Glufosinate Ammonium-Induced Pathogen Inhibition and Defense Responses Culminate in Disease Protection in bar-Transgenic Rice^,

Abstract

Glufosinate ammonium diminished developments of rice (Oryza sativa) blast and brown leaf spot in 35S:bar-transgenic rice. Pre- and postinoculation treatments of this herbicide reduced disease development. Glufosinate ammonium specifically impeded appressorium formation of the pathogens Magnaporthe grisea and Cochliobolus miyabeanus on hydrophobic surface and on transgenic rice. In contrast, conidial germination remained unaffected. Glufosinate ammonium diminished mycelial growth of two pathogens; however, this inhibitory effect was attenuated in malnutrition conditions. Glufosinate ammonium caused slight chlorosis and diminished chlorophyll content; however, these alterations were almost completely restored in transgenic rice within 7 d. Glufosinate ammonium triggered transcriptions of PATHOGENESIS-RELATED (PR) genes and hydrogen peroxide accumulation in transgenic rice and PR1 transcription in Arabidopsis (Arabidopsis thaliana) wild-type ecotype Columbia harboring 35S:bar construct. All transgenic Arabidopsis showed robust hydrogen peroxide accumulation by glufosinate ammonium. This herbicide also induced PR1 transcription in etr1 and jar1 expressing bar; however, no expression was observed in NahG and npr1. Fungal infection did not alter transcriptions of PR genes and hydrogen peroxide accumulation induced by glufosinate ammonium. Infiltration of glufosinate ammonium did not affect appressorium formation of M. grisea in vivo but inhibited blast disease development. Hydrogen peroxide scavengers nullified blast protection and transcriptions of PR genes by glufosinate ammonium; however, they did not affect brown leaf spot progression. In sum, both direct inhibition of pathogen infection and activation of defense systems were responsible for disease protection in bar-transgenic rice.

Rice (Oryza sativa) is one of the most important crops worldwide. Farmers have applied integrated crop management and governments have implemented environmental regulations to reduce chemical applications to a desirable level; however, synthetic chemicals are still required for stable cereal production. In addition, development of genetically modified rice plants resistant to nonselective herbicides like glufosinate ammonium is expected to improve crop productivity (Delannay et al., 1995). Agricultural chemicals frequently induce side effects on their target crops and agroecosystems. For example, some herbicides are known to induce morphological and physiological alterations in crops and higher incidence of plant diseases (Campbell and Altman, 1977; Sanogo et al., 2000). Glyphosate occasionally suppressed disease defense responses and enhanced disease development (Keen et al., 1982; Brammall and Higgins, 1988). Interestingly, glufosinate ammonium treatment onto transgenic rice expressing bar gene controls rice blast and sheath blight progressions by Magnaporthe grisea and Rhizoctonia solani (Uchimiya et al., 1993; Tada et al., 1996). This herbicide also induced resistance against brown patch and dollar spot, caused by R. solani and Sclerotinia homoeocarpa, in bar-transgenic bentgrass (Agrostis spp.; Higgins et al., 2003).

The bar gene, which codes for the enzyme phosphinothricin (PPT) acetyl transferase (PAT), is one of the most prevalent selectable markers of genetically modified crops and confers tolerance against glufosinate ammonium, an active ingredient of the nonselective herbicide Basta. Glufosinate ammonium is an ammonium salt of PPT and efficiently kills various kinds of plants, including rice. PAT inactivates PPT by acetylating it (Botterman et al., 1991). Glufosinate ammonium is a Glu analog that inhibits Gln synthetase by irreversible binding (Manderscheid and Wild, 1986). Gln synthetase catalyzes an ATP-dependent incorporation of ammonium into the amide position of Glu, resulting in the formation of Gln. This is indispensable for capturing toxic ammonium produced during photorespiration and inorganic nitrogen assimilation (Wallsgrove et al., 1983). The inhibition of Gln synthetase by glufosinate ammonium results in an accumulation of toxic ammonium derived from photorespiration (Martin et al., 1983; Wild et al., 1987; Sechley et al., 1992; Last, 1993). Accumulation of toxic ammonium disturbs electron transport systems of both chloroplasts and mitochondria and induces production of hazardous free radicals (Krogmann et al., 1959; Puritch and Barker, 1967). Free radicals in turn cause lipid peroxidation (especially on membranes), damage of other cellular constituents, and eventually, cell death (Hess, 2000). Because of this nature, glufosinate ammonium is often termed a “pro-oxidative herbicide” (Strobel and Kuc, 1995). Herbicidal mechanisms and target sites of glufosinate ammonium are definitive in nontransgenic plants; however, physiological alterations induced by this chemical in bar gene-harboring plants still remain unclear.

Plant defense activators induce systemic acquired resistance and condition the plant in a resistant state. These chemicals trigger defense-related responses like augmented and accelerated transcription of PATHOGENESIS-RELATED (PR) genes, defense-related materials, and burst of active oxygen species (AOS; Friedrich et al., 1996; Benhamou and Belanger, 1998; Cohen, 2002; Iriti and Faoro, 2003). Systemic resistance-acquired plants show broad-spectrum disease resistance against multiple pathogens because of the activation of defense mechanisms including hypersensitive response and systemic translocation of resistance via secondary signal messengers like salicylic acid and jasmonic acid. Transcriptional activation of PR genes and rapid accumulation of hydrogen peroxide have been recognized as reliable molecular and cellular markers of disease resistance in rice and other plants (Ganesan and Thomas, 2001; Kachroo et al., 2003; Tsukamoto et al., 2005). Augmented and accelerated mRNA synthesis of PR1 has been reported in incompatible interaction between rice and avirulent M. grisea and in rice treated with plant defense activators (Kim et al., 2001b; Ahn et al., 2005b). Recently, novel functions of AOS have been investigated. Besides arresting pathogen proliferation in planta, AOS are involved in cell wall reinforcement (Olivain et al., 2003), callose deposition (Huckelhoven et al., 1999), and acts to signal molecules in systemic translocation of acquired resistance (Bolwell et al., 1995, 1998; Tenhaken et al., 1995; Wojtaszek, 1997; Alvarez et al., 1998; Chamnongpol et al., 1998). Similar to previously known plant defense activators, methyl viologen and mercuric chloride promote oxidative damage and chlorosis-induced plant defense responses (van Loon, 1975; Lund et al., 1993; Strobel and Kuc, 1995). Phosphates and oxalates inducing localized chlorosis and necrosis and AOS accumulation also trigger systemic acquired resistance in cucumber (Cucumis sativus; Doubrava et al., 1988; Gottstein and Kuc, 1989; Mucharromah and Kuc, 1991; Orober et al., 2002).

