Systemic Acquired Tolerance to Virulent Bacterial Pathogens in Tomato

Virulent and Avirulent Xcv Induce SAT in Tomato

To investigate systemic responses to Xcv, wild-type tomato plants at the three-leaf stage were mock inoculated, inoculated with virulent Xcv strain 93-1 or avirulent Xcv strain 87-7 (Bonas et al., 1993) on their lowest two leaves. The 87-7 strain contains the avirulence gene avrBs3-2 and is avirulent on all tomato varieties. The inoculations were then permitted to run their full course of approximately 14 d, at which point those leaves inoculated with virulent Xcv were fully necrotic and those inoculated with avirulent Xcv had developed lesions associated with the hypersensitive response (Fig. 1A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Symptom development of Xcv-infected tomato. Disease symptoms at 16 dpi of wild-type tomato plants (A) inoculated with avirulent and virulent Xcv and (B) following challenge with virulent Xcv. These pictures are representative of at least two independent biological experiments.

To determine if this inoculation with Xcv affected responses to subsequent pathogen exposure, a challenge was performed with virulent Xcv on uninoculated systemic leaves. Prior inoculations with either virulent or avirulent Xcv, but not mock inoculations, reduced the necrosis resulting from the challenge (Fig. 1B). The primary symptoms such as lesion formation were unaffected and secondary symptom development such as chlorosis and some confluent necrosis were still apparent. The major difference between challenged and unchallenged plants was reduced necrosis in plants with prior Xcv inoculation. As the response consists of two independent interactions between two biological entities, a high degree of variation is to be expected. With this in mind the level of necrosis was determined in large population groups by measuring ion leakage in leaf five at 16 d after challenge (Fig. 2). Percent ion leakage, an indicator of cell death, was 2-fold higher upon challenge in plants previously mock inoculated than those with prior Xcv inoculations. Prior virulent or avirulent Xcv inoculations led to similar reductions in ion leakage upon challenge with virulent Xcv.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Cell death of tomato 16 d after challenge with virulent Xcv. Cell death was measured in the form of percent ion leakage in plants treated with a mock challenge or challenged with virulent Xcv. The plants were exposed to a mock inoculation or inoculation with virulent (vir) or avirulent (avr) Xcv prior to treatment. These data are representative of two independent biological experiments. Bars equal se, n = 30.

The reduction of symptom development due to previous pathogen exposure is consistent with SAR generation. Bacterial growth measurements confirmed that UC82B is resistant to avirulent Xcv but not virulent Xcv, as growth of avirulent Xcv was 10-fold lower than that of virulent Xcv (Fig. 3A). This difference is due to a gene-for-gene interaction (Bonas et al., 1993) and is consistent with levels of growth reported previously (Ciardi et al., 2000). However, when bacterial populations in challenge infections were determined, there was no difference in growth caused by prior inoculation with either virulent or avirulent Xcv when compared to plants with prior mock inoculations (Fig. 3B). This result leads to the conclusion that both virulent and avirulent Xcv induced SAT rather than SAR, as they reduced symptom development but not bacterial growth during subsequent challenge with virulent Xcv. Note that there is an inherent age-dependent resistance in tomato with the first leaves supporting higher bacterial growth than later ones.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Bacterial growth during inoculation and challenge of tomato with Xcv. A, The growth of virulent (vir) and avirulent (avr) Xcv was measured during primary infections. B, A systemic challenge with virulent Xcv was then performed on these plants as well as mock-inoculated controls and the bacterial growth was measured. These data are representative of two independent biological experiments. Bars equal se, n = 5.

