- © 1998 American Society of Plant Physiologists
Abstract
Changes in gene expression induced in tobacco leaves by the harpin HrpN protein elicitor were examined, and a new cDNA, piox (for pathogen-induced oxygenase), with homology to genes encoding cyclooxygenase or prostaglandin endoperoxide synthase (PGHS), was identified. In addition to the amino acid identity determined, the protein encoded by piox is predicted to have a structural core similar to that of ovine PGHS-1. Moreover, studies of protein functionality demonstrate that the PIOX recombinant protein possesses at least one of the two enzymatic activities of PGHSs, that of catalyzing the oxygenation of polyunsaturated fatty acids. piox transcripts accumulated after protein elicitor treatment or inoculation with bacteria. Expression of piox was induced in tissues responding to inoculation with both incompatible and compatible bacteria, but RNA and protein accumulation differed for both types of interactions. We show that expression of piox is rapidly induced in response to various cellular signals mediating plant responses to pathogen infection and that activation of piox expression is most likely related to the oxidative burst that takes place during the cell death processes examined. Cyclooxygenase catalyzes the first committed step in the formation of prostaglandins and thromboxanes, which are lipid-derived signal molecules that mediate many cellular processes, including the immune response in vertebrates. The finding of tobacco PIOX suggests that more similarities than hitherto expected will be found between the lipid-based responses for plant and animal systems.
INTRODUCTION
The hypersensitive response (HR) is characterized by a rapid, localized death of tissues at the site of microbial attack and is associated with the defense of plants against invading microorganisms. This response forms part of a general inducible defense reaction that includes production of active oxygen species (AOS), cell wall modifications, and synthesis of antimicrobial compounds (phytoalexins), hydrolytic enzymes, and antimicrobial proteins, all of which contribute to overall resistance in the plant (reviewed in Hammond-Kosack and Jones, 1996). Many of these responses are due to transcriptional activation of specific defense-related genes; however, several processes involve activation of preexisting components rather than changes in gene expression. An example of the latter is the striking production of AOS (Sutherland, 1991; Mehdy, 1994). Studies of the early steps of infection by incompatible HR-inducing pathogens have shown that H2O2 and oxygen radicals play key roles as local triggers of this programmed cell death (Levine et al., 1994; Tenhaken et al., 1995; Chandra et al., 1996). Recent studies with Arabidopsis have shown that it is also possible to identify mutations that cause spontaneous lesions to form in the absence of pathogen attack (Dietrich et al., 1994; Greenberg et al., 1994) and that the rapid generation of reactive oxygen species plays a role in lesion formation (Jabs et al., 1996).
Certain microbial molecules can produce a phenotype that mimics the cell death response induced by incompatible pathogens. In particular, a protein termed harpin from Erwinia amylovora has been reported to elicit an HR-like cell death when infiltrated into the leaf laminae of appropriate non-host plants (Wei et al., 1992). Furthermore, the addition of harpin to plant cell cultures or inoculation into the leaf apoplast stimulates defense-related responses that are associated with the bacterial induction of the HR (Baker et al., 1993; Gopalan et al., 1996a). Of special significance is the E. amylovora harpin HrpN that stimulates AOS production in suspension cells (Baker et al., 1993). The hrpN gene, encoding the HrpN protein, is a structural gene of the hrp (for hypersensitive response and pathogenicity) cluster, which enables E. amylovora to elicit the HR on non-host plants (Wei et al., 1992). hrp genes are thought to be involved primarily in protein secretion, with nine of them encoding highly conserved components of a type III protein secretion system (hrc genes) (Alfano and Collmer, 1996). This Hrp secretion pathway is used by the bacterium to secrete harpin into the leaf intercellular spaces and is also required to cause disease in susceptible plants (Alfano and Collmer, 1996). Despite the recent data accumulated on the importance of different bacterial harpins to elicit HR symptoms, neither the biological function of these proteins during bacterial infection nor the mechanism that allows plants to recognize them is known (Alfano and Collmer, 1996).
Harpin proteins are able to trigger HR responses on a range of host plants. Bacteria, on the other hand, often show a species- and even cultivar-specific host range in which an HR can be induced. This specificity is due in part to bacterial avirulence (avr) genes shown to be implicated in host specificity for HR induction at the race–cultivar level. Bacterial avr genes have been cloned on the basis of their ability to trigger the HR when expressed in a pathogen that is normally compatible with a particular host, and an Hrp-dependent translocation of Avr proteins into the plant cytoplasm appears to be required for its activity (Gopalan et al., 1996b; Van den Ackerveken et al., 1996). As was the case with harpins, the primary function of Avr proteins is, in general, unclear. In Pseudomonas syringae pv syringae, Avr proteins appear to be more important than harpins in eliciting the HR (Alfano et al., 1997).
To understand better the cellular mechanisms controlling cell death in plants, we have examined the plant response to inoculation with a protein elicitor that induces an HR-like cell death. By using differential mRNA display (Liang and Pardee, 1992), we identified a new plant gene whose expression was induced in non-host tobacco leaves after inoculation with the E. amylovora HrpN protein. We show that gene expression is induced in response to various biotic and abiotic treatments related to the plant defense mechanisms and that protein accumulates after pathogen inoculation. Sequence and functional analysis of the cDNA-encoded protein showed significant homology with cyclooxygenase (COX), also referred to as prostaglandin endoperoxide synthase (PGHS). COX is a key enzyme in the production of prostanoids, which are critical regulators of many cellular processes, including the immune response, in vertebrates. The significance of this finding in the context of the plant response to pathogen infection is discussed.
RESULTS
Identification of a Harpin-Induced Tobacco Gene
We used differential mRNA display (Liang and Pardee, 1992) to identify plant genes induced after inoculation with HrpN, a protein elicitor from E. amylovora that elicits an HR-like cell death response in non-host tobacco leaves. The recombinant HrpN protein was produced in Escherichia coli as a fusion protein by cloning the E. amylovora hrpN gene into the pMAL-C expression vector. The hrpN gene was amplified by polymerase chain reaction (PCR) from total bacterial DNA and inserted downstream of the malE gene into the pMAL-C vector. The malE gene encodes a maltose binding protein (MBP), and expression of the recombinant plasmid pMAL-C/hrpN results in the production of an MBP–HrpN fusion protein that was purified by retention on an amylose column. In addition, the MBP protein expressed from the pMAL-C vector was similarly purified and used as a control for the analysis of HR-like induction. Tobacco leaves infiltrated with the purified MBP–HrpN protein fraction at concentrations >2.5 μM showed a cell death reaction within 24 hr, whereas leaves infiltrated with the control MBP protein did not induce any visible response (data not shown).