M. grisea and Cochliobolus miyabeanus, the teleomorphs of Pyricularia grisea and Bipolaris oryzae, are hemibiotrophic and necrotrophic fungal pathogens that cause blast and brown leaf spot, the most devastating diseases of rice worldwide. Yield losses due to blast have been estimated at 11% to 30%. Moreover, 10% incidence of neck blast in the field could reduce rice production by 5% to 6%. Brown leaf spot is one of the most common diseases and is observed at any stage of growth in the field. The outbreak of rice brown leaf spot caused the Bengal famine and was the major cause of two million fatalities in 1943 (Stuthman, 2002). Prior to infection by both pathogenic fungi, a series of infection-related morphological changes, initiated by spore adhesion to the host surface, spore germination, germ tube elongation, appressorium formation, and penetration by an infection peg, were observed (Goto, 1958; Ou, 1985). Conidial germination and appressorium formation have been recognized as important target sites for screening and developing novel fungicides (Oh and Lee, 2000). For example, tricyclazole, an inhibitor of melanin biosynthesis prerequisite for turgor generation, effectively inhibits rice blast disease. Inhibition of appressorium formation by methylglyoxal-bis-guanyl hydrazone (a polyamine biosynthesis inhibitor) in C. miyabeanus resulted in a significant reduction of rice brown leaf spot (Ahn and Suh, 2007a).

This research shows the inhibitory effects of glufosinate ammonium on the developments of blast and brown leaf spot in bar gene-expressing rice. Direct and indirect effects of glufosinate ammonium on the disease progressions by pathogens and defense-related cellular and molecular responses in rice were examined.

RESULTS

Effects of Glufosinate Ammonium on Rice Blast and Brown Leaf Spot

Rice ‘Dongjin’ (nontransgenic control [NC]) and 35S:bar-transgenic ‘Dongjin’ (transgenic) inoculated with M. grisea strain KJ201 exhibited typical rice blast symptoms and normal disease development (Fig. 1). Water-soaked lesions began to form 3 d postinoculation (dpi). Invasive mycelial growth was apparent within and around lesions. At 5 dpi, almost one-half of the inoculated leaf surface was covered with blast lesions, and massive sporulation began to appear in the center of the lesion at 10 dpi. When NC and transgenic rice were inoculated with C. miyabeanus strain HIH-1, approximately 93% of conidia germinated within 12 h postinoculation (hpi). Visible spot, representing host cell death, was evident around 16 hpi. At 3 dpi, the average size of lesions was approximately 0.6 × 3 mm (width × length), and chlorosis began to appear around the margin of lesions.

Figure 1.

Effects of glufosinate ammonium (G) on the developments of rice blast and brown leaf spot caused by M. grisea strain KJ201 and C. miyabeanus strain HIH-1, respectively. Glufosinate ammonium or mock was applied 24 h before fungal inoculation. Disease progression was estimated according to the number and size of lesions. A, Disease progression in ‘Dongjin’ (−, NC) and bar-transgenic ‘Dongjin’ (+) pretreated with mock (−, 250 μg mL⁻¹ Tween 20) or glufosinate ammonium (+, 100 μg mL⁻¹) and inoculated with M. grisea strain KJ201 and C. miyabeanus strain HIH-1. Each data point is mean ± se. B, Blast and brown leaf spot developments in ‘Dongjin’ (NC) and bar-transgenic ‘Dongjin’ (bar) pretreated with glufosinate ammonium (+) or mock (−). Photographs depicting representative symptoms were taken 10 d after fungal inoculation. Experiments were repeated more than three times with three replicates consisting of 15 plants; almost similar tendencies were obtained. [See online article for color version of this figure.]

The effect of glufosinate ammonium on disease development was evaluated. Transgenic rice was highly susceptible to KJ201; however, treatment of 100 μg mL⁻¹ glufosinate ammonium 24 h prior to infection greatly increased blast protection. The number and size of blast lesions were significantly decreased. In addition, their developments were defined within the initial infection sites or retarded. Sporulation was barely observed on any of the lesions on the glufosinate ammonium-treated leaves. Moreover, glufosinate ammonium treatment noticeably hampered brown leaf spot disease development. The number of lesions was also decreased, and symptom development was defined distinctively.

Glufosinate ammonium treatments 5 d or 1 d prior to (preinoculation treatment) and 12 h or 1 d after (postinoculation treatment) inoculation significantly reduced the developments of both diseases. Protection effect against rice blast was retained at the same level in all testing periods; however, that against brown leaf spot reached a maximum level 1 d prior to fungal inoculation and slightly decreased thereafter (Fig. 2).

Figure 2.

Rice blast and rice brown leaf spot disease progressions of glufosinate ammonium-treated bar-transgenic (bar) or NC rice plants. Plants were inoculated by M. grisea strain KJ201 or C. miyabeanus strain HIH-1. In the meantime, 100 μg mL⁻¹ of glufosinate ammonium was sprayed 5 d or 1 d prior to (preinoculation treatments; −5 and −1, respectively) fungal inoculation. Treatments were also performed 12 h or 1 d after (postinoculation treatments; +0.5 and +1, respectively) inoculation. In the analysis of preinoculation treatment effects, glufosinate ammonium-treated rice leaves were carefully washed several times by spraying distilled water at 24 h after treatment. Disease progression was determined 7 dpi. Each data point is mean ± se. Different letters indicate statistically significant differences between treatments (Duncan's multiple range test; P < 0.05). Experiments were repeated more than three times with three replicates consisting of 15 rice plants; almost similar tendencies were observed.

