- Email: [email protected]
Crystal Structure of the Type III Effector AvrB from Pseudomonas syringae
Structure, Vol. 12, 487–494, March, 2004, 2004 Elsevier Science Ltd. All rights reserved.
DOI 10.1016/j.str.2004.02.013
Crystal Structure of the Type III Effector AvrB from Pseudomonas syringae Christian C. Lee,1,4,* Michelle D. Wood,2,3,4,5 Kenneth Ng,1 Carsten B. Andersen,1 Yi Liu,1 Peter Luginbu¨hl,2,6 Glen Spraggon,1 and Fumiaki Katagiri 2,7 1 Genomics Institute of the Novartis Research Foundation 10675 John Jay Hopkins Drive San Diego, California 92121 2 Torrey Mesa Research Institute and Syngenta Research and Technology 3115 Merryfield Row San Diego, California 92121 3 Department of Biological Sciences University of Maryland Baltimore County 1000 Hilltop Circle Baltimore, Maryland 21250
Summary AvrB is a Pseudomonas syringae type III effector protein that is translocated into host plant cells during attempted pathogenesis. Arabidopsis harboring the corresponding resistance protein RPM1 can detect AvrB and mount a rapid host defense response, thus avoiding active infection. In the plant cell, AvrB induces phosphorylation of RIN4, a key component in AvrB/ RPM1 recognition. Although the AvrB/RPM1 system is among the best characterized of the numerous bacterial effector/plant resistance protein systems involved in plant disease resistance and pathogenesis, the details of the molecular recognition mechanism are still unclear. To gain further insights, the crystal structure of AvrB was determined. The 2.2 A˚ structure exhibits a novel mixed ␣/ bilobal fold. Aided by the structural information, we demonstrate that one lobe is the determinant of AvrB/RPM1 recognition specificity. This structural information and preliminary structure-function studies provide a framework for the future understanding of AvrB function on the molecular level. Introduction Many Gram-negative bacterial pathogens of plants and animals use the type III secretion system to deliver protein factors into host cells (Gala´n and Collmer, 1999). Although no underlying common mechanistic themes appear to exist, these type III effectors generally contribute to bacterial virulence by perturbing host cellular *Correspondence: [email protected] 4 These authors contributed equally to this work. 5 Present address: 9939 Azuaga Street, H104, San Diego, California 92129. 6 Present address: Diversa Corporation, 4955 Directors Place, San Diego, California 92121. 7 Present address: Department of Plant Biology, University of Minnesota, 1445 Gortner Avenue, St. Paul, Minnesota 55108.
functions. Plants have evolved a defense mechanism that is rapidly induced upon specific recognition of certain pathogen factors, including some type III effectors. Plant resistance (R ) gene products are central in recognition of such pathogen factors, which are the direct or indirect products of avirulence (avr) genes of pathogens (Dangl and Jones, 2001). Thus, many type III effector genes of phytopathogenic bacteria were initially discovered and named as avr genes, such as avrB. The correspondence between an R gene and an avr gene is usually highly specific; hence this type of disease resistance is termed gene-for-gene resistance. It remains unclear whether the recognition event for gene-for-gene resistance involves direct interactions between plant R proteins and the corresponding Avr type III effector proteins. AvrB is one of several type III effectors of the phytopathogenic bacterium Pseudomonas syringae (Tamaki et al., 1988). Arabidopsis carrying the R gene, RPM1, are able to detect the presence of AvrB upon delivery inside the plant cell, rapidly induce defense mechanisms upon detection, and consequently are resistant to P. syringae strains expressing AvrB (Grant et al., 1995). AvrB is one of several P. syringae type III effector proteins that are myristoylated in planta. This AvrB modification is essential for its membrane localization, RPM1-mediated resistance, and an RPM1-independent cytotoxic effect (Nimchuk et al., 2000). The RPM1-independent cytotoxic effect may reflect a virulence function of AvrB. RPM1-mediated resistance to P. syringae expressing AvrB is known to require at least one other plant protein, RIN4 (RPM1-interacting protein) (Mackey et al., 2002). Like AvrB, both RPM1 and RIN4 localize to the plasma membrane (Boyes et al., 1998; Mackey et al., 2002). Furthermore, RIN4 interacts with both RPM1 and AvrB, suggesting that these three proteins may form a higher order complex upon delivery of AvrB into the plant cell, and that these interactions are critical for RPM1-mediated resistance (Mackey et al., 2002). When AvrB is delivered into the plant cell, phosphorylation of RIN4 occurs in an RPM1-independent manner (Mackey et al., 2002). However, it is not known whether RIN4 phosphorylation results from a direct interaction with AvrB or from an indirect mechanism. Recently, the guard hypothesis was postulated to explain many gene-for-gene resistance interactions (Bonas and Lahaye, 2002; Dangl and Jones, 2001; Van Der Biezen and Jones, 1998). According to this hypothesis, R proteins function to monitor and respond to changes in the state of virulence targets of Avr proteins. The notion that RPM1 responds to the AvrB-induced phosphorylation of RIN4 is consistent with this hypothesis. Another type III effector protein AvrRpt2, unrelated to AvrB, also appears to target RIN4 as a virulence mechanism, inducing the posttranscriptional elimination of RIN4 (Axtell and Staskawicz, 2003; Mackey et al., 2003). The Arabidopsis RPS2 gene product, which mediates resistance to P. syringae expressing AvrRpt2, is believed to respond to the disappearance of RIN4 as a signal to
Crystal Structure of the Type III Effector AvrB 489
initiate plant defense responses. Thus, RIN4 appears to represent a common virulence target that is “guarded” by at least two different R gene products in Arabidopsis. Here, we report the three-dimensional crystal structure of AvrB. AvrB assumes an apparently novel ␣/ bilobal fold, with no significant structural homologs in the Protein Data Bank (PDB). A functional analysis of chimeras composed of AvrB and an AvrB homolog, AvrPphC, identified one of the lobes as the specificity determinant in the AvrB/RPM1 system. A simple explanation for the fact that introduction of AvrB into a plant cell induces RIN4 phosphorylation is that AvrB is a RIN4 protein kinase. However, AvrB lacks sequence and functional motifs common to canonical protein kinase catalytic domains. Thus, our structural and preliminary functional studies suggest either AvrB does not function as a kinase and is not involved in the direct phosphorylation of RIN4 in the AvrB/RPM1/RIN4 system or alternatively represents a protein kinase domain fold previously undescribed. Results Crystal Structure of AvrB Reveals a Novel Fold In efforts to gain insight into the molecular details behind AvrB function and host recognition, the crystal structure was solved to a resolution of 2.2 A˚. The structure reveals a bilobal fold with a smaller mixed ␣/ lobe and a larger predominantly helical lobe separated by a deep interlobe cleft (Figure 1A). We refer to the two lobes as the upper (amino acid residues 123–217) and lower (residues 28–122 and 218–317) lobes, respectively. The smaller upper lobe consists of three helices (␣5, ␣6, and ␣7) surrounding a central five-stranded antiparallel  sheet that separates the three upper lobe helices from the helices of the lower lobe. The adjoining helices ␣6 and ␣7 are oriented approximately 50⬚ relative to each other and extend from strands 4 and 5 of the central  sheet. Helix ␣5 is positioned between strands 1 and 2 and packs perpendicularly against the midportion of the  sheet. The strand order for the antiparallel  sheet is 1-5-4-3-2 with strands 2, 3, and 4 forming two long opposing  turns. The lower lobe consists of a major five helix bundle comprised of helices ␣2, ␣3, ␣8, ␣9, and ␣10 with four shorter flanking helices (␣1, ␣4, ␣11, and ␣12). The two lobes are joined by two segments forming a hinge-like region, the first between helix ␣4 and strand 1, and the second between strand 5 and helix ␣8. The dimensions of this large, solventaccessible interlobe cleft are approximately 24 A˚ wide and 18 A˚ in length, which corresponds to over 900 A˚3 in volume as determined by CastP (Liang et al., 1998). Dali searches (Holm and Sander, 1993) of the PDB using