Changes in gene expression associated with the cell death response induced when leaves were inoculated with MBP–HrpN were then investigated. To study these changes, we simultaneously subjected total RNA extracted from leaves 4 hr after treatment with MBP–HrpN and RNA extracted from healthy untreated leaves (control) to differential display analysis. From the analysis of two separate experiments on a single gel, a 211-bp cDNA fragment, which was shown to be reproducibly induced in response to protein inoculation, was selected for further characterization. This partial cDNA, designated A5.2, was used for initial expression studies. A5.2 did not hybridize with RNA obtained from healthy untreated tobacco leaves. On the other hand, A5.2 hybridized with a single major transcript of ~2.3 kb in leaves harvested 4 hr after protein infiltration (data not shown).
Isolation of the A5.2 cDNA and Analysis of Gene Expression
A cDNA library was prepared with total RNA extracted from tobacco leaves 4 hr after inoculation with the MBP–HrpN protein and screened by using an A5.2 probe. Three positive clones were isolated and analyzed by sequencing. The longest cDNA, A5.2-1 (1576 bp), was sequenced completely. Sequences corresponding to the two remaining clones were shorter and identical to A5.2-1 within their overlapping regions, with the exception of clone A5.2-2, which contained an extension of 69 nucleotides in the 3′ end noncoding region. The rapid amplification of cDNA ends (RACE) protocol was used to clone the 5′ end region of A5.2 by using the Marathon cDNA amplification kit. A 622-nucleotide DNA fragment corresponding to the 5′ end of A5.2-1 was identified, and the entire A5.2 cDNA (2128 bp) was constructed. The complete DNA sequence of A5.2 is shown in Figure 1.
The expression of A5.2 was investigated at different time intervals in response to treatment with harpin as well as in response to inoculation with the incompatible bacterium E. amylovora carrying the hrpN gene. In addition, the expressions of two well-known defense-related genes—gn2, a basic β-1,3-glucanase gene from Nicotiana plumbaginifolia (Gheysen et al., 1990), and PR-1, encoding an acidic extracellular PR-1 isoform (Ward et al., 1991)—were examined simultaneously. As shown in Figure 2, no A5.2 transcripts were detected in untreated tobacco leaves used as a control, whereas A5.2 hybridized with transcripts present in RNA samples extracted 4 hr after protein inoculation. The maximum level of A5.2 RNA was found 24 hr after treatment, and it decreased to low levels 4 days after protein injection. Transcripts followed a similar pattern of accumulation in plants inoculated with E. amylovora, although the levels of RNA rose slightly higher than those detected in response to protein injection. No induction of gene expression was observed in response to two control treatments, MBP infiltration and water inoculation (Figure 2A). As found for A5.2, both gn2 and PR-1 were induced in response to protein treatment and bacterial inoculation (Figures 2B and 2C). gn2 and PR-1 transcripts followed a similar pattern of accumulation after the two treatments, but activation of gene expression was delayed when compared with the activation of the A5.2 cDNA. In these studies, we found that PR-1 and gn2 transcripts could first be observed at 8 and 24 hr after treatment, respectively. Activation of these two genes, therefore, was significantly slower than activation of A5.2, in which high levels of transcripts had accumulated 4 hr after bacterial inoculation.
Nucleotide and Deduced Amino Acid Sequence of the Entire A5.2 cDNA (GenBank Accession Number AJ007630).
Nucleotides are numbered from the first nucleotide of the cloned cDNA. The stop codon is marked with an asterisk. In addition, a nonsense codon present upstream from the initiating ATG is indicated by a dot.
The specificity of A5.2 induction in tobacco leaves responding to inoculation with an incompatible strain of P. s. syringae was compared with that in leaves treated with the compatible bacterium P. s. pv tabaci. Necrotic lesions characteristic of an HR were observed between 24 and 48 hr after P. s. syringae inoculation, whereas the compatible bacterium produced small areas of disease necrosis and tissue chlorosis on infected tissue. As shown in Figures 3A and 3B, A5.2 expression was induced in response to both bacteria; however, a different pattern of mRNA accumulation was clearly observed. Thus, high levels of A5.2 transcripts were observed 4 hr after inoculation with the HR-inducing bacterium, whereas the compatible interaction initially was accompanied by weak and somewhat variable A5.2 expression until 16 hr after bacterial inoculation. It is possible that this fluctuation in the level of RNA follows changes in the progress of the pathogenic infection, but the significance of this result is currently not known. In both infections, A5.2 expression persisted for more than 4 days but declined to barely undetectable levels 8 days after inoculation with the HR-inducing bacterium. In contrast, high levels of transcripts were still detected at this time for the compatible interaction.
Analysis of Gene Expression.
RNA was extracted at different intervals from leaves inoculated with the purified MBP–HrpN protein fraction at a concentration of 5 μM and with the HR-inducing bacterium E. amylovora. The controls were inoculated with purified MBP at a concentration of 5 μM and with water.
(A) Blots were hybridized with riboprobes derived from the A5.2 cDNA.
(B) RNA was hybridized with riboprobes derived from a basic β-1,3-glucanase gn2.
(C) Blots were hybridized with a PR-1 cDNA encoding a secreted acidic PR-1 isoform.
Specificity of A5.2 Induction in Response to Bacterial Infection.
(A) RNA was extracted from tobacco leaves in response to inoculation with an incompatible strain of P. s. syringae.
(B) RNA was extracted in response to inoculation with the virulent bacterium P. s. tabaci.
Ethidium bromide staining used as a control for equal loading of RNA is shown.
The A5.2 cDNA Encodes a Protein with Homology to COX
The A5.2 cDNA insert is 2128 bp long, and its larger open reading frame (1929 nucleotides) encodes a polypeptide of 643 amino acids with a predicted molecular weight of ~73,000. The putative amino acid sequence of A5.2 shows high similarity (75% identity) to the protein encoded by an Arabidopsis expressed sequence tag clone of unknown function (GenBank accession number N38086) and 50 to 60% identity to the FB protein encoded by the tomato feebly gene (van der Biezen et al., 1996). The complete sequence of the Arabidopsis clone was determined, and the amino acid identities with the A5.2 protein are shown in Figure 4. A precise comparison between A5.2 and the tomato FB protein was hampered by the uncertainties found in the feebly cDNA sequence in the GenBank database (accession number U35643).
In addition to the described plant genes, the deduced amino acid sequence of A5.2 shares sequence similarity with COX, the key enzyme in the production of prostaglandins and other eicosanoids in vertebrates. Sequence alignment of A5.2 with that of murine PGHS-2 (COX-2; O'Banion et al., 1992) and ovine PGHS-1 (COX-1; DeWitt and Smith, 1988) is shown in Figure 4. Overall, A5.2 shares 22 and 20% amino acid identity with PGHS-2 and PGHS-1, respectively, and 26 and 25% when the protein region corresponding to the catalytic domain is compared (domain-spanning residues 103 to 604 in PGHS-2 and 117 to 600 in PGHS-1). Moreover, if conservative replacements are considered, the homology between A5.2 and the two PGHS isoforms shows a 33% similarity for the overall proteins and a 39% similarity when the catalytic domains are compared.