Direct Effects of Glufosinate Ammonium on Fungal Developments

Glufosinate ammonium treatment reduced the number and size of lesions in transgenic rice plants. To investigate direct, antimicrobial activities of this herbicide, the effects of glufosinate ammonium on conidial germination and appressorium formation in M. grisea and C. miyabeanus were examined on the artificial substratum, hydrophobic surface of GelBond, and on transgenic rice leaves. Appressorium formations in M. grisea and C. miyabeanus on the GelBond were inhibited by glufosinate ammonium in a dose-dependent manner (Fig. 3, A and B). Glufosinate ammonium reduced appressorium formation in M. grisea and C. miyabeanus by 71% and 83% at 100 μg mL⁻¹ concentration, respectively. In contrast, the same treatment did not affect conidial germination of both pathogens. To evaluate the effect of glufosinate ammonium on the prepenetration morphogenesis in vivo, conidial suspension along with or without glufosinate ammonium was placed on detached NC and transgenic rice leaves. For more precise observation, chlorophyll was removed from the samples, and fungal cells were stained. Twenty-four hours after placements, most of the water-treated conidia germinated and formed appressoria. Glufosinate ammonium inhibited 61% of appressorium formation in M. grisea and 85% in C. miyabeanus. Conidial germination was reduced by 22% in M. grisea and 13% in C. miyabeanus.

Figure 3.

Effects of glufosinate ammonium on conidial germination and appressorium formation in M. grisea strain KJ201 and C. miyabeanus strain HIH-1. Conidial germination and appressorium formation were estimated microscopically 24 h after placement onto the tested surface. A, Effects of glufosinate ammonium on prepenetration development of M. grisea and C. miyabeanus on the hydrophobic surface of GelBond. Each data point is mean ± se. B, Effects of glufosinate ammonium (100 μg mL⁻¹) on conidial germination (white bar) and appressorium formation (black bar) of M. grisea and C. miyabeanus on rice leaves from bar-transgenic plants. Different letters indicate statistically significant differences between treatments (Duncan's multiple range test; P < 0.05). C, Prepenetration developments of M. grisea and C. miyabeanus on the hydrophobic surface of GelBond (artificial surface) or transgenic (bar) rice leaves in the presence (+) or absence (−) of 100 μg mL⁻¹ glufosinate ammonium. a, Appressorium; c, conidium; g, germ tube. Bars = 50 μm. Experiments were repeated more than three times with three replicates; almost similar tendencies were obtained. [See online article for color version of this figure.]

Phytotoxic effects of glufosinate are induced by accumulation of toxic ammonium derived from nitrogen assimilation. To investigate the effects of glufosinate on the pathogen growth in the plant mimic conditions, the inhibition rates of pathogen growth by glufosinate on the nutrient-rich and nutrient-deprived media were compared. After infection into host tissues, pathogen was encountered to a malnutrition conditions deficient in nitrogen or carbon source. Transcriptions of several fungal genes, MPG1 in M. grisea and ccSNF1 in Cochliobolus carbonum infecting maize (Zea mays), specifically expressed during in planta ramification, were also induced in the malnutrition conditions (Beckerman and Ebbole, 1996; Talbot et al., 1996; Tonukari et al., 2000). In addition, the same treatment was expected to attenuate accumulation of toxic ammonia derived from nitrogen assimilation. Similar results were described previously (Snoeijers et al., 2000). Glufosinate ammonium supplementation inhibited mycelial growth of M. grisea and C. miyabeanus by 91% and 88% on the complete media (CM; Fig. 4). In contrast, glufosinate ammonium inhibited mycelial growth of M. grisea and C. miyabeanus by about 20% and 8% on the nutrient-deficient media (CM lack of nitrogen and carbon sources).

Figure 4.

Effects of nutrient starvation and/or glufosinate ammonium on the fungal growth. Prior to colony contact on the CM, colony morphologies of M. grisea (A) and C. miyabeanus (B) on the CM containing carbon and nitrogen sources (+) or mock (−) and supplemented with glufosinate (+) or mock (−) were photographed and colony areas were measured (C). Different letters indicate statistically significant differences between treatments (Duncan's multiple range test; P < 0.05). Experiments were repeated more than three times with three replicates; almost similar results were obtained. [See online article for color version of this figure.]

Indirect Effects of Glufosinate Ammonium on Disease Development

Glufosinate ammonium distinctively inhibited symptom development of rice blast and brown leaf spot. Preinoculation treatment also significantly diminished disease progression. Further, this chemical did not show sufficient pathogen growth inhibition in the nutrient-deficient condition. These results implied that glufosinate might induce other unknown disease-inhibiting mechanisms except its direct, antimicrobial activities. To investigate effects of glufosinate ammonium on host disease resistance, the activity of glufosinate ammonium on the transcription of PR1 in transgenic rice and Arabidopsis (Arabidopsis thaliana) was analyzed. Previous results showed that pathogen infection, plant defense activators, and environmental stresses induce transcriptions of the PR genes (Lawton et al., 1996; Kim et al., 2001b). Glufosinate ammonium higher than 10 μg mL⁻¹ per se induced transcription of PR1 gene in transgenic rice (Fig. 5A). A large amount of PR1 transcript was accumulated locally and systemically (Fig. 5B). Further, fortified PR1 transcription was retained for more than 15 d after glufosinate spray (Fig. 5C). In NC and transgenic rice infected with KJ201 after mock (250 μg mL⁻¹ Tween 20) treatment, the inductions of PR1, PBZ1 (a probenazole-inducible PR gene), and POX22.3 (a gene encoding PEROXIDASE22.3) were significantly delayed (by 1–2 d), and their transcriptions reached maximum levels at 3 dpi (Fig. 5D). This coincided with the formation of water-soaked lesions (Fig. 1B). Beyond 3 dpi, the level of transcripts decreased slightly. Transcriptions of PR1, PBZ1, and POX22.3 peaked within 1 dpi with HIH-1 in NC and transgenic rice. Glufosinate ammonium treatment triggered robust transcriptions of tested marker genes in transgenic rice. Robust transcription remained unaffected and pathogen infection did not alter this pattern.

Figure 5.

