INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Induced plant resistance to pathogens involves a complex array of biochemical and structural alterations in the plant cell. Induced expression of a large number of defense-related genes is essential for plants to counter pathogen infections. This has been appreciated since the identification of pathogenesis-related (PR) proteins in virus-infected plants (Van Loon and Van Kammen, 1970). Many defense-related genes encode proteins possessing antifungal or antibacterial activities or enzymes that catalyze defense metabolites (Bowles, 1990). Others encode regulatory proteins important for defense signal transductions (Eulgem et al., 1999; Glazebrook, 1999). The isolation and characterization of these genes are essential for our understanding of plant disease resistance mechanisms.

The tomato (Lycopersicon esculentum) disease resistance gene Pto confers gene-for-gene resistance to the bacterial pathogen Pseudomonas syringae pv tomato (avrPto). When constitutively expressed under the control of the CaMV 35S promoter, Pto induces the development of spontaneous microscopic lesions and PR gene expression in tomato plants in the absence of pathogen infections (Tang et al., 1999). The 35S::Pto plants display broad resistance to both bacterial and fungal pathogens. To facilitate the study of mechanisms underlying lesion formation and disease resistance, we have used suppression subtractive hybridization (SSH; Diatchenko et al., 1996) to isolate tomato cDNA corresponding to 35S::Pto-induced transcripts. Characterization of 82 unique cDNA clones indicated that a large number of genes belonging to diverse pathways are induced in 35S::Pto plants. It is notable that there were a large number of genes encoding PR proteins, proteins with a putative role in hypersensitive reaction (HR)/lesion development, and a number of proteins related to the oxidative burst, proteolysis, signal transduction, and lipid metabolism. Furthermore, 10 cDNA clones have no match with the tomato expressed sequence tag (EST) database, which contains 107,000 entries as of February 2001. Consistent with the role of Pto in disease resistance, many of these genes induced by the 35S::Pto transgene were also induced by pathogen inoculation. The isolation of these cDNA clones that are related to a diverse array of functions should assist in the characterization of defense pathways regulated by Pto.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Characterization of the Subtracted cDNA Library

To isolate cDNA for transcripts that accumulated in 35S::Pto transgenic plants, an SSH library (Diatchenko et al., 1996) was made with the tester cDNA from 35S::Pto line 48 (Tang et al., 1999), which consistently produces uniform microscopic lesions on leaves, and driver cDNA from isogenic non-transgenic tomato cv Money Maker plants. The library comprised approximately 2,000 clones, the majority of which carried fragments of 300 to 1,000 bp. Northern-blot analysis was conducted with six randomly selected clones from the cDNA library (Fig. 1). Transcripts corresponding to all six clones were more abundant in the 35S::Pto line, indicating that the majority of clones in the library represent transcripts with increased abundance in plants that overexpress Pto. Initial sequencing of the library showed that several cDNA clones were highly redundant in the library. These encode catalases, a Gly-rich cell wall protein, and several members of the PR1 family of proteins. These cDNA clones, together with several known tomato PR gene cDNA clones (PR1a1, PR1b1, GluB, Osmotin, and Chia; cited in Tang et al., 1999), were used as probes in colony hybridization in the subsequent experiments, and non-hybridizing clones were sequenced. This effectively removed the majority of redundant clones except for clone 440 that encodes GluB (Table I). A total of 190 cDNAs were sequenced. CLUSTALW analysis (Thompson et al., 1994) showed that they belonged to 82 nonredundant cDNA fragments.

View larger version (51K): [in this window] [in a new window]   Figure 1.   RNA expression of randomly selected clones from the SSH cDNA library confirms differential expression in 35S::Pto-transgenic plants. Ten micrograms of total RNA from tomato cv Money Maker plants with (+) or without () the 35S::Pto transgene was separated in a denaturing agarose gel, and duplicated RNA blots were hybridized to radiolabeled cDNA for each clone. Ethidium bromide staining from one of the duplicates indicates equal loading of the RNA (rRNA).