Comparison of the three-dimensional structure of crystallized prostaglandin synthase (COX-1; Picot et al., 1994) with the predicted structural features of A5.2 and PGHSs reveals greater structural similarity than that expected from their sequence similarity alone. In the first place, the predicted secondary structure of the tobacco A5.2 protein was almost entirely composed of α helices, similar to PGHS, which also has a high helical content. A diagram representing the distribution of the predicted α helices (red boxes) is shown in Figure 5A. Furthermore, with highly significant probability, studies of sequence–structure fitness (Rost, 1995) identified PGHSs as the most likely structure to match the tobacco sequence. The alignment between the tobacco protein and the three-dimensional structure of ovine PGHS-1 (COX-1) showed a good correlation between the predicted structural features and the secondary structural elements of PGHS-1 (Figure 5B). The structural similarity (thick line) extends over the entire catalytic domain, where the spatial organization of the identified α helices (shown in red) was found to be strikingly conserved between PGHS-1 (Picot et al., 1994) and the tobacco protein. The few discrepancies found correspond to external protein regions around residues 180, 360, and 450 (Figure 5B, thin gray line). Of particular significance, the amino acid residues His-207, Tyr-385, and His-388 (shown in dark blue), which are keys for enzyme activity and are conserved completely within the PGHS family, were found to be present in the tobacco sequence at similar positions (His-167, Tyr-389, and His-392). Moreover, Ser-530 (shown in green), which is known to be acetylated by aspirin, has a possible equivalent in the tobacco protein at position 567 (Ser-567). The absence of other conserved residues around this region in the PGHS family precludes a more definitive assignment to this residue.
Sequence Alignment of Tobacco A5.2.
Sequences shown correspond to the tobacco A5.2 protein (tobacco), Arabidopsis protein encoded by the expressed sequence tag clone N38086 (Arabidopsis), murine prostaglandin synthase-2 (mur PGHS-2), and ovine prostaglandin synthase-1 (ovi PGHS-1). Regions of identity to tobacco A5.2 are highlighted in black. Alignment analyses were performed using the MegAlign program. Dashes denote gaps introduced to optimize alignment.
In contrast to the structural conservation found within the catalytic domain, the regions of PGHS falling outside of this domain (thin line in light blue), that is, the EGF-like (for epidermal growth factor) domain and the membrane-binding domain, do not seem to have equivalent residues in the tobacco protein.
Protein Functionality Analyses in a Baculovirus Expression System
To examine whether A5.2 encodes a protein with enzymatic activity similar to that of COX, we used a baculovirus expression system to express the A5.2 cDNA in insect cells. To this end, the A5.2 cDNA was cloned behind the polyhedrin promoter into the pFASTBAC1 donor vector, and the recombinant plasmid pFASTBAC/A5.2 was transferred into competent DH10Bac E. coli cells containing the baculovirus shuttle vector (bacmid) bMON14272 and the helper plasmid pMON7124. Recombinant bacmid DNA was then prepared and used to transfect insect cells. Appropriate viral stocks were harvested and used to infect fresh cells for subsequent expression studies.
In addition to A5.2, recombinant baculovirus containing the Arabidopsis N38086 cDNA was similarly prepared. Furthermore, a cDNA encoding the prostaglandin endoperoxide synthase-2 from rat (Feng et al., 1993) was cloned into the pFASTBAC1 vector and used as a positive control for enzyme activity analyses. As shown in Figure 6, infection of insect cells with recombinant virus pFASTBAC/A5.2 (lane 3) and pFASTBAC/N38086 (lane 4) resulted in the production of an abundant protein with a molecular mass of ~75 kD, corresponding to the size predicted for the two plant proteins. In a similar manner, infection with recombinant virus pFASTBAC/PGHS-2 resulted in the accumulation of an abundant protein of ~70 kD in accordance with the molecular mass of the PGHS-2 protein (lane 5). None of the three proteins described was found to accumulate in extracts obtained from mock-inoculated cells (lane 6) or in cells infected with recombinant baculovirus prepared with an empty pFASTBAC vector (lane 2).
PGHS possesses two enzymatic activities, a cyclooxygenase that catalyzes the oxygenation of polyunsaturated substrates, such as arachidonic acid, to form prostaglandin G2 (PGG2) and a peroxidase that can use a variety of electron donors to reduce PGG2 to form prostaglandin H2 (PGH2). Both enzymatic activities were assayed in the protein extracts described above, and the results obtained from these analyses are shown in Table 1. Cyclooxygenase activity was measured in the presence of arachidonic acid (20:4) and other plant unsaturated fatty acids, such as linolenic acid (18:3) and linoleic acid (18:2). Oxygen consumption was demonstrated for the three substrates examined when protein extracts were prepared from insect cells infected with the pFASTBAC/A5.2, pFASTBAC/N38086, or pFASTBAC/PGHS-2 recombinant baculovirus; however, very weak or no activity was detected in samples prepared from mock-inoculated virus or in cells infected with the pFASTBAC recombinant virus.
Sequence–Structure Alignment between the Tobacco A5.2 Protein and Ovine PGHS-1.
(A) Schematic representation of aligned structural features. Distribution of α helices (red boxes) is shown. Positions corresponding to nonaligned PGHS-1 protein domains, the EGF-like domain (EGF), and the membrane-binding domain (MB) are shown by light blue boxes. The position of the PGHS-1 signal peptide (SP) is also indicated. ovi PGHS-1, ovine PGHS-1.
(B) Alignment of the predicted A5.2 structural features and the three-dimensional structure of ovine PGHS-1 was performed using the Topits program (Rost, 1995). Homology regions, represented by thick lines, extend over the entire catalytic domain, where distribution of the α helices is shown in red. Amino acid residues found to be important for enzyme activity (dark blue) and the aspirin target (green) are marked. Nonaligned regions within the catalytic domain are represented as thin gray lines. Ovals (short regions with a thicker shape) represent insertions in the tobacco protein. The position of two short predicted β sheets is shown in yellow. Positions corresponding to nonaligned PGHS-1 protein domains (the EGF-like domain and the membrane-binding domain) are shown by thin light blue lines.
The enzymatic activity determined for the plant-encoded products reached levels similar to that of PGHS-2; however, the substrate specificity varied between them. Thus, cells expressing the A5.2 and N38086 proteins showed maximum activity with the plant polyunsaturated substrates linolenic and linoleic acid, whereas activity was reduced by ~70 to 50% when using arachidonic acid. The activity determined for the PGHS-2 protein with arachidonic acid was similar to that demonstrated in transfected cells for these enzymes (DeWitt et al., 1990; Fletcher et al., 1992). In addition, we show here that the PGHS-2 enzyme has weak activity when using linolenic acid (18:3) or linoleic acid (18:2) as a substrate and that activity was ~22% of the activity determined with arachidonic acid.