Transcriptions of PR genes induced by glufosinate ammonium (100 μg mL⁻¹) treatment and/or pathogen inoculation at varying dpt and dpi. Pathogen inoculation and glufosinate ammonium treatment were performed as described in Figure 1A. A, PR1 gene transcription induced by varying concentrations of glufosinate ammonium treatment at 1 dpt. B, Transcription of PR1 gene induced by glufosinate ammonium in bar-transgenic rice. Glufosinate ammonium was sprayed on second leaves. Leaves sprayed with glufosinate ammonium (L, local) or left untreated (S, systemic) were harvested from the same plants 1 dpt. Transgenic rice leaves treated with mock (C) were also harvested at the same time. C, PR1 gene expression was retained up to 15 d after glufosinate treatment. D, Transcriptions of PR1, PBZ1, and POX22.3 in NC and transgenic (bar) rice challenged with M. grisea strain KJ201 or C. miyabeanus strain HIH-1 1 d after mock or 100 μg mL⁻¹ glufosinate ammonium. Total RNA was extracted from the leaves of five rice plants recovered 0, 1, 2, 3, and 4 dpi. Experiments were repeated more than three times; almost similar results were obtained.

To further confirm indirect effects of glufosinate and investigate defense signaling pathways induced by this herbicide, we investigated the effects of glufosinate ammonium on the hydrogen peroxide accumulation and PR1/PDF1.2 transcriptions in ecotype Columbia-0 (Col-0), bacterial NahG-expressing Col-0, and defense-defective mutants like npr1, etr1, and jar1 harboring 35S:bar construct. There was no discrete hydrogen peroxide accumulation in rosette leaves treated with mock (Fig. 6A). In contrast, glufosinate ammonium on rosette leaves induced robust hydrogen peroxide accumulation in treated rosette leaves (local) of transgenic Col-0 and all transgenic plants expressing bar. Further, hydrogen peroxide accumulation was also observed in the cauline leaves (systemic) of all transgenic lines tested. Although there were some differences, quantitative analyses of hydrogen peroxide production also corroborated these phenomena (Fig. 6B). Mock spray induced no PR1 and PDF1.2 transcriptions in all tested lines (Fig. 6C). Discrete PR1 transcription was observed in glufosinate-treated local (rosette) and nontreated systemic (cauline) leaves of transgenic Col-0, etr1, and jar1 lines; however, NahG and npr1 did not show these transcriptions. Salicylic acid spray triggered local PR1 transcription in the rosette leaves of transgenic Col-0. Glufosinate did not induce PDF1.2 transcriptions in all transgenic lines tested. However, jasmonic acid treatment triggered local PDF1.2 transcriptions in the rosette leaves from transgenic Col-0 line. Transgenic Arabidopsis was produced using the floral dip method and it possessed single PR1:eGFP (enhanced GFP gene) and 35S:bar construct (data not shown). Similar with the above result, strong green fluorescence was induced on the transgenic Col-0 seedlings by glufosinate treatment (Fig. 6D). Mock-treated transgenic plants did not show detectable fluorescence. Glufosinate triggered eGFP mRNA synthesis in transgenic plants (Fig. 6E).

Figure 6.

Effects of glufosinate ammonium on the accumulation of hydrogen peroxide and transcriptions of PR1 and PDF1.2 in 35S:bar-transgenic Arabidopsis Col-0 and its mutants. Glufosinate (+, 100 μg mL⁻¹) was sprayed onto rosette leaves and treated rosette and nontreated cauline leaves were harvested 24 h later. A, Hydrogen peroxide accumulation in treated rosette (local) and nontreated cauline (systemic) leaves of 35S:bar-harboring transgenic Arabidopsis Col-0, NahG, npr1, etr1, and jar1 lines exposed to glufosinate. Harvested leaves were stained with DAB (0.1%, w/v). B, Quantification of hydrogen peroxide accumulation in treated rosette (local) and nontreated (systemic) cauline leaves of transgenic Arabidopsis Col-0, NahG, npr1, etr1, and jar1 lines exposed to glufosinate. Each bar represents the mean ± se. Different letters indicate statistically significant differences between treatments (Duncan's multiple range test; P < 0.05). C, Transcriptions of PR1 and PDF1.2 in treated rosette (L; local) and nontreated cauline (S; systemic) leaves of transgenic Arabidopsis Col-0, NahG, npr1, etr1, and jar1. Data were from Arabidopsis sprayed with glufosinate (+) or 250 μg mL⁻¹ Tween 20 only (−, mock). In addition, leaves of transgenic Col-0 were harvested 1 d after salicylic acid (SA, 500 μm) or jasmonic acid (JA, 100 μm) spray. D, Expression of eGFP by glufosinate treatment in transgenic Arabidopsis containing PR1:eGFP and 35S:bar gene construct. Bar = 1 mm. E, Transcription of eGFP transcript under the control of PR1 promoter by glufosinate treatment in transgenic Arabidopsis. Experiments were repeated more than three times; almost similar results were obtained.

Glufosinate ammonium induced severe chlorosis and wilting on NC (Fig. 7A). The transgenic plant was highly resistant to glufosinate ammonium; however, slight chlorosis was evident 1 to 2 d posttreatment (dpt), and this alteration was almost completely restored within 7 to 10 d. Glufosinate ammonium diminished maximum photochemical efficiency (F_v/F_m) of PSII by 28% in NC and 26% in transgenic rice at 3 dpt. F_v/F_m value of transgenic plants was restored within 7 dpt; however, that within NC decreased continuously and finally died (Fig. 7B). Hydrogen peroxide accumulation and pathogen growth were analyzed in transgenic rice infected by two pathogens. In NC inoculated with M. grisea strain KJ201, hydrogen peroxide was not accumulated at the infection site 24 hpi and began to be observed at 72 hpi (Fig. 6C; data not shown). Active fungal ramification was observed at 72 hpi. In transgenic rice, glufosinate ammonium induced hydrogen peroxide accumulation within 24 h posttreatment; however, the same treatment did not trigger host cell alteration. No hydrogen peroxide was observed at 24 hpi and fungal ramification developed normally in the transgenic rice leaves. In the transgenic rice pretreated with glufosinate ammonium, most rice cells exhibited hydrogen peroxide accumulation and KJ201 infection did not affect this cellular response.

Figure 7.

