Overexpression of NtERF5, a New Member of the Tobacco Ethylene Response Transcription Factor Family Enhances Resistance to Tobacco mosaic virus

Overexpression of NtERF5, a New Member of the Tobacco Ethylene Response Transcription Factor Family Enhances Resistance to Tobacco mosaic virus

Fischer, Ute

Plants have to respond to environmental changes, such as biotic and abiotic stresses, in order to reprogram gene expression in an appropriate manner. Pathogen-responsive gene regulation has been shown to be mediated by specific transcription factors. Members of the bZIP (Jakoby et al. 2002), WRKY (Eulgem et al. 2000), Myb (Daniel et al. 1999), or AP2/ERF (ethylene response factor) (Riechmann and Meyerowitz 1998) transcription Factors have been implicated in pathogen-induced transcription. In particular, AP2/ERF transcription factors belong to one of the largest plant transcription factor families, comprising 145 members in the model plant Arabidopsis thaliana (Sakuma et al. 2002). They are characterized by a conserved DNA-binding domain, the AP2/ERF domain, which is 57 to 66 amino acids in size (Okamuro et al. 1997). Interestingly, this DNA-binding domain (Okamuro et al. 1997; Hao et al. 1998) is found only in the plant kingdom, which suggests plant-specific functions (Riechmann and Meyerowitz 1998). Allen and associates (1998) have analyzed the 3D solution structure of an AP2/ERF domain and have found that it consists of a three-stranded antiparallel [beta]-sheet and an [alpha]-helix packed approximately parallel to the [beta]-sheet. The [beta]-sheet interacts with specific motifs located inside the major groove of the DNA helix.

The founding member of AP2/ERF transcription factor family, Arabidopsis APETALA2 (AP2) is a well-defined transcription factor regulating flower development (Riechmann and Meyerowitz 1998). Whereas the AP2-like proteins harbor two AP2/ERF DNA-binding domains, the RAV1-like subfamily contains one AP2/ERF and a B3 DNA-binding domain (Kagaya et al. 1999; Sakuma et al. 2002). A third subfamily is characterized by one AP2/hiRF domain only. Based on DNA-binding data, this group has been subdivided into dehydration-responsive element-binding (DREB)-like proteins, which interact with the dehydration-responsive, or cold-repeat, element (consensus sequence TACCGACAT) (Liu et al. 1998; Stockingcr et al. 1997) and the EREBP-like (ethylenc responsive clement binding protein) proteins binding the GCC box (consensus sequence AGCCGCC) (Ohme-Takagi and Shinshi 1995; Sakuma et al. 2002). Recently, the latter group has been renamed as ERF. The ERF founding members, the tobacco proteins MERFl to NtERF4, have been isolated, due to their binding to GCC boxes found in several PR promoters (Ohme-Takagi and Shinshi 1995; Ohta et al. 2000). NtERF1 to MERF4 are transcriptionally activated by ethephon, a compound leading to the release of the plant hormone ethylene. Therefore, MERF1 to MERF4 have been implicated in ethylene signaling. However, the biological function of these factors is not, as yet, well-characterized.

Up to now, ERF transcription factors have been described in different plant species, such as tomato (Zhou et al. 1997), rice (Dubouzet et al. 2003), Catharanthus roseus (van der Fits and Memelink 2000), and Arabidopsis thaliana (Fujimoto et al. 2000; Ohta et al. 2001). In those cases in which functions have been assigned, these proteins have been found to be participating in plant responses to abiotic and biotic stresses (Kizis et al. 2001; Ohme-Takagi et al. 2000; Riechmann and Meyerowitz 1998). As examples, Arabidopsis DREB1 and DREB2 are involved in drought and cold stress response, ORCA3, isolated from Catharanthus roseus, regulates genes in jasmonic acid (JA)-induced secondary metabolism (van der Fits and Memelink 2000), and tomato Pti4, Pti5, and Pti6 regulate PR genes (Wu et al. 2002; Zhou et al. 1997). Following Pseudomonas infection, Pti4 is phosphorylated by the Pto kinase, which is thought to mediate transcriptional activation of PR genes (Gu et al. 2000).