Prior Inoculation with Xcv Leads to Early Production of Ethylene upon Challenge with Virulent Xcv

Ethylene and SA are involved in the development of systemic responses, including SAR and induced systemic resistance in Arabidopsis (Lawton et al., 1995; Knoester et al., 1999). In order to examine potential roles for these hormones in the systemic response of tomato to Xcv, levels of ethylene and SA were measured during inoculation and challenge with Xcv. Inoculation with virulent Xcv led to ethylene accumulation in local tissues at 5 d post infection (dpi), while avirulent Xcv caused a greater accumulation of ethylene at 2 dpi (Fig. 4A). Virulent Xcv induced SA accumulation at 10 dpi, while avirulent Xcv induced SA accumulation at 4 dpi, peaking at 10 dpi (Fig. 4B). The accumulation of ethylene and SA was later and at reduced magnitude in response to virulent Xcv than avirulent Xcv. No systemic accumulation of ethylene was observed in response to Xcv inoculations and at 16 dpi no systemic accumulation of SA was observed in plants inoculated with Xcv (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Local ethylene and SA accumulation following inoculation and challenge of tomato with Xcv. Tomato plants were mock inoculated or inoculated with virulent (vir) or avirulent (avr) Xcv and (A) ethylene and (B) SA accumulation following inoculation were measured. Fourteen days later, a challenge with virulent Xcv was performed on uninfected leaves and (C) ethylene and (D) SA accumulation following challenge was measured. These data are representative of two independent biological experiments. Bars equal se, n = 4.

Accumulation of ethylene and SA were also measured following challenge with virulent Xcv. Results indicated that ethylene accumulated earlier and to a greater extent upon challenge during SAT than in equivalent plants that were previously mock inoculated (Fig. 4, C and D). The ethylene accumulation in response to a challenge with virulent Xcv, following prior mock inoculation, reached its maximum at 10 dpi. However, plants with prior Xcv inoculations accumulated additional ethylene around 5 dpi when challenged (Fig. 4C). Although at a reduced magnitude, this accumulation resembles a response to avirulent Xcv. Challenge with virulent Xcv caused ethylene accumulation at 8 to 10 dpi in all plants, the magnitude of which was dependent on the prior inoculation. Plants with prior avirulent Xcv inoculation accumulated the most ethylene when challenged. This ethylene accumulation was lower in plants with prior virulent Xcv inoculations and lowest in the mock-inoculated controls. It can be concluded that challenged leaves from plants previously infected with either avirulent or virulent Xcv synthesized ethylene earlier and to a greater magnitude than those previously mock inoculated.

Plants with prior mock or virulent Xcv inoculations accumulated similar amounts of SA in response to challenge with virulent Xcv. However, reduced SA accumulation was observed in plants with a prior avirulent Xcv inoculation. These plants also produced an early SA peak (Fig. 4D) that resembled the response to avirulent Xcv although reduced in magnitude. Subsequent accumulation of SA was delayed and resembled that caused by virulent Xcv although at reduced magnitude. That there is no significant difference in SA accumulation between the prior mock and the prior virulent Xcv inoculation in the challenged leaves seems to rule out a role in SAT for SA synthesized in the systemic leaves.

Pathogenesis-Related Gene Expression Demonstrates the Necessity of Ethylene and SA in Systemic Signal Generation

As ethylene- or SA-deficient plants develop less secondary disease symptoms in response to virulent Xcv than plants displaying SAT, a direct approach to determining any roles for these phytohormones in SAT is difficult. Therefore, pathogenesis-related (PR) gene expression was used as a marker for defense responses in ethylene- and SA-deficient plants. Induction of PR genes is often used as an indicator of early defense responses. They show both local and systemic induction during SAR (Ward et al., 1991). To determine if ethylene and SA are involved in systemic signal transduction, two transgenic lines were used. Loss of ethylene was achieved with transgenic plants expressing 1-aminocyclopropane-1-carboxylic acid deaminase (ACD). These plants do not accumulate the ethylene precursor 1-aminocyclopropane-1-carboxylic acid and thus underproduce ethylene. The ACD line was compared to its isogenic parent UC82B. The role of SA was determined using transgenic tomato plants expressing the bacterial salicylate hydroxylase, nahG, which does not accumulate SA. The NahG line was compared to its isogenic parent MoneyMaker (MM).