Effects of glufosinate ammonium and M. grisea strain KJ201 on hydrogen peroxide accumulation and fungal ramification in bar-transgenic and nontransgenic rice. The plants were sprayed with glufosinate ammonium (+) in 250 μg mL⁻¹ Tween 20 or mock only (−). One day after treatment, rice plants were inoculated with virulent M. grisea strain KJ201. A, Effects of glufosinate ammonium on the NC and transgenic (bar) rice plants. The 4-week-old plants were sprayed with 100 μg mL⁻¹ glufosinate ammonium (+) or mock (−). Representative leaves were photographed 10 dpt. B, Effects of glufosinate ammonium on the F_v/F_m in NC and transgenic (bar) rice. Each data point is mean ± se. Effects of glufosinate ammonium on hydrogen peroxide accumulation (C) and fungal ramification (D). Microscopic observation of hydrogen peroxide and in planta growth was performed on leaves recovered at 24 and 72 hpi. *, Hydrogen peroxide accumulation. a, Appressorium; c, conidium; m, invasive mycelia. Bars = 50 μm. Experiments were repeated more than three times; almost similar tendencies were obtained.

Effects of Hydrogen Peroxide Scavengers on the Disease Resistance Induced by Glufosinate Ammonium

Hydrogen peroxide scavengers, ascorbic acid or catalase, were infiltrated 24 h after glufosinate ammonium spray and treated transgenic plants were inoculated with M. grisea or C. miyabeanus. Both hydrogen peroxide scavengers significantly diminished blast disease protection induced by glufosinate ammonium; however, the same treatment did not affect the incidence of rice brown leaf spot (Fig. 8, A and B). Further, ascorbic acid or catalase alone did not affect rice brown leaf spot developments (data not shown). Infiltration of glufosinate ammonium did not affect appressorium formation in M. grisea and C. miyabeanus (Fig. 8B). Infiltration of glufosinate ammonium induced robust hydrogen peroxide accumulation; however, catalase or ascorbic acid almost completely inhibited hydrogen peroxide production induced by glufosinate ammonium (Fig. 8C; data not shown). In addition, glufosinate ammonium spray and catalase infiltration resulted in the vigorous in planta M. grisea ramification at 72 hpi (Fig. 8D) and abolishment of PR1, PBZ1, and POX22.3 transcriptions (Fig. 8E).

Figure 8.

Effects of ascorbic acid and catalase on rice blast and rice brown leaf spot diseases inhibited by glufosinate ammonium on bar-transgenic rice and transcriptions of PR genes. A, Transgenic rice was sprayed with 100 μg mL⁻¹ glufosinate ammonium (+) in 250 μg mL⁻¹ Tween 20 or mock only (−, 250 μg mL⁻¹ Tween 20) and then mock (−, distilled water), ascorbic acid (+), or catalase (+) was infiltrated into the carefully washed leaves 24 h after herbicide treatment. M. grisea or C. miyabeanus was inoculated 3 h after treatment with hydrogen peroxide scavengers. Photographs were taken 7 dpi. B, Quantification of rice blast (white bar) and rice brown leaf spot (black bar) disease developments and effects of catalase and ascorbic acid on the appressorium formation in M. grisea (white bar) and C. miyabeanus (black bar). Different letters indicate statistically significant differences between treatments (Duncan's multiple range test; P < 0.05). C, Effects of catalase (+) on hydrogen peroxide accumulation in glufosinate ammonium-treated plants. *, Hydrogen peroxide accumulation. a, Appressorium; c, conidium. Bars = 50 μm. D, Invasive mycelial growth in 35S:bar-transgenic rice pretreated with 100 μg mL⁻¹ glufosinate ammonium (+) and infiltrated with 5,000 units mL⁻¹ catalase (+). M. grisea was inoculated 24 h after final treatment. Photograph depicting representative infection hyphae in aniline blue staining 72 hpi. m, Invasive mycelial growth. Bars = 50 μm. E, Analyses of PR1, PBZ1, and POX22.3 transcriptions in 35S:bar-transgenic rice leaves sprayed with glufosinate ammonium (+) and/or infiltrated with catalase (−). Total RNA was prepared from five plants 24 h after infiltration, separated using denaturing gel electrophoresis, and transferred to nylon membrane. The blots were hybridized with radiolabeled PR1, PBZ1, and POX22.3 probes. All experiments were done at least three times; almost similar results were obtained.

DISCUSSION

Glufosinate Ammonium Confers Disease Protection

Glufosinate ammonium treatment protected 35S:bar-transgenic rice from blast and brown leaf spot. Number and size of lesions were significantly lower than the mock-treated plants. Effects of glufosinate ammonium on disease development in bar-expressing crops have been investigated previously. Similar with our results, glufosinate ammonium treatment protected bar-transgenic rice from blast and sheath blight caused by R. solani (Uchimiya et al., 1993; Tada et al., 1996). Similar disease-protecting ability against R. solani and S. homoeocarpa was observed in bar-transgenic bentgrass (Liu et al., 1998; Higgins et al., 2003; Wang et al., 2003). Hence, appropriate application of glufosinate ammonium could be helpful in inhibiting disease as well as in controlling weeds.

Glufosinate Ammonium Inhibits Prepenetration Morphogenesis

Disease protection by glufosinate ammonium is a result of direct inhibition of pathogens or indirect inhibition via plant-mediated responses. We examined direct effects of glufosinate ammonium on the conidial germination and appressorium formation in M. grisea and C. miyabeanus on the hydrophobic surface of GelBond (Fig. 3). The hydrophobic side of GelBond is known to be suitable to examine the chemical effects on the prepenetration development of both fungi in vitro (Lee and Dean, 1994; Ahn and Suh, 2007b). Our results showed that glufosinate ammonium (100 μg mL⁻¹) effectively inhibited appressorium formation of both fungi, while conidial germination remained unaffected. Similar results were generated for M. grisea in vitro (Tada et al., 1996). Further, this specific inhibition was also observed in the leaves of transgenic rice. Inhibition of appressorium formation is one of the strong presumptions why glufosinate ammonium diminishes both diseases. In contrast, bean (Phaseolus vulgaris) roots treated with a low dose of glufosinate ammonium enhanced sporangial germination of Pythium ultimum (Liu et al., 1997). These variable effects on development of fungal species might be due more to the alternatives of biochemical signaling pathway(s) than to the structural differences of Gln synthetase in these fungi. This is supported by the fact that Gln synthetase genes are highly conserved in their amino acid sequences in several plant pathogenic fungi and play a crucial role in ammonium assimilation and Glu biosynthesis in prokaryotes and eukaryotes, including fungi (Filetici et al., 1996; Stephenson et al., 1997). To confirm the role of inhibition of appressorium formation on the disease protection, glufosinate ammonium was washed 24 h prior to inoculation and the remaining chemicals were completely removed 1 h prior to inoculation. In spite of the normal appressorium formation, blast and brown leaf spot diseases were distinctively decreased (Fig. 7). These results suggested that glufosinate ammonium exerts disease-controlling activity via multiple routes, including inhibition of appressorium formation in M. grisea and C. miyabeanus.