Participating in pathogen-induced signaling networks, molecules like salicylic acid (SA) are involved in mediating local defense responses and establishing systemic acquired resistance (SAR) (Klessig et al. 2000). Ethylene in combination with JA also mediates pathogen responses in a SA-independent pathway. Depending on the pathogen, different pathways are triggered, leading to the activation of defined sets of target genes. For example, resistance to the necrotrophic fungus Botrytis cinerea utilizes ethylene/JA signaling networks, and accordingly, overexpression of the Arabidopsis ERF1 transcription factor enhances resistance to B. cinerea. In contrast, tolerance to Pseudomonas infection is decreased in ERF1-overexpressing plants (Berrocal-Lobo et al. 2002). These findings support the model that cross-talk and cross-inhibition between pathogen-induced signaling pathways exist (Lorenzo et al. 2003). As it has been shown in a number of publications, overexpression of ERF genes results in enhanced tolerance to pathogen attack (Berrocal-Lobo et al. 2002; Gu et al. 2002; He et al. 2001; Park et al. 2001; Shin et al. 2002). Whereas, overexpression of the tobacco Tsi1 gene leads to constitutive expression of PR genes and a broad-spectrum resistance (Park et al. 2001), overexpression of Pti5 does not lead to constitutive PR gene expression but accelerates pathogen-induced transcription (He et al. 2001). In summary, defined members of the ERF family seem to play specific roles in pathogen defense.

In this work, we describe a new member of the tobacco ERF family, designated NtERF5. Transcription of NtERF5 is enhanced after wounding and pathogen attack. Overexpression of NtERF5 does not lead to constitutive activation of PR genes or enhanced resistance to Pseudomonas infection. In contrast, NtERF5-overexpressing plants suppress TMV proliferation, leading to enhanced viral resistance.

RESULTS

Isolation of a new tobacco ERF gene.

A yeast one-hybrid approach has been set up to isolate tobacco transcription factors that specifically bind promoter cis-elements known to mediate pathogen-induced transcription (discussed below). Unexpectedly, a cDNA encoding a DNA-binding protein has been isolated, which does not bind to any of the well-characterized m-elements used in this study, but interacts with a GC-rich stretch of DNA originating from the multiple cloning site of the pUK21 vector, which has been included in the one-hybrid screening construct due to the cloning strategy used. The isolated cDNA encodes a formerly undescribed member transcription factor of the ERF gene family, which bind GC-rich cis-elements, the so-called GCC boxes. Since ERF proteins have been shown to be involved in pathogen-regulated gene control, we characterized this clone in more detail. The full-length cDNA clone was isolated by 5′-RACE (rapid amplification of cDNA ends) and was found to encode a 27-kDa protein (Fig. 1A). Despite the highly conserved AP2/ERF domain (Fig. 1A), this protein differs considerably from the previously described tobacco ERF family members MERF1 to MERF4 (Ohme-Takagi and Shinshi 1995) and Tsil (King et al. 1996; Park et al. 2001) and was therefore designated NtERF5. The most closely related ERF is ERFl from A. thaliana (Solano et al. 1998), showing an amino acid identity of 40% (Fig. 1B).

In vitro, MERFS binds GCC boxes with low affinity.

NtERF5 has been isolated by binding to an artificial GC-rich sequence in a yeast one-hybrid screen. This sequence (TGG ATC CGA TAT CGC CGT CGC GGC CGC TCT) is related to the typical GCC-box binding site, which is frequently found in PR promoters. To analyze DNA binding, a His-tagged NtERF5 gene was expressed in Escherichia coli. Protein preparations were monitored by Western analysis and Coomassie staining (Fig. 2A and B). HIS-NtERF5 protein was compared with HIS-tagged AtERF1 protein, which has been extensively analyzed with respect to DNA binding (Allen et al. 1998). Under identical conditions, both proteins were purified, and DNA binding was studied using electrophoretic mobility shift assays (EMSA) (Fig. 2C). Whereas purified HIS-AtERF1 binds efficiently to GCC boxes, no binding of HIS-ERF5 was observed with several putative ERF target sites (Fig. 2D). However, with crude protein extract of E. coli expressing HIS-NtERF5, a weak GCC box-specific binding activity was observed that was not found using mutated GCC box (Fig. 2D, GCCmut). Several attempts to modify EMSA conditions failed to enhance DNA binding. Furthermore, binding sites of ERF-related DREB factors or the artificial GC-rich sequence that has been found to be an efficient binding site in yeast were not recognized by NtERF5 in vitro (Fig. 2D). We therefore conclude that the GCC box is a low-affinity binding site in vitro. Most likely, posttranslational modifications or cofactors might enhance binding efficiency under in vivo conditions.