In contrast to these results, hydroperoxidase activity was present only in protein extracts from pFASTBAC/PGHS-2–infected cells in which the level of activity was similar to that assigned to other PGHSs from different animal sources (DeWitt et al., 1990; Fletcher et al., 1992). The absence of hydroperoxidase activity in extracts prepared from the pFASTBAC/A5.2– and pFASTBAC/N38086–infected cells suggested that the plant-encoded proteins lack this activity. Alternatively, these data could reflect some functional alteration associated with the expression system utilized. However, we believe it unlikely because the activity assayed in the PGHS-2–expressing cells demonstrated that at least in this case, the expression in insect cells does not interfere with protein functionality.
According to the structural analysis shown in Figure 5, the Ser-567 of the tobacco A5.2 protein could be the amino acid residue equivalent to the Ser-530 of PGHS-1, which is known to be acetylated by aspirin, causing irreversible inactivation of the cyclooxygenase activity (DeWitt et al., 1990; Picot et al., 1994). This observation was not supported by our functional analyses in which we found that the oxygenase activity of the two plant proteins under study was not reduced by aspirin. Thus, whereas enzyme activity in protein samples prepared from pFASTBAC/PGHS-2–infected cells was inhibited to ~34% when extracts were incubated for 10 min with 0.1 mM aspirin before activity determination, no enzyme activity was lost when samples containing the plant proteins were preincubated for at least 30 min with either 0.1 or 1 mM aspirin (data not shown). In a similar manner, no reduction in enzyme activity was detected when extracts from insect cells expressing the plant proteins were incubated with salicylic acid (SA) at the same concentrations (data not shown).
Analysis of the Protein Enzymatic Activity in Insect Cells.
The results of SDS-PAGE analysis of insect cell extracts after infection with recombinant baculovirus are shown. Lane 1 contains molecular mass markers (indicated in kilodaltons); lane 2, cells containing the pFASTBAC vector; lane 3, cells containing the pFAST-BAC/A5.2 construct; lane 4, cells containing the pFASTBAC/N38086 construct; lane 5, cells containing the pFASTBAC/PGHS-2 construct; and lane 6, mock-infected cells.
Enzymatic Activity Assaya
Despite the differences described, the results given above reveal that in addition to the structural similarity, A5.2 and PGHSs are functionally related proteins and that A5.2 encodes a functional oxygenase enzyme. Therefore, we designated A5.2 as PIOX for pathogen-induced oxygenase.
Induction of piox Expression in Response to Bacterial Inoculation Is Correlated with Protein Accumulation
To determine whether induction of piox expression in response to bacterial inoculation could be correlated with an increase of protein levels in the pathogen-responding plant tissues, we evaluated the abundance of the corresponding protein by using protein gel blot analyses with an antiserum raised against the piox-encoded protein. As shown in Figures 7A to 7C, a protein band of ~75 kD, corresponding to the size predicted for the piox product, was found to accumulate in plants responding to bacterial inoculation. Immunoreactive protein was detected in healthy untreated leaves, increasing significantly 24 hr after inoculation with the incompatible bacterium P. s. syringae. Protein accumulation was slightly delayed during the compatible interaction in which a strong increase in protein level was observed 48 hr after P. s. tabaci inoculation. In both interactions, protein accumulation persisted for at least 8 days after inoculation, reaching higher levels during the compatible interaction. In contrast to these results, the level of protein detected in untreated tissues did not show any significant variation in plant extracts prepared after water inoculation, which was used as a control in these experiments (Figure 7C).
Analysis of PIOX Protein Accumulation in Response to Bacterial Infection.
Shown are protein gel blots of fractionated extracts probed with the anti-PIOX antiserum. Extracts were prepared at different intervals in response to the following treatments.
(A) Inoculation with the incompatible bacterium P. s. syringae.
(B) Inoculation with the virulent bacterium P. s. tabaci.
(C) Water inoculation.
The position of an immunoreactive protein band of ~75 kD detected in the plant extracts is indicated. The position of molecular mass markers is indicated at left in kilodaltons.
Analysis of piox Expression in Response to Signal Molecules Mediating the Plant Response to Pathogen Infection
The role of SA as a cellular signal affecting the plant response to pathogen infection has been well documented (reviewed in Chen et al., 1995; Ryals et al., 1996). In addition, jasmonic acid (JA) may mediate an SA-independent pathway leading to induction of specific defense-related genes (Penninckx et al., 1996). To evaluate whether either of these two cellular signals had any effect on piox expression, we examined the level of piox transcripts in leaves of tobacco plants after exogenous application of SA (1 mM) or JA (50 μM). As shown in Figure 8, expression of piox was induced in response to the two treatments examined. Weak levels of transcripts were detected 1 hr after JA inoculation and reached maximum RNA accumulation when plants were analyzed 4 hr after inoculation. Activation of gene expression by SA was slightly delayed, and a weak hybridization signal was initially observed in RNA samples extracted 2 hr after treatment, whereas maximum accumulation was seen only after 8 hr.
In addition to pathogen infection, JA is known to play an important role in the signaling pathway controlling the plant response to mechanical stress (Farmer and Ryan, 1990; Bergey et al., 1996). Therefore, it was of interest that activation of piox expression preceded the response to JA treatment. Whereas the expression of this gene reached its highest measured point 4 hr after JA treatment, it took only 2 hr to accumulate to that level after mechanical damage (Figure 8).
Analysis of piox Expression in Response to Various Cellular Signals Known to Mediate the Plant Defense Response to Pathogen Infection.
Induction was examined in response to SA, wounding (Wound), JA, paraquat, 3AT, xanthine–xanthine oxidase (X-XO), or glucose–glucose oxidase (G-GO). See Methods for a detailed description.
The AOS generated during the plant defense response also have been proposed to play a significant role as early signal molecules that trigger defense gene expression. In particular, both H2O2 and the superoxide anion have been found to act as important components of the HR reaction and as inducers of gene expression (Sutherland, 1991; Levine et al., 1994; Tenhaken et al., 1995). To determine whether these chemical species also induced piox expression, we subjected tobacco leaves to various abiotic treatments that generate reactive oxygen derivatives. Gene expression was analyzed after leaves were injected with paraquat or 3-amino-1,2,4-triazole (3AT) to produce intracellular superoxide and H2O2, respectively. In addition, plants were subjected to treatment with the extracellular superoxide- and H2O2–generating systems xanthine–xanthine oxidase and glucose–glucose oxidase, respectively. As concluded from the increase in RNA level shown in Figure 8, induction of piox expression occurred in response to the four treatments examined. In all cases, high levels of piox transcripts were found 4 hr after initiation of each particular treatment. The level of RNA was maintained for the first 8 hr and decreased to almost undetectable levels in plants examined 16 hr after chemical inoculation. The only exception was the response found to paraquat, which maintained high levels of RNA for at least the first 24 hr after treatment.