PR1a and PR1b were induced in local and systemic tissues of tomato during infection with virulent Xcv (Fig. 5). PR1a expression was lower than PR1b. The local expression of both genes was induced in all lines in response to virulent Xcv. The level of local PR1a induction varied between different cultivars, with an earlier and greater induction in MM than UC82B. The local expression of PR1b was also higher in MM than UC82B, although the timing of induction was the same. When compared to their isogenic parents, infected ACD and NahG plants had reduced local expression of both genes. However, this reduction did not affect pathogen growth (O'Donnell et al., 2001).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Local and systemic PR gene induction in ethylene- and SA-deficient tomato lines in response to inoculation with virulent Xcv. Ethylene-deficient ACD, SA-deficient NahG, and their isogenic parents were inoculated with virulent Xcv. PR1a and PR1b expression levels were determined by real-time RT-PCR in local and systemic tissues. These data are representative of two independent biological experiments. Bars equal se, n = 4.

Virulent Xcv induced systemic expression of both genes in MM and UC82B. In MM systemic induction of PR1a and PR1b occurred at the same time, whereas in UC82B the expression of PR1b was later than PR1a. ACD and NahG showed no systemic induction of PR gene expression in response to virulent Xcv. Disease symptoms in ACD and NahG are comparable to their isogenic parents up to 8 dpi (O'Donnell et al., 2001). As there is no widespread necrosis before 8 dpi, the lack of systemic PR gene expression must be related to loss of ethylene or SA accumulation rather than to the loss of widespread necrosis in these lines. The lack of systemic PR gene induction indicates that ethylene- and SA-deficient lines may be compromised in their systemic signal transduction. However, the tolerant phenotype of these lines makes direct investigation difficult.

Greater PR Gene Expression Was Induced by Avirulent Than Virulent Xcv

The accumulation of ethylene and SA is more rapid in response to avirulent than virulent Xcv, yet the resulting systemic response to these two pathogens is the same. To determine if the early defense responses follow this trend, induction of PR gene expression by avirulent and virulent Xcv inoculations was assayed. Avirulent Xcv induced higher and earlier local expression of PR genes than virulent Xcv (Fig. 6). The timing of systemic PR1b induction was similar in response to each pathogen, but systemic PR1a induction was faster in response to avirulent Xcv. Faster and stronger local PR gene response to avirulent than virulent Xcv may be an indication of increased defense responses that limit pathogen growth.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Local and systemic PR gene expression during inoculations with virulent or avirulent Xcv. Wild-type (UC82B) plants were mock inoculated or inoculated with virulent (vir) or avirulent (avr) Xcv and the local and systemic expression of PR1a and PR1b was determined by real-time RT-PCR. These data are representative of two independent biological experiments. Bars equal se, n = 4.

Prior avirulent and virulent Xcv inoculations produced similar inductions of ethylene and reductions in symptom development during challenge. The two prior Xcv inoculations induced similar PR gene expression upon challenge. Both caused an early peak of PR gene expression at 1 dpi upon challenge that was absent in plants with prior mock inoculation (Fig. 7). These results led to the hypothesis that inoculation with Xcv sensitized systemic defenses and caused their faster induction upon challenge, leading to tolerance. Sensitization or priming to the presence of pathogens is also observed in the generation of SAR (Conrath et al., 2002), therefore SAT may use similar mechanisms to SAR.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Induction of PR genes during challenge with virulent Xcv in the presence or absence of SAT. Wild-type plants were mock inoculated or inoculated with virulent (vir) or avirulent (avr) Xcv. Fourteen days later, a challenge was performed with virulent Xcv. The expression of PR1a and PR1b was determined with real-time RT-PCR following challenge. These data are representative of two independent biological experiments. Bars equal se, n = 4.

