Harpin, the product of the hrpN gene of Erwinia amylovora, elicits the hypersensitive response and disease resistance in many plants. Harpin and known inducers of systemic acquired resistance (SAR) were tested on five genotypes of Arabidopsis thaliana to assess the role of SAR in harpin-induced resistance. In wild-type plants, harpin elicited systemic resistance to Peronospora parasitica and Pseudomonas syringae pv. tomato, accompanied by induction of the SAR genes PR-1 and PR-2. However, in experiments with transgenic Arabidopsis plants containing the nahG gene which prevents accumulation of salicylic acid (SA), harpin neither elicited resistance nor activated SAR gene expression. Harpin also failed to activate SAR when applied to nim1 (non-inducible immunity) mutants, which are defective in responding to SA and regulation of SAR. In contrast, mutants compromised in responsiveness to methyl jasmonate and ethylene developed the same resistance as did wild-type plants. Thus, harpin elicits disease resistance through the NIM1-mediated SAR signal transduction pathway in an SA-dependent fashion. The site of action of harpin in the SAR regulatory pathway is upstream of SA.
Certain interactions between harpins and plants must occur for the induction of resistance (Hoyos et al. 1996). In tobacco suspension cell cultures treated with HrpZ or HrpN protein, a variety of responses occur that lead to ion influxes across the membranes, alkalization of the growth medium, cell membrane depolarization and production of active oxygen species (Baker et al. 1993; He et al. 1994; Popham et al. 1995; Wei et al. 1992). Programmed cell death occurs in Arabidopsis suspension cultured cells in response to HrpZ (Desikan et al. 1998). A potential role for phosphorylation in the regulation of these responses is suggested by observations that they are sensitive to K252a, an inhibitor of protein kinases (Baker et al. 1993), and that tobacco leaves infiltrated with harpin or water accumulate a mitogen-activated protein kinase (MAPK) (Adám et al. 1997). Together, these data suggest that application of harpin to plants initiates signal transduction events that lead to defence responses (see review by Boller & Felix 1996).
To elucidate the parts of the signal transduction pathway downstream of SA that lead to SAR, a variety of mutant screens have been used. These have involved assays for defective induction of pathogen resistance (Delaney et al. 1995) and screens using SAR-gene-promoter fusions to reporter or selectable marker genes, coupled with treatment with SAR-inducing compounds like INA (Cao et al. 1994; Shah et al. 1997). In each case, mutations were discovered in the same gene, called NIM1, NPR1 or SAI1, respectively, by each group (Cao et al. 1994; Delaney et al. 1995; Shah et al. 1997). nim1 mutants were shown to be unresponsive to SA, INA or BTH for induction of SAR (Delaney et al. 1995; Lawton et al. 1996) demonstrating that like SA, the synthetic compounds also act through a plant signaling pathway defined by the NIM1/NPR1. Because mutants at this locus retain the ability to accumulate SA, yet fail to respond to this compound, the NIM1/NPR1 gene is believed to act between the site of action of SA and the induction of SAR-associated defence genes. The NIM1/NPR1 gene was cloned in two laboratories (Cao et al. 1997; Ryals et al. 1997) and was shown to encode a protein that contains ankyrin repeats, motifs present in a variety of proteins and believed to mediate protein–protein interactions. The protein NPR1 or NIM1 is suggested to regulate PR gene expression by interacting with a transcription factor (Kim & Delaney 1999; Zhang et al. 1999).
Other induced resistance pathways exist that are independent of SAR (Niki et al. 1998; Penninckx et al. 1996; Vijayan et al. 1998; Xie et al. 1998). For example, certain growth promoting rhizobacteria can elicit a form of systemic resistance called induced systemic resistance (ISR) (Hoffland et al. 1995; Liu et al. 1995; van Loon et al. 1998), which is distinct from SAR because it is not dependent upon SA accumulation and is not linked to the accumulation of SAR-associated gene products (Pieterse et al. 1996; van Loon et al. 1998). Furthermore, also unlike SAR, ISR appears to depend on signaling by jasmonic acid and ethylene, based on the inability to induce ISR in Arabidopsis mutants insensitive to these compounds (Pieterse et al. 1998). Thus, at least two distinct pathways contribute to the suite of pathogen-induced resistance systems. These involve distinct signaling pathways that require either the accumulation of SA or the action of ethylene and jasmonic acid for SAR and ISR, respectively. Activation of SAR requires function of the NIM1/NPR1 gene product. Curiously, however, although the induction of ISR occurs independently of SA, it is reported to depend upon action of the NIM1/NPR1 gene product, as npr1-2 mutants are unable to induce ISR (Pieterse et al. 1998).