the AvrB structure model as a whole and as independent lobes found no existing structural homologs, suggesting that AvrB assumes a unique fold (Figure 1B). Molecular and Electrostatic Surface Examination of the electrostatic surface potential of AvrB mapped onto the molecular surface reveals a striking degree of negative surface charge that encompasses a vast majority of the molecular surface (Figure 1C). This extensive surface electronegativity may reflect interactions with RPM1 and/or RIN4 are largely electrostatic in nature. A localization of more basic electrostatic surface appears to adorn the mouth of the interlobe cavity. Ignoring any clues implicating AvrB function from existing biological data, the sheer breadth of the interlobe cavity may suggest AvrB functions in a yet undetermined enzymatic capacity. This speculation is especially tempting when sequence conservation among other AvrB family members is taken into consideration in the context of the AvrB structure. The Interlobe Cleft May Be Important for an Enzymatic Function Common among AvrB Family Members. AvrB homologs from three pathovars of P. syringae and one pathovar of Xanthomonas campestris were identified by a BLAST search of the GenBank protein sequence database (Altschul et al., 1997). The amino acid sequence alignment of AvrB and its homologs reveals 30%–44% sequence identity and 47%–58% similarity when conservative substitutions are considered (Figure 2). Examination of the multiple sequence alignment among AvrB family members reveals several contiguous stretches of amino acid conservation that correspond to helical and surrounding regions of helices ␣1–␣3 and helices ␣8–␣12 (Figure 2). When mapped onto the threedimensional structure of the AvrB model, these stretches of high degree of sequence conservation localize predominantly to the lower lobe (Figures 2 and 3A). Moreover, many of the conserved solvent-accessible residues that do not appear to play a structural role in maintaining the integrity of the AvrB fold appear to localize in the interlobe cleft. This phylogenetically conserved deep cleft may reflect a common enzymatic active site responsible for substrate and/or cofactor binding of AvrB family members. The AvrB Upper Lobe Is Important in Specificity Determination for RPM1-Mediated Resistance Guided by the AvrB crystal structure, a series of chimeric proteins consisting of AvrB and a homolog, AvrPphC, were generated to delineate the portion of AvrB that determines its specificity in Arabidopsis RPM1-medi-
Figure 1. AvrB Crystal Structure (A) Ribbon cartoon diagram of the AvrB model shown in two orthogonal views. Residues 28–317 are shown. Residues 28 (“N”) and 317 (“C”) along with successive secondary structural elements are labeled. The three helices that are part of the upper lobe are depicted in blue. The five-stranded antiparallel  sheet connecting the helices of the upper lobe and the lower lobe is shown in orange. The remaining eight helices constituting the lower lobe are colored green. (B) Topology diagram of the AvrB model. Circles represent helices and triangles  strands and are colored as in (A). (C) Electrostatic potential surface representation of AvrB. Electropositive, blue; electronegative, red. The central and right orientations correspond to those of (A), whereas the leftmost orientation represents another orthogonal view.
Structure 490
Figure 2. Amino Acid Sequence Alignment of AvrB Family Members AvrB (GenBank accession, P13835) (Tamaki et al., 1988) and AvrC (P13836) (Tamaki et al., 1988) originate from P. syringae pv. glycinea. AvrPphC (AAA83419) (Yucel et al., 1994) is from P. syringae pv. phaseolicola. Avr_Xcc (NP_637473) (da Silva et al., 2002) represents a homolog from Xanthomonas campestris pv. campestris. Avr_Pss (AAF71496) (Alfano et al., 2000) represents a homolog from P. syringae pv. syringae. Invariant amino acid residues among all the sequences are highlighted in red with yellow lettering. Residues invariant in four of five sequences in the alignment are highlighted in blue with yellow lettering. Conserved residues based upon amino acid size and charge among four of the five sequences are highlighted in green with yellow lettering. Helical and  strand secondary structural elements according to the AvrB structure are represented by blue bars and purple arrows, respectively. They are numbered according to Figure 1A.
ated recognition (Figure 2). Previously, Tamaki et al. reported a similar chimera study consisting of AvrB and AvrC, which is essentially identical to AvrPphC (Figure 2), to identify the specificity-determining region of AvrB in resistance mediated by the soybean R gene Rpg1 (Tamaki et al., 1991). In that study, however, chimeras were designed according to ease of construction because the AvrB three-dimensional structure was not known and because it predated PCR. In our studies, we designed one chimera, B54-225, based on the smallest AvrB region Tamaki et al. identified for recognition by Rpg1. The constructs were named according to the amino acid residues from AvrB that replaced the corresponding region of AvrPphC. For example, B54-225 was generated in an AvrPphC background with amino acid residues 54–225 of AvrB replacing the corresponding
region of AvrPphC. Construct B54-225 was chosen as a starting chimera because it was anticipated that recognition of AvrB by Arabidopsis RPM1 and soybean Rpg1 were similar. Three additional chimeras, B102123, B102-216, and B126-216, were designed to determine the minimal region of AvrB required to elicit RPM1mediated resistance and at the same time to minimize any potential structural perturbation as predicted by the AvrB structure. The functional assay for RPM1-mediated resistance was based on the hypersensitive cell death that occurs pursuant to “gene-for-gene” recognition, and the consequential reduction of reporter activity (Leister et al., 1996). When AvrB was coexpressed with the renilla luciferase (rLUC) reporter in plants, a strong reduction of rLUC activity was observed in a manner dependent upon
Crystal Structure of the Type III Effector AvrB 491
Figure 3. Stereoviews of C␣ Backbone Trace of AvrB Model and Representative Electron Density (A) Stereoview of the C␣ backbone trace of AvrB in a similar orientation as in Figure 1A. Amino acid residues are indicated using their single letter codes and numbered as labeled. Residues conserved among AvrB family members are colored red and light blue. Red, invariant; light blue, conserved; yellow, unconserved. (B) Stereoview of representative 2fo-fc electron density contoured at 1 . Electron density is shown in purple. The salt bridge between Arg137 of helix ␣5 and Glu299 of the interconnecting loop between helix ␣1 and ␣2 is shown along with neighboring residues.
coexpression of RPM1 (Figure 4). Expression of RPM1 alone moderately reduces rLUC activity (compare ⫹RPM1 and ⫺RPM1 with Vector). This may be due to a weak basal activity of RPM1 in the absence of specific stimulation. Similar but stronger basal activity was observed with another Arabidopsis resistance protein RPS2 in this assay (Tao et al., 2000). Nevertheless, reduction of rLUC activity in the presence of both AvrB and RPM1 was clearly greater than its reduction in the presence of RPM1 alone, and, therefore, there is no ambiguity in interpretation of the results presented in Figure 4. In contrast to AvrB, AvrPphC expression did not result in a strong reduction of rLUC activity, regardless of the presence of RPM1. These results indicate
that AvrB, but not AvrPphC, was able to induce RPM1mediated resistance. B54-225, B102-216, and B126-216 functioned like AvrB, while B102-123 did not. The B126216 chimera represents a precise substitution of the upper lobe of AvrPphC with AvrB. Based on immunoblot analysis, AvrPphC and the four chimeras were all able to accumulate in rpm1 mutant cells to a level comparable to that of AvrB (data not shown). Therefore, a failure of RPM1-mediated resistance is not attributable to decreased expression of the chimera in planta. We conclude that the region of AvrB involved in eliciting an RPM1-mediated resistance response maps within the upper lobe. The amino acid region of AvrB from 54–225 was previously shown to determine the specificity to
Structure 492
Figure 4. The Region Determining Specificity to RPM1 Is Mapped to the Upper Lobe of AvrB The RPM1-mediated recognition of chimeric proteins B54-225, B102-123, B102-216, and B126-216, was detected by a large reduction of rLUC reporter activity. Each bar represents the mean value of four bombardment events (error bar ⫽ standard deviation). This experiment was repeated with similar results.
