A Xanthomonas transcription activator-like effector is trapped in nonhost plants for immunity

Abstract

Xanthomonas oryzae pv. oryzae (Xoo), the causal agent of bacterial leaf blight in rice, delivers transcription activator-like effector (TALE) proteins into host cells to activate susceptibility or resistance (R) genes that promote disease or immunity, respectively. Nonhost plants serve as potential reservoirs of R genes; consequently, nonhost R genes may trap TALEs to trigger an immune response. In this study, we screened 17 Xoo TALEs for their ability to induce a hypersensitive response (HR) in the nonhost plant Nicotiana benthamiana (Nb); only AvrXa10 elicited an HR when transiently expressed in Nb. The HR generated by AvrXa10 required both the central repeat region and the activation domain, suggesting a specific interaction between AvrXa10 and a potential R-like gene in nonhost plants. Evans blue staining and ion leakage measurements confirmed that the AvrXa10-triggered HR was a form of cell death, and the transient expression of AvrXa10 in Nb induced immune responses. Genes targeted by AvrXa10 in the Nb genome were identified by transcriptome profiling and prediction of effector binding sites. Using several approaches (in vivo reporter assays, electrophoretic mobility-shift assays, targeted designer TALEs, and on-spot gene silencing), we confirmed that AvrXa10 targets NbZnFP1, a C2H2-type zinc finger protein that resides in the nucleus. Functional analysis indicated that overexpression of NbZnFP1 and its rice orthologs triggered cell death in rice protoplasts. An NbZnFP1 ortholog was also identified in tomato and was specifically activated by AvrXa10. These results demonstrate that NbZnFP1 is a nonhost R gene that traps AvrXa10 to promote plant immunity in Nb.

Keywords: AvrXa10; Xanthomonas oryzae pv. oryzae; hypersensitive response; nonhost plant; zinc finger protein.

Figure 1

HR-like cell death induced by Xoo TALE AvrXa10 in N. benthamiana. (A) Diagram showing RVDs in AvrXa10 and the construction of mutants AvrXa10ΔCRR and AvrXa10ΔAD, which lack the central repeat region (CRR) and activation domain (AD), respectively. (B) Phenotypes of Nb leaves expressing AvrXa10, AvrXa10ΔCRR, and AvrXa10ΔAD. Agrobacterium strains containing constructs pHB-AvrXa10, pHB-AvrXa10ΔCRR, pHB-AvrXa10ΔAD, pHB-Hpa1 (positive control), and empty pHB were used to transform Nb leaves at OD₆₀₀ 1.0. Photographs were taken at 3 dpi. (C) Detection of AvrXa10, AvrXa10ΔCRR, and AvrXa10ΔAD expression in Nb leaves by western blotting. Actin was used as a loading control. (D) Detection of cell death in Nb leaves infiltrated with Agrobacterium containing AvrXa10, AvrXa10ΔCRR, AvrXa10ΔAD, and empty pHB by Evans blue staining. (E) Measurement of ion leakage in Nb leaves expressing AvrXa10, AvrXa10ΔCRR, AvrXa10ΔAD, or empty pHB vector. Error bars show means ± SD (n = 3), and columns labeled with an asterisk are significantly different (∗P < 0.01) from the pHB control. The experiment was repeated at least three times with similar results.

Figure 2

Experimental design for identification of candidate AvrXa10 target genes in N. benthamiana. (A) Schematic flowchart of experimental design. RNA was sequenced from Nb leaves transiently expressing AvrXa10 or AvrXa10ΔCRR at 32 and 48 hpi. Differentially regulated genes were identified at the two time points, and upregulated DEGs common to both time points were identified. The TALgetter computational tool was then used to predict AvrXa10 EBEs in the promoter regions of the upregulated genes. (B) Venn diagram of DEGs upregulated by AvrXa10 but not by AvrXa10ΔCRR at 32 and 48 hpi. (C) The predicted theoretical EBE of AvrXa10 RVDs. The AvrXa10 RVDs and their associated nucleotides are shown. The logo was produced using TALgetter (Galaxy v.1.1 http://galaxy.informatik.uni-halle.de/).

