Pathogen-induced ERF68 regulates hypersensitive cell death in tomato

Abstract

Ethylene response factors (ERFs) are a large plant-specific transcription factor family and play diverse important roles in various plant functions. However, most tomato ERFs have not been characterized. In this study, we showed that the expression of an uncharacterized member of the tomato ERF-IX subgroup, ERF68, was significantly induced by treatments with different bacterial pathogens, ethylene (ET) and salicylic acid (SA), but only slightly induced by bacterial mutants defective in the type III secretion system (T3SS) or non-host pathogens. The ERF68-green fluorescent protein (ERF68-GFP) fusion protein was localized in the nucleus. Transactivation and electrophoretic mobility shift assays (EMSAs) further showed that ERF68 was a functional transcriptional activator and was bound to the GCC-box. Moreover, transient overexpression of ERF68 led to spontaneous lesions in tomato and tobacco leaves and enhanced the expression of genes involved in ET, SA, jasmonic acid (JA) and hypersensitive response (HR) pathways, whereas silencing of ERF68 increased tomato susceptibility to two incompatible Xanthomonas spp. These results reveal the involvement of ERF68 in the effector-triggered immunity (ETI) pathway. To identify ERF68 target genes, chromatin immunoprecipitation combined with high-throughput sequencing (ChIP-seq) was performed. Amongst the confirmed target genes, a few genes involved in cell death or disease defence were differentially regulated by ERF68. Our study demonstrates the function of ERF68 in the positive regulation of hypersensitive cell death and disease defence by modulation of multiple signalling pathways, and provides important new information on the complex regulatory function of ERFs.

Keywords: ERF; cell death; defence; disease; tomato.

Figure 1

Tomato ERF68 is induced by pathogens and hormones. Reverse transcription‐quantitative polymerase chain reaction (RT‐qPCR) was used to analyse ERF68 expression in leaves of 3–4‐week‐old tomato H7996 after the indicated treatments. (a) Leaves vacuum infiltrated with Ralstonia solanacearum (Rs) Pss4 [10⁷ colony‐forming units (cfu)/mL]. (b) Plants treated with ethylene (ET) gas (10 ppm) or air in a sealed chamber. (c) Leaves sprayed with jasmonic acid (JA), salicylic acid (SA) or 0.01% ethanol (mock). (d) Leaves vacuum infiltrated with different bacterial pathogens (10⁶ cfu/mL). Pst, Pseudomonas syringae pv. tomato; Xcc, Xanthomonas campestris pv. campestris; Xvt, Xanthomonas euvesicatoria (Xeu) Xvt28; hpi, h post‐inoculation. The expression levels were normalized using tomato ELONGATION FACTOR 1α (EF1α) as the internal control. Values are means ± standard deviation. The data are from a single experiment that was repeated at least three times with similar results.

Figure 2

ERF68 is localized in the nucleus and functions as a transcription factor. (a) Subcellular localization of ERF68. The vector carrying the ERF68‐green fluorescent protein (ERF68‐GFP) fusion was introduced into Arabidopsis protoplasts by polyethylene glycol (PEG) transfection. Photographs were taken at 16 h after transfection. 4',6‐Diamidino‐2‐phenylindole (DAPI) staining indicates the location of the nucleus. The experiment was repeated at least three times with similar results. (b) Transactivation assay of ERF68. Top panel: schematic diagram of constructions for transactivation assay. Bottom panel: activation of transcriptional activity of GCC‐box‐containing promoter by ERF68 and the positive control AtERF5. Total proteins were extracted from Arabidopsis protoplasts 16 h after co‐transfection with reporter, reference and effector plasmids. ‘Relative fold’ indicates the relative fluorescence of firefly luciferase (LUC)/renilla luciferase (REL). For each independent experiment, three technical replicates were conducted. The data are from a single experiment that was repeated at least three times with similar results. Values are means ± standard deviation. Pair‐wise comparisons between ERF68 or AtERF5 and vector control were made using Student's t‐test. **Highly significant difference (P < 0.01). *Significant difference (P < 0.05). (c) Electrophoretic mobility shift assay (EMSA) for the physical interaction between a biotin‐labelled probe containing the GCC‐box and the C‐terminal ERF (ethylene response factor) domain of ERF68 (GST‐ERF68C). The experiment was repeated at least three times with similar results. GCC, unlabelled probe. mGCC, probe containing a mutated GCC box (AGAAGAA).