It is of interest that with the exception of paraquat and wounding, none of the treatments assayed in our study produced any significant symptoms in the tissues inoculated; therefore, the activation of piox expression is not necessarily associated with the formation of visible cell death necrotic lesions.
DISCUSSION
Here, we have characterized a new plant gene, piox, encoding a protein with homology to COX, a key enzyme in the synthesis of lipid signal molecules that mediate different cellular processes in vertebrates, including the immune response to pathogen infection. piox was identified by investigating the cell death response induced by injecting tobacco leaves with the harpin HrpN protein from E. amylovora. HrpN, like other characterized harpin proteins (reviewed in Alfano and Collmer, 1996), can elicit an HR-like necrosis in the leaves of several plant species (Wei et al., 1992) and the induction of some HR defenses characteristically produced in response to bacterial infection (Baker et al., 1993; Gopalan et al., 1996a). Despite these properties, the primary function of the various harpins that are characterized is still unclear.
We have shown that expression of piox in tobacco is highly induced in response to protein elicitor infiltration and to inoculation with the HR-inducing pathogenic bacteria E. amylovora. In addition, we showed that expression of two well-known defense-related genes, PR-1 and gn2 (encoding a basic β-1,3-glucanase), was also induced in response to protein inoculation and to bacterial infection. Interestingly, we observed that each particular gene (piox, PR-1, and gn2) showed a very similar pattern of expression in the two cell death responses that were examined, which suggests that irrespective of the specific role of the HrpN harpin in bacterial infection, the cell death induced by this elicitor is similar to bacterial-induced HR. This observation is in accordance with previous data that report changes in gene expression in plants responding to inoculation with harpins. Thus, the expression of two previously identified harpin-induced (hin) tobacco genes has been shown to be activated similarly by HR-eliciting bacteria (Gopalan et al., 1996a). Moreover, further similarities in gene activation could be inferred from the data that show the induction of plant resistance by the harpin protein of E. amylovora (Wei and Beer, 1996) and by the HrpZPss from P. s. syringae 61 (Strobel et al., 1996).
An additional conclusion from these experiments is derived from the observation that induction of PR-1 and gn2 expression was significantly slower than that of piox, suggesting that the PIOX protein could be part of the early response of plants to pathogen infection.
We have shown that expression of piox was induced not only by incompatible HR-inducing bacteria E. amylovora and P. s. syringae but also by the compatible strain P. s. tabaci. Similarly, the PIOX protein was found to accumulate in plant tissues responding to inoculation with the two types of bacterial strains. The fact that both RNA and protein accumulated more rapidly when infection resulted in an HR indicates that piox could play a more significant role in this type of plant defense. A major difference between resistance and susceptibility is the timing required for gene induction, and in this respect, the expression of various genes involved in plant defense responses has been shown to be induced early after inoculation with avirulent pathogens (Zhou et al., 1997). On the other hand, the observation that activation of piox expression persisted for longer periods of time during the compatible interaction, combined with the fact that protein levels rose slightly higher, might reflect a more significant participation of piox in this type of interaction. Further data on the role of piox in the plant response to pathogen infection will be obtained by generating transgenic plants with altered levels of the PIOX protein.
We found that the tobacco piox gene is induced in response to bacterial infection, whereas no mRNA was detected in healthy untreated leaves. In contrast to the absence of detectable mRNA, low levels of protein were detected in untreated control leaves. Currently, it is not known whether this protein is derived from the piox gene characterized here or corresponds to a distinct isoform that is weakly expressed in untreated control plants. Consistent with this observation, a low level of expression was reported for the tomato feebly gene in healthy tissues (van der Biezen et al., 1996). Gel blot analysis of plant DNA with a radiolabeled piox probe under high-stringency hybridization conditions revealed piox to be present as a single-copy gene in tobacco (data not shown). Low-stringency hybridization conditions will now be required to investigate the existence of additional piox genes in tobacco plants.
Our sequence analysis showed that piox encodes a protein with homology to COXs, or PGHSs. Our analysis also showed that the predicted structure of PIOX has an overall good fit with the sequences of animal PGHSs. Structural similarity was found within the catalytic domain and the amino acid residues required for activity. It is presumably significant that the two PGHS protein domains falling outside of the catalytic region, the EGF-like domain and the membrane-binding domain, do not have counterparts in the tobacco protein.
By expressing piox in insect cells, we have determined directly that the encoded product possesses at least one of the two enzymatic activities of PGHSs catalyzing the oxygenation of polyunsaturated fatty acids. This activity is similar to that of PGHS-2, which reached maximum levels with arachidonic acid (20:4), known to be the substrate of PGHSs in vertebrates. The plant proteins examined showed maximum enzymatic activity with linolenic acid (18:3) and linoleic acid (18:2), suggesting that these plant compounds could be endogenous substrates for the plant PIOX. This suggestion is consistent with the fact that arachidonic acid is not found in plant membranes. These data suggest that indeed the plant proteins identified here could be functional equivalents of PGHSs. Further analyses are necessary, however, to investigate the absence of hydroperoxidase activity in extracts from insect cells expressing the plant proteins. Studies with plant-purified PIOX will help to examine its enzymatic properties, providing definitive evidence on whether the endogenous protein lacks this activity.
COX as well as lipoxygenase are key enzymes that mediate the conversion of polyunsaturated substrates, such as arachidonic acid, to prostaglandins or leukotrienes and lipoxins, respectively (Nelson and Seitz, 1994; Serhan et al., 1996). Both types of molecules serve as intracellular and/or extracellular lipid-derived signals and mediate many cellular responses in mammalian species, such as the immune response, fever, pain, and inflammation. They most likely play roles in mitogenesis and apoptosis as well. Lipoxygenases have been found in both plant and animal kingdoms. In plants, lipoxygenase catalyzes the dioxygenation of polyunsaturated octadecanoid fatty acids, such as linoleic (18:2) and linolenic acid (18:3), to produce a series of signal molecules important in host–pathogen interactions (Croft et al., 1993; Véronési et al., 1996). Many of those molecules are similarly induced by mechanical wounding, and some of them have been shown to act as potent activators of defense gene transcription (Farmer and Ryan, 1990; Reinbothe et al., 1994).
In contrast to the well-characterized plant lipoxygenases, no protein equivalent to COX has been found before in plants, and the lipoxygenase pathway has been considered to be the only one of the two pathways shared between plant and animal cells (Bergey et al., 1996). That PIOX does not correspond to a lipoxygenase was concluded by its low similarity (~11% identity) with known plant lipoxygenases (Véronési et al., 1996). In addition, the increase in absorbance at 234 nm typical for the conjugated-diene structure that results from the activity of lipoxygenase on its fatty acid substrates (Shimizu et al., 1990) was not detected when PIOX and PGHS-2 activities were assayed (data not shown).