Figure 3

AvrXa10 activates the expression of potential target genes by binding EBEs in promoter regions. (A) Functional maps of effector and reporter plasmid constructs. The effector constructs contained FLAG-tag fused AvrXa10 or PthXo1 in vector pHB under the control of the CaMV 35S promoter. Reporter constructs contained gusA reporter cassettes that were driven by the candidate gene promoters (∼1 kb in length); these were cloned in pCAMBIA1381. Abbreviations: rbcS, ribulose-1,5-bisphosphate carboxylase, small subunit; polyA, polyadenylation site; NOS, NOS terminator site. (B)Nb promoters from six genes direct the AvrXa10-dependent, transient expression of GUS in N. benthamiana. The photographs of qualitative GUS assays are shown above the bars. The TALE PthXo1 and the Os8N3 promoter were used as a control. Samples were collected at 36 hpi, and GUS activity was calculated. Error bars indicate means ± SD (n = 3), and asterisks indicate significant differences (∗P ≤ 0.05; ∗∗P ≤ 0.01). The experiment was performed at least three times with similar results. (C) Schematic map of effector and reporter constructs used to identify target genes that promote an HR in Nb. The effector constructs contained FLAG-tag fused pthXo1 under the control of the CaMV 35S promoter. The reporter constructs contained the coding sequences of candidate target genes fused with the Os8N3 promoter, and the Os8N3-hpa1 construct served as a positive control. (D) HR assay in N. benthamiana. Agrobacterium strains containing the effector construct (pHB-pthXo1) and one of the six reporter constructs were infiltrated into fully expanded Nb leaves, which were evaluated for the HR at 4–7 dpi. Representative results were chosen from five independent experiments. Legend: 1, PthXo1 + pOs8N3::29135g; 2, PthXo1 + pOs8N3::36259g; 3, PthXo1 + pOs8N3::20731g; 4, PthXo1 + pOs8N3::44252g; 5, PthXo1 + pOs8N3::45656g; 6, PthXo1 + pOs8N3::3692g; and 7, PthXo1 + pOs8N3::Hpa1. (E) Oligonucleotide sequences of the EBE probes used in electrophoretic mobility-shift assays. Xa10 EBE was used as a positive control, and the mutated 20731g EBE was used as a negative control. The AvrXa10 RVDs corresponding to the EBE sequences are shown above. (F) Electromobility shift assays using biotin-labeled putative EBE fragments derived from the promoter regions of 20731g, 44252g, and 45656g. The XA10 EBE probe was used as a positive control. (G) Binding specificity of AvrXa10 to the target EBEs. Competition of biotinylated probes with unlabeled probes that were used at increasing concentrations (0, 5, 20, and 50×). The experiments were repeated three times.

Figure 4

Overexpression of 20731g (NbZnFP1) results in the HR in N. benthamiana. (A) Designer TALE assays. Promoter sequences of Nb20731g, Nb44252g, and Nb45656g; bold, underscored bases indicate sites for dTALE insertion. The gray highlighted region shows the putative AvrXa10 EBE, and the repeat variable di-residues of dTAL-A, dTAL-B, and dTAL-C and target EBEs are shown. (B) Detection of dTAL-A, dTAL-B, and dTAL-C by western blot analysis. Fully expanded Nb leaves were infiltrated with pHB-TAL constructs, and FLAG-tagged dTALs were detected at 48 hpi. The pHB vector and actin protein were used as negative and loading controls, respectively. (C) Expression analysis of 20731g, 44252g, and 45656g in Nb leaves infiltrated with Agrobacterium carrying the pHB constructs dTAL-A, dTAL-B, and dTAL-C; pHB and pHB-AvrXa10 served as negative and positive controls, respectively. Four-week-old Nb leaves were infiltrated and collected at 36 hpi for qRT-PCR. Error bars indicate means ± SD (n = 3), and asterisks indicate significant differences (∗P ≤ 0.01). The results shown are representative of three independent replicates. (D) dTAL-A induces the HR in N. benthamiana. Constructs were transiently expressed in fully expanded Nb leaves via infiltration with Agrobacterium (OD₆₀₀ 0.8). Constructs included pHB-dTAL-A, pHB-dTAL-B, pHB-dTAL-C, pHB-AvrXa10 (positive control), pHB-AvrXa10ΔCRR, and pHB. Nb leaves were photographed at 3 dpi. All experiments were repeated three times with similar results. (E) Transient overexpression of NbZnFP1 in Nb leaves. The Agrobacterium strains containing pHB-AvrXa10, pHB-NbZnFP1, or empty pHB vector were infiltrated into Nb leaves with a needleless syringe. The phenotype was photographed at 3 dpi. (F) For the western blotting assay, the center region of infiltrated leaf samples was collected at 2 dpi.

Figure 5

AvrXa10 causes cell death when delivered by X. axonopodis pv. glycines (Xag) in N. benthamiana. (A) Detection of AvrXa10 production in Xag by western blotting using an anti-FLAG primary antibody (see methods). RNA polymerase subunit alpha (RNAP) from E. coli was used as a loading control. (B) HR phenotype at 24 hpi. The Xag strains containing AvrXa10 or empty vector (EV) were inoculated into Nb leaves. The simple buffer MgCl₂ was used as a mock control. The experiment was repeated three times with similar results. (C) Quantification of bacterial growth at different time points. The Xag strains containing AvrXa10 or empty vector (EV) were inoculated into Nb leaves. The infiltrated leaf samples were collected with a 1-cm diameter cork borer at different time points (0, 12, 24, and 32 hpi). Three leaf discs from three different plants were used as a single replicate, and three replicates were used for this experiment. Error bars indicate means ± SD (n = 3), and asterisks indicate significant differences (∗P ≤ 0.01). (D) qRT-PCR analysis of NbZnFP1 in Nb leaves upon inoculation with derivatives of the Xag strain. Fully expanded leaves were infiltrated with Xag strains carrying AvrXa10 or pHM1 (EV), then collected at 24 hpi for RNA isolation. NbEF1α was used as an internal control. Error bars represent means ± SD (n = 3), and asterisks indicate significant differences (∗P ≤ 0.05). The results shown are representative of three independent replicates.

Figure 6

NbZnFP1 is the biologically relevant target of AvrXa10 in N. benthamiana. (A) On-spot VIGS-mediated silencing of NbZnFP1 partially inhibits AvrXa10-induced HR. Nb leaves were co-infiltrated with Agrobacterium carrying pHB-AvrXa10 + pTRV-RNA1 + pYL156-NbZnFP1 or pHB-AvrXa10 + pTRV-RNA1 + pYL156 (EV); these two combinations are labeled AvrXa10 + VIGS-NbZnFP1 and AvrXa10 + VIGS-EV, respectively. Nb leaves were photographed at 3 dpi. (B) RT-PCR analysis of NbZnFP1 expression in Nb after on-spot VIGS. Infiltrated leaves were collected at 3 dpi and analyzed for NbZnFP1 expression by RT-PCR; NbEF1α served as a reference gene. (C) Transient expression of NbZnFP1 induces cell death in rice protoplasts. Constructs pRTVcHA-NbZnFP1 (NbZnFP1) and pRTVcHA (EV, empty vector) were co-expressed with the LUC reporter construct pRTVcVC-LUC in rice protoplasts. Co-transfection of pRTVcHA-Xa10 (XA10) with the LUC construct was used as a positive, cell-death-inducing control. LUC activity was measured after 24 h of transfection using the Promega LUC assay system. The images on the left side of the graph show microtiter plates containing protoplasts expressing the constructs. The image of LUC fluorescence was taken with a CCD imaging system (IVIS Spectrum, PerkinElmer, USA). The graph on the right shows the relative LUC activity measured with a luminometer (Tecan, M200). Cell death in protoplasts was monitored by the reduction in luciferase activity. Error bars represent means ± SD (n = 3), and asterisks indicate significant differences (∗P ≤ 0.01). The results are representative of three replicates.

Figure 7

AvrXa10 triggers HR in tomato, probably by activating a zinc finger-like protein gene (ZnFP) (A) Putative AvrXa10 EBEs in the promoters of ZnFP homologs in tomato. The AvrXa10-EBE in the promoter of NbZnFP is also shown. (B) HR phenotypes at 4 dpi. (C) qRT-PCR analysis of the four ZnFP homologs. Leaves of 4-week-old tomato cv. Ailsa Craig were infiltrated with Agrobacterium strains containing pHB-AvrXa10, pHB-AvrXa10ΔCRR, pHB-AvrXa10ΔAD, or pHB (empty vector). Samples were collected at 32 hpi for RNA isolation and qRT-PCR analysis. The SolyEF1α gene was used to normalize the data. Error bars represent means ± SD (n = 3), and asterisks indicate significant differences (∗P ≤ 0.01). The results shown are representative of three independent replicates.