Induction of resistance by harpin could result from the activation of a variety of defence pathways. To better understand the mechanisms underlying harpin-induced disease resistance, we examined harpin-treated plants for accumulation of SAR-associated gene products, which would suggest that harpin functions through the SAR pathway. To assess the role of specific signaling pathways in harpin-induced resistance, we examined its effectiveness in several Arabidopsis genetic backgrounds, including SA-non-responsive nim1-1, jasmonate-insensitive jar1-1 and ethylene-insensitive etr1-1 and etr1-3 mutants, and SA-non-accumulating NahG plants (Delaney et al. 1994; Delaney et al. 1995; Guzmàn & Ecker 1990; Schaller & Bleecker 1995; Staswick et al. 1992; Staswick et al. 1998). Harpin was found to be an effective inducer of resistance to Peronospora parasitica and Pseudomonas syringae and caused induction of SAR genes in all genotypes except nim1-1 and NahG. The present data indicate that harpin-induced resistance acts specifically through the SAR pathway, and does not depend upon JAR1 or ETR1 gene products.
Harpin induces expression of SAR genes
In Arabidopsis, several PR genes, including PR-1, PR-2 and PR-5 are expressed co-ordinately with SAR (Ryals et al. 1994; Uknes et al. 1992). We monitored the expression of PR-1 and PR-2 in plants treated with the HrpN protein (harpin) from Erwinia amylovora. Harpin was obtained from a cell-free elicitor preparation (CFEP) made from cultured Escherichia coli cells containing a cloned hrpN gene. As a negative control a cell-free empty vector preparation (CFVP) was similarly prepared from E. coli cells that contain the vector without the hrpN insert. Plants sprayed with INA were used as a positive control for SAR in most experiments. Infiltration of harpin into the older lower leaves of Arabidopsis caused accumulation of PR-1 transcripts in the untreated apices and the youngest leaves (Fig. 1), showing systemic induction of the gene. Expression of PR-1 occurred in a time-dependent manner, first being detected after 2 days, and increasing through 6 days after harpin application. Harpin-induced PR-1 mRNA accumulation exceeded that mediated by treatment with 0.3 mm INA. No PR-1 induction was observed in plants treated with the empty vector control extract (CFVP) (Fig. 1).
Gene induction in SAR-compromised genotypes
To test whether harpin-induced gene expression involves the SAR pathway, we assayed for PR-1 and PR-2 mRNA accumulation in salicylate hydroxylase (NahG)-expressing plants and in the nim1-1 mutant (Fig. 2). Both genotypes are unable to express SAR, which depends upon SA accumulation and signaling through the NIM1 pathway (Delaney et al. 1994; Delaney et al. 1995; Gaffney et al. 1993). Harpin and control solutions were sprayed onto wild-type, NahG and nim1-1 plants; leaf tissues were collected for RNA analysis 1, 3 and 5 days after treatment. In wild-type (Col-O and Ws-O) plants, both PR-1 and PR-2 showed strong induction by harpin at days 3–5, while NahG and nim1-1 plants showed no accumulation of these mRNAs at any timepoint assayed. Plants treated with the positive control, INA, showed induction of PR-1 and PR-2 in wild-type and NahG plants. INA is capable of inducing resistance in salicylate hydroxylase plants because it is not a substrate for this enzyme (Delaney et al. 1994; Vernooij et al. 1995). These data indicate that harpin-induced PR-1 and PR-2 expression requires a functional SAR signal transduction pathway.
Harpin-induced resistance to Peronospora parasitica
To determine if harpin can induce disease resistance in Arabidopsis, the growth of the oomycete pathogen P. parasitica in harpin-treated plants was determined. Three lower leaves of 20-day-old wild-type Col-O and Ws-O seedlings were infiltrated with CFEP or CFVP. Five days later, the plants were inoculated with P. parasitica strains Noco2 and Emwa, which are virulent on Arabidopsis ecotypes Col-O and Ws-O, respectively. The development of infection was observed macroscopically and by staining leaves with lactophenol-trypan blue and microscopic examination. The data in Fig. 3 clearly show the effectiveness of harpin-induced resistance against the oomycete pathogen. Control plants treated with CFVP or water were obviously infected and supported growth of large numbers of conidiospores. In contrast, plants treated with harpin were less infected, as indicated by few conidiospores growing on the leaves. Thus, the severity of infection, based on the numbers of conidiospores per leaf, was remarkably reduced by treatment with harpin, suggesting that harpin induces systemic resistance to the oomycete in two ecotypes of Arabidopsis.
Resistance also developed following spray application of harpin to leaves of 14-day-old wild-type Arabidopsis seedlings (Fig. 4). Substantial pathogen growth was observed in leaves of plants treated with CFVP. In contrast, only a few conidiophores grew on the leaf surfaces and a few oospores and hyphae were produced within the leaf tissues of harpin-treated plants.
Harpin-induced resistance to Pseudomonas syringae pv. tomato DC3000
We then assayed for specificity of harpin-induced resistance in Arabidopsis. We found that harpin elicited resistance against P. syringae pv. tomato DC3000 (Fig. 5a,b). Bacterial growth was reduced in the untreated upper leaves after infiltration of harpin into the lower three leaves of the plant. At each timepoint, the bacterial population in harpin-treated plants was less than that in control plants. Thus, disease resistance elicited by harpin occurs systemically in leaves not directly treated with harpin. When harpin was applied by spraying, bacterial multiplication was also reduced. The bacterial population increased in 4 days approximately 5000-fold and 3000-fold in CFVP-treated plants of ecotypes Col-O and Ws-O, respectively, while in harpin-treated plants an approximate 400-fold increase was observed in the two ecotypes during the same period (Fig. 5a,b). This reduction is similar to that caused by synthetic inducers of SAR (Uknes et al. 1992).
Harpin-induced disease resistance requires SA accumulation and NIM1 function
Unlike wild-type plants, harpin treatment of NahG and nim1-1 plants failed to induce resistance to P. parasitica (Fig. 4) and P. syringae pv. tomato DC3000 (Fig. 5c,d). These results are consistent with the failure of harpin to induce SAR gene expression in these SAR-defective genotypes. Because harpin fails to induce resistance in these SAR-disabled genotypes, it is not likely to be directly antimicrobial, but rather to act through induction of an endogenous plant defence pathway requiring SA and defined by the NIM1 gene. Harpin, SA and INA produced similar levels of resistance to P. parasitica and P. syringae pv. tomato DC3000 (Figs 3 and 5). All three compounds also failed to induce resistance in nim1 mutants, highlighting the central role of the NIM1 gene in defence signaling.
Harpin induces resistance in jasmonate and ethylene response mutants
To determine whether harpin-induced resistance requires signaling by jasmonate or ethylene, we tested the response to harpin in the methyl jasmonate insensitive mutant jar1-1 (Staswick et al. 1992), ethylene insensitive mutants etr1-1 and etr1-3 (formerly ein1-1) (Guzmán & Ecker 1990; Schaller & Bleecker 1995), and isogenic wild-type Col-O plants. Plants were sprayed with harpin, subsequently inoculated with P. parasitica, and disease development was monitored. Macroscopic and microscopic observations showed that a similar level of resistance developed in wild-type and three mutant lines following application of harpin (Fig. 6). All CFVP-treated plants supported vigorous growth of P. parasitica within leaf tissues. In contrast, harpin-treated plants were nearly free of infection. Seven days after inoculation, 70–90% of the CFVP-treated mutant and wild-type plants were infected, compared to 5–10% of harpin-treated mutant and wild-type plants infected. Thus, jasmonate and ethylene signaling systems do not appear to affect the function of harpin in inducing resistance to the oomycete pathogen.
The aim of this study was to determine the mode of action through which the HrpN protein (harpin) of E. amylovora elicits disease resistance in Arabidopsis. We examined wild-type Arabidopsis (Col-O and Ws-O), two SAR defective genotypes (NahG and nim1-1), a jasmonate-insensitive mutant (jar1-1), and two ethylene-insensitive mutants (etr1-1 and etr1-3) for their responsiveness to harpin and established chemical elicitors of SAR. Both phenotypic and molecular data support our conclusion that harpin-induced resistance in Arabidopsis functions through activation of SAR that requires accumulation of SA and regulation by the NIM1/NPR1 gene product. The demonstration of alternative resistance signaling pathways mediated by jasmonic acid and ethylene indicates the possibility for multiple actions of an inducer in triggering resistance signal transduction. Because methyl jasmonate and ethylene-insensitive mutants developed resistance following the application of harpin, the involvement of jasmonic acid and ethylene signaling mechanisms in the action of harpin for pathogen resistance seems unlikely, and further supports the conclusion that harpin acts through the SAR pathway.
An important understanding from this study is that harpin induces resistance by a signaling process that begins upstream of SA, leads to activation of PR genes, and is regulated by NIM1 gene product. This distinguishes harpin from other elicitors. First, SA dependence and PR gene activation distinguishes harpin-induced resistance from ISR, which is neither dependent on SA nor associated with PR gene expression (Pieterse et al. 1996; Pieterse et al. 1998). Second, SA-dependence also distinguishes harpin from INA and BTH that act downstream of SA (Lawton et al. 1996; Vernooij et al. 1995). Finally, harpin-induced resistance to P. parasitica isolate Noco2 is different from the constitutive, NPR1-independent and defensin gene (PDF1.2) expression-associated resistance to the same isolate in Arabidopsis mutant cpr5 that expresses both NPR1-dependent and NPR1-independent resistance (Bowling et al. 1994; Bowling et al. 1997). Otherwise, because harpin induces resistance to bacteria, it may also activate antibacterial genes that have not been defined and are suggested to require regulation by NPR1/NIM1 (Bowling et al. 1997; Clarke et al. 1998).
This study presents a preliminary understanding of a harpin-triggered signal transduction process. The SA-dependent and NIM1/NPR1-mediated signal transduction pathway may be only one signaling pathway used by harpin for resistance induction, although we were able to rule out dependence on the JAR1 and ETR1 signaling pathways. In addition to resistance against pathogens (Qiu et al. 1997; Wei & Beer 1996; Wei et al. 1998; this work), several other beneficial effects occur in plants treated with harpin, including the enhancement of plant growth (H. Dong and S.V. Beer, unpublished results; Qiu et al. 1997; Wei et al. 1998) and the repellency of insects (Zitter & Beer 1998). The mechanisms that underlie these diverse beneficial effects of harpin are not known. Nevertheless, previous data suggest the involvement of reactive oxygen intermediates (for its implication see Alvarez et al. 1998; Dangl et al. 1996; Jabs et al. 1996) and programmed cell death, calcium ion channels and protein kinase cascades in interactions of Hrp proteins with plants (Adám et al. 1997; Baker et al. 1993; Dong et al. 1999; He et al. 1994; Popham et al. 1995; Wei et al. 1992). These may together account for the pleiotropic effects of harpins in plants. Further explorations of the relationships between signaling pathways that affect the several beneficial effects of harpin are underway.
Plant growth and pathogen maintenance
Arabidopsis thaliana ecotypes Columbia (Col-O) and Wassilewskija (Ws-O) were used in all experiments. NahG transgenics (Delaney et al. 1995; Gaffney et al. 1993) and nim1-1 mutant plants (Delaney et al. 1995) were previously produced from the Col-O and Ws-O ecotypes. Methyl jasmonate response mutants jar1-1 and ethylene response mutants etr1-1 and etr1-3 were derived from Col-O (Guzmàn & Ecker 1990; Schaller & Bleecker 1995; Staswick et al. 1992), and their seeds (accession numbers CS8072, CS237and CS3037) were provided by the Arabidopsis Biological Resource Center at the Ohio State University (Columbus, OH, USA). All the genotypes were grown in greenhouse soil mix at 21°C and 14 h light per day for vegetative growth, at 18°C and 12 h day length for infection by Peronospora parasitica, and at 24°C and 14 h day length for infection by Pseudomonas syringae pv. tomato DC3000(Koncz et al. 1992).
Peronospora parasitica strain Emwa and Noco2 were maintained by weekly culture on Arabidopsis ecotypes Ws-O and Col-O. Conidial suspensions were made from infected leaves and inoculated as described previously (Uknes et al. 1992). Pseudomonas syringae pv. tomato DC3000 was cultured on L-Agar medium (Gerhardt et al. 1981) prior to inoculation of plants.
Preparation of elicitors and treatment of plants
INA (2,6-dichloroisonicotinic acid) was kindly provided as 25% wettable powder by Dr Kay Lawton (Novartis Crop Protection, Inc., Research Triangle Park, North Carolina, USA). INA was used at 0.3 mm in water except when otherwise noted. Salicylic acid (SA) was used at 0.3 or 0.5 mm in water as described. Harpin was prepared as a cell-free elicitor preparation (CFEP) from E. coli strain DH5-α harboring plasmid pCPP2139, which contains the hrpN of E. amylovora (Wei et al. 1992) in the expression vector pCPP50 (Bauer et al. 1997). The cell-free empty vector preparation (CFVP) was similarly made, except that the DH5-α strain contained only the vector, pCPP50. The HR-eliciting activity of CFEP and CFVP was determined by infiltrating opposite leaf panels of Xanthi NN tobacco leaves with dilutions of both preparations. The undiluted CFVP did not elicit HR, and was used as a negative control for harpin-containing CFEP. The concentration of harpin in CFEP was determined by HPLC by Eden Bioscience Corporation (Bothell, WA, USA). Inducing compounds, elicitor preparations and controls were applied by spraying the plants to run-off with an atomizer (Devilbiss no. 15) 14 days after sowing except when otherwise noted. Five days later, plants were inoculated using an atomizer with a P. parasitica conidial suspension containing 5 × 104 conidiospores per ml, or with a suspension of P. syringae pv. tomato DC3000 at 5 × 108 cfu per ml of water. Ecotype Col-O and genotypes derived from it (NahG, jar1-1, etr1-1 and etr1-3) were inoculated with the virulent P. parasitica isolate Noco2; Ws-O and nim1 were inoculated with the virulent isolate Emwa. Inoculated plants were maintained under the conditions described above for 5 days before infection was assessed. Each induction–inoculation combination included six pots, and each pot contained 15–25 seedlings.
Evaluation of infection
Infection by P. parasitica was judged based on the presence of conidiophores on the leaf surfaces (Koncz et al. 1992). Conidiospores on leaves were estimated by counting spores in leaf washes using a haemocytometer under the microscope and expressed as conidiospores per leaf. Oomycete growth in leaves was examined using an Olympus BX60 microscope following staining with lactophenol trypan blue and clearing with chloral hydrate (Uknes et al. 1992). To monitor in planta bacterial multiplication, leaves of inoculated plants were detached at designated times, sterilized with 70% ethanol and homogenized in sterile water; bacteria were recovered from the resulting homogenates by culturing on l-agar medium (Gerhardt et al. 1981).
RNA blot analyses
RNA was prepared from experimental and control plants using the RNeasy Plant Mini Kit (Qiagen, Chatsworth, CA, USA), size-fractionated by agarose gel electrophoresis (Clark 1997), and transferred to Immobilo-N transfer membrane (Millipore). Replicate blots were hybridized to 32P[dCTP]-labeled Arabidopsis SAR gene cDNA probes PR-1 and PR-2 as described previously (Church & Gilbert 1984). Loadings were standardized by calculating the total RNA (4 μg per lane) of samples and verified by ethidium bromide (EtBr) staining of gels.
We thank Dr Kay Lawton of Novartis Crop Protection, Inc. (Research Triangle Park, North Carolina, USA) for the gifts of INA and NahG Arabidopsis; Dr Zhongmin Wei of Eden Bioscience Corporation (Bothell, Washington, USA) for the determination of harpin concentration in the CFEP preparations; and the Arabidopsis Biological Resource Center at Ohio State University for providing mutant plant lines. This research was supported by Eden Bioscience Corporation, a USDA Special Grant (99–34367–7390) to S.V.B., and an NSF-CAREER Award (IBN-9722377) to T.P.D.