Effects of Glufosinate Ammonium in Vitro and in Vivo

Glufosinate ammonium almost completely inhibited mycelial growth of both fungal pathogens on the CM containing carbon and nitrogen sources (Fig. 4). Similar with this result, glufosinate ammonium significantly inhibited mycelial growth of R. solani and S. homoeocarpa infecting bentgrass (Wang et al., 2003). Glufosinate also inhibited soil fungi and bacteria beneficial for plant growth (Ismail et al., 1995). Glufosinate reduced Pythium blight caused by Pythium aphanidermatum in transgenic bentgrass expressing bar. However, it did not alter mycelial growth in vitro (Liu et al., 1998). The inhibitory effects of glufosinate on fungal mycelial growth were significantly diminished on the medium deficient in carbon and nitrogen sources. Glufosinate did not affect germ tube elongation, another form of mycelial growth, in the absence of nutrient supplementation (Fig. 3). Our results suggest that nitrogen or carbon starvation could mimic the malnutrition conditions of host plants. Therefore, direct inhibitory effects of glufosinate on the fungal growth could be diminished or abolished in the host cells or tissues. In spite of the similarities in nutrient deficiency, glufosinate distinctively inhibited fungal growth in vivo (Fig. 7C). These results imply that glufosinate ammonium might induce multiple disease-protecting mechanisms, including inhibition of appressorium formation in M. grisea and C. miyabeanus.

Glufosinate Ammonium Induces Plant Defense Responses

Glufosinate ammonium significantly diminished the size of lesions of rice blast and brown leaf spot on transgenic rice. Other reports also showed that glufosinate ammonium treatment confers resistance against rice sheath blight on bar-transgenic rice (Uchimiya et al., 1993). Our results and this disease protection suggest that inhibition of appressorium formation is not the only mechanism to protect plants from disease as conferred by glufosinate ammonium and alteration of plant defense status. To examine this speculation, we compared the effects of pre- and postinoculation treatment on the rice blast and brown leaf spot developments. In addition, glufosinate ammonium remaining on rice leaves was washed carefully 24 h after spray to rule out its direct effects on the pathogen infection. As shown in Figure 2, preinoculation treatment also significantly inhibited progressions of both diseases. The same results were observed in the bar-transgenic rice (Tada et al., 1996). Further, the most distinctive rice brown leaf spot protection was accomplished by the treatment 1 d prior to C. miyabeanus inoculation. These results imply that indirect effects like modulation of host defense responses might be responsible for the disease protection by glufosinate ammonium.

We analyzed molecular and cellular defense-related responses in transgenic rice. Glufosinate ammonium treatment induced transcriptions of PR1, PBZ1, and POX22.3 in transgenic rice and PR1 transcription in transgenic Arabidopsis. This transcription was dose dependent, translocated systemically, and retained for a long period. Both M. grisea and C. miyabeanus did not alter the transcriptions of PR genes triggered by glufosinate ammonium. Similar results were also observed in transgenic Col-0 containing PR1:eGFP and 35S:bar. The role of PR1 remains unclear; however, transcript accumulation of PR1 has been recognized as a molecular marker to determine whether the plant is in a resistant state or not. Constitutive transcriptional activation of PR genes and distinctive disease protection were observed in the benzothiadiazole-treated tobacco (Nicotiana tabacum; Friedrich et al., 1996), Pto-overexpressing tomato (Solanum lycopersicum) plants (Tang et al., 1999), and Arabidopsis cpr and dnd mutants (Bowling et al., 1997; Clough et al., 2000; van Hulten et al., 2006). These results imply that glufosinate ammonium could induce resistance in the transgenic rice and Arabidopsis. Similar to our results, treatment with lactofen, a herbicide belonging to the diphenylether class, induced robust transcriptions of PR-1a, PR-5, and PR-10 in soybean (Glycine max) leaves (Graham, 2005). In addition, glufosinate ammonium induced systemic transcriptions of three PR genes within the leaves of transgenic plants. Therefore, the effect of glufosinate ammonium mobilized in other parts of the plant. Systemic translocation of defense-related signaling has been described in tobacco, rice, and Arabidopsis treated with plant defense activators like benzothiadiazole, probenazole, and salicylic acid (Friedrich et al., 1996; Midoh and Iwata, 1997; Maleck et al., 2000).

We investigated the effects of glufosinate on the hydrogen peroxide accumulation and PR1 and PDF1.2 transcriptions in Arabidopsis Col-0, bacterial NahG-expressing Col-0, and three mutants harboring 35S:bar construct (Fig. 6). These lines are unable to metabolize salicylic acid, synthesize mRNAs of PR genes, or perceive jasmonic acid or ethylene-dependent signaling. Glufosinate ammonium triggered robust accumulation of hydrogen peroxide in all tested lines. Glufosinate-treated transgenic Col-0 showed high PR1 transcription; however, no PDF1.2 transcription was observed. Glufosinate did not trigger PR1 transcription in the NahG and npr1 lines. These results strongly suggest that glufosinate exerts its effects through the hydrogen peroxide and salicylic acid-dependent signaling pathways and there is no tight correlation between hydrogen peroxide accumulation by glufosinate and jasmonic acid-dependent signaling pathways. Similar dependencies on hydrogen peroxide and salicylic acid and tight correlation between them were also observed in β-aminobutyric acid-treated tobacco and Arabidopsis (Siegrist et al., 2000; Zimmerli et al., 2000; Ganesan and Thomas, 2001; Orozco-Cardenas et al., 2001; Zimmerli et al., 2001; Kachroo et al., 2003). In addition, correlation between resistance against rice blast and hydrogen peroxide and/or salicylic acid has been described (Ganesan and Thomas, 2001; Uchimiya et al., 2002; Agrawal et al., 2003).