the soybean R protein Rpg1 (Tamaki et al., 1991). Since this amino acid region encompasses the entire upper lobe of AvrB, the determinant of specificity to Rpg1 may also map within the upper lobe. Our finding that the AvrB upper lobe is involved in recognition by the RPM1mediated resistance response is especially poignant when amino acid conservation among AvrB family members is taken into account. The overall limited sequence conservation among AvrB family members mapping to the upper lobe may further support this finding that this region is involved in specific molecular recognition by the plant R gene products. Discussion AvrB induces phosphorylation of RIN4 in Arabidopsis plant cells (Mackey et al., 2002). The simplest hypothesis for this biological phenomenon is that AvrB directly phosphorylates RIN4. The fact that a Dali search of the PDB with the AvrB coordinates did not yield any similar folds indicates that AvrB does not possess a fold similar to those of known protein kinases. Similarly, structural analysis using programs such as PINTS and PROMOTIF to detect local structural similarities and functional motifs were unsuccessful in finding any similarities with AvrB (Stark et al. 2003; Hutchinson and Thornton, 1996) and did not afford any insights to AvrB function. These findings suggest that AvrB may not be a protein kinase or if it is a protein kinase it is one that assumes a fold previously undescribed for proteins capable of phosphotransfer to protein substrates. Nonetheless, we attempted to obtain experimental biochemical evidence in support of AvrB protein kinase function. These efforts were unfruitful in that we failed to detect in vitro phosphorylation of nonspecific substrates, including myelin basic protein, casein, histone H1, and poly-glutamate-tyrosine peptide, by AvrB (data not shown). Furthermore, we were unable to detect autophosphorylation or demonstrate ATP-coordination of AvrB in vitro (data not shown). Attempts at demonstrating AvrB binding to ATP via capture on ATP-sepharose resin were likewise unconvincing in our hands. Thus, we failed to detect any evidence that AvrB is a protein kinase or bind ATP by these conventional approaches. Attempts at obtaining an ATP-bound AvrB structure were also unsuccessful. However, it should be noted that many type III effectors of bacteria require host factors for their activation and/or modification. For example, the Yersinia pestis Ser/Thr protein kinase YpkA re-
quires actin for its activation (Juris et al., 2000). It is conceivable that AvrB may also require a host factor to be active. Taken together, our findings and preliminary functional studies suggest that AvrB is not a protein kinase as existing biochemical evidence may suggest, and if it does represent a protein kinase capable of direct phosphorylation of the RIN4 plant target protein, the structure and mechanism for this potential protein kinase is of a nature previously undescribed and warrants further biochemical investigation. Thus, the precise functional role of AvrB in RIN4 phosphorylation may involve a mechanism that does not involve direct phosphorylation of RIN4 by AvrB. In conclusion, we have successfully crystallized and solved the structure of AvrB from P. syringae. To our knowledge, this is the first crystal structure to be solved among type III effector proteins from phytopathogenic bacteria and represents a fold unique to AvrB. This bilobal mixed ␣/ fold is comprised of a smaller upper lobe and larger predominantly helical lower lobe separated by a deep interlobe cleft. Primary sequence and structural analysis of AvrB and family homologs suggest these effector proteins may possess a common enzymatic function involving the interlobe cleft region that may be related to virulence function. This information, in conjunction with our results from preliminary structurefunction studies of AvrB chimeras implicating the AvrB upper lobe as the RPM1-specificity determinant, provides initial insight to AvrB function on the molecular level. The crystal structure will undoubtedly continue to be useful as a scaffold for design and interpretation of future structure-function studies aimed at interrogating the precise function of AvrB as a type III effector and a more detailed understanding of the AvrB-RIN4-RPM1 bacterial phytopathogen/plant host defense system. Solving the AvrB crystal structure is a first step toward elucidating the AvrB/RPM1 molecular recognition mechanism. A more comprehensive understanding will be achieved when the crystal structures of other components involved in the recognition both alone and in complex with AvrB, such as RPM1 and RIN4, are determined. We believe that further appreciation of the three-dimensional structure of this type III effector protein will provide insights that lead to novel intervention strategies to control crop diseases. Experimental Procedures AvrB Expression and Purification The full-length C-terminally hexahistidine-tagged AvrB protein was cloned, expressed, and purified from E. coli and found to function
Crystal Structure of the Type III Effector AvrB 493
Table 1. Crystallographic Data Data Collection Data Set
Native
SeMet
Space group Cell a ⫽ b (A˚) c (A˚) ␣ ⫽  (⬚) ␥ (⬚) Wavelength (A˚) Resolution (A˚) Completeness (%) Rsymm I/ Number of sites
P65
P65
125.028 63.246 90 120 1.00 50–2.2 93.4 (64.6) 0.066 (0.493) 13.0
125.75 63.90 90 120 0.97840 20–2.65 99.4 (96.5) 0.136 (0.808) 8.1 4
Refinement Resolution Range Number of reflections (working/free) Number of atoms Protein Solvent Rwork/Rfree Rms deviation from ideality Bond lengths (A˚) Bond angles (⬚) Number of residues (monomer/asymmetric unit)
500–2.2 26,473/1,419
2,254 189 22.2/27.5 0.0104 1.54 289/331
Functional Assay of Chimeras C-terminally FLAG-tagged AvrB, AvrPphC, and chimeric constructs were transiently expressed in Arabidopsis leaf tissue, essentially as described (Leister et al., 1996). Because the response mediated by RPM1 is very strong in this assay, normalization of the transformation efficiency was not necessary and was omitted. To assay protein accumulation, the same constructs were transiently expressed, along with a firefly luciferase (LUC) reporter, in rpm1 mutant Arabidopsis protoplasts as described (Leister and Katagiri, 2000). The cell lysates were assayed for LUC activity in order to normalize the transformation efficiency prior to sample loading. Samples were then subjected to SDS-PAGE and immunoblot analysis. Acknowledgments
Ramachandran analysis (%) Most favored Additionally allowed Generously allowed Disallowed Average B factors (main chain/side chain) (A˚2)
(Otwinowski, 1997). Native diffraction data (Rmerge, 6.6%) are 85.6% complete to 2.2 A˚. Derivative data (Rmerge, 13.6%) are complete to 99.4% overall. Initial phases and substructure for four of the five Se sites were calculated using SOLVE and SHARP giving a mean figure of merit of 0.328 to 3 A˚ (de La Fortelle and Bricogne, 1997). Phases were refined and extended to 2.6 A˚ resolution using RESOLVE. Iterative cycles of model building and refinement were performed using “O” (Jones et al., 1991) and Refmac (Murshudov et al., 1997). Final rounds of refinement were conducted using CNS (Bru¨nger et al., 1998). The R factor and free R factor converged at 22.2% and 27.5%, respectively. The final AvrB model consists of 290 amino acids (amino acids 28–317) and 189 water molecules. The electron density for the first 27 amino acids was disordered and uninterpretable in 2fo-fc electron density maps. The overall geometry and stereochemistry of the final model was excellent with no residues in disallowed regions as analyzed using PROCHECK. Data and refinement statistics are summarized in Table 1.
88.9 10.7 0.4 0.0 54.6/58.1
Numbers in parentheses indicate values for the highest resolution shell.
We thank Paul Haynes, Donna Eckert, and Dave Schieltz for verifying SeMet incorporation into the protein and Jane Glazebrook (TMRI), Scott Lesley (GNF), and Michael DiDonato (GNF) for critical reading of the manuscript. The Torrey Mesa Research Institute has been closed. Received: June 20, 2003 Revised: November 12, 2003 Accepted: November 28, 2003 Published: March 9, 2004 References
in Arabidopsis protoplasts as described (Wu et al., 2003). Briefly, nickel affinity chromatography fractions containing partially purified recombinant AvrB were pooled and buffer exchanged into 20 mM Tris-HCl, pH 7.9, 50 mM NaCl, and 0.25 mM TCEP prior to anionexchange chromatography (Resource-Q) using a 0–0.5 M linear salt gradient over 15 column volumes. Peak fractions containing recombinant AvrB were pooled and concentrated followed by size exclusion chromatography (HiLoad 16/60 Superdex 200) in 20 mM Tris, pH 7.9, 150 mM NaCl, 0.25 mM TCEP. The AvrB elution peak from size exclusion was pooled and buffer exchanged into 20 mM Tris, pH 7.9, 50 mM NaCl, 0.25 mM TCEP, and concentrated to ⵑ9.5 mg/ml prior to crystallization trials by vapor diffusion. Structure Determination Crystals grew as hexagonal rods (space group P65, cell constants a ⫽ b ⫽ 125.03 A˚, c ⫽ 63.25 A˚) at 4⬚C from 40% PEG 600, 100 mM CHES, pH 9.5–10.0, by hanging drop vapor diffusion. Seleno-Lmethionine (SeMet)-labeled protein was produced in E. coli B834 (DE3) pLysS according to published methods (LeMaster and Richards, 1985). The labeled protein was purified and crystallized under the same conditions used for the native protein, and derivative crystals obtained were isomorphous. Native data from a single crystal were collected to ⵑ2.2 A˚ on beamline 5.03 at ALS (Berkeley, CA) using an ADSC Quantum 4 CCD detector. Prior to data collection, crystals were flash-frozen in a nitrogen cold stream at 100 K. Single wavelength anomalous dispersion (SAD) data from Se-incorporated crystals were collected on beamline 5.02 to ⵑ2.6 A˚ at the selenium K absorption edge (⫽ 0.97856) using an inverse beam strategy with 30⬚ wedges. Data were reduced using DENZO/SCALEPACK
Alfano, J.R., Charkowski, A.O., Deng, W.-L., Badel, J.L., PetnickiOcwieja, T., van Dijk, K., and Collmer, A. (2000). The Pseudomonas syringae Hrp pathogenicity island has a tripartite mosaic structure composed of a cluster of type III secretion genes bounded by exchangeable effector and conserved effector loci that contribute to parasitic fitness and pathogenicity in plants. Proc. Natl. Acad. Sci. USA 97, 4856–4861. Altschul, S.F., Madden, T.L., Scha˚ffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. Axtell, M.J., and Staskawicz, B.J. (2003). Initiation of RPS2-specified disease resistance in Arabidopsis is coupled to the AvrRpt2-directed elimination of RIN4. Cell 112, 369–377. Bonas, U., and Lahaye, T. (2002). Plant disease resistance triggered by pathogen-derived molecules: refined models of specific recognition. Curr. Opin. Microbiol. 5, 44–50. Boyes, D.C., Nam, J., and Dangl, J.L. (1998). The Arabidopsis thaliana RPM1 disease resistance gene product is a peripheral plasma membrane protein that is degraded coincident with the hypersensitive response. Proc. Natl. Acad. Sci. USA 95, 15849–15854. Bru¨nger, A.T., Adams, P.D., Clore, G.M., Delano, W.L., Gros, P., Grosse-Kuntsleve, R.W., Jiang, J.-S., Kuszewski, J., Nilges, N., Pannu, N.S., et al. (1998). Crystallography and NMR system (CNS): a new software system for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921. da Silva, A.C.R., Ferro, J.A., Reinach, F.C., Farah, C.S., Furlan, L.R.,
Structure 494
Quaggio, R.B., Monteiro-Vitorello, C.B., Van Sluys, M.A., Almeida, N.F., Jr., Alves, L.M.C., et al. (2002). Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417, 459–463. Dangl, J.L., and Jones, J.D.G. (2001). Plant pathogens and integrated defense responses to infection. Nature 411, 826–833. de La Fortelle, E., and Bricogne, G. (1997). Maximum-likelihood heavy-atom parameter refinement in the MIR and MAD methods. Methods Enzymol. 276, 472–494. Gala´n, J.E., and Collmer, A. (1999). Type III secretion machines: bacterial devices for protein delivery into host cells. Science 284, 1322–1328. Grant, M.R., Godiard, L., Straube, E., Ashfield, T., Lewald, J., Sattler, A., Innes, R.W., and Dangl, J.L. (1995). Structure of the Arabidopsis RPM1 gene enabling dual specificity disease resistance. Science 269, 843–846. Holm, L., and Sander, C. (1993). Protein structure comparison by alignment of distance matrices. J. Mol. Biol. 233, 123–138. Hutchinson, E.G., and Thornton, J.M. (1996). PROMOTIF: program to identify and analyze structural motifs in proteins. Protein Sci. 5, 212–220. Jones, T.A., Zou, J.-Y., Cowan, S.W., and Kjeldgaard, M. (1991). Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, 905–921. Juris, S.J., Rudolph, A.E., Huddler, D., Orth, K., and Dixon, J.E. (2000). A distinctive role for the Yersinia protein kinase: actin binding, kinase activation, and cytoskeleton disruption. Proc. Natl. Acad. Sci. USA 97, 9431–9436. Leister, R.T., and Katagiri, F. (2000). A resistance gene product of the nucleotide binding site-leucine rich repeats class can form a complex with bacterial avirulence proteins in vivo. Plant J. 22, 345–354. Leister, R.T., Ausubel, F.M., and Katagiri, F. (1996). Molecular recognition of pathogen attack occurs inside of plant cells in plant disease resistance specified by the Arabidopsis genes RPS2 and RPM1. Proc. Natl. Acad. Sci. USA 93, 15497–15502. LeMaster, D.M., and Richards, F.M. (1985). 1H-15N heteronuclear NMR studies of Escherichia coli thioredoxin in samples isotopically labeled by residue type. Biochemistry 24, 7263–7268. Liang, J., Edelsbrunner, H., and Woodward, C. (1998). Anatomy of protein pockets and cavities: measurement of binding site geometry and implications for ligand design. Protein Sci. 7, 1884–1897. Mackey, D., Holt, B.F., III, Wiig, A., and Dangl, J.L. (2002). RIN4 interacts with Pseudomonas syringae type III effector molecules and is required for RPM1-mediated resistance in Arabidopsis. Cell 108, 743–754. Mackey, D., Belkhadir, Y., Alonso, J.M., Ecker, J.R., and Dangl, J.L. (2003). Arabidopsis RIN4 is a target of the type III virulence effector AvrRpt2 and modulates RPS2-mediated resistance. Cell 112, 379–389. Murshudov, G., Vagin, A.A., and Dodson, E.J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53, 240–255. Nimchuk, Z., Marois, E., Kjemtrup, S., Leister, R.T., Katagiri, F., and Dangl, J.L. (2000). Eukaryotic fatty acylation drives plasma membrane targeting and enhances function of several type III effector proteins from Pseudomonas syringae. Cell 101, 353–363. Otwinowski, Z.M., W. (1997). Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–326. Stark, A., Sunyaev, S., and Russell, R.B. (2003). A model for statistical significance of local similarities in structure. J. Mol. Biol. 326, 1307– 1316. Tamaki, S., Dahlbeck, D., Staskawicz, B., and Keen, N.T. (1988). Characterization and expression of two avirulence genes cloned from Pseudomonas syringae pv. glycinea. J. Bacteriol. 170, 4846– 4854. Tamaki, S.J., Kobayashi, D.Y., and Keen, N.T. (1991). Sequence
domains required for the activity of avirulence genes avrB and avrC from Pseudomonas syringae pv. glycinea. J. Bacteriol. 173, 301–307. Tao, Y., Yuan, F., Leister, R.T., Ausubel, F.M., and Katagiri, F. (2000). Mutational analysis of the Arabidopsis nucleotide binding site-leucine-rich repeat resistance gene RPS2. Plant Cell 12, 2541–2554. Van Der Biezen, E.A., and Jones, J.D.G. (1998). Plant disease-resistance proteins and the gene-for-gene concept. Trends Biochem. Sci. 23, 454–456. Wu, Y., Wood, M.D., Tao, Y., and Katagiri, F. (2003). Direct delivery of bacterial avirulence proteins into resistant Arabidopsis protoplasts leads to hypersensitive cell death. Plant J. 33, 131–137. Yucel, I., Slaymaker, D., Boyd, C., Murillo, J., Buzzell, R.I., and Keen, N.T. (1994). Avirulence gene avrPphC from Pseudomonas syringae pv. phaseolicola 3121: a plasmid-borne homologue of avrC closely linked to an avrD allele. Mol. Plant Microbe Interact. 7, 677–679. Accession Numbers Coordinates for the AvrB model were deposited in the Protein Data Bank with PDB code 1NH1.
DOI 10.1016/j.str.2004.02.013
Crystal Structure of the Type III Effector AvrB from Pseudomonas syringae Christian C. Lee,1,4,* Michelle D. Wood,2,3,4,5 Kenneth Ng,1 Carsten B. Andersen,1 Yi Liu,1 Peter Luginbu¨hl,2,6 Glen Spraggon,1 and Fumiaki Katagiri 2,7 1 Genomics Institute of the Novartis Research Foundation 10675 John Jay Hopkins Drive San Diego, California 92121 2 Torrey Mesa Research Institute and Syngenta Research and Technology 3115 Merryfield Row San Diego, California 92121 3 Department of Biological Sciences University of Maryland Baltimore County 1000 Hilltop Circle Baltimore, Maryland 21250
Summary AvrB is a Pseudomonas syringae type III effector protein that is translocated into host plant cells during attempted pathogenesis. Arabidopsis harboring the corresponding resistance protein RPM1 can detect AvrB and mount a rapid host defense response, thus avoiding active infection. In the plant cell, AvrB induces phosphorylation of RIN4, a key component in AvrB/ RPM1 recognition. Although the AvrB/RPM1 system is among the best characterized of the numerous bacterial effector/plant resistance protein systems involved in plant disease resistance and pathogenesis, the details of the molecular recognition mechanism are still unclear. To gain further insights, the crystal structure of AvrB was determined. The 2.2 A˚ structure exhibits a novel mixed ␣/ bilobal fold. Aided by the structural information, we demonstrate that one lobe is the determinant of AvrB/RPM1 recognition specificity. This structural information and preliminary structure-function studies provide a framework for the future understanding of AvrB function on the molecular level. Introduction Many Gram-negative bacterial pathogens of plants and animals use the type III secretion system to deliver protein factors into host cells (Gala´n and Collmer, 1999). Although no underlying common mechanistic themes appear to exist, these type III effectors generally contribute to bacterial virulence by perturbing host cellular *Correspondence: [email protected] 4 These authors contributed equally to this work. 5 Present address: 9939 Azuaga Street, H104, San Diego, California 92129. 6 Present address: Diversa Corporation, 4955 Directors Place, San Diego, California 92121. 7 Present address: Department of Plant Biology, University of Minnesota, 1445 Gortner Avenue, St. Paul, Minnesota 55108.
functions. Plants have evolved a defense mechanism that is rapidly induced upon specific recognition of certain pathogen factors, including some type III effectors. Plant resistance (R ) gene products are central in recognition of such pathogen factors, which are the direct or indirect products of avirulence (avr) genes of pathogens (Dangl and Jones, 2001). Thus, many type III effector genes of phytopathogenic bacteria were initially discovered and named as avr genes, such as avrB. The correspondence between an R gene and an avr gene is usually highly specific; hence this type of disease resistance is termed gene-for-gene resistance. It remains unclear whether the recognition event for gene-for-gene resistance involves direct interactions between plant R proteins and the corresponding Avr type III effector proteins. AvrB is one of several type III effectors of the phytopathogenic bacterium Pseudomonas syringae (Tamaki et al., 1988). Arabidopsis carrying the R gene, RPM1, are able to detect the presence of AvrB upon delivery inside the plant cell, rapidly induce defense mechanisms upon detection, and consequently are resistant to P. syringae strains expressing AvrB (Grant et al., 1995). AvrB is one of several P. syringae type III effector proteins that are myristoylated in planta. This AvrB modification is essential for its membrane localization, RPM1-mediated resistance, and an RPM1-independent cytotoxic effect (Nimchuk et al., 2000). The RPM1-independent cytotoxic effect may reflect a virulence function of AvrB. RPM1-mediated resistance to P. syringae expressing AvrB is known to require at least one other plant protein, RIN4 (RPM1-interacting protein) (Mackey et al., 2002). Like AvrB, both RPM1 and RIN4 localize to the plasma membrane (Boyes et al., 1998; Mackey et al., 2002). Furthermore, RIN4 interacts with both RPM1 and AvrB, suggesting that these three proteins may form a higher order complex upon delivery of AvrB into the plant cell, and that these interactions are critical for RPM1-mediated resistance (Mackey et al., 2002). When AvrB is delivered into the plant cell, phosphorylation of RIN4 occurs in an RPM1-independent manner (Mackey et al., 2002). However, it is not known whether RIN4 phosphorylation results from a direct interaction with AvrB or from an indirect mechanism. Recently, the guard hypothesis was postulated to explain many gene-for-gene resistance interactions (Bonas and Lahaye, 2002; Dangl and Jones, 2001; Van Der Biezen and Jones, 1998). According to this hypothesis, R proteins function to monitor and respond to changes in the state of virulence targets of Avr proteins. The notion that RPM1 responds to the AvrB-induced phosphorylation of RIN4 is consistent with this hypothesis. Another type III effector protein AvrRpt2, unrelated to AvrB, also appears to target RIN4 as a virulence mechanism, inducing the posttranscriptional elimination of RIN4 (Axtell and Staskawicz, 2003; Mackey et al., 2003). The Arabidopsis RPS2 gene product, which mediates resistance to P. syringae expressing AvrRpt2, is believed to respond to the disappearance of RIN4 as a signal to
Crystal Structure of the Type III Effector AvrB 489
initiate plant defense responses. Thus, RIN4 appears to represent a common virulence target that is “guarded” by at least two different R gene products in Arabidopsis. Here, we report the three-dimensional crystal structure of AvrB. AvrB assumes an apparently novel ␣/ bilobal fold, with no significant structural homologs in the Protein Data Bank (PDB). A functional analysis of chimeras composed of AvrB and an AvrB homolog, AvrPphC, identified one of the lobes as the specificity determinant in the AvrB/RPM1 system. A simple explanation for the fact that introduction of AvrB into a plant cell induces RIN4 phosphorylation is that AvrB is a RIN4 protein kinase. However, AvrB lacks sequence and functional motifs common to canonical protein kinase catalytic domains. Thus, our structural and preliminary functional studies suggest either AvrB does not function as a kinase and is not involved in the direct phosphorylation of RIN4 in the AvrB/RPM1/RIN4 system or alternatively represents a protein kinase domain fold previously undescribed. Results Crystal Structure of AvrB Reveals a Novel Fold In efforts to gain insight into the molecular details behind AvrB function and host recognition, the crystal structure was solved to a resolution of 2.2 A˚. The structure reveals a bilobal fold with a smaller mixed ␣/ lobe and a larger predominantly helical lobe separated by a deep interlobe cleft (Figure 1A). We refer to the two lobes as the upper (amino acid residues 123–217) and lower (residues 28–122 and 218–317) lobes, respectively. The smaller upper lobe consists of three helices (␣5, ␣6, and ␣7) surrounding a central five-stranded antiparallel  sheet that separates the three upper lobe helices from the helices of the lower lobe. The adjoining helices ␣6 and ␣7 are oriented approximately 50⬚ relative to each other and extend from strands 4 and 5 of the central  sheet. Helix ␣5 is positioned between strands 1 and 2 and packs perpendicularly against the midportion of the  sheet. The strand order for the antiparallel  sheet is 1-5-4-3-2 with strands 2, 3, and 4 forming two long opposing  turns. The lower lobe consists of a major five helix bundle comprised of helices ␣2, ␣3, ␣8, ␣9, and ␣10 with four shorter flanking helices (␣1, ␣4, ␣11, and ␣12). The two lobes are joined by two segments forming a hinge-like region, the first between helix ␣4 and strand 1, and the second between strand 5 and helix ␣8. The dimensions of this large, solventaccessible interlobe cleft are approximately 24 A˚ wide and 18 A˚ in length, which corresponds to over 900 A˚3 in volume as determined by CastP (Liang et al., 1998). Dali searches (Holm and Sander, 1993) of the PDB using
the AvrB structure model as a whole and as independent lobes found no existing structural homologs, suggesting that AvrB assumes a unique fold (Figure 1B). Molecular and Electrostatic Surface Examination of the electrostatic surface potential of AvrB mapped onto the molecular surface reveals a striking degree of negative surface charge that encompasses a vast majority of the molecular surface (Figure 1C). This extensive surface electronegativity may reflect interactions with RPM1 and/or RIN4 are largely electrostatic in nature. A localization of more basic electrostatic surface appears to adorn the mouth of the interlobe cavity. Ignoring any clues implicating AvrB function from existing biological data, the sheer breadth of the interlobe cavity may suggest AvrB functions in a yet undetermined enzymatic capacity. This speculation is especially tempting when sequence conservation among other AvrB family members is taken into consideration in the context of the AvrB structure. The Interlobe Cleft May Be Important for an Enzymatic Function Common among AvrB Family Members. AvrB homologs from three pathovars of P. syringae and one pathovar of Xanthomonas campestris were identified by a BLAST search of the GenBank protein sequence database (Altschul et al., 1997). The amino acid sequence alignment of AvrB and its homologs reveals 30%–44% sequence identity and 47%–58% similarity when conservative substitutions are considered (Figure 2). Examination of the multiple sequence alignment among AvrB family members reveals several contiguous stretches of amino acid conservation that correspond to helical and surrounding regions of helices ␣1–␣3 and helices ␣8–␣12 (Figure 2). When mapped onto the threedimensional structure of the AvrB model, these stretches of high degree of sequence conservation localize predominantly to the lower lobe (Figures 2 and 3A). Moreover, many of the conserved solvent-accessible residues that do not appear to play a structural role in maintaining the integrity of the AvrB fold appear to localize in the interlobe cleft. This phylogenetically conserved deep cleft may reflect a common enzymatic active site responsible for substrate and/or cofactor binding of AvrB family members. The AvrB Upper Lobe Is Important in Specificity Determination for RPM1-Mediated Resistance Guided by the AvrB crystal structure, a series of chimeric proteins consisting of AvrB and a homolog, AvrPphC, were generated to delineate the portion of AvrB that determines its specificity in Arabidopsis RPM1-medi-
Figure 1. AvrB Crystal Structure (A) Ribbon cartoon diagram of the AvrB model shown in two orthogonal views. Residues 28–317 are shown. Residues 28 (“N”) and 317 (“C”) along with successive secondary structural elements are labeled. The three helices that are part of the upper lobe are depicted in blue. The five-stranded antiparallel  sheet connecting the helices of the upper lobe and the lower lobe is shown in orange. The remaining eight helices constituting the lower lobe are colored green. (B) Topology diagram of the AvrB model. Circles represent helices and triangles  strands and are colored as in (A). (C) Electrostatic potential surface representation of AvrB. Electropositive, blue; electronegative, red. The central and right orientations correspond to those of (A), whereas the leftmost orientation represents another orthogonal view.
Structure 490
Figure 2. Amino Acid Sequence Alignment of AvrB Family Members AvrB (GenBank accession, P13835) (Tamaki et al., 1988) and AvrC (P13836) (Tamaki et al., 1988) originate from P. syringae pv. glycinea. AvrPphC (AAA83419) (Yucel et al., 1994) is from P. syringae pv. phaseolicola. Avr_Xcc (NP_637473) (da Silva et al., 2002) represents a homolog from Xanthomonas campestris pv. campestris. Avr_Pss (AAF71496) (Alfano et al., 2000) represents a homolog from P. syringae pv. syringae. Invariant amino acid residues among all the sequences are highlighted in red with yellow lettering. Residues invariant in four of five sequences in the alignment are highlighted in blue with yellow lettering. Conserved residues based upon amino acid size and charge among four of the five sequences are highlighted in green with yellow lettering. Helical and  strand secondary structural elements according to the AvrB structure are represented by blue bars and purple arrows, respectively. They are numbered according to Figure 1A.
ated recognition (Figure 2). Previously, Tamaki et al. reported a similar chimera study consisting of AvrB and AvrC, which is essentially identical to AvrPphC (Figure 2), to identify the specificity-determining region of AvrB in resistance mediated by the soybean R gene Rpg1 (Tamaki et al., 1991). In that study, however, chimeras were designed according to ease of construction because the AvrB three-dimensional structure was not known and because it predated PCR. In our studies, we designed one chimera, B54-225, based on the smallest AvrB region Tamaki et al. identified for recognition by Rpg1. The constructs were named according to the amino acid residues from AvrB that replaced the corresponding region of AvrPphC. For example, B54-225 was generated in an AvrPphC background with amino acid residues 54–225 of AvrB replacing the corresponding
region of AvrPphC. Construct B54-225 was chosen as a starting chimera because it was anticipated that recognition of AvrB by Arabidopsis RPM1 and soybean Rpg1 were similar. Three additional chimeras, B102123, B102-216, and B126-216, were designed to determine the minimal region of AvrB required to elicit RPM1mediated resistance and at the same time to minimize any potential structural perturbation as predicted by the AvrB structure. The functional assay for RPM1-mediated resistance was based on the hypersensitive cell death that occurs pursuant to “gene-for-gene” recognition, and the consequential reduction of reporter activity (Leister et al., 1996). When AvrB was coexpressed with the renilla luciferase (rLUC) reporter in plants, a strong reduction of rLUC activity was observed in a manner dependent upon
Crystal Structure of the Type III Effector AvrB 491
Figure 3. Stereoviews of C␣ Backbone Trace of AvrB Model and Representative Electron Density (A) Stereoview of the C␣ backbone trace of AvrB in a similar orientation as in Figure 1A. Amino acid residues are indicated using their single letter codes and numbered as labeled. Residues conserved among AvrB family members are colored red and light blue. Red, invariant; light blue, conserved; yellow, unconserved. (B) Stereoview of representative 2fo-fc electron density contoured at 1 . Electron density is shown in purple. The salt bridge between Arg137 of helix ␣5 and Glu299 of the interconnecting loop between helix ␣1 and ␣2 is shown along with neighboring residues.
coexpression of RPM1 (Figure 4). Expression of RPM1 alone moderately reduces rLUC activity (compare ⫹RPM1 and ⫺RPM1 with Vector). This may be due to a weak basal activity of RPM1 in the absence of specific stimulation. Similar but stronger basal activity was observed with another Arabidopsis resistance protein RPS2 in this assay (Tao et al., 2000). Nevertheless, reduction of rLUC activity in the presence of both AvrB and RPM1 was clearly greater than its reduction in the presence of RPM1 alone, and, therefore, there is no ambiguity in interpretation of the results presented in Figure 4. In contrast to AvrB, AvrPphC expression did not result in a strong reduction of rLUC activity, regardless of the presence of RPM1. These results indicate
that AvrB, but not AvrPphC, was able to induce RPM1mediated resistance. B54-225, B102-216, and B126-216 functioned like AvrB, while B102-123 did not. The B126216 chimera represents a precise substitution of the upper lobe of AvrPphC with AvrB. Based on immunoblot analysis, AvrPphC and the four chimeras were all able to accumulate in rpm1 mutant cells to a level comparable to that of AvrB (data not shown). Therefore, a failure of RPM1-mediated resistance is not attributable to decreased expression of the chimera in planta. We conclude that the region of AvrB involved in eliciting an RPM1-mediated resistance response maps within the upper lobe. The amino acid region of AvrB from 54–225 was previously shown to determine the specificity to
Structure 492
Figure 4. The Region Determining Specificity to RPM1 Is Mapped to the Upper Lobe of AvrB The RPM1-mediated recognition of chimeric proteins B54-225, B102-123, B102-216, and B126-216, was detected by a large reduction of rLUC reporter activity. Each bar represents the mean value of four bombardment events (error bar ⫽ standard deviation). This experiment was repeated with similar results.
the soybean R protein Rpg1 (Tamaki et al., 1991). Since this amino acid region encompasses the entire upper lobe of AvrB, the determinant of specificity to Rpg1 may also map within the upper lobe. Our finding that the AvrB upper lobe is involved in recognition by the RPM1mediated resistance response is especially poignant when amino acid conservation among AvrB family members is taken into account. The overall limited sequence conservation among AvrB family members mapping to the upper lobe may further support this finding that this region is involved in specific molecular recognition by the plant R gene products. Discussion AvrB induces phosphorylation of RIN4 in Arabidopsis plant cells (Mackey et al., 2002). The simplest hypothesis for this biological phenomenon is that AvrB directly phosphorylates RIN4. The fact that a Dali search of the PDB with the AvrB coordinates did not yield any similar folds indicates that AvrB does not possess a fold similar to those of known protein kinases. Similarly, structural analysis using programs such as PINTS and PROMOTIF to detect local structural similarities and functional motifs were unsuccessful in finding any similarities with AvrB (Stark et al. 2003; Hutchinson and Thornton, 1996) and did not afford any insights to AvrB function. These findings suggest that AvrB may not be a protein kinase or if it is a protein kinase it is one that assumes a fold previously undescribed for proteins capable of phosphotransfer to protein substrates. Nonetheless, we attempted to obtain experimental biochemical evidence in support of AvrB protein kinase function. These efforts were unfruitful in that we failed to detect in vitro phosphorylation of nonspecific substrates, including myelin basic protein, casein, histone H1, and poly-glutamate-tyrosine peptide, by AvrB (data not shown). Furthermore, we were unable to detect autophosphorylation or demonstrate ATP-coordination of AvrB in vitro (data not shown). Attempts at demonstrating AvrB binding to ATP via capture on ATP-sepharose resin were likewise unconvincing in our hands. Thus, we failed to detect any evidence that AvrB is a protein kinase or bind ATP by these conventional approaches. Attempts at obtaining an ATP-bound AvrB structure were also unsuccessful. However, it should be noted that many type III effectors of bacteria require host factors for their activation and/or modification. For example, the Yersinia pestis Ser/Thr protein kinase YpkA re-
quires actin for its activation (Juris et al., 2000). It is conceivable that AvrB may also require a host factor to be active. Taken together, our findings and preliminary functional studies suggest that AvrB is not a protein kinase as existing biochemical evidence may suggest, and if it does represent a protein kinase capable of direct phosphorylation of the RIN4 plant target protein, the structure and mechanism for this potential protein kinase is of a nature previously undescribed and warrants further biochemical investigation. Thus, the precise functional role of AvrB in RIN4 phosphorylation may involve a mechanism that does not involve direct phosphorylation of RIN4 by AvrB. In conclusion, we have successfully crystallized and solved the structure of AvrB from P. syringae. To our knowledge, this is the first crystal structure to be solved among type III effector proteins from phytopathogenic bacteria and represents a fold unique to AvrB. This bilobal mixed ␣/ fold is comprised of a smaller upper lobe and larger predominantly helical lower lobe separated by a deep interlobe cleft. Primary sequence and structural analysis of AvrB and family homologs suggest these effector proteins may possess a common enzymatic function involving the interlobe cleft region that may be related to virulence function. This information, in conjunction with our results from preliminary structurefunction studies of AvrB chimeras implicating the AvrB upper lobe as the RPM1-specificity determinant, provides initial insight to AvrB function on the molecular level. The crystal structure will undoubtedly continue to be useful as a scaffold for design and interpretation of future structure-function studies aimed at interrogating the precise function of AvrB as a type III effector and a more detailed understanding of the AvrB-RIN4-RPM1 bacterial phytopathogen/plant host defense system. Solving the AvrB crystal structure is a first step toward elucidating the AvrB/RPM1 molecular recognition mechanism. A more comprehensive understanding will be achieved when the crystal structures of other components involved in the recognition both alone and in complex with AvrB, such as RPM1 and RIN4, are determined. We believe that further appreciation of the three-dimensional structure of this type III effector protein will provide insights that lead to novel intervention strategies to control crop diseases. Experimental Procedures AvrB Expression and Purification The full-length C-terminally hexahistidine-tagged AvrB protein was cloned, expressed, and purified from E. coli and found to function
Crystal Structure of the Type III Effector AvrB 493
Table 1. Crystallographic Data Data Collection Data Set
Native
SeMet
Space group Cell a ⫽ b (A˚) c (A˚) ␣ ⫽  (⬚) ␥ (⬚) Wavelength (A˚) Resolution (A˚) Completeness (%) Rsymm I/ Number of sites
P65
P65
125.028 63.246 90 120 1.00 50–2.2 93.4 (64.6) 0.066 (0.493) 13.0
125.75 63.90 90 120 0.97840 20–2.65 99.4 (96.5) 0.136 (0.808) 8.1 4
Refinement Resolution Range Number of reflections (working/free) Number of atoms Protein Solvent Rwork/Rfree Rms deviation from ideality Bond lengths (A˚) Bond angles (⬚) Number of residues (monomer/asymmetric unit)
500–2.2 26,473/1,419
2,254 189 22.2/27.5 0.0104 1.54 289/331
Functional Assay of Chimeras C-terminally FLAG-tagged AvrB, AvrPphC, and chimeric constructs were transiently expressed in Arabidopsis leaf tissue, essentially as described (Leister et al., 1996). Because the response mediated by RPM1 is very strong in this assay, normalization of the transformation efficiency was not necessary and was omitted. To assay protein accumulation, the same constructs were transiently expressed, along with a firefly luciferase (LUC) reporter, in rpm1 mutant Arabidopsis protoplasts as described (Leister and Katagiri, 2000). The cell lysates were assayed for LUC activity in order to normalize the transformation efficiency prior to sample loading. Samples were then subjected to SDS-PAGE and immunoblot analysis. Acknowledgments
Ramachandran analysis (%) Most favored Additionally allowed Generously allowed Disallowed Average B factors (main chain/side chain) (A˚2)
(Otwinowski, 1997). Native diffraction data (Rmerge, 6.6%) are 85.6% complete to 2.2 A˚. Derivative data (Rmerge, 13.6%) are complete to 99.4% overall. Initial phases and substructure for four of the five Se sites were calculated using SOLVE and SHARP giving a mean figure of merit of 0.328 to 3 A˚ (de La Fortelle and Bricogne, 1997). Phases were refined and extended to 2.6 A˚ resolution using RESOLVE. Iterative cycles of model building and refinement were performed using “O” (Jones et al., 1991) and Refmac (Murshudov et al., 1997). Final rounds of refinement were conducted using CNS (Bru¨nger et al., 1998). The R factor and free R factor converged at 22.2% and 27.5%, respectively. The final AvrB model consists of 290 amino acids (amino acids 28–317) and 189 water molecules. The electron density for the first 27 amino acids was disordered and uninterpretable in 2fo-fc electron density maps. The overall geometry and stereochemistry of the final model was excellent with no residues in disallowed regions as analyzed using PROCHECK. Data and refinement statistics are summarized in Table 1.
88.9 10.7 0.4 0.0 54.6/58.1
Numbers in parentheses indicate values for the highest resolution shell.
We thank Paul Haynes, Donna Eckert, and Dave Schieltz for verifying SeMet incorporation into the protein and Jane Glazebrook (TMRI), Scott Lesley (GNF), and Michael DiDonato (GNF) for critical reading of the manuscript. The Torrey Mesa Research Institute has been closed. Received: June 20, 2003 Revised: November 12, 2003 Accepted: November 28, 2003 Published: March 9, 2004 References
in Arabidopsis protoplasts as described (Wu et al., 2003). Briefly, nickel affinity chromatography fractions containing partially purified recombinant AvrB were pooled and buffer exchanged into 20 mM Tris-HCl, pH 7.9, 50 mM NaCl, and 0.25 mM TCEP prior to anionexchange chromatography (Resource-Q) using a 0–0.5 M linear salt gradient over 15 column volumes. Peak fractions containing recombinant AvrB were pooled and concentrated followed by size exclusion chromatography (HiLoad 16/60 Superdex 200) in 20 mM Tris, pH 7.9, 150 mM NaCl, 0.25 mM TCEP. The AvrB elution peak from size exclusion was pooled and buffer exchanged into 20 mM Tris, pH 7.9, 50 mM NaCl, 0.25 mM TCEP, and concentrated to ⵑ9.5 mg/ml prior to crystallization trials by vapor diffusion. Structure Determination Crystals grew as hexagonal rods (space group P65, cell constants a ⫽ b ⫽ 125.03 A˚, c ⫽ 63.25 A˚) at 4⬚C from 40% PEG 600, 100 mM CHES, pH 9.5–10.0, by hanging drop vapor diffusion. Seleno-Lmethionine (SeMet)-labeled protein was produced in E. coli B834 (DE3) pLysS according to published methods (LeMaster and Richards, 1985). The labeled protein was purified and crystallized under the same conditions used for the native protein, and derivative crystals obtained were isomorphous. Native data from a single crystal were collected to ⵑ2.2 A˚ on beamline 5.03 at ALS (Berkeley, CA) using an ADSC Quantum 4 CCD detector. Prior to data collection, crystals were flash-frozen in a nitrogen cold stream at 100 K. Single wavelength anomalous dispersion (SAD) data from Se-incorporated crystals were collected on beamline 5.02 to ⵑ2.6 A˚ at the selenium K absorption edge (⫽ 0.97856) using an inverse beam strategy with 30⬚ wedges. Data were reduced using DENZO/SCALEPACK
Alfano, J.R., Charkowski, A.O., Deng, W.-L., Badel, J.L., PetnickiOcwieja, T., van Dijk, K., and Collmer, A. (2000). The Pseudomonas syringae Hrp pathogenicity island has a tripartite mosaic structure composed of a cluster of type III secretion genes bounded by exchangeable effector and conserved effector loci that contribute to parasitic fitness and pathogenicity in plants. Proc. Natl. Acad. Sci. USA 97, 4856–4861. Altschul, S.F., Madden, T.L., Scha˚ffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. Axtell, M.J., and Staskawicz, B.J. (2003). Initiation of RPS2-specified disease resistance in Arabidopsis is coupled to the AvrRpt2-directed elimination of RIN4. Cell 112, 369–377. Bonas, U., and Lahaye, T. (2002). Plant disease resistance triggered by pathogen-derived molecules: refined models of specific recognition. Curr. Opin. Microbiol. 5, 44–50. Boyes, D.C., Nam, J., and Dangl, J.L. (1998). The Arabidopsis thaliana RPM1 disease resistance gene product is a peripheral plasma membrane protein that is degraded coincident with the hypersensitive response. Proc. Natl. Acad. Sci. USA 95, 15849–15854. Bru¨nger, A.T., Adams, P.D., Clore, G.M., Delano, W.L., Gros, P., Grosse-Kuntsleve, R.W., Jiang, J.-S., Kuszewski, J., Nilges, N., Pannu, N.S., et al. (1998). Crystallography and NMR system (CNS): a new software system for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921. da Silva, A.C.R., Ferro, J.A., Reinach, F.C., Farah, C.S., Furlan, L.R.,
Structure 494
Quaggio, R.B., Monteiro-Vitorello, C.B., Van Sluys, M.A., Almeida, N.F., Jr., Alves, L.M.C., et al. (2002). Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417, 459–463. Dangl, J.L., and Jones, J.D.G. (2001). Plant pathogens and integrated defense responses to infection. Nature 411, 826–833. de La Fortelle, E., and Bricogne, G. (1997). Maximum-likelihood heavy-atom parameter refinement in the MIR and MAD methods. Methods Enzymol. 276, 472–494. Gala´n, J.E., and Collmer, A. (1999). Type III secretion machines: bacterial devices for protein delivery into host cells. Science 284, 1322–1328. Grant, M.R., Godiard, L., Straube, E., Ashfield, T., Lewald, J., Sattler, A., Innes, R.W., and Dangl, J.L. (1995). Structure of the Arabidopsis RPM1 gene enabling dual specificity disease resistance. Science 269, 843–846. Holm, L., and Sander, C. (1993). Protein structure comparison by alignment of distance matrices. J. Mol. Biol. 233, 123–138. Hutchinson, E.G., and Thornton, J.M. (1996). PROMOTIF: program to identify and analyze structural motifs in proteins. Protein Sci. 5, 212–220. Jones, T.A., Zou, J.-Y., Cowan, S.W., and Kjeldgaard, M. (1991). Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, 905–921. Juris, S.J., Rudolph, A.E., Huddler, D., Orth, K., and Dixon, J.E. (2000). A distinctive role for the Yersinia protein kinase: actin binding, kinase activation, and cytoskeleton disruption. Proc. Natl. Acad. Sci. USA 97, 9431–9436. Leister, R.T., and Katagiri, F. (2000). A resistance gene product of the nucleotide binding site-leucine rich repeats class can form a complex with bacterial avirulence proteins in vivo. Plant J. 22, 345–354. Leister, R.T., Ausubel, F.M., and Katagiri, F. (1996). Molecular recognition of pathogen attack occurs inside of plant cells in plant disease resistance specified by the Arabidopsis genes RPS2 and RPM1. Proc. Natl. Acad. Sci. USA 93, 15497–15502. LeMaster, D.M., and Richards, F.M. (1985). 1H-15N heteronuclear NMR studies of Escherichia coli thioredoxin in samples isotopically labeled by residue type. Biochemistry 24, 7263–7268. Liang, J., Edelsbrunner, H., and Woodward, C. (1998). Anatomy of protein pockets and cavities: measurement of binding site geometry and implications for ligand design. Protein Sci. 7, 1884–1897. Mackey, D., Holt, B.F., III, Wiig, A., and Dangl, J.L. (2002). RIN4 interacts with Pseudomonas syringae type III effector molecules and is required for RPM1-mediated resistance in Arabidopsis. Cell 108, 743–754. Mackey, D., Belkhadir, Y., Alonso, J.M., Ecker, J.R., and Dangl, J.L. (2003). Arabidopsis RIN4 is a target of the type III virulence effector AvrRpt2 and modulates RPS2-mediated resistance. Cell 112, 379–389. Murshudov, G., Vagin, A.A., and Dodson, E.J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53, 240–255. Nimchuk, Z., Marois, E., Kjemtrup, S., Leister, R.T., Katagiri, F., and Dangl, J.L. (2000). Eukaryotic fatty acylation drives plasma membrane targeting and enhances function of several type III effector proteins from Pseudomonas syringae. Cell 101, 353–363. Otwinowski, Z.M., W. (1997). Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–326. Stark, A., Sunyaev, S., and Russell, R.B. (2003). A model for statistical significance of local similarities in structure. J. Mol. Biol. 326, 1307– 1316. Tamaki, S., Dahlbeck, D., Staskawicz, B., and Keen, N.T. (1988). Characterization and expression of two avirulence genes cloned from Pseudomonas syringae pv. glycinea. J. Bacteriol. 170, 4846– 4854. Tamaki, S.J., Kobayashi, D.Y., and Keen, N.T. (1991). Sequence
domains required for the activity of avirulence genes avrB and avrC from Pseudomonas syringae pv. glycinea. J. Bacteriol. 173, 301–307. Tao, Y., Yuan, F., Leister, R.T., Ausubel, F.M., and Katagiri, F. (2000). Mutational analysis of the Arabidopsis nucleotide binding site-leucine-rich repeat resistance gene RPS2. Plant Cell 12, 2541–2554. Van Der Biezen, E.A., and Jones, J.D.G. (1998). Plant disease-resistance proteins and the gene-for-gene concept. Trends Biochem. Sci. 23, 454–456. Wu, Y., Wood, M.D., Tao, Y., and Katagiri, F. (2003). Direct delivery of bacterial avirulence proteins into resistant Arabidopsis protoplasts leads to hypersensitive cell death. Plant J. 33, 131–137. Yucel, I., Slaymaker, D., Boyd, C., Murillo, J., Buzzell, R.I., and Keen, N.T. (1994). Avirulence gene avrPphC from Pseudomonas syringae pv. phaseolicola 3121: a plasmid-borne homologue of avrC closely linked to an avrD allele. Mol. Plant Microbe Interact. 7, 677–679. Accession Numbers Coordinates for the AvrB model were deposited in the Protein Data Bank with PDB code 1NH1.