SAT Is Induced by Both Xcv and P. syringae pv tomato DC3000

We have demonstrated that infection with either virulent or avirulent Xcv leads to the production of tolerance to subsequent infections with virulent Xcv. It may be that SAT is due to an inability of tomato to induce resistance to virulent Xcv. To test this hypothesis, tomato plants were treated with the SAR inducer Actigard at 10 and 4 d prior to challenge with virulent Xcv. Ion leakage and bacterial population measurement following challenge with virulent Xcv (Fig. 8, A and C) demonstrated that treatment with Actigard reduced Xcv growth and tissue damage following challenge. These data demonstrate that tomato can be induced to mount systemic resistance to Xcv. However, under normal circumstances SAT is the default response to Xcv.

• Download figure
• Open in new tab
• Download powerpoint

Figure 8.

SAT induction by Xcv and Pst. Wild-type plants were mock inoculated, treated with the SAR inducer Actigard, or inoculated with virulent Xcv or Pst. Fourteen days later, a challenge was performed with virulent Xcv or Pst. Populations of (A) Xcv and (B) Pst were determined at 5 d after challenge. Bars equal se, n = 5. Percent ion leakage was determined 16 d after challenge in plants challenged with (C) Xcv or (D) Pst. Bars equal se, n = 30. Each experiment was repeated at least twice.

To assess if SAT is a specific response to Xcv, we tested the ability of virulent Xcv to induce tolerance to subsequent challenge with the virulent bacterial pathogen P. syringae pv tomato DC3000 (Pst). We first confirmed that tomato was capable of inducing resistance to Pst. This was accomplished by treatment with Actigard at 10 and 4 d prior to challenge. Ion leakage and bacterial growth were measured following challenge with Pst (Fig. 8, B and D). Both tissue damage and bacterial growth were reduced compared to mock-treated controls, demonstrating that tomato can induce resistance to Pst. Ion leakage and bacterial growth were then measured following challenge with Pst in plants with prior exposure to virulent Xcv (Fig. 8, B and D). These data show that there is no significant difference in tissue damage or bacterial growth following challenge with Pst between plants with prior mock inoculations and prior inoculation with virulent Xcv. Therefore, virulent Xcv does not induce tolerance to Pst.

We then examined if Pst can itself induce SAT. Mock-inoculated and Pst-inoculated plants were challenged with either virulent Xcv or Pst. Ion leakage and bacterial populations following challenge were determined (Fig. 8). Prior exposure to Pst reduced tissue damage but not bacterial growth following both Pst and virulent Xcv challenge. Therefore, Pst can induce SAT both to itself and to virulent Xcv, indicating that this response is not specific to Xcv.

Plant Materials and Treatments

Tomato (Lycopersicon esculentum) cultivars MM and UC82B are the parental lines for NahG (Oldroyd and Staskawicz, 1998) and ACD (Klee et al., 1991), respectively. Wild type refers to UC82B in all experiments except those associated with Figure 5. Plant growth and treatments were performed under ambient temperature and lighting in a greenhouse. To avoid confusion, we use the terms “inoculation” for the first treatment of a tomato plant and “challenge” for the subsequent treatment on distal leaves. Plants were inoculated by submersion of the leaves for 15 s in a bacterial suspension of 1 × 10⁷ colony forming units (cfu) mL⁻¹ of Xcv strain 93-1 (virulent) or 87-7 (avirulent) containing 10 mm MgCl₂ and 0.02% (v/v) Silwet L-77. Mock inoculations were performed by dipping plants in buffered Silwet. For analysis of the systemic response to infection, primary inoculations were performed on 3-week-old plants by dipping the first and second leaves in infection media. Challenges were then performed 14 d later by dipping leaves 5 and 6 in the infection media containing 1 × 10⁷ cfu mL⁻¹ of virulent Xcv. For PR gene analysis, the first and second leaves of 6-week-old plants were inoculated, and tissue from the local inoculation (first and second leaves) and the systemic response (fifth and sixth leaves) were removed at the indicated time points and flash frozen in liquid nitrogen. Pst inoculations were performed with 1 × 10⁷ cfu mL⁻¹ Pst DC3000. Actigard (0.03 gL⁻¹) treatments were sprayed on the first and second leaves of tomato plants at 10 and 4 d prior to challenge. Each experiment was repeated at least twice.

Bacterial Culture

Xcv and Pst cultures were grown as previously described by O'Donnell et al. (2001). Leaf colony counts were determined on two independent biological repeats as previously described by Lund et al. (1998). Briefly, five 1-cm² leaf discs were sampled from each line at each time point indicated. The discs were ground in 10 mm MgCl₂, and serial dilutions were incubated at room temperature (Pst) or at 30°C (Xcv) for 2 d on solid media. The average cfu per square centimeter (cfu cm⁻²) for each sample was determined by counting individual colonies (n = 5).

Ion Leakage

Amount of cell death was estimated by measuring percentage of ion leakage on two independent biological repeats. For challenge inoculations at 16 dpi the fourth leaf of each plant was placed in 6-mL deionized water and a vacuum of 20 psi applied for 5 min. The samples were then shaken at room temperature for 1 h. Three milliliters of the water was then removed and its conductivity was measured. The samples were then placed in a boiling water bath for an hour and the conductance of the remaining 3 mL of water measured. Percent ion leakage was determined by conductivity of first 3 mL divided by conductivity of second 3 mL multiplied by 100. n = 30 for challenge inoculations with a pathogen and n = 10 for mock challenges. Each experiment was repeated at least twice.

Ethylene Measurements

Ethylene measurements were performed on a minimum of two independent biological replicates with n = 4. Ethylene production was determined by sampling the headspace above a single leaf enclosed in 5-cm³ tubes for 1 h as described by Lund et al. (1998). Ethylene concentration in a 1-mL sample was determined by gas chromatography (model 5890; Hewlett-Packard, Palo Alto, CA).

SA Measurements

SA was extracted from tomato tissue and derivatized using trimethylsilyl-diazomethane. The volatile SA methyl ester was collected from the complex matrix using vapor phase extraction and quantified by isobutene chemical ionization gas-chromatography/mass spectrometry as described in Schmelz et al. (2004). Each experiment was repeated at least twice; n = 4.

Real-Time RT-PCR

PR1a (M69247) and PR1b (M69248) mRNA levels were quantified by real-time quantitative RT-PCR using Taqman one-step RT-PCR reagents (Applied Biosystems, Foster City, CA) and an Applied Biosystems GeneAmp 5700 sequence detection system. Each determination was performed using 250 ng of DNase-1 treated total RNA isolated using RNeasy Plant mini kit (Qiagen, Valencia, CA) in a 25-μL reaction volume. RT-PCR conditions were: 48°C for 30 min, 95°C for 10 min followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Absolute mRNA levels were quantified using synthesized sense strand RNAs as standards. Primers and probes were designed using PRIMER EXPRESS software (Applied Biosystems) and were as follows: PR1b probe 5′-/56-FAM/CAACGGATGGTGGTTCATTTCTTGCA/3BQH_1/-3′; PR1a probe 5′-/56-FAM/TGTGGGTGTCCGAGAGGCCAGA/3BHQ_1/-3′; PR1b forward primer 5′-GGTCGGGCACGTTGCA-3′; PR1b reverse primer 5′-GATCCAGTTGCCTACAGGACATA-3′; PR1a forward primer 5′-GAGGGCAGCCGTGCAA-3′; PR1a reverse primer 5′-CACATTTTTCCACCAACACATTG-3′ (Intergrated DNA Technologies, Coralville, IA). Each experiment is representative of two independent biological replicates; n = 2.