NtERF5 transcription is induced by wounding and pathogen attack.

Many transcription factors belonging to the ERF family are involved in plant stress responses, in particular in pathogen-regulated gene expression. Tissue-specific NtERF5 expression was found in parts of the flower but not in vegetative tissues (Fig. 3A). In contrast, NtERF5 transcription in leaves was strongly enhanced after infection with the bacterial pathogen Pseudomonas syringae. Depending on the type of interaction, using avirulent P. syringae pv. pisi (Tronchet et al. 2001) or virulent P. syringae pv. Tabaci (Salch and Shaw 1988), we observed a rapid or retarded increase of NtERF5 steady-state transcript level, respectively (Fig. 3B). It has to be noted that wounding due to the injection of the bacteria rapidly and transiently induced NtERF5 transcription (Fig. 38). Other wound treatments, like rubbing the leaf surface with Carborundum also led to NtERF5 induction (data not shown).

In contrast to the susceptible tobacco cultivar W38 (nn), a TMV-induced signaling cascade is triggered in the resistant tobacco cultivar Samsun (NN), which harbors the TV resistance gene. Plant defense responses include the establishment of a hypersensitive response (HR) and the induction of PR genes. Moreover, it is well-established that the N gene-mediated plant defense is inactivated at elevated temperatures (32°C) (Malamy et al. 1992). Therefore, TMV-induced plant defense response can easily be synchronized by a temperature shift experiment. Plants cultivated at 32°C were infected with TMV to allow systemic spread of the virus. After 4 days, temperature was shifted to 22°C, and TV gene-mediated defense responses like enhanced NtERF5 transcription could be monitored in directly and systemically infected leaves (Fig. 3C). The PR1a gene was used to monitor the onset of the plant resistance response (Cutt et al. 1988).

Tobacco transcription factors NtERF5 to NtERF4 have been named with respect to their responsiveness to ethephon. In further experiments, ethephon, which releases ethylene after application, was used, (Ohme-Takagi and Shinshi 1995). However, NtERF5 was not induced following ethephon treatment (data not shown). Furthermore, signaling molecules like SA (Fig. 3D) or JA (data not shown) were not found to enhance NtERF5 transcription. The transcriptional responses observed both in mock- and hormone-treated samples are caused by wounding due to the cutting of leaf disks.

Overexpression of NRF5 does not change plant defense responses toward the bacterial pathogen Pneudomonas syringae.

Ectopic overexpression of several ERF genes resulted in stress-resistant transgenic plants (Berrocal-Lobo et al. 2002; Gu et al. 2002; He et al. 2001; Park et al. 2001; Shin et al. 2002). NtERFS was overexpressed in the tobacco cultivai’ Wisconsin W38 under control of the Cauliflower mosaic virus (CaMV) 35S promoter (ERF5-Oex), and the level of overexpression was confirmed by Northern analysis (Fig. 4A). In comparison with wild type, no visible phenotypic alterations in plant growth were observed in ERF5-Oex transgenic tobacco lines. Since Pseudomonas infection induces NtERFS transcription, we investigated whether constitutive NtERFS expression increases resistance to bacterial infection. However, bacterial growth of virulent P. syringae pv. tabaci and avirulent P. syringae pv. pisi bacteria (Fig. 4B) and induction of typical PR genes (data not shown) was comparable with that of wild type. The establishment of SAR in uninfected leaves was measured 10 days after challenge with P. syringae pv. pisi. However, with reference to bacterial growth of the virulent pathogen P. syringae pv. tabaci, NtERF5 overexpression has no effect on induction of SAR against re-infection with P. syringtie (data not shown).

Overexpression of NtERF5 enhances resistance to TMV.