Glufosinate ammonium triggered slight chlorosis and reduction of chlorophyll content (F_v/F_m levels) in transgenic plants (Fig. 7, A and B). Although transgenic plants contain PAT, detoxifying glufosinate, the above phenomena imply that glufosinate induces temporary phytotoxicity on bar-transgenic rice. Similar phytotoxic responses were observed on other transgenic rice plants expressing the bar gene as the selectable marker (data not shown). Glufosinate ammonium induced hydrogen peroxide accumulation and pathogen inoculation did not affect this defense-related cellular response (Fig. 7C). Therefore, there should be correlation between AOS production by glufosinate and activation of disease defense responses. Hydrogen peroxide is involved in cell wall reinforcement, pathogen abolishment, cell death, and modulates plant hypersensitive disease resistance (Levine et al., 1994). Because of its relatively long half-life and good permeability, hydrogen peroxide is generally accepted as the major AOS messenger (Bowler and Fluhr, 2000).

To confirm the role of hydrogen peroxide, the effects of hydrogen peroxide scavengers on disease resistance triggered by glufosinate ammonium were investigated (Fig. 8, A and B). Both hydrogen peroxide scavengers significantly attenuated resistance against blast induced by glufosinate ammonium; however, the same treatment did not affect resistance against brown leaf spot. These results indicate that hydrogen peroxide accumulation, indispensable for blast disease resistance, is not necessary for expression of resistance against rice brown leaf spot. In addition, inhibition of appressorium formation is not the only mechanism of protection against rice blast and brown leaf spot, because glufosinate ammonium infiltration successfully inhibited this disease in bar-expressing rice in spite of the vigorous appressorium formation by M. grisea and C. miyabeanus (Fig. 8, B and C). Microscopic observation of M. grisea growth in planta also corroborates these results (Fig. 8D). Infiltration of catalase with glufosinate abolished inhibitory effects of glufosinate ammonium on fungal ramification. Further, the same treatment nullified PR1, PBZ1, and POX22.3 transcriptions induced by infiltration with glufosinate (Fig. 8E). Glufosinate ammonium-induced disease protection against M. grisea and C. miyabeanus is due to the induction of multiple defense responses and some of them are dependent on hydrogen peroxide accumulation (against M. grisea), while others are not (against C. miyabeanus). As described above, hydrogen peroxide is one of the general AOS messengers and its accumulation is observed in the host resistant for infected pathogens (Chamnongpol et al., 1998). They are also observed on plants pretreated with plant defense activators or resistance-promoting rhizobacteria and inoculated with compatible pathogens (Park et al., 2000; Ahn et al., 2007b). This defense-related response effectively fends off pathogen invasion and in planta pathogen growth as hydrogen peroxide accumulation is often culminated in host cell death. In contrast, hydrogen peroxide accumulation did not affect infection and disease development by C. miyabeanus. This finding is consistent with one of the most representative characteristics of disease progression by necrotrophic pathogens, such as Botrytis cinerea and Pectobacterium carotovorum (Govrin and Levine, 2000; Govrin et al., 2006). There should be other unknown defense mechanisms triggered by glufosinate, which is independent from hydrogen peroxide accumulation or host cell death and effective in inhibition of rice brown leaf spot.

In Figure 9, an overall proposed model is presented to show possible mechanisms of glufosinate ammonium in the bar-transgenic rice. This model is based on previous literature and findings from this study. One of the most important observations is distinctive accumulation of hydrogen peroxide by glufosinate ammonium in the transgenic rice and Arabidopsis. Free radical production by exogenous stimuli resulted in the burst of AOS like hydrogen peroxide (Rao et al., 1996; Karpinski et al., 1999; Fryer et al., 2002). Hydrogen peroxide acts as a signaling messenger in the systemic translocation of defense responses, and part of them effectively fends off M. grisea infection (Chamnongpol et al., 1998). All resistance mechanisms by glufosinate are not fully dependent on hydrogen peroxide, because glufosinate-induced resistance against C. miyabeanus was retained in the absence of hydrogen peroxide. In addition, the importance of direct effects of glufosinate ammonium on the inhibition of prepenetration development in both pathogens should not be ruled out in the disease protection caused by this herbicide.

Figure 9.

Proposed model for glufosinate-induced disease resistance. Glufosinate initiates accumulation of free radicals by irreversible binding with and inactivation of Gln synthetase. Toxic ammonia derived from photorespiration or nitrogen assimilation is increased within the cell and disturbed electron transport system within chloroplast. Free radicals were produced and in turn, this molecule triggers disease resistance against M. grisea and C. miyabeanus.

MATERIALS AND METHODS

Fungal Isolates, Plants, and Chemicals

Magnaporthe grisea strain KJ201 and Cochliobolus miyabeanus isolate HIH-1, virulent on rice (Oryza sativa) ‘Dongjin’, were obtained from the National Institute of Agricultural Science and Technology, Rural Development Administration and recovered from rice leaves showing typical symptoms of brown leaf spot, respectively. Conidia of M. grisea were harvested from 10- to 12-d colonies grown on oatmeal agar (Ahn et al., 2005a). Conidia of C. miyabeanus were harvested from 7-d-old cultures grown on Suc Pro agar (Dhingra and Sinclair, 1985) at 22°C under continuous fluorescent light. Nontransgenic and transgenic ‘Dongjin’ harboring bar gene under the regulation of 35S promoter were also obtained from the National Institute of Agricultural Biotechnology, Rural Development Administration. Rice seeds were surface sterilized by immersion in a 100 μg mL⁻¹ solution of thiophanate methylthiram for 16 h and sown in a commercial soil mixture at a density of three plants per 5- × 5-cm pot. Rice plants were propagated in the greenhouse for 4 weeks. Glufosinate ammonium (99% pure) was purchased from Riedel-de Haen and resuspended in distilled water. Arabidopsis (Arabidopsis thaliana) wild-type Col-0, transgenic Col-0 expressing bacterial NahG gene, and mutants (npr1, etr1, and jar1) from this line were obtained from The Arabidopsis Information Resource. Arabidopsis was grown in a growth chamber at 22°C, 65% to 70% relative humidity, and 16 h of illumination daily. Four-week-old Arabidopsis was used for chemical treatment.

Generation of Transgenic Arabidopsis and Treatments

To examine the effects of glufosinate ammonium on the PR1 transcription in Arabidopsis, the 2,000-bp upstream region of PR1 in Arabidopsis Col-0 genome was amplified by PCR using forward primer (5′-ggggacaagtttgtacaaaaaagcaggctATCTCATTTTATCCGTTCGC-3′) and reverse primer (5′-ggggaccactttgtacaagaaagctgggtTTTTCTAAGTTGATAATGGT-3′). This fragment was introduced into destination vector pBGWFS7 (Karimi et al., 2002), harboring 35S:bar, by GATEWAY cloning system according to the manufacturer's recommendations (Invitrogen). The resulting expression vector contained eGFP and GUS gene adjacent to the introduced PR1 promoter and was introduced into Agrobacterium tumefaciens strain GV3101. This bacterial strain was used for transformation of Arabidopsis wild-type Col-0 via floral dip (Clough and Bent, 1998). In addition, pBGWFS7 was transformed into the same bacterial strain and used for transformation of Arabidopsis wild-type Col-0, NahG (transgenic Col-0 line expressing the bacterial salicylate hydroxylase), npr1 (a mutant that does not accumulate PR1 in response to salicylic acid), etr1 (an altered perception of ethylene mutant), and jar1 (a mutant that displays reduced sensitivity to methyl jasmonate). T₃ seeds from independent transgenic lines were surface sterilized with 70% ethanol and 5% sodium hypochlorite and were grown on Murashige-Skoog medium. After incubation for 7 d at 22°C, the seedlings were submerged in distilled water (mock) or 100 μg mL⁻¹ glufosinate ammonium for 3 h and washed three times with distilled water. Green fluorescence was observed using Confocal Laser Microscope (Olympus, Fluoview TM 300) 6 h after treatment. In addition, total RNA was extracted from the transgenic Col-0 plants and PR1 activity was assayed by northern analysis as described later.

Conidial Germination and Appressorium Formation in Vitro and in Vivo

To examine the effects of glufosinate ammonium on the prepenetration development on artificial surface and on rice leaves, conidia of both fungi were harvested and dropped with or without glufosinate ammonium, as described previously (Oh and Lee, 2000; Ahn and Suh, 2007a). Percentages of germinated and germinating conidia to form appressoria were determined from at least 100 conidia with three replicates per treatment. Experiments were done independently at least three times.

In Vivo Effect of Glufosinate Ammonium on Rice Blast and Brown Leaf Spot

Wild-type and transgenic rice plants were grown in commercial soil mix in plastic pots (5 cm in diameter) for 4 weeks in the greenhouse. Mock (250 μg mL⁻¹ Tween 20) or glufosinate ammonium (100 μg mL⁻¹ in 250 μg mL⁻¹ Tween 20) was sprayed on 10 rice plants 5 d or 1 d prior to and 12 h and 1 d after M. grisea (2 × 10⁵ conidia mL⁻¹) or C. miyabeanus (1 × 10⁵ conidia mL⁻¹) inoculation. Rice leaves were sprayed with distilled water 24 h after glufosinate ammonium treatment. The inoculated rice plants were placed in a dew chamber (25°C, 100% relative humidity) for 16 to 24 h and transferred to the greenhouse. Progression of rice blast was assessed 10 d after inoculation and brown leaf spot was assessed 7 d after inoculation. The diseases were estimated according to the method developed by the International Rice Research Institute (1988). In vivo assays were done independently more than three times.

Nutrient Deficiency on the Pathogen Growth Inhibited by Glufosinate

Mycelial blocks (6 mm in diameter) from actively growing colony edges of M. grisea and C. miyabeanus were placed on CM or CM without nitrogen and carbon sources and supplemented with mock (distilled water) or 100 μg/mL glufosinate ammonium (Yang et al., 1994; Talbot et al., 1996). The colony area was estimated after incubation at 25°C in the dark for 7 d.

Hydrogen Peroxide Accumulation and Invasive Mycelial Growth

To investigate the effect of glufosinate ammonium on hydrogen peroxide accumulation, 10 transgenic rice plants were applied with 100 μg mL⁻¹ glufosinate ammonium. In addition, the same chemical was sprayed onto rosette leaves of Arabidopsis transformed with pBGWFS7 containing 35S:bar. All tested plants were grown on the soil. Rice plants were infected 1 d later with virulent KJ201. Histochemical detection of hydrogen peroxide was performed as described previously (Wohlgemuth et al., 2002) with minor modification. To determine the effects of glufosinate ammonium on the accumulation of hydrogen peroxide, rice leaves were recovered at 24 hpi and stained with 0.1% (w/v) diaminobenzidine (DAB; Sigma). In addition, rosette and cauline leaves of transgenic Arabidopsis treated with glufosinate were stained with DAB. Stained plant leaves were cleared with 96% (v/v) ethanol, preserved in 50% (v/v) ethanol, and observed under the light microscope. DAB staining was a red-brown color under the light microscope. Quantitative determination of hydrogen peroxide within transgenic Arabidopsis was performed as described (Ahn et al., 2007a). Briefly, debris was removed by perchloric acid extraction, purified using AG1-X8 resin (Bio-Rad Laboratories), and hydrogen levels were determined using Autolumat LB953 luminometer (EG & G Derthod). To observe fungal growth within rice plants, leaves were recovered 96 hpi, fixed with lactophenol, stained with 0.1% (w/v) aniline blue, and mycelial growth was observed under the light microscope (Peng et al., 1986). More than 15 leaves from five randomly selected plants were observed in each experiment. These experiments were done independently at least three times.

Estimation of Chlorophyll Fluorescence

Rice was grown as described above. Green parts of 10 seedlings were cut by scissors and floated on 100 μg mL⁻¹ glufosinate ammonium under continuