with three alanines. The numbers indicate the location of the three residues replaced with alanines. (G) NopT cleavage assay using seven truncated variants of NFR5CD with 10 adjacent amino acids deleted. The numbers indicate the location of the 10 deleted residues.

Mass spectrometry analysis of cleavage site of NFR5 by NopT.

(A) The sequence of NFR5 juxtamembrane domain (286-320 a.a.). (B) Mass spectrometry analysis of peptides of NFR5CD after proteolysis by NopT. Blue lines indicated the peptides characterized by Mass spectrometry. Red frame delineated cleavage site of NFR5 identified using Mass spectrometry.

NopT cleaves the NFR5 homolog proteins at the juxtamembrane domain.

NopT but not NopTC93S could cleave the CDs of AtLYK5 and LjLYS11 expressed in E. coli. NopT could cleave both AtLYK5JM–NFR5KC, LjLYS11JM–NFR5KC and NFR1JM–NFR5KC, three recombinant proteins expressed in E. coli. AtLYK5JM–NFR5KC, LjLYS11JM–NFR5KC, and NFR1JM–NFR5KC recombinant NFR5 proteins with the the JM was replaced with the JMs from AtLYK5, LjLYS11 and LjNFR1, respectively. KC: Kinase domain and C-terminal domain as ilustrated in Fig. S5A.

NopT cleaves the JM of MtNFP.

NopT cleaves a recombinant protein where His–tagged SUMO and GFP is bridged with the JM domain of NFP. F.T. represents flow through sample after Ni–beads purification. IB: immunoblotting.

NopT cleaves recombinant proteins NFR5268-445-NFP458-595 and NFP270-457-NFR5456-595.

NopT but not NopTC93S could cleave recombinant proteins NFR5268-445-NFP458-595 and NFP270-457-NFR5456-595 detected by immunobltting (IB).

NFR1CD phosphorylates NFR5CD in vitro.

The cytoplasmic domains (CDs) of NFR1, NFR5, and kinase–dead NFR1CD-K351E were used for kinase assays in E. coli. Proteins were detected by immunoblotting using anti–HA and anti–Myc antibodies. Asterisk indicated the band retardation on the gel representing the phosphorylated NFR5CD proteins.

Rhizobial infection in the mutant versions of NFR5.

NFR55m failed to form rhizobial infection (n=7, Student’s t–test: P < 0.01). The transgenic roots expressing NFR5 or NFR55m in the nfr5-3 mutant under the control of native promoter were inoculated with GFP-labelled wild type NGR234 and NGR234ΔnopT respectively. The data was collected 9 days post inoculation. NFR55m represents a mutant version of NFR5 with S283Y, G294Q, Y304S, A310I, T311Y point mutation.

Phylogenetic tree based on the amino acid sequence f NopT homologs from different bacterial species.

The cleavage efficiency of NFR5 in different system.

The relative cleavage efficiency for experiment displayed in Fig. 3A (N. enthamiana), Fig. 3B (L. japonicus) and Fig. 3D (E. coli) quantified from three ological replicates using ImageJ software. Values are means ± SE (n=3, tudent t-test, letters represent significant differences, p<0.05).

Transient expression of NopT triggers cell death in rabidopsis thaliana and Nicotiana tabacum.

The relative cleavage efficiency for experiment displayed in Fig. 3A (N) Pseudomonas syringae pv tomato DC3000 harboring control plasmids (left) or xpressing NopT (right) were infiltrated into Arabidopsis leaves. Pictures were ken three days post inoculation. (B) Agrobacterium strains harboring NopT, opT with an N-terminal 50–amino acid deletion, NopTC93S, or NopTC93S with an -terminal 50–amino acid deletion were infiltrated into N. tabacum leaves. ictures were taken three days post inoculation.

Phosphopeptides identified by liquid romatography–mass spectrometry.

An in vitro kinase assay was performed using the D of NFR1 and NopT. NopT was separated by SDS-PAGE, and the digested gel slices presenting the phosphorylated NopT were used for analysis. The deduced hosphopeptides are listed.

Primers used in this study.