Furthermore, NtERF5 was overexpressed in the tobacco cultivar Samsun (NN), which harbors the N resistance gene. Infection with TMV induces the formation of HR lesions. In order to quantify plant resistance responses, HR lesion size was determined according to Vernooij and associates (1994). In comparison with TMV-infected wild-type SNN plants, ERF5-Oex plants display lesions of reduced size (Fig. 5A and B). The reduction of lesion size corresponds to the level of NtERF5 steady-state transcripts as monitored by Northern analysis (Fig. 5C).

Resistance to TMV might be due either to reduced viral spread, replication, or both. To further assess the mechanism of ERF5-mediated TMV resistance, a temperature shift experiment was performed. Plants cultivated at 32°C were infected with TMV. After 4 days, temperature was shifted to 22°C (Fig. 6A), which initiates a massive N gene-mediated plant defense response. Directly infected leaves (local infected leaves) of wild-type and ERF5-Oex plants collapse within several hours after temperature shift (data not shown). During four days of culture at 32°C, the virus has spread systemically, and after shifting temperature, systemically infected leaves at the apex of wild-type plants collapse as well. Surprisingly, systemically infected leaves of ERF5-Oex plants showed no sign of cell death, indicating that the virus had not entered this tissue (Fig. 6B). Since the N gene-mediated resistance is inactivated at high temperature, the observed resistance is due to an N gene-independent mechanism.

To examine this phenomenon on the molecular level, we analyzed TMV-induced PR1a expression in directly and systemically infected leaves at the RNA and protein levels. Whereas, no difference in PR1a expression was observed in the directly infected leaves of wild-type and ERF5-Oex plants, PR1a steady-state RNA and protein accumulation was reduced in systemically infected leaves of ERF5-Oex plants (Fig. 6C). Accordingly, as monitored by ethidium bromide staining of viral RNA, TMV was observed in systemically infected leaves of wild-type plants but not in ERF5-Oex plants. These data suggest that the amount of TMV or systemic movement of the virus, or both, is reduced in ERF5-Oex plants. In agreement with this notion, the local response of several defense or PR genes in ERF5-Oex plants was comparable to wild type, but no induction of these genes was observed in systemically infected ERF5-Oex leaves (Fig. 6D). Overexpression of several ERF genes in transgenic plants resulted in constitutive expression of PR and defense genes. In contrast, no constitutive activation of PR (e.g., PR1a, PR1b, PR3) or defense-related genes (e.g., SAR8.2, HMGR, PAL) was observed in ERF5-Oex plants (Fig. 6D). Moreover, the kinetics of pathogen-induced transcription in directly infected leaves were comparable in ERF5-Oex and wild-type plants.

To analyze the mechanism that leads to TMV resistance in ERF5-Oex plants, we measured the steady-state TMV-RNA level in locally infected leaves. Immediately after temperature shift (0 h) and after 8 and 24 h, respectively, TMV-specific RNA was quantified (Fig. 7). In comparison with wild-type plants, we detected only 10 to 30% of TMV-RNA in the infected ERF5-Oex leaves. Since TMV-RNA levels have been measured at 32°C, these data indicate that the N gene-mediated defense is not involved in reducing viral RNA. In conclusion, overexpression of NtERF5 is sufficient to enhance resistance to TMV, most likely by regulating viral propagation.

DISCUSSION

In this study we have identified NtERF5, a new member of the tobacco AP2/ERF transcription factor family that enhances plant resistance to TMV when overexpressed. Overexpression of ERF genes has been shown to result in resistance to bacterial and fungal pathogens (Berrocal-Lobo et al. 2002; Gu et al. 2002; He et al. 2001). Here, we further extend strategies to engineer plants displaying an enhanced resistance to viral attack by overexpressing NtERF5.

Tobacco ERFs have been shown to act as activators or repressers of transcription. Whereas NtERF3 is characterized by the well-defined EAR repressor domain (Hiratsu et al. 2003; Ohta et al. 2001), NtERF1, NtERF2, and NtERF4, as well as Tsi1 have been shown to function as transcriptional activators in plant cells or yeast, respectively (Ohta et al. 2000; Park et al. 2001). Stretches of acidic amino acids are located at the N and C termini of NtERF5, which have been shown to possess activation properties (Lemon and Tjian 2000). Hence, NtERF5 might act as a transcriptional activator in planta.

Whereas the NtERF5 DNA-binding domain is highly conserved, other parts of the protein differ considerably from other described tobacco ERFs, like NtERF1 to NtERF4 or Tsi1 (Fig. 1). The most related protein is ERF1 from A. thaliana (Solano et al. 1998). In particular, NtERF5 and ERF1 share acidic domains in the N and C termini. However, since only a limited number of the tobacco ERF family members have been isolated, NtERF5 orthologs are difficult to assign.

As far as functions are described, members of the AP2 subfamily are involved in developmental aspects (Riechmann and Meyerowitz 1998), whereas ERFs mediate plant stress responses. Developmental or tissue-specific expression of NtERF5 has been detected only in sepals of the tobacco flower, indicating a possible function during flower development. PR genes have been shown to be ERF targets (Chakravarthy et al. 2003; Ohme-Takagi and Shinshi 1995; Ohta et al. 2000), and expression of PR1b and PR2 has been detected in parts of the flower (Cote et al. 1991). Whether PR proteins have specific functions during flower development or whether they defend flowers and developing seeds against pathogens is not well-investigated. Since ntERF5 overexpression does not lead to changes in vegetative or generative plant growth, there is no indication that NtERF5 plays a role in plant development.

NtERF5 transcription is strongly enhanced by infection with the bacterial pathogen P. syringae. In comparison with the compatible interaction, NtERF5 transcription is enhanced with an accelerated kinetic during incompatible interaction, indicating that the transcriptional response is triggered by a specific pathogen-derived avirulence factor. NtERF5 transcription is also induced by TMV. These data imply that NtERF5 plays a more general role in plant pathogen defense. Abiotic stresses like cold and salt treatment do not induce NtERF5 transcription (U. Fischer and W. Dröger-Laser, unpublished data). As has been described for NtERF2, NtERF3, and NtERF4 (Nishiuchi et al. 2002), NtERF5 transcription is transiently induced by various wound treatments, which might also be related to a function in pathogen defense.

Two well-defined signaling pathways are involved in pathogen-defense responses that make use of the plant hormones SA or ethylene/JA, respectively (Dong 2001). Whereas ERF1 from Ambiclopsis is induced by ethylene or JA, transcription of NtERF5 and Pti5 are not enhanced by any of these hormones (Gu et al. 2000; Thara et al. 1999). These data suggest that these proteins, although closely related, have divergent functions. Moreover, our results suggest that pathogeninduced NtERF5 transcription is either mediated independently of the SA or ethylene/JA pathways or single hormone action is not sufficient to trigger the response.

Overexpression of several ERF genes has been shown to result in tolerance to abiotic (Stockinger et al. 1997; Liu et al. 1998) or biotic (Park et al. 2001) stresses. Analysis of several PR promoters reveals GCC cis-elements to be crucial for pathogen-induced gene activation. However, different ERF proteins induce only a subset of PR genes. For example, overexpression of Ambidopsis ERF1 results in constitutive activation of [beta]-Glu, b-CHI, or PDF1.2, whereas expression of tomato Pti4 in Arabidopsis leads to activation of PRI, PR3 (chilinase), PR4, and PDF1.2, which are typical markers for the SA or the ethylene/JA pathway (Gu et al. 2002; Lorenzo et al. 2003). Only slight induction of PR1 was observed by Pti5 overexpressing plants. In tomato, Pti5 expression causes no constitutive gene activation but potentiates pathogen-induced activation of [beta]-glucanase and catalase genes (He et al. 2001). ERF5-Oex plants show no constitutive PR gene expression. Furthermore, NtERF5 does not potentiate pathogen-induced PR gene expression, as has been shown for Pti5-overexpressing plants. In conclusion, these data support the view that the huge number of ERF genes reflects not redundant but distinct functions of the plant to deal with specific environmental stress conditions.

The amino acids of the ERF DNA-binding domain that are crucial for in vitro binding to GCC or DREB boxes have been defined (Hao et al. 1998, 2002, 2003; Sakuma et al., 2002). According to these data, NtERF5 harbors typical amino acids (A14, D19) that should mediate a GCC box-specific binding activity. Although NtERF5 was isolated by binding to GC-rich sequences in a yeast one-hybrid screen, rccombinant HIS-tagged NtERF5 protein only weakly interacted with GCC box-containing promoter fragments in vitro. Moreover, no in vitro binding to DREB elements was observed. In contrast, AtERF1 recombinant protein used as a control efficiently binds GCC boxes under identical conditions. In the case of the barley ERF protein HvCBF2, in vitro binding was enhanced by low temperature (Xue 2003). However several attempts varying conditions of incubation, binding buffers, and protein preparation failed to enhance in vitro DNA binding. Moreover, since NtERF5 can activate transcription mediated by GC-rich cis elements in yeast, we propose that additional cofactors or posttranslational modifications are necessary for efficient binding to GCC boxes. As has been shown by Gu and associates (2000), phosphorylation of Pti4 by the serine-threonine kinase Pto enhances its DNA binding in vitro. Anticipating that GCC boxes are target sites of NtERF5, a posttranslational mechanism to regulate DNA binding of NtERF5 in vivo has to be postulated. In a recent study, in vivo binding of the ERF protein Pti4 was shown by means of chromatin immuno precipitation even to promoters harboring no GCC box (Chakravarthy et al. 2003). Hence, we cannot exclude the possibility that other unknown promoter cis elements might be targeted by ERFs. Furthermore, protein-protein interaction has been shown between AlEBP and a member of the TGA class of bZIP transcription factors (Buttner and Singh 1997). In many PR promoters, GCC boxes are clustered with W boxes or as-1 elements, which are supposed to be regulated by WRKY or TGA transcription factors, respectively (Maleck et al. 2000). Hence, it is tempting to speculate that a synergistic action of these transcription factors might affect DNA binding and selection of the subset of PR promoters that are regulated by a specific ERF. These data support the view, that the presence of a GCC box itself is not sufficient for ERF-mediated gene activation. Chromatin immuno precipitation will provide further insight into ERF5 in vivo binding sites.

In contrast to Tsi1, which when overexpressed, leads to broad spectrum resistance (Shin et al. 2002), MERF5 overexpression specifically resulted in enhanced resistance to TMV. Resistance is due to a significantly reduced accumulation of TMV-RNA in infected leaves, implying that NtERF5-regulated genes are involved in suppressing viral replication. Since viral RNA accumulation is reduced also at 32°C, this resistance is not clue to an N gene-mediated mechanism. The reduced lesion size might also be a result of an inhibition of viral movement. Therefore, infeeted plants were incubated at 32°C, to allow systemic spread of the virus. In comparison to wild-type plants, no TMV-RNA, TMV-induced HR, or defense gene activation was observed in the upper part of the plants. These data indicate that viral spread is strongly suppressed in ERF5-Oex plants. However, we cannot rule out that the suppression of viral spread is a consequence of the reduced TMV propagation in the primary infected leaves. Thus, the mechanism of exactly how NtERF5 is mounting a plant defense to TMV remains elusive. Treatment with SA (Murphy and Carr 2002) as well as the presence of several host factors have been implicated in resistance to TMV replication and viral spread (Abbink et al. 2002; Akad et al. 1999; Valentine et al. 2002). It must be noted that the observed resistance phenotype is based on overexpression studies, and in order to demonstrate whether ERF5 is part of a natural-occurring resistance pathway, RNAi plants will need to be analyzed. Furthermore, the use of gene array analysis will assist in defining NtERF5 target genes, to further elucidate the mechanism of how NtERF5 is mounting a resistant response to viral infection.

MATERIALS AND METHODS

Plant material, plant cultivation, plant infection.

Tobacco cultivars Samsun (NN) and Wisconsin W38 (nn) were grown in a growth chamber under a 16-h-light and 8-h-dark regime at 22°C and 85% humidity. For TMV infection, fully expanded leaves of 6- to 8-week-old, soil-grown tobacco plants were used. Inoculation with TMV strain U1 (5 mg per leaf) was performed in 50 mM potassium phosphate buffer, pH 7.5, by gently rubbing the leaves with Carborundum, using a method which has been described previously (Yalpani et al. 2001). Mock treatment was done with Carborundum and buffer. HR was induced by incubating TMV-inoculated plants at 32°C for 4 days and then shifting temperature to 22°C, as has been described previously (Malamy et al. 1992). For bacterial infection, leaves from 6-week-old tobacco plants were infiltrated with a P. syringae pv. pisi (Cournoyer et al. 1995) or P. syringae pv. tabaci (Salch and Shaw 1988) suspension (1 × 10^sup 5^ CFU/ml). Infection inoculum was calculated to be 500 CFU per infection. For each timepoint, bacteria from 20 independent infection sites were harvested from two to three independent plants and were titcred by plating on King’s B medium (King et al. 1954). With respect to Northern blot experiments, leaves were inoculated with 100 µl of a bacterial suspension (10^sup 7^ CFU/ml).

For application of signal molecules, leaves from 6-week-old tobacco plants were sprayed with a solution containing 1 mM ethephon (Sigma, Taufkirchen, Germany) in 50 mM sodium phosphate buffer (pH 7.0) or 0.1 mM JA in water. Leaf disks were floated on 50 mM sodium phosphate buffer (pH 5.8) with 1 mM SA.

Standard molecular biological techniques.

Standard DNA techniques were used according to Sambrook and associates (1989). DNA sequence analysis was performed on an ABI310 sequencer, using an ABI PRISM BigDye terminator cycle sequencing reaction kit. RACE has been performed using a SMART-RACE cDNA amplification kit (Clontech, Palo Alto, CA, U.S.A.).

Yeast one-hybrid screening.

The commercial “matchmaker” yeast one-hybrid system (Clontech) was used according to the manufacturer’s protocol. Yeast one-hybrid vectors are based on pHISi-1 and pLacZi coding for HIS3 and lacZ reporter gene, respectively (Clontech). Various cis elements have been used for the yeast one-hybrid screening. For example, oligonuclcolides carrying a C-box (5’TAGATACAATTGACGTCATGTATCAT 3′, 3’CTATG TTAACTGCAGTACATAGTAAT 5′) were annealed, multimerized, and inserted into the NdeI site of pUK21. As a control K box, oligonuclcotides were used harboring specific mutations in the cix element (5’TAGATACAATTaAgtaaATGT ATCAT 3′, 3’CTATGTTAAlTcattTACATAGTAAT 5′). Fragments carrying four copies of the C box and K box were obtained by EcoRI and XbaI restriction and were inserted into pHISi-1. Due to this cloning step, a GC-rich sequence (TGG ATC CGA TAT CGC CGT GGC GGC CGC TCT AGA) from the mes of pUK21 has been included into the pHIS-1 screening vector. The HIS reporter genes were integrated into the yeast chromosomal background, using the procedure described in the manufacturer’s protocol. For yeast one-hybrid screening, a tobacco SR1 cDNA library (Stralhmann et al. 2001) has been constructed in pGAD424 (Clontech). Yeast transformants were selected on medium lacking histidine and were supplemented with 45 mM 3-AT (3-aminotriazol). The vector pGAD-ERF5, which encodes an activation domain NtERF5 fusion, activates the HIS reporter genes independently of whether they are driven by the C or K box cix elements. To test whether ERF5 binds to the artificial GC-rich sequence, which has been included in the screening construct due the cloning strategy, C- and K-box-containing EcoRI-EcoRV fragments were inserted into the vector pLacZi. The vectors obtained do not harbor the GC-rich sequence, and expression of GADERF5 does not result in lacZ reporter gene activation. We therefore conclude that ERF5 binds the GC-rich sequence and none of the specifically introduced cix elements. [beta]-Galactosidase assays were performed as described in Strathmann and associates (2001).

Plant transformation.

The NtERF5 cDNA was inserted into the EcoRI site of pBluescriptIISK (Stratagene, La Jolla, Ca, U.S.A.). After restriction with AccI and XbaI, NtERF5 was obtained from pBluescriptIISK and inserted into the plant transformation vector pBINHygTx (Rieping et al. 1994). Agrobacterium-mediated transformation of the Nicotiana tabacum cultivars Samsun (NN) and Wisconsin W38 (nn) was performed as previously described (Strathmann et al. 2001).

Protein techniques.

Recombinant NtERF5 protein was prepared by inserting an EcoRI fragment obtained from pGAD-ERF5 into the HIS-tag vector pET28a (Novagcne, Abingdon, U.K.). AtERF1 was PCR-amplified using the primers 5′ GGG GAC AAG TTT GTA CAA AAA AGC AGG CTT CAT GAC GGC GGA TTC TCA ATC TG 3′ and GGG GAC CAC TTT GTA CAA GAA AGC TGG GTT TAT AAA ACC AAT AAA CGA TCG 3′. The fragment was inserted into the GATEWAY vector pDONR201 (Invitrogen, Karlsruhe, Germany). Via recombination, AtERF1 was integrated into the HIS-tag vector pDEST17 (Invitrogen). Expression studies were performed in E. coli B121 (Stratagene, La Jolla, CA, U.S.A.). After induction with 1 mM isopropyl-[beta]-D-thiogalactoside, recombinant proteins were purified, using the Ni-NTA resin according to the manufacturer’s protocol (Qiagen, Hildesheim, Germany).

Proteins were separated using a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Sambrook et al. 1989) and were transferred to a polyvinylidene diflouride membrane (Millipore, Braunschweig, Germany), following the procedure described previously (Strathmann et al. 2001). For immuno-detection, a HIS-tag-specific antibody (M. Benley, Göttingen, Germany), a PR-specific monoclonal antibody (D. Klessig, Boyce Thompson Institute, Ithaca, NY, U.S.A.), and an anti-rabbit immunoglobin horseradish peroxidase-linked antibody from donkey (Amersham Bioscience, Freiburg, Germany) were applied. Visualization was performed using the ECL+ system (Amersham-Bucher, Braunschweig, Germany).

EMSA.

EMSA has been performed as described by Lyss and associates (1998). The following oligonucleotides were used for the binding reactions:

GCC box (Fujimoto et al. 2000):

5′ ATG AGT TAA CGC AGA CAT AGC CGC CAT TT 3′

5′ TAC TCA ATT GCG TCT GTA TCG GCG GTA AA 3′

GCCmut box:

5′ ATG AGT TAA CGC AGA CAT AtC CtC CAT TT 3′

5′ TAC TCA ATT GCG TCT GTA TaG GaG GTA AA 3′

DREB box (Sakuma et al. 2002):

5′ ATG AGT TAA GCG AGA CAT ATA CCG ACA TAT TT 3′

5′ TAG TCA ATT GCG TCT GTA TAT GGC TGT ATA AA 3′

GC rich (pUK21):

5′ TGG ATC CGA TAT CGC CGT GGC GGC CGC TCT AGA 3′

5′ ACC TAG GCT ATA GCG GCA CCG CCG GCG AGA TCT 3′

Northern blot analysis.

Total RNA was purified from tobacco tissue by using the RNeasy plant mini kit (Qiagen). Northern hybridization was performed as described in Strathmann and associates (2001). The following hybridization probes have been used: a 1,000-bp EcoRI fragment from NtERF5, a 531-bp EcoRI/BamHI fragment from PR1a (Cutt et al. 1988), a 1,000-bp PstI fragment from PR3 (Lawton et al. 1992), an 800-bp SalI/BamHI fragment from the movement protein of TMV (D. Hofius, unpublished data), a 440-bp PCR fragment coding for ACC oxidase (Kim et al. 1998), a 1,000-bp EcoRI fragment from HMGR (Kang et al. 1998), and a 1,000-bp EcoRI fragment from SAR 8.2 (Herbers et al. 1996). Quantification of hybridization signals was achieved by using Bio-imager analysis (BAS-1000, Fuji, Tokyo) and Tina software (Raytest, Straubenhardt, Germany).

ACKNOWLEDGMENTS

We thank C. Gatz and R. Weigel for critically reading the manuscript and D. Hofius (IPK, Gatersleben, Germany), M. Ohme-Takagi (Japan), D. Klessig (Boyce Thompson Institute, Ithaca, NY, U.S.A.) and M. Benley (Biosystems, Göttingen, Germany) for providing DNA constructs and antibodies, respectively.

Copyright American Phytopathological Society Oct 2004

Provided by ProQuest Information and Learning Company. All rights Reserved.