View this table: [in this window] [in a new window]   Table I.   cDNA clones isolated from 35S::Pto plants , No induction; +, 2- to 4-fold induction; ++, 5- to 9-fold induction; +++, greater than 10-fold induction; blank, not tested. Relative signal on the x-ray film was quantified by densitometry and normalized to constitutive controls (see "Materials and Methods") before fold induction was calculated.

The clones were characterized by "reverse northern" analysis (see "Materials and Methods"). Southern blots containing PCR products of the cDNA clones were probed with radiolabeled cDNA probes made from RNA extracted from non-transgenic tomato leaves or the 35S::Pto tomato leaves. Of 77 clones tested, 64 showed increased transcript expression in 35S::Pto plants (Table I). A representative example of the hybridization is shown in Figure 2. The remaining five clones were either not tested or tested with no informative results. The tight correlation with results from northern analysis (Fig. 1 and Table I) validates the "reverse northern" experiments.

View larger version (25K): [in this window] [in a new window]   Figure 2.   Reverse northern analysis of the subtracted library. PCR-amplified cDNA inserts were separated on agarose gels and transferred to nylon filters. Lane a contained the cDNA insert of clone 93 that encodes a chlorophyll a/b-binding protein. Lanes b through t contained cDNA inserts from 19 randomly selected clones. The duplicated filters contained equal amounts of the PCR products and were hybridized with cDNA probes synthesized from non-transgenic tomato cv Money Maker (MM) or tomato cv 35S::Pto transgenic mRNA.

We used the BLASTX program to search the GenBank database for proteins that are homologous to those encoded by the 82 clones (Table I). In a similar manner, the BLASTN program was used to search for tomato EST sequences identical to these cDNAs. The clones were grouped according to functions of the putative protein products. Sixty-two clones had significant similarity with genes encoding known proteins. These were placed into several functional categories including plant defense, oxidative burst, proteolysis, signal transduction, gene expression and regulation, lipid metabolism, stress response, ribosomal function, and others. Seven clones shared homology only with those encoding proteins of unknown functions. Fourteen clones did not match any protein sequences in the database. Furthermore, 18 clones were novel tomato cDNAs that did not have a match with the tomato EST database or previously reported tomato sequences.

Genes Induced by Bacterial Inoculation

We tested whether the clones from our cDNA library are relevant to defense responses in non-transgenic plants. Reverse northern analysis indicated that 35 of the 52 clones examined showed increased accumulation when tomato cv RioGrande PtoR plants that carry the resistance gene Pto were inoculated with P. syringae pv tomato carrying the corresponding avirulence gene avrPto. Northern analysis was conducted for 40 clones to further determine the bacterial inducibility. Twenty-four clones exhibited increased transcripts when plants were infiltrated with bacteria, whereas the remaining clones showed either no bacterial induction or did not yield detectable signals in the northern analysis (Table I). Figure 3 shows northern analysis of four clones that were tested repeatedly and showed early induction of transcripts following bacterial inoculation (1-3 h).

View larger version (91K): [in this window] [in a new window]   Figure 3.   Pathogen-induced expression of selected genes. Tomato cv Rio Grande PtoR plants were inoculated with 106 colony forming units (cfu) mL1 P. syringae pv tomato (avrPto) or 10 mM MgCl buffer (Mock). RNA was isolated at indicated times, and 10 µg RNA was separated in a denaturing agarose gel. Duplicated RNA blots were hybridized with indicated probes. Ethidium bromide staining from one of the duplicates indicates equal loading of the RNA (rRNA).

