Figure 1.
Structural model of the AvrBs2 phosphodiesterase (GDE) domain and in vitro GDE activity.
A. CLUSTALW alignment of the GDE domain shared by Xe AvrBs2 (amino acids 274 to 328) and its homologs from Xanthomonas pathogens of tomato, euvesicatoria (Xe); alfalfa, campestris pv. alfalfa (Xca); citrus, axonopodis pv. citri (Xac); cabbage, campestris pv. campestris (Xcc); and rice, oryzae pv. oryzae (Xoo) and oryzae pv. oryzicola (Xoc). Also includes selected GDE (orputative GDE) proteins from planta (AtGDE and OsGDE), animala (HsMIR16), fungi (ScGDE1), and other bacteria (Tm GDPD, BhGlpQ and AgtACS). The conserved putative catalytic sites of the human GDE, HsMIR16 are in red. B. The three-dimensional structural model for AvrBs2 (amino acids 274 to328) using the solved crystal structure of TmGPDO (1o1z A) as a template. TmGPDO 1o1z A is displayed in gray, and AvrBs2 is in green. The GDE catalytic sites ExD and H are blue in 1o1z A and red in AvrBs2. Shown in yellow are the amino acid mutations in the AvrBs2 putative catalytic sites and three additional sites that disrupt AvrBs2 activation of Bs2. C. Multiple sequence alignment of a 23 amino acid area of the GDE, BhGlpQ, catalytic site (amino acid 61–84) for two chimeric BhGlpQ proteins with the underlined sequences of the GDE catalytic site from either wild-type AvrBs2or AvrBs2 GDE mutants. D. In vitro GDE activity. The GDE enzyme activity of purified GST:BhGlpQ (positive control), GST:BhGlpQ-AvrBs2-WT, and BhGlpQ-AvrBs2-E304A/D306A were analyzed using an indirect coupled enzyme assay. A higher absorbance indicates increased GDE enzyme activity.
Figure 2.
The GDE catalytic sites are required for AvrBs2 virulence function.
A. The map of the AvrBs2 coding region with numbers representing amino acid positions. AvrBs2 mutations for the GDE catalytic site and the AvrBs2 mutations from Xe field strains isolated from diseased BS2 pepper. B. In planta pathogen growth assay for Xanthomonas strains GM98-38 Xe (w/o avrBs2), GM 98-38-1 Xe (avrBs2), and GM98-38-1 exchange mutants for the putative GDE catalytic site Xe (avrBs2-E304A/D306A) and Xe (avrBS2-H319A) and two previously published control exchange mutants Xe (avrBs2-R403P) and Xe (avrBs2-A410E). Host plants include the near-isogenic pepper (w/o Bs2) and pepper (Bs2) along with the tomato line VF36 (w/o Bs2) and the previously published transgenic line VF36 (Bs2). Student t-test was used to compare different growth assays with the most virulent case (Xe (avrBs2) on non-Bs2 plants) for both Pepper and Tomato hosts; p-values were <0.01 for all other combination when compared to (Xe (avrBs2) on non-Bs2 plants). The Xe strains carrying AvrBs2 mutations for the GDE catalytic site still activate full Bs2 resistance but do not maintain the full virulence in the absence of Bs2. There is a corresponding wild type HR brown necrosis phenotype for Bs2 pepper inoculated with high-density suspensions (2×108 CFU/ml) for these two Xe mutant strains (Supplemental Figure S1). The Xe strains carrying AvrBs2 mutations from Xe field strains isolated from diseased Bs2 pepper grow to similar levels in the presence or absence of Bs2. There is a corresponding loss of the HR brown necrosis phenotype for Bs2 pepper inoculated with high-density suspensions of Xe avrBs2-R403P and only a weak HR brown necrosis phenotype for Xe avrBs2-A410E (Supplemental Figure S1). C. Plot of amino acid substitution rate analysis using a sliding window calculation of non-synonymous (KA) and synonymous (Ks) changes between avrBs2 homologs from Xe and Xcc. The KA/Ks ratios less than 0.5 indicate that much of AvrBs2 is under purifying selection, including the region homologous to GDE that is required for AvrBs2 virulence function. The numbers below the plot represent amino acid positions.
Figure 3.
Characterization of the minimal AvrBs2 domain required for Bs2 activation.
A. Various AvrBs2 deletions and mutations of the minimal AvrBs2 domain required for Bs2 activation are expressed by the 35S promoter transiently on Bs2 Nicotiana benthamiana by Agrobacterium (48 hpi at 3×108 CFU/ml). B. AvrBs2 coding region map with numbers representing amino acid positions. Deletion analysis defined AvrBs2 amino acids 271–520 as the minimal domain for Bs2-activation. In Supplemental Figure S3B we confirm the HR response of these AvrBs2 clones on Bs2 pepper. Also we detect similar protein expression for all clones using C-terminal HA epitope tags and immunoblot analysis (Supplemental Figure S3A). C. Mutational analysis of the C-terminal region of the minimal Bs2-HR activation domain. Single amino acid mutations identify amino acid Y419 as needed for BS2 activation. In Supplemental Figure S3B we confirm the HR response of these AvrBs2 clones on Bs2 pepper. Also we detect similar protein expression for all clones using C-terminal HA epitope tags and immunoblot analysis (Supplemental Figure S3A).
Figure 4.
The slower Bs2-HR (48 hpi) from high-density (1.5×108 CFU/ml) inoculation of Xe (avrBs1, avrBs2) strain was epistatic to the faster Bs1-HR (18 hpi) for pepper (Bs1, Bs2).
Near-isogenic pepper lines with bacterial spot resistance genes (Bs1, Bs2 and the combination of Bs1 and Bs2), at 18 and 48 hours post-inoculation (hpi) with the strains Xe (avrBs1), Xe (avrBs2) and Xe (avrBs1, avrBs2). When the strain Xe (avrBs1, avrBs2)) was inoculated on pepper (Bs1, Bs2) the fast Bs1/AvrBs1 HR was not detected at 18 hpi.
Figure 5.
Bs2 activation by AvrBs2 blocks TTSS delivery of two independent effector reporters to host cells.
The TTSS effector delivery reporter constructs consisted of the effector promoter and the secretion and translocation signal peptide translationally fused to adenylate cyclase (Cya). These constructs were introduced into Xe in tandem with the native effector by single homologous recombination. Pepper (no R genes), pepper (Bs1) and pepper (Bs2) were sampled 8 hours post-inoculation to avoid in planta multiplication of the reporter strain pairs (with and without AvrBs2) and assayed for cyclic AMP (cAMP). TTSS delivered effector-Cya translational fusions into the plant cell and calmodulin from the plant cell leads to elevated levels of cAMP. A. AvrBs11-212-Cya reporter in strains Xe (avrBs2 and avrBs1) and Xe (avrBs1) were inoculated into pepper plants (no R genes, Bs1 or Bs2). In planta cAMP levels were assayed. Student t-test was used to compare TTSS delivery of effector reporter in an Xe strain with and without avrBs2 on Pepper (+) Bs2; p-values were <0.05. B. XopX1-183-Cya reporter in strains Xe (avrBs2 and avrBs1) and Xe (avrBs1) were inoculated into pepper plants (no R genes, Bs1 or Bs2). In planta cAMP levels were assayed. Student t-test was used to compare TTSS delivery of effector reporter in an Xe strain with and without avrBs2 on Pepper (+) Bs2; p-values were <0.01.