Finding a new plant protein with homology to COX allows us to propose the likelihood of similarities, in addition to those previously reported, between the lipid-based cellular responses of plant and animal systems. In a manner similar to COXs and lipoxygenases, the PIOX plant protein could be involved in the synthesis of lipid-derived signal molecules that mediate the plant response to pathogen infection. In addition, a role in the metabolism of polyunsaturated fatty acids released during membrane deterioration also could be consistent with the expression of piox in pathogen-infected tissues.
We have shown that expression of piox is induced in response to various chemicals known to act as or to generate cellular signals mediating the plant's response to pathogen infection. Thus, piox was induced rapidly in response to SA or JA inoculation and to the production of AOS, such as superoxide radicals or H2O2. This rapid and generalized induction was not found, however, for PR-1 and gn2, whose expression was examined after similar and simultaneous treatments (data not shown). According to previously reported results (Chen et al., 1993), PR-1 transcripts accumulate in response to SA, paraquat, or 3AT; however, we found that the level of PR-1 mRNA did not increase significantly when extracellular superoxide or H2O2 was generated, and the level was increased only slightly when the plant was inoculated with xanthine–xanthine oxidase. Expression of gn2 was induced by SA or JA, but no significant expression was found by generation of AOS (data not shown).
The rapid response of piox to all of the cellular signals examined in this study might indicate that PIOX synthesis participates in the initial response of plants to pathogen infection. Based on our studies, piox expression could be activated by formation of the AOS known to be generated in response to most of the treatments examined here. Thus, in addition to the chemical treatments that specifically generate AOS, it is known that SA binds to and inhibits a catalase and that the increase of H2O2 produced could serve to potentiate the oxidative burst in tissues undergoing an HR. In addition, because peroxide treatment can lead to SA accumulation, the existence of a runaway cycle resulting in very high free radical concentrations has been proposed (Neuenschwander et al., 1995). Mechanical stimuli, such as direct physical pressure, can also evoke a characteristic oxidative burst in soybean cells (Yahraus et al., 1995). The fact that expression of piox was activated by formation of oxygen species could account for the transient accumulation of transcripts found in response to some of the treatments examined here. This suggestion is based on the data reported by Chandra et al. (1996), which show that the oxidative burst produced in response to pathogen infection or elicitor treatment can occur in one or two phases and that the timing and the intensity of the response differ for each particular interaction.
Production of AOS is known to occur during early steps of pathogenesis and during later stages associated with lipid peroxidation. The strong reactivity of the oxygen species leads to nonenzymatic membrane lipid peroxidation, which in turn results in electrolyte leakage, disruption of cell membranes, and subsequent removal of unsaturated fatty acids from the bilayer (Keppler and Baker, 1989; May et al., 1996). Interestingly, a similar mechanism is known to occur during the course of neutrophil and macrophage activation (Segal and Abo, 1993; Jacobson, 1996). During activation, the increase of extracellular oxygen derivatives gives rise to the release of arachidonic acid from the membrane and to the synthesis of lipid-derived signal molecules, such as prostaglandins and other related compounds, through an enzymatic pathway involving phospholipases, COXs, and lipoxygenases (Serhan et al., 1996). As in animal systems, the synthesis of lipid-derived signal molecules is thought to be initiated by release of polyunsaturated fatty acids from plant membranes by the action of phospholipases and lipoxygenases (Bergey et al., 1996). In addition, specific lipid-derived molecules mediating the plant response to pathogen infection could be generated through the enzymatic activity of the PIOX protein identified here.
METHODS
Plant Material
Tobacco plants used in this study were wild-type Nicotiana tabacum cv Petit Havana SR1 grown on soil under a 14-hr-light and 10-hr-dark photoperiod. Plants were treated and examined at ~8 weeks after seed germination.
Recombinant DNA Techniques
Standard DNA techniques were performed as described by Ausubel et al. (1988) and Sambrook et al. (1989). Plasmid vectors pT3T7 and pBluescript SK+ (Stratagene, La Jolla, CA) were used for DNA cloning, and Escherichia coli DH5α was used for all of the in vivo DNA transformation experiments. Competent cells for bacterial transformation were prepared according to Inoue et al. (1990). Restriction fragments for cloning were purified from agarose gels by centrifugal filtration (Zhu et al., 1985). DNA sequence was determined by using the chain termination method (Sanger et al., 1977). Oligonucleotides were synthesized using a DNA synthesizer (model 380B; Applied Biosystems, Foster City, CA).
Construction and Purification of an MBP–HrpN Fusion Protein
Erwinia amylovora PMV6089 (Brisset and Paulin, 1992), producing an incompatible hypersensitive response (HR) in tobacco leaves, was used as a source of DNA to amplify by polymerase chain reaction (PCR) the hrpN gene. Bacterial DNA was isolated as described by Ausubel et al. (1988). Based on the reported nucleotide sequence (Wei et al., 1992; GenBank accession number M92994), two oligonucleotides primers, 5′-CCGGGGATCCATCGAGGGTAGGATGAGTCTGAATACAAGTGGG-3′ and 5′-CAGGATCCATCATGGCATCAATACC-3′, containing convenient BamHI sites, were synthesized and used to amplify the hrpN gene. After BamHI digestion, the PCR-amplified product was purified from an agarose gel and inserted into the BamHI site of the E. coli expression vector pMAL-C (New England Biolabs, Beverly, MA). The cloned gene was inserted downstream from the malE gene, which encodes a maltose binding protein (MBP), resulting in the expression of an MBP–HrpN fusion protein. The amplified hrpN gene was verified as correct by sequencing. The vector pMAL-C contains the strong Ptac promoter, which is induced by adding isopropyl β-d-thiogalactopyranoside. The recombinant plasmid pMAL-C/hrpN was transferred to E. coli TB1 and grown at 37°C to a cell density of 0.5 (A600). After the addition of isopropyl β-d-thiogalactopyranoside (0.3 mM final concentration), incubation was continued for 2 hr more, and cells were harvested by centrifugation. Bacterial cells were resuspended in lysis buffer (10 mM phosphate, 30 mM NaCl, 0.25% Tween 20, 10 mM β-mercaptoethanol, 10 mM EDTA, and 10 mM EGTA), frozen in a dry ice–ethanol bath, thawed in cold water, and sonicated. Protein expression was monitored by SDS-PAGE, and the resulting MBP–HrpN fusion protein was purified by retention on an amylose column followed by protein elution with maltose, according to the manufacturer's instructions (New England Biolabs). The MBP protein expressed from the pMAL-C vector was similarly purified and used as a control for the analysis of the HR induction. The ability of the purified MBP–HrpN fusion to induce the HR was examined by treating tobacco leaves. Before inoculation, protein was dialyzed against 5 mM phosphate buffer, pH 6.5, and 0.1 mM phenylmethylsulfonyl fluoride (PMSF). Plant leaves were infiltrated by pressuring the protein fraction into the apoplast, using a syringe.
Differential Display
Differential display analysis (Liang and Pardee, 1992) was performed using total RNA extracted from tobacco leaves, according to Jones et al. (1985), 4 hr after inoculation with a purified MBP–HrpN protein fraction at a concentration of 5 μM. Independently prepared RNA samples were subjected to this analysis to select the cDNA bands that were reproducibly induced in response to the treatment examined. RNA from healthy untreated leaves was used as a control. After DNase treatment, 2 μg of each total RNA sample was used for reverse transcription and subsequent PCR by using an RNAmap kit (GenHunter Corporation, Brookline, MA). Each RNA sample was reverse transcribed using T12MA (GenHunter) followed by PCR amplification in the presence of T12MA and any of each of the five arbitrary 10mers, AP-1, AP-2, AP-3, AP-4, and AP-5, provided with the RNAmap kit. Forty cycles of PCR were performed at 94°C for 30 sec, 40°C for 2 min, and 72°C for 30 sec. Reaction products were separated on 6% sequencing gels and visualized by autoradiography. Bands reproducibly induced in the two independent RNA samples examined were excised and reamplified according to the manufacturer's protocol. Reamplified products were purified from an agarose gel, cloned into the pT3T7 vector (Stratagene), sequenced, and used as probes for RNA blot analysis and cDNA library screening.
cDNA Library Construction and Screening
Poly(A)+ mRNA was prepared from total RNA extracted from tobacco leaves 4 hr after inoculation with a purified MBP–HrpN protein fraction. Complementary DNA was synthesized and cloned into the EcoRI-XhoI sites of the λ ZAP II vector (Stratagene), according to the manufacturer's instructions. Ligation mix was packaged using a Giga-pack II Gold packaging kit (Stratagene). Approximately 106 clones were screened with a radioactive riboprobe that was generated using the Stratagene RNA transcription kit. Filters (Hybond N; Amersham) were hybridized overnight at 42°C in 50% formamide, 5 × SSC (1 × SSC is 0.15 M NaCl and 0.015 M sodium citrate), 5 × Denhardt's solution (1 × Denhardt's solution is 0.02% Ficoll, 0.02% PVP, and 0.02% BSA), 0.5% SDS, and 20 μg/mL denatured herring sperm DNA and washed twice with 5 × SSC and 0.1% SDS at 42°C for 30 min and then twice with 2 × SSC and 0.1% SDS at 42°C for 30 min. Filters were exposed to Hyperfilm MPfilm (Amersham) with an intensifying screen at −80°C for 24 hr. Positives clones were excised with the helper phage and recircularized to generate a subclone in the pBluescript SK− phagemid vector (Stratagene).
The rapid amplification of cDNA ends (RACE) protocol was used to clone the 5′ end region of the longest A5.2 cDNA by using the Marathon cDNA amplification kit (Clontech, Palo Alto, CA). Total RNA was isolated from induced tobacco leaves (16 hr after bacterial inoculation) that showed HR, and poly(A)+ RNA was purified from total RNA with oligo(dT)–Latex (Qiagen, Chatsworth, CA). One microgram of mRNA was reverse transcribed using the cDNA synthesis primer according to the manufacturer's instructions. Thirty cycles of PCR were performed at 94°C for 30 sec and 68°C for 4 min with 10 pmol of 5′ end Clontech AP1 adapter and 10 pmol of 5′ end amplification primer (nucleotides between 621 and 602; 5′-CTTTAGTGGGCATTGACTTG-3′). The amplified product (622 nucleotides) was purified from an agarose gel and cloned into the pT3T7 vector for sequence analysis. To avoid PCR artifacts, we performed two separate PCRs, and the sequence of three independent clones from each reaction was determined for both strands by using appropriated internal primers and an automated DNA sequencer (model 373A; Applied Biosystems). After our analysis, only bases present in six of the PCR products were considered to be genuine. A complete A5.2 cDNA clone was constructed by using a unique AflII restriction site present in an overlapping region of the two cDNA fragments.
Sequence Analysis
Sequences were analyzed using the Genetics Computer Group (Madison, WI) package (version 9.0). Database searches involved BLAST (Altschul et al., 1990), FASTA (Pearson and Lipman, 1988), and the GenQuiz system (Scharf et al., 1994). Sequence–structure alignment was performed by using the Topits program (Rost, 1995). Topits optimizes the match between a given sequence and its corresponding two-dimensional prediction with a known three-dimensional structure.
Generation of Recombinant Baculoviruses: Analysis of Protein Expression and Enzyme Activity in Insect-Infected Cells
Recombinant baculoviruses expressing the genes of interest were generated by using the Bac-to-Bac baculovirus expression system (Gibco BRL). A 2.1-kb DNA fragment containing the entire A5.2 cDNA was purified from an agarose gel and inserted, behind the polyhedrin promoter, into the BamHI and EcoRI restriction sites of the pFASTBAC1 vector, and a recombinant plasmid, pFASTBAC/A5.2, was generated. A second recombinant plasmid, pFASTBAC/N38086, was similarly created by cloning the Arabidopsis thaliana N38086 cDNA (2.3 kb) into the PstI and SphI restriction sites of the pFASTBAC1 vector. A third recombinant plasmid, pFASTBAC/PGHS-2, was obtained by inserting a prostaglandin endoperoxide synthase (PGHS) cDNA (4 kb) from rat into the EcoRI restriction site of pFASTBAC. Recombinant plasmids were transferred into competent DH10Bac E. coli cells (Gibco BRL) containing the baculovirus shuttle vector bMON14272 and the helper plasmid pMON7124. In addition, the pFASTBAC vector was also transferred to competent DH10Bac cells. Recombinant bacmid DNA was then prepared from positive bacterial colonies in which the construct of interest had been transposed to the bacmid vector, and recombinant baculoviruses were obtained by transfecting the corresponding bacmids into High Five insect cells (Invitrogen), according to the manufacturer's instructions. Recombinant viruses were grown and titrated as previously described (O'Reilly et al., 1992).
Expression analyses were performed by infecting cultures of High Five cells, grown in TC100 medium supplemented with 10% fetal calf serum (Gibco BRL and Life Technologies), with the recombinant viruses at a multiplicity of infection of five plaque-forming units per cell. At 48 hr after infection, cells were collected and subjected to centrifugation at 3000g for 5 min. Cell pellets were washed twice with ice-cold PBS resuspended in 10 mM Tris-HCl, pH 8, 1 mM EDTA, 5% glycerol, 1 mM PMSF, 10 μg/mL pepstatin, 10 μg/mL chymostatin, and 10 μg/mL leupeptin and disrupted by sonication. After centrifugation at 12,000g for 5 min, supernatants were used for protein expression analysis by SDS-PAGE and for enzyme activity determination. Protein concentration was determined by the method of Bradford (1976), using the Bio-Rad reagent. Cyclooxygenase assays were performed at 30°C by monitoring O2 uptake with a Clark O2 electrode. For enzyme activity determination, cell extracts typically were incubated in 0.1 M Tris-HCl, pH 8, 1 mM phenol, 25 μg of hemoglobin, and 100 μM substrate (DeWitt et al., 1990). Consumption of O2 was evaluated with substrates such as arachidonic (20:4), linolenic (18:3), and linoleic (18:2) acids, and reactions were initiated by adding the enzyme preparation. Cyclooxygenase activity was expressed as the nanomoles of O2 consumed per minute per milligram of cell extract. Hydroperoxidase assays were performed spectrophotometrically to measure the oxidation of tetramethylphenylenediamine at 611 nm (DeWitt et al., 1990). Each assay mixture contained 0.1 M Tris-HCl, pH 8, 200 μM H2O2, 1 μM hematine, 100 μM tetramethylphenylenediamine, and 50 μg of protein extract.
Antibody Production and Immunoblot Analysis
Polyclonal antibodies generated against the A5.2-encoded product were raised in rabbits by injecting them with protein electroeluted from SDS–polyacrylamide gels as described by Harlow and Lane (1988). Total protein extracts from A5.2 insect-expressing cells were separated by gel electrophoresis, visualized by Coomassie Brilliant Blue R 250 staining, and eluted by using a Bio-Rad Electro-Eluter apparatus. Antigen was dialyzed against 0.2 M sodium bicarbonate and 0.02% SDS and lyophilized, and purity was verified by electrophoresis and Coomassie blue staining. Eluted protein was injected into New Zealand white rabbits according to standard procedures. Acetone powder prepared from control noninfected insect cells according to Harlow and Lane (1988) was used to remove any nonspecific binding from the immunosera previous to its use.
Protein extracts of plant tissues for immunoblot analyses were prepared as follows. Leaves from healthy, untreated, and inoculated plants were collected at the indicated times and immediately frozen in liquid nitrogen before protein extraction. Leaves were ground to a fine powder in liquid nitrogen by using a mortar and pestle. Extraction buffer (100 mM Tris-HCl, pH 8, 5% glycerol, 150 mM NaCl, and 1 mM PMSF) was then added at a ratio of 2 mL/g fresh tissue. Supernatants were clarified by centrifugation for 10 min at 3000g, and 25 μg of total protein was separated on 10% SDS–polyacrylamide gels, transferred to nitrocellulose membranes by electroblotting, and used for immunoblot analyses. Membranes were washed three times with TBS buffer (20 mM Tris, pH 7.5, 150 mM NaCl, and 0.1% Tween 20) and blocked for at least 1 hr at room temperature in the same solution with 5% nonfat milk. Blots were incubated for 4 hr at room temperature in TBS buffer with 1% nonfat milk and anti-A5.2 antibodies (1:1000 dilution). Membranes were then washed three times in TBS buffer and incubated with a horseradish peroxidase–conjugated secondary antibody (1:5000 dilution; Sigma). Complexes were visualized using an enhanced chemiluminescence kit (Amersham) and following the manufacturer's instructions.
Analysis of Gene Expression
RNA from leaves was prepared according to Logemann et al. (1987). Total RNA (10 μg per lane) was electrophoresed in 1.5% agarose–formaldehyde gels according to Sambrook et al. (1989) and transferred to Hybond N membranes. The amount of loaded RNA was verified by the addition of ethidium bromide to the samples. Photography was under UV light after electrophoresis. Single-stranded riboprobes derived from the tobacco cDNA characterized were generated using the Stratagene RNA transcription kit. In addition, riboprobes were prepared from an AhaIII fragment of gn2, a basic β-1,3-glucanase gene from N. plumbaginifolia (Gheysen et al., 1990), and from tobacco PR-1 (Ward et al., 1991). Blots were hybridized in 50% formamide, 5 × SSC, 5 × Denhardt's solution, 0.5% SDS, and 20 μg/mL denatured herring sperm DNA. Hybridizations were washed twice with 5 × SSC and 0.1% SDS for 30 min, twice with 1 × SSC and 0.1% SDS for 30 min, and twice with 0.2 × SSC and 0.1% SDS for 15 min. Hybridization and washes were performed at 65°C.
Study of Expression in Response to Plant Stress
Plant leaves were infiltrated in the apoplast by injecting the protein fraction MBP–HrpN or MBP at a concentration of 5 μM. Bacterial inoculation was performed by injecting leaves with a late-logarithmic culture by using a syringe. The bacteria were grown in King's medium (2% proteose peptone, 2% glycerol, 6.5 mM K2HPO4, and 6 mM MgSO4 7H2O) at 27°C. After centrifugation, pellets were resuspended in sterile water to reach a final concentration of 106 bacteria per mL. The following bacterial species were used: E. amylovora PMV6089 (Brisset and Paulin, 1992), Pseudomonas syringae pv syringae (NCPPB2686), and P. s. pv tabaci (NCPPB1427). Plants were wounded by gently rubbing the upper epidermis of leaves with wet Carborundum. For chemical treatment, leaves were injected with water, salicylic acid (SA; 1 mM), jasmonic acid (JA; 50 μM), 3-amino-1,2,4-triazole (3AT; 4 mM), paraquat (100 μM), xanthine–xanthine oxidase (2 mM to 0.5 units per mL), or glucose–glucose oxidase (6 mM to 250 units per mL). Tissues for RNA extraction were harvested at the indicated times and immediately frozen in liquid nitrogen before RNA extraction.
ACKNOWLEDGMENTS
We are indebted to Dr. Jose F. Rodriguez for help with virus infections. We thank Florencio Pazos and Dr. Alfonso Valencia (Centro Nacional de Biotecnología, Protein Design Group) for protein structural analysis. We thank Dr. Allan Caplan for critical reading of the manuscript and Dr. Jose J. Sanchez-Serrano for helpful discussions. We also thank Dr. Jean-Pierre Paulin for the PMV6089 strain of E. amylovora; Dr. John Ryals for the PR-1 cDNA and Dr. Daniel Hwang for the rat PGHS-2 cDNA; Tomas Cascón for excellent technical assistance; and Angel Sanz and Ines Poveda for expert photography. The expressed sequence tags were obtained from the Arabidopsis Biological Resource Center (Ohio State University, Columbus). This work was supported by a grant from the Education and Science Ministry (CICYT, BIO94-1355); A.S. was supported by a CSIC fellowship.
- Received May 21, 1998.
- Accepted July 17, 1998.
- Published September 1, 1998.
REFERENCES