Involvement of Salicylic Acid (SA) in Gene Regulation

SA is globally involved in defense responses during plant-pathogen interactions (Ryals et al., 1996). 35S::Pto transgenic tomato plants accumulate high levels of SA and exhibit increased resistance to both bacterial and fungal pathogens (Tang et al., 1999). To test if any genes described here are induced by an SA-dependent pathway, we crossed transgenic tomato plants carrying the bacterial nahG gene, which encodes a salicylate hydroxylase, to the 35S::Pto transgenic tomato plants (Brading et al., 2000) and examined gene expression in F₁ plants. The hemizygous nahG/35S::Pto plants were indistinguishable from the hemizygous 35S::Pto plants in the expression of spontaneous lesions, and all plants displayed resistance to both the virulent strain and avirulent strain (avrPto) of P. syringae pv tomato (J. Zhou and X. Tang, unpublished data). Reverse northern hybridizations of nylon filter arrays containing the 82 cDNA clones showed that majority of the clones hybridized equally with the cDNA probe generated from 35S::Pto plants and that from 35S::Pto/nahG plants. However, four clones (14, 267, 554, and 561) showed reduced signal when probed with cDNA derived from 35S::Pto/nahG plants (data not shown). These clones encode a Gly-rich cell wall protein, sterol reductase, Pro-rich protein, and PR1a1, respectively. Northern-blot analysis confirmed that the expression of the four genes in 35S::Pto plants was reduced by the presence of the nahG transgene (Fig. 4).

View larger version (58K): [in this window] [in a new window]   Figure 4.   Effects of nahG transgene on the 35S::Pto-induced expression of genes. Homozygous 35S::Pto plants (line 48) were crossed to homozygous nahG plants to produce 35S::Pto/nahG hemizygous plants. As a control, homozygous 35S::Pto plants were crossed to non-transgenic tomato cv Money Maker plants to produce hemizygous 35S::Pto plants. RNA from the hemizygous plants was separated in a denaturing agarose gel, and duplicated RNA blots were hybridized with the indicated cDNA probes.

We further examined the involvement of SA in the P. syringae-induced expression of two genes (clones 271 and 398) that were induced early by bacterial inoculation. Repeated northern analyses indicated that transcripts of clones 271, which encodes a Cys protease, were induced more strongly following bacterial inoculation in nahG plants (Fig. 5), suggesting a negative role of SA in the expression of this gene. In contrast, the expression of transcripts of clone 398, encoding a fatty acid desaturase, was not significantly affected by nahG (Fig. 5), indicating that its induction is independent of SA accumulation.

View larger version (62K): [in this window] [in a new window]   Figure 5.   Effects of nahG on pathogen-induced expression of clones 271 and 398. Non-transgenic tomato cv Money Maker or tomato cv Money Maker plants containing the nahG transgene were inoculated with 106 cfu mL1 P. syringae pv tomato, and RNA was isolated at the indicated times. Duplicated RNA blots were hybridized with the indicated cDNA probes.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In this report, we describe the isolation and characterization of a significant number of cDNA clones corresponding to genes expressed at a elevated level in 35S::Pto plants. Two lines of evidence indicate that the majority of these genes are bona fide defense-related genes. First, a large number of clones (19) encode proteins with an apparent role in plant defense against pathogens. Considering that the cDNA library had been prescreened with six PR gene probes prior to sequencing, the actual number of clones coding for antimicrobial proteins could be larger than 20. This is a clear indication that the library is highly enriched for genes involved in defense responses. In addition, most of the clones that showed an increased expression in the 35S::Pto plants appeared to be induced by pathogen inoculation.

Most clones in the "defense-related protein" category had a match with known proteins that encode structural proteins for plant defense. These include 16 clones encoding various PR proteins that may play a direct role in inhibiting pathogens. Except for clones 60 and 502, which appear to be two fragments of the same gene for a PRb-1b-like protein, the remaining clones encode 14 distinct PR proteins (clones 255, 454, 561, 248, 307, 440, 480, 540, 569, 718, 243, 106, 503, and 616). Proteins encoded by clones 14 and 554 are cell wall proteins that probably are involved in cell wall fortification in plants. In addition, lipid transfer proteins encoded by clones 27 and 101 in the "lipid metabolism" category may also play a direct role in defense. It has been reported that lipid transfer proteins possess antibacterial activities (Caaveiro et al., 1997). Clone 301 encodes a protein similar to maize HR-associated protein. This protein is also similar to the tobacco NG8 that has been implicated to play a role in tobacco mosaic virus-induced HR (Karrer et al., 1998). An Arabidopsis homolog (accession no. 7269612) accumulates its transcripts in the mpk4 mutant that exhibits constitutively activated defense responses (Petersen et al., 2000). It appears that this class of proteins maybe involved in HR development in a variety of plants.

In addition to antimicrobial proteins, several classes of proteins encoded by the cDNA clones are potentially important for plant defense. Three clones encode proteases. It is notable that the Cys protease encoded by clone 271 and the FtsH protease encoded by clone 441 may be related to programmed cell death. Caspases, a group of Cys proteases, are important components in animal programmed cell death pathway (Green, 2000). In plants, inhibitors of Cys proteases can inhibit cell death triggered by avirulent P. syringae bacteria (Solomon et al., 1999). The Xanthomonas campestris Avr protein AvrBsT shares homology with Cys proteases, and mutation of the protease catalytic site of AvrBsT eliminates its ability to activate the hypersensitive response in plants (Orth et al., 2000). Clone 271 is also similar to Arabidosis SAG12 that is expressed during leaf senescence, a cell death program invoked during normal development of plants (Gan and Amasino, 1995). Clone 441 is similar to the FtsH class of metalloproteases that is conserved in both prokaryotes and eukaryotes. In tobacco, a chloroplast FtsH protein has been shown to be a negative regulator of tobacco mosaic virus-induced HR (Seo et al., 2000). Overexpression of this gene in tobacco attenuates the HR. Clone 728 encodes a subtilisin-like Ser protease that is induced by pathogen infection (P69B; Tornero et al., 1996, 1997). It will be important to determine if any of these genes are involved in programmed cell death during plant disease resistance.

A group of proteins encoded by the cDNA clones are related to lipid metabolism. These include lipid transfer proteins (clones 27 and 101), microsomal lipase (clone 487), -6 fatty acid desaturase (clone 398), and sterol delta-7 reductase (clone 267). In parsley (Petroselinum crispum), the treatment of a peptide elicitor from Phytophthora sojae greatly alters fatty acid profiles of the plant cell. This correlates with the elicitor-induced expression of transcripts of several fatty acid desaturases (Trezzini et al., 1993; Kirsch et al., 1997). Some of these events maybe involved in the accumulation of jasmonic acids that are important defense hormones. EDS1 and PAD4, two proteins that function in disease resistance gene-mediated pathways, are homologous to lipases, and their transcripts are induced upon pathogen infection (Falk et al., 1999; Jirage et al., 1999). The accumulation of transcripts related to lipid metabolism in 35S::Pto plants reinforces the significance of lipids in plant defense.

Several clones encode proteins involved in the generation or scavenging of oxidative stress. The NADH dehydrogenase encoded by clone 8 may have a role in generating reactive oxygen species, whereas the catalases encoded by clones 12, 20, and 568 may help the host cell to cope with reactive oxygen species accumulated during defense reactions. The transcripts represented by clones 8, 12, 20, and 568 accumulated in the 35S::Pto plants. At least those represented by clones 12 and 20 were induced by bacterial infection. Clone 486 encoding annexin was also induced by both pathogen and the 35S::Pto transgene. The expression of annexin also may play a role in scavenging reactive oxygen species. An Arabidopsis annexin-like protein has been shown to possess peroxidase activity, and the expression of its cDNA restored H₂O₂ tolerance to the Escherichia coli oxyR mutant that is unable to express catalase in response to oxidative stress (Gidrol et al., 1996). It is interesting that the Arabidosis Annexin-like gene is induced by H₂O₂ and SA, suggesting a role in defense responses. In contrast to clones 12, 20, and 486, clone 533, which encodes an ascorbate peroxidase, was expressed at a higher level in 35S::Pto plants but was not induced by pathogen infection. The accumulation of the ascorbate peroxidase transcripts maybe an indirect adaptive response to the reactive oxygen species accumulated during the lesion development of the 35S::Pto plants rather than a defense response activated by Pto.

The cDNA clones identified in this report represent genes that are induced by pathogen via a variety of signaling pathways. For example, clones 271, 301, 398, and 486 were induced early (1-3 h) after P. syringae inoculation, whereas clones 12, 14, 307, and 480, encoding catalase, Gly-rich cell wall protein, -1,3-glucanase, and class III acidic chitinase, respectively, are induced late (12 h) after bacterial inoculation (data not shown). A detailed study of two early induced genes showed that the pathogen induction of RNA corresponding to clone 271 appeared to be regulated negatively by SA. This could be a result of the antagonistic interaction between SA and ethylene. In fact, the pathogen-induced expression of transcripts corresponding to clone 271 is delayed in never ripe plants that are insensitive to ethylene (data not shown). In contrast, the transcripts corresponding to clones 301 and 398 did not appear to be affected by nahG. In addition, the induction of genes corresponding to clones 14, 267, 554, and 561, encoding Gly-rich cell wall protein, sterol reductase, Pro-rich cell wall protein, and PR1, respectively, by the 35S::Pto transgene required SA signaling. The availability of a large number of genes that are differentially regulated by distinct defense pathways and encode proteins with diverse functions provides an excellent tool to study the cell death and disease resistance mechanisms in tomato plants.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plants and Inoculation

Tomato (Lycopersicon esculentum) cultivars used were: Rio Grande PtoR, which carries a native Pto gene; Money Maker, which contains no Pto gene; a transgenic line carrying nahG (Money Maker background; Brading et al., 2000); and 35S::Pto transgenic line 48 (Money Maker background; Tang et al., 1999). Plants were grown in the greenhouse at 28°C (day) and 20°C (night). Healthy and well-expanded leaves from 6-week-old plants were used for experiments.

Pseudomonas syringae pv tomato strain T1 (avrPto) was grown in King's B (King et al., 1954) medium containing 50 mg L¹ rifampicin and 50 mg L¹ kanamycin. For inoculation, overnight bacterial culture was washed twice with 10 mM MgCl₂, and the density was determined by turbidity at 600 nm. Bacteria were diluted to 10⁶ cfu mL¹ in 10 mM MgCl₂ plus 0.04% (v/v) silwet L-77 (Osi, Danbury, CT) for vacuum infiltration of six-week-old plants. Well-expanded leaf tissue was collected at different time points after inoculation for RNA isolation. The 0-h sample was harvested immediately prior to vacuum infiltration.

Suppression Subtractive Cloning

A cDNA subtraction kit (PCR-Select; CLONTECH, Palo Alto, CA) was used for isolating cDNA clones from 35S::Pto plants. The subtractive cDNA library was constructed by following the user manual with slight modification. The cDNA from the 35S::Pto line was used as tester, and the cDNA from the non-transgenic tomato cv Money Maker plants was used as driver. The driver/tester ratio was increased to 4-fold over the suggested ratio. The subtracted cDNA was fractionated through an agarose gel and fragments greater than 300 bp were inserted into the TA-cloning vector (CLONTECH). To eliminate the common PR genes and redundant clones from the subtractive library, a pool of cDNA clones for five PR genes (PR1a1, PR1b1, GluB, Osmotin, and Chia; Tang et al., 1999) and other redundant clones from the library were used as probes to hybridize to the nylon membrane containing clones from the cDNA library. Single-pass sequencing with the T7 primer was carried out for non-hybridizing clones.

Reverse Northern and Northern-Blot Analysis

cDNA inserts were individually PCR amplified from the plasmid with T3 and T7 primers, separated in an agarose gel, and transferred to nylon membranes. The DNA blot contained cDNA inserts from the library and a chlorophyll a/b-binding protein gene (clone 93) or an actin cDNA (accession no. AW737353) as a constitutive control. The actin cDNA was PCR amplified from tomato cDNA with the following primers: 5'-GAAGAAGAAGAAAGAGSGCTTTTC-3' and 5'-AGCCTGAATAGCAACATACATAG-3'. Duplicated filters were hybridized with ³²P-labeled, oligo (dT)-primed cDNA probes derived from uninoculated tomato cv Money Maker or tomato cv 35S::Pto plants, or probes from uninoculated or bacterial-inoculated tomato cv RioGrande PtoR plants. Each cDNA probe was synthesized from 1 µg mRNA by using a cDNA synthesis kit (Life Technologies, Rockville, MD). The mRNA was removed by RNaseH treatment. RNA gel-blot analysis was carried out as described (Goldsbrough et al., 1990; Tang et al., 1999). Relative signal on the x-ray film was quantified by densitometry and normalized to constitutive controls before fold induction was calculated.