A Xanthomonas uridine 5[prime]-monophosphate transferase inhibits plant immune kinases
Feng Feng,Chaozu He,Fan Yang,Xiaogang Wu,Jie Zhang,She Chen,Jian-Min Zhou,Wei Rong
DOI: https://doi.org/10.1038/nature10962
IF: 64.8
2012-01-01
Nature
Abstract:Xanthomonas campestris pathovar campestris (Xcc) is a causal agent of black rot diseases on numerous crucifer plants such as Brassica and Arabidopsis. AvrAC exists in all three sequenced strains of Xcc and a newly sequenced Xanthomonas campestris pathovar raphani strain, but not other Xanthomonas species. (http://www.xanthomonas.org/t3e.html). In Xcc strain 8004 (Xcc8004), AvrAC is recognized by Arabidopsis plants to trigger a weak effector-triggered immunity (ETI) in vascular, but not mesophyll tissues9. To determine if AvrAC has a role in virulence, we generated an avrAC knockout strain (Xcc8004ΔavrAC) and performed competitive index assays on cabbage plants (Methods). The competitive index values (Xcc8004ΔavrAC to Xcc8004 ratios) were 0.5–0.3 in 3/5 cultivars tested in multiple experiments (Fig. 1a), demonstrating a role of AvrAC in virulence on these plants. Well known PAMPs, including the bacterial flagellar peptide flg22, bacterial elongation factor (EF-Tu) peptide elf18, and fungal cell wall component chitin10, are recognized by pattern recognition receptors (PRRs) FLS2, EFR and CERK1, respectively4, to induce the expression of a PAMP-responsive reporter gene FRK1::LUC. The presence of the AvrAC transgene largely inhibited FRK1::LUC induction in protoplasts (Fig. 1b and Supplementary Fig. 1), indicating that AvrAC is a potent inhibitor of the PAMP-triggered immunity (PTI) signalling pathway. Two AvrAC transgenic lines were generated to test if AvrAC impedes mesophyll resistance to Xcc8004, a nonpathogenic mutant bacteria Xcc8004ΔhrcV and Pseudomonas syringae pathovar tomato (Pst) ΔhrcC mutant bacteria. In all three cases, the transgenic lines supported 0.7–1.1 log more bacterial growth than did the wild-type plants (Fig. 1c, d). Thus AvrAC can markedly compromise PTI resistance to these bacteria. The AvrAC transgenic lines were also inoculated with Pst strains carrying effector genes avrB, avrRpt2 or avrPphB, which elicit ETI resistance specified by the cytoplasmic immune receptors RPM1, RPS2 and RPS5, respectively3. The transgenic lines supported 0.9–1.0 log more Pst(avrB) bacterial growth than did wild-type plants (Fig. 1e). In contrast, the transgenic lines supported less than 0.3 log increase in bacterial growth when Pst(avrRpt2), Pst(avrPphB) and Pst bacteria lacking an avr gene were used (Fig. 1e and Supplementary Fig. 2). Thus, although the AvrAC transgene compromised RPM1-dependent ETI, it did not appear to affect RPS2- and RPS5-dependent ETI. To determine further if bacterially delivered AvrAC similarly interferes with RPM1 activation, we infiltrated Arabidopsis plants with Xcc8004 before Pst(avrB) inoculation. Whereas the control plants inoculated with Pst(avrB) alone developed strong a hypersensitive response, the plants pre-treated with Xcc8004 developed no hypersensitive response (Fig. 1f), indicating a strong interference of RPM1 activation by Xcc8004. The Xcc8004ΔavrAC mutant strain was less capable of preventing hypersensitive response development (Fig. 1f and Supplementary Fig. 3), indicating that AvrAC interferes with RPM1-mediated ETI when delivered from Xcc. To identify host targets and the mechanism by which AvrAC inhibits PTI, we systematically investigated various molecular events of PTI signalling pathways11. flg22 perception triggers H2O2 production and MAP kinase (MAPK) activation in wild-type plants within several minutes; both events were completely abolished or diminished in AvrAC transgenic plants (Fig. 2a, b). However, AvrAC was unable to inhibit MAPK activation triggered by the constitutively active MKK5DD (ref. 10; Supplementary Fig. 4). These results suggested that AvrAC acts upstream of MAPK cascades and H2O2 production to inhibit PTI signalling. PAMP detection rapidly induces phosphorylation of BIK1 (refs 6, 7), which acts immediately downstream of FLS2, EFR and CERK1 (ref. 7). The expression of AvrAC in protoplasts prevented the flg22-induced BIK1 phosphorylation, as indicated by a slow migrating form on SDS–polyacrylamide gel electrophoresis (SDS–PAGE; Fig. 2c). Instead, BIK1 uniformly migrated between the unphosphorylated and phosphorylated forms independent of flg22 treatment. Upstream of BIK1 phosphorylation is the flg22-induced association between FLS2 and a receptor-like kinase called BAK1, forming an active receptor complex12, 13. This interaction occurred normally in protoplasts irrespective of the presence or absence of AvrAC (Fig. 2d). The results described above suggested that AvrAC directly target BIK1, FLS2 or BAK1. Co-immunoprecipitation assays showed that AvrAC can interact with BIK1, but not FLS2 or BAK1 (Fig. 2e). Glutathione S-transferase (GST) pull-down assays indicated that this interaction is direct (Fig. 2f). RIPK, an RLCK related to BIK1, can phosphorylate the RPM1-interacting protein RIN4 at T166 (ref. 8), and the phosphorylated RIN4 is responsible for triggering RPM1-specified ETI8, 14. AvrAC also interacted with RIPK in vitro and in protoplasts (Figs 2f and 4b). We conclude that BIK1 and RIPK are specific targets of AvrAC, a conclusion consistent with its inhibition of PTI and the RPM1-specific ETI. The AvrAC-induced electrophoresis mobility shift of BIK1 was resistant to phosphatase treatment (Supplementary Fig. 5a). Co-expression of BIK1 and AvrAC in Escherichia coli detected the same phosphatise-resistant, slow migrating form of BIK1 (Supplementary Fig. 5b, c), suggesting an AvrAC-induced BIK1 modification that is not caused by protein phosphorylation. AvrAC contains a filementation-induced by c-AMP domain (Fic) in the carboxy terminus. More than 4,300 Fic-domain-containing proteins have been found in bacteria and animals, mostly with unknown biochemical function15, 16. Two Fic-domain-containing, animal bacterial pathogen effectors Vibrio VopS and Histophilus IbpA catalyse adenylylation (AMPylation) on threonine or tyrosine residues of Rho family GTPases to disrupt cytoskeleton in the host cell17, 18. Another bacterial effector protein, Legionella pneumophila DrrA (also known as SidM) uses a glutamine synthetase adenylyl transferase domain to adenylylate another host GTPase called Rab1b19. However, in vitro labelling and mass spectrometric assays with AvrAC and BIK1 recombinant proteins failed to detect any AMPylation on BIK1. Surprisingly, mass spectrometry of the AvrAC recombinant protein (expressed alone in E. coli) revealed multiple modified protein components with serial additions of 306 Da, indicative of uridine 5′-monophosphate (UMP) modifications (Supplementary Fig. 6a). Tandem mass spectrometry (MS/MS) revealed multiple serine and threonine residues in AvrAC containing mono-UMP modifications (Supplementary Fig. 6b). The results suggested that AvrAC is a UMP transferase. In vitro assays with α32P-UTP showed that both BIK1 and RIPK were uridylylated only when incubated with AvrAC (Fig. 3a). AvrAC also labelled itself in this assay, explaining the UMP modifications on AvrAC described above. PBL1, a homologue of BIK1 that acts additively with BIK1 in PTI signalling7, was also modified by AvrAC (Supplementary Fig. 7). Interestingly, another RLCK VII subfamily member, PBS1, was not modified under the same condition. The same assay failed to label the kinase domain of FLS2 and BAK1, indicating that certain members of the RLCK VII subfamily are specific substrates of AvrAC (Fig. 3a). MS/MS analysis on the recombinant BIK1 protein that had been co-expressed with AvrAC uncovered a tryptic peptide (DGPMGDLSYVSTR) derived from the activation loop containing UMP at Ser 236 and Thr 237 (Supplementary Fig. 8a). Likewise, RIPK that had been co-expressed with AvrAC also produced a tryptic peptide (DGPEGDDTHVSTR) located in the activation loop in which Ser 251 and Thr 252 were UMP modified (Supplementary Fig. 8b). BIK1 and RIPK proteins that had been co-expressed with AvrAC in protoplasts also contained the same UMP modifications (Fig. 3b, c), indicating that AvrAC can uridylylate BIK1 and RIPK in the plant cell. The vast majority of the BIK1 and RIPK activation loop peptides were uridylylated (Supplementary Fig. 9). Together these results unequivocally demonstrated that AvrAC is a potent enzyme that uridylylates BIK1 and RIPK both in vitro and in plant cells. ATP is the preferred nucleotide substrate for IbpA and VopS to modify Rho GTPases in vitro16, although GTP, UTP and CTP can also be used. Whether nucleotides other than ATP are used by these effectors to modify Rho GTPases in host cells remains unknown. AvrAC failed to AMPylate BIK1, and we observed no AMPylation on any peptides of BIK1, RIPK or AvrAC expressed in E. coli or plant cells. Another Fic-domain protein, AnkX, from Legionella pneumophila was recently shown to covalently attach a phosphocholine moiety to the Rho GTPase Rab1 (ref. 20). The commonality shared by the three modifications suggests that Fic-domain proteins are phosphotransferases with different substrates. BIK1 Ser 236 and Thr 237 and RIPK Ser 251 and Thr 252 are highly conserved among RLCK VII subfamily members (Fig. 3d). Phosphorylation of BIK1 Ser 236 and Thr 237 has a crucial role in PTI signalling6, 7, 21. AvrAC failed to induce the BIK1(S236A) and BIK1(T237A) mutant protein migration in protoplasts (Fig. 3e), indicating that these residues are required for UMP modification. A similar result was obtained with a BIK1(S233A,Y234A,S236A,T237A) quadruple mutant7 (Supplementary Fig. 10). In contrast, mutations in Lys 105, an invariant residue required for ATP binding, and Ser 233 and Tyr 234, two less conserved amino acid residues in the activation loop, did not affect the AvrAC-induced migration (Supplementary Fig. 10 and Fig. 3e). RIPK and BIK1 that had been UMP modified by AvrAC in E. coli displayed reduced autophosphorylation (Fig. 3f). Taken together, these results demonstrated that AvrAC inhibits BIK1 and RIPK kinase activities by directly uridylylating the conserved phosphorylations sites in the activation loop, thereby preventing their phosphorylation and subsequent signal transduction. AvrAC contains an N-terminal domain (NTD) of unknown function, a leucine-rich-repeat domain (LRR), and a C-terminal region (CTR) harbouring the Fic domain (Fig. 4a). The homology of the AvrAC Fic domain with other Fic domains mainly resides in the C-terminal subdomain (Supplementary Fig. 11), in which the highly conserved HPFx(D/E)GN(G/K)R motif is responsible for nucleotide binding and catalysis22, 23. In particular, the invariant His residue corresponding to AvrAC His 469 has a key role in the catalysis. AvrAC constructs lacking the Fic catalytic motif (ΔFic) or CTR (ΔCTR) were fully capable of interacting with BIK1 and RIPK, whereas the AvrAC CTR alone was completely unable to interact with BIK1 and RIPK (Fig. 4b and Supplementary Fig. 12). Furthermore, the AvrAC NTD alone, the construct lacking NTD, and the construct lacking LRR were all able to interact with BIK1, indicating that NTD and LRR constitute the kinase-binding region. Removing the kinase-binding region or CTR, or substitution of His 469 with Ala, rendered AvrAC completely inactive in RIPK uridylylation in vitro (Fig. 4c), and induced a BIK1 mobility shift in protoplasts (Supplementary Figs 12 and 13). These results are consistent with the notion that the AvrAC Fic domain is responsible for catalysis, whereas NTD and LRR allow docking of substrate proteins. We next determined the role of individual AvrAC domains in the inhibition of PTI signalling and virulence function. All the AvrAC deletion constructs were unable to inhibit FRK1::LUC expression, whereas the AvrAC(H469A) mutant was partially affected (Fig. 4d, Supplementary Figs 14 and 15). Only the wild-type, but not the truncated versions of avrAC, was able to promote Xcc8004ΔavrAC bacterial growth on Zhonggan15 plants (Fig. 4e). Together these results indicated that the ability of AvrAC to inhibit immune signalling and enhance virulence not only required its binding to, but also UMP modification of, the substrate proteins. Xcc8004 and Xcc8004ΔavrAC are weakly virulent on Arabidopsis plants9, preventing a test of AvrAC virulence function in Arabidopsis. We found that Xcc strain B186, which is highly virulent on Arabidopsis plants24, contains AvrAC with a single amino acid substitution (AvrAC(T358A)). We then generated an XccB186ΔavrAC mutant and conducted competitive index assays in Arabidopsis mesophyll tissues. XccB186ΔavrAC grew to ~60% of XccB186 in wild-type Arabidopsis plants, and the difference was statistically significant (Fig. 4f). In contrast, the two strains grew to nearly identical levels in bik1 mutant plants. The bik1 mutant is known to accumulate high levels of the defence hormone salicylate25, which may potentially affect the virulence assay. We therefore carried out the assays on the sid2 mutant, which has greatly reduced salicylate production. XccB186ΔavrAC grew to ~40% of XccB186 on sid2 plants (Fig. 4f). In contrast, the two strains grew to nearly identical levels in the bik1/sid2 double-mutant plants, again indicating that avrAC promotes XccB186 bacterial growth in a BIK1-dependent manner. Together these results demonstrated that BIK1 is a key target for AvrAC virulence activity. In this study we demonstrated that AvrAC inhibits both PTI and RPM1-specified ETI by targeting two related RLCKs, BIK1 and RIPK, as illustrated in the model (Fig. 4g). BIK1 and PBL1 are not required for PAMP-induced MAPK activation (Supplementary Fig. 16), suggesting that additional host targets account for the MAPK inhibition by AvrAC. Although Xcc is not known to trigger the RPM1-specific ETI in Arabidopsis, the Xcc bacterium carries XopAH (also known as AvrXccC), which is highly homologous to AvrB26. It is formally possible that XopAH triggers ETI specified by an unknown resistance protein in some plants through RIPK. Inactivation of RIPK by AvrAC may allow the Xcc bacterium to inhibit such a potential ETI resistance in plants. AvrAC is the only Fic-domain effector among all sequenced phytopathogenic bacteria. However, the Fic and Doc (death on curing) domains contain a core structure, designated Fido, which is shared by AvrB and XopAH27. AvrB and XopAH do not carry the catalytic His residue conserved in Fic and Doc proteins, and it remains to be determined whether they also possess phosphotransferase activities. AvrAC uridylylates and masks the conserved phosphorylation sites in the activation loop of BIK1 and RIPK, preventing the activation of these kinases and impairing plant immunity. Interestingly, a similar strategy has been used by the Yersinia effector YopJ, which selectively acetylates serine and threonine residues in the activation loop of human MEK6, thereby blocking the phosphorylation of the target protein and signalling28. Multiple P. syringae type III effectors are known to target plant protein kinases and inhibit PTI signalling by using diverse biochemical mechanisms5. The work described here illustrates a unique biochemical mechanism by which the Xcc bacterium combats the plant innate immune system. Arabidopsis plants were grown in a growth room at 23 °C and 70% relative humidity with a 10/14 h day/night light cycle for 5 weeks before protoplast isolation or bacterial inoculation. Cabbage cultivars Zhonggan15, Chungan45, Jingfeng1, Zhonggan12, Zhonggan8 were grown in greenhouses at 25 °C. The bacterial strains used in this study included Xcc8004 (ref. 9), Xcc8004ΔhrcV (ref. 29), Xcc strain B186 (XccB186) (ref. 23), Pst strain DC3000, PstΔhrcC (ref. 30), Pst(avrB), Pst(avrRpt2), Pst(avrPphB) and Agrobacterium GV3101. To generate recombinant protein constructs, full-length AvrAC and truncated mutants were PCR-amplified from Xcc8004 genomic DNA and inserted between EcoRI and XhoI sites of pGEX-6p-1 (Amersham Biosciences) to generate GST–AvrAC, GST–CTR (C-terminal region), GST–ΔCTR and GST–ΔFic. A GST–AvrAC(H469A) mutant construct was generated by site-directed mutagenesis. The coding sequences of Arabidopsis RIPK, BIK1 and PBL1 were amplified from cDNA and inserted into pET28a or pET28b (Novagen) to generate His–RIPK, His–BIK1 and His–PBL1. To generate constructs for protoplast transfection assay, RIPK, AvrAC and its derivatives were PCR-amplified and inserted between XhoI and Csp45I sites of pUC19-35S-Flag-RBS vector31 to generate RIPK–Flag, AvrAC–Flag, NTD–Flag, ΔNTD–Flag, CTR–Flag, ΔCTR–Flag, ΔFic–Flag, ΔLRR–Flag and AvrAC(H469A)–Flag. RIPK and AvrAC were cloned into KpnI and SalI sites of pUC19-35S-HA-RBS31 to generate RIPK–HA and AvrAC–HA, respectively. BIK1–Flag, BIK1–HA and BIK1 mutants were described previously7. To generate avrAC knockout mutants in Xcc8004 and XccB186, an internal sequence corresponding to nucleotides 30–500 of the avrAC coding region was PCR-amplified from genomic DNA of Xcc8004 and cloned into the suicide plasmid pK18mob32. The recombinant plasmid was transferred into the wild-type Xcc strains by triparental conjugation, and the single crossover mutants were confirmed by PCR and designated Xcc8004ΔavrAC and XccB186ΔavrAC. To complement the Xcc8004ΔavrAC mutant, the 400 bp upstream sequence of the avrAC coding region was PCR-amplified from the genomic DNA of Xcc8004 and ligated between the XbaI and HindIII sites of pLAFR6 (ref. 9), generating an intermediate plasmid carrying avrAC promoter sequence. A full-length avrAC, avrACΔLRR or avrACΔFic fragments were PCR-amplified using the pUC19-35S-Flag-RBS plasmids containing the corresponding avrAC fragments as templates (see above) before integrated into the intermediate plasmid. The resulting constructs were introduced into the Xcc8004ΔavrAC strain by triparental conjugation. The avrAC coding region was PCR-amplified from Xcc8004 genomic DNA, ligated into a modified pER8 vector31, 33. The resulting clone containing AvrAC-Flag under the control of the oestrogen-inducible promoter was transformed into Arabidopsis (Col-0) by Agrobacterium-mediated transformation according to standard protocols. Transgenic plants were selected on Murashige and Skoog (MS) plates containing hygromycin. Two independent transgenic lines were selected for experiments. The transgenic plants were sprayed with 50 mM oestradiol in a 0.01% silwet L-77 solution for 12 h to induce AvrAC protein expression. AvrAC and its truncation mutants in the pUC19-35S-Flag-RBS vector were co-transfected with FRK1::LUC (firefly luciferase) and 35S::RLUC (Renilla luciferase) into Arabidopsis protoplasts. The protoplasts were incubated overnight under low light, treated with 1 µM flg22, 1 µM elf18, or 200 µg ml−1 chitin (Sigma) for 3 h. Protein was then isolated, and LUC activity was recorded by using the Dual-Luciferase Reporter system (Promega) according to the manufacturer’s instructions. Oestradiol-induced leaves were sliced into 1 mm strips, incubated in 200 µl water in a 96-well plate for 12 h before the addition of 1 µM flg22 in 200 µl reaction buffer supplement with 20 mM luminol and 1 µg horseradish peroxidase (Sigma). Luminescence was recorded with a Luminometer (Promega) for 20 min. Two-week-old wild-type and transgenic seedlings grown on 1/2 MS medium were sprayed with 10 µM flg22 or water containing 0.02% Silwet L-77 for 10 min and frozen in liquid nitrogen. The seedlings were homogenized in an extraction buffer containing 50 mM HEPES-KOH (pH 7.5), 150 mM KCl, 1 mM EDTA, 0.2% Triton-X 100, 1 mM DTT, complete protease inhibitors (Roche), and phosphatase inhibitors (Fisher). The protein concentration was determined using a Bio-Rad Bradford protein assay kit, and equal amounts of total protein were electrophoresed on 10% SDS–PAGE. An anti-pERK antibody (no. 4370S, Cell Signaling) was used to determine phosphorylation state of MPK3, MPK4 and MPK6 in an immunoblot. A duplicate blot was reacted with anti-MPK3 antibodies (Sigma) to determine the amount of total MPK3. For MAPK activation by MKK5DD, protoplasts prepared from wild-type plants were transfected with the indicated plasmids, and protein was isolated 12 h later for immunoblot analyses. The identities of phosphorylated MAPKs were labelled as described previously34, 35. Five-week-old Arabidopsis leaves were infiltrated with the indicated Xcc or Pst bacteria at 106 c.f.u. ml−1, and the bacterial population in the leaf was counted at the indicated times. Each data point consisted of at least four replicates. For the hypersensitive response assay, wild-type plants were first infiltrated with Xcc8004 or Xcc8004ΔavrAC mutant bacteria at 107 c.f.u. ml−1 1 day before infiltrating 5 × 107 c.f.u. ml−1 Pst or Pst(avrB). Plants were photographed 6 h after the second inoculation. A competitive index assay36 was used to compare virulence between wild-type and ΔavrAC mutant strains of Xcc. Briefly, equal amounts (c.f.u.) of wild-type and ΔavrAC mutant Xcc bacteria were mixed and inoculated into plants, and leaf bacterial number for each strain was determined by plating on NYG agar medium supplemented with appropriate antibiotics. The ratio between the mutant and wild-type bacteria (competitive index) was calculated at the desired times. For virulence assay on cabbage, fully expanded leaves of 4-week-old plants were clipping inoculated37 with a mixture of Xcc8004 and Xcc8004ΔavrAC bacteria at an equal concentration of 108 c.f.u. ml−1, maintained at high humidity for 48 h under a transparent dome. Leaf bacterial number and competitive index were determined 6 days after inoculation. For virulence assay on Arabidopsis plants, 5-week-old Arabidopsis leaves were infiltrated with a mixture of wild-type XccB186 and XccB186ΔavrAC mutant bacteria at an equal concentration of 106 c.f.u. ml−1. Leaf bacterial number and competitive index were determined 3 days after inoculation. The plasmid pCPP3221 containing an AvrPto-Cya fusion sequence was introduced into the wild-type Pst (DC3000) and its ΔhrcC mutant derivative. The derived strains were inoculated at 1 × 108 c.f.u. ml−1 into Arabidopsis leaves that had been pre-treated with 107 c.f.u. ml−1 Xcc8004 or Xcc8004ΔavrAC for 24 h. Amounts of AvrPto-Cya secretion were determined as described previously38. Briefly, leaf samples were ground in liquid nitrogen and resuspended in 0.1 M HCl. The total protein content of each sample was determined using the Bradford assay (Bio-Rad). A direct cAMP (cAMP) immunoassay kit (Sigma) was used to measure cAMP levels following the manufacturer’s instruction. The recombinant proteins were affinity purified following the manufacturer’s instruction. For pull-down assays, 10 µg each of His–RIPK, His–BIK1 and GST–AvrAC were incubated on a rotator with glutathione agarose beads (GE Healthcare) in a buffer containing 25 mM Tris-Hcl (pH 7.5), 100 mM NaCl and 1 mM DTT for 3 h, washed five times with a buffer containing 25 mM Tris-Hcl (pH 7.5), 100 mM NaCl and 1 mM DTT. The bound protein was eluted with 15 mM GSH and immunoblotted with a mouse anti-His monoclonal antibody (Sigma-Aldrich). The protoplasts were transfected with the indicated constructs, incubated for 12 h, treated with either H2O or 1 µM flg22 for 3–10 min, and total protein was isolated with an extraction buffer (see MAPK activity assay). For anti-Flag immunoprecipitation, total protein was incubated with an agarose-conjugated anti-Flag antibody (Sigma-Aldrich) for 4 h, washed six times with a buffer containing 50 mM HEPES-KOH (pH 7.5), 150 mM KCl, 1 mM EDTA, 0.2% Triton-X 100, 1 mM DTT, and the bound protein was eluted with 0.5 mg ml−1 3×Flag peptide. For anti-HA immunoprecipitation, total protein was pre-cleared with protein A agarose (Millipore) for 1 h, followed by precipitation with 2 µg anti-HA antibody (TianGen) together with protein A agarose for 4 h. Immunoprecipitates were separated by a 10% NuPAGE gel (Invitrogen) and detected by anti-HA or anti-Flag immunoblot. His–RIPK, His–BIK1 and GST–AvrAC were co-expressed in E. coli (strain BL21), and the His-tagged protein was purified by Ni-NTA affinity chromatography. Kinase assays were performed in 25 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 100 µM ATP, 1 mM DTT, γ-32P ATP (5 µCi), and 3 µg recombinant protein in a total volume of 20 µl. Reactions were allowed to proceed for 30 min at 30 °C and terminated by adding 5× protein loading buffer and boiling for 5 min at 100 °C. The proteins were separated on a 12% SDS–PAGE gel, and incorporated radiolabel was visualized by autoradiography. As an equal loading control for proteins used in the kinase reactions, the SDS–PAGE gels were stained by Coomassie blue. Arabidopsis protoplasts were transfected with HA-tagged BIK1 or BIK1 mutants7 together with AvrAC–Flag or truncation mutants of AvrAC, treated with 1 µM flg22 and total protein was extracted at various time points. Samples were separated by 10% SDS–PAGE gels followed by anti-HA immunoblot. For phosphatase treatment, total protein from protoplasts or affinity-purified recombinant protein was treated with λ protein phosphatase (New England Biolabs) according to the manufacturer’s instructions. GST-tagged AvrAC derivative proteins, His-tagged RIPK, BIK1 and PBL1 proteins were affinity purified and directly used for the uridylylation assay. The GST–PBS1 recombinant protein (a gift from J. Chai) was cleaved with PreScission Protease (GE) to remove GST before the enzymatic assay. Approximately 400 ng of GST–AvrAC, GST–ΔCTR, GST–CTR, GST–ΔFic or GST–AvrAC(H469A) were incubated with 2 µg of His–RIPK, His–BIK1, PBS1, His–PBL1, or His-tagged kinase domain of FLS2 or BAK1 (ref. 39) in 20 µl reaction buffer containing 25 mM Tris-HCl (pH 7.5), 25 mM MgCl2, 500 µM UTP, 1 mM DTT, α-32P UTP (5 µCi) for 30 min at 30 °C. Reactions were stopped by addition of loading buffer. The products were separated on 12% NuPAGE gels (Invitrogen) and visualized by autoradiography. An affinity-purified GST–AvrAC protein solution was loaded into an analytical capillary column (75 μm × 2 cm) packed with Poros 20 R2 packing material (AppliedBiosystems). An Agilent 1100 binary pump was used to generate HPLC gradient as follows: 0–100% B in 60 min (A = 0.1 M acetic acid in water; B = 0.1 M acetic acid/70% acetonitrile). The eluted proteins were sprayed into a QSTAR XL mass spectrometer (AB Sciex). The spray voltage was set at 2,100 V and the data was acquired in MS mode. The protein charge envelop from the raw spectrum was de-convoluted into non-charged form by the BioAnalyst software provided by the manufacturer. Affinity-purified GST–AvrAC, His–BIK1, His–RIPK, BIK1–Flag and RIPK–Flag proteins were separated by SDS–PAGE. Protein bands on the SDS–PAGE gel were de-stained, and then reduced in 10 mM DTT at 56 °C for 30 min followed by alkylation in 55 mM iodoacetamide at dark for 1 h. After that the protein bands were in-gel digested with sequencing grade trypsin (10 ng μl−1 trypsin, 50 mM ammonium bicarbonate, pH 8.0) overnight at 37 °C. Peptides were extracted with 5% formic acid/50% acetonitrile and 0.1% formic acid/75% acetonitrile sequentially and then concentrated to ~20 μl. The extracted peptides were separated by an analytical capillary column (50 μm × 10 cm) packed with 5 μm spherical C18 reversed phase material (YMC). An Agilent 1100 binary pump was used to generate the HPLC gradient as follows: 0–5% B in 5 min, 5–40% B in 25 min, 40–100% B in 15 min (A = 0.1 M acetic acid in water; B = 0.1 M acetic acid/70% acetonitrile). The eluted peptides were sprayed into a QSTAR XL mass spectrometer equipped with a nano-ESI ion source. The mass spectrometer was operated in information-dependent mode with one MS scan followed by three MS/MS scans for each cycle. Database searches were performed on an in-house Mascot server (Matrix Science) and the following variable modifications were included: oxidation on methionine, carbamidomethylation on cysteine, UMP modification on serine or threonine residues, AMP modification on serine, threonine or tyrosine residues. Download references The authors thank J. Chai for sharing plasmids before publication, A. Bent for the XccB186 strain, S. Y. He and F. White for helpful comments. J.-M.Z. was supported by grants from the Chinese Ministry of Science and Technology (2011CB100700; 2010CB835301). S.C. was supported by a grant from the Chinese Ministry of Science and Technology (2010CB835204) C.H. was supported by Funds from Hainan University.