Figure 3

Overexpression of ERF68 causes spontaneous lesions and induces the expression of defence‐related genes in tomato H7996. (a) Overall growth of the test plants. One‐week‐old plants were agroinfiltrated with the indicated virus‐induced gene silencing (VIGS) or virus‐mediated gene overexpression (VMGO) constructs at bacterial doses with an optical density at 600 nm (OD₆₀₀) of 2.0 and 0.8, respectively. Validation assays for ERF68 expression were carried out in all experiments prior to further analyses to ensure the reliability of the test materials. The photograph was taken 2 weeks after agroinfiltration. (b, c) Ten‐day‐old plants were agroinfiltrated with the indicated VMGO constructs and the plants were used for further phenotypic and gene expression analyses 3 weeks after agroinfiltration. (b) Development of spontaneous lesions. (c) Expression of marker genes (see ‘Results’ for gene definitions) involved in defence pathways in leaves measured by reverse transcription‐quantitative polymerase chain reaction (RT‐qPCR). The samples are random combinations of at least four leaflets taken from at least two leaves of two tomato plants. The levels of expression are normalized using tomato ELONGATION FACTOR 1α (EF1α) as an internal control. The expression level of each gene in green fluorescent protein (GFP)‐overexpressing plants is indicated as 1.0. Values are means ± standard deviation from three technical repeats in a single experiment that was repeated at least three times with similar results. ET, ethylene; GUS, β‐glucuronidase; HR, hypersensitive response; JA, jasmonic acid; PDS, phytoene desaturase; SA, salicylic acid.

Figure 4

Induced ERF68 overexpression leads to cell death. Leaves of 4‐week‐old Nicotiana benthamiana were infiltrated with Agrobacterium carrying ERF68‐HA driven by the β‐estradiol inducible promoter. Two days after agroinfiltration, the leaves were infiltrated with β‐estradiol (+ or + β‐estradiol) or 0.005% ethanol (– or – β‐estradiol). The samples were harvested at different time points after induction for the following analyses. (a) Development of cell death on leaf sections with induced ERF68‐HA expression. (b) Detection of ERF68‐HA by western blotting using a mouse anti‐haemagglutinin (anti‐HA) monoclonal antibody (H3663, Sigma‐Aldrich). Fifty micrograms of each protein sample were used for western blotting. (c) Cell conductivity determined by the measurement of ion leakage. (d) H₂O₂ production detected by 2′,7′‐dichlorodihydrofluorescein diacetate (DCFDA) staining. The accumulation of H₂O₂ was determined by relative fluorescence intensity with the level at 0 h as 1.0. (c, d) Values are means ± standard deviation. Pair‐wise comparisons between β‐estradiol‐treated and ethanol‐treated samples were made using Student's t‐test. **Highly significant difference (P < 0.01). *Significant difference (P < 0.05). The data are from a single experiment that was repeated at least three times with similar results.

Figure 5

Silencing of ERF68 alters tomato defence responses to Xanthomonas spp. One‐week‐old tomato plants were agroinfiltrated with virus‐induced gene silencing (VIGS) constructs. Three weeks after VIGS treatment, plants were vacuum infiltrated with the indicated avirulent and virulent Xanthomonas pathogens [optical density at 600 nm (OD₆₀₀) = 0.0004, 2 × 10⁵ colony‐forming units (cfu)/mL]. In planta proliferation of bacteria was determined. (a) Tomato cultivar H7998 was inoculated with X. euvesicatoria Xvt28 (incompatible, labelled as ‘IC’) and X. vesicatoria Xvt45 (compatible, labelled as ‘C’). (b) Tomato cultivar H7981 was inoculated with X. perforans Xtn50 (incompatible, labelled as ‘IC’) and X. vesicatoria Xvt45 (compatible, labelled as ‘C’). Values are means ± standard deviation from at least three independent experiments [18 × 4 = 72 data for (a), 18 × 3 = 54 data for (b)]. Statistical analysis was performed using Duncan's test.

Figure 6

ERF68 differentially regulates the target genes. Leaves of 4‐week‐old Nicotiana benthamiana were infiltrated with Agrobacterium carrying ERF68‐HA driven by the β‐estradiol inducible promoter. Two days after agroinfiltration, the leaves were infiltrated with β‐estradiol (+ β‐estradiol) or 0.01% ethanol (– β‐estradiol). Reverse transcription‐quantitative polymerase chain reaction (RT‐qPCR) was performed to assess the expression level of selected target genes of ERF68. (a) Genes up‐regulated by ERF68: GSTU8, GLUTATHIONE S‐TRANSFERASE; COPA, COATMER PROTEIN SUBUNIT α; AOS, ALLENE OXIDE SYNTHASE; Sw‐5a, TOSPOVIRUS RESISTANT PROTEIN 5a; OFP2, OVATE FAMILY PROTEIN 2. (b) Genes down‐regulated by ERF68: CAB, CHLOROPHYLL a/b BINDING PROTEIN; SCF 35 E3, SOLUTE CARRIER FAMILY MEMBER 35 E3. The relative expression (fold change) was normalized using Actin 1. Values are means ± standard deviation from three technical repeats in a single experiment that was repeated at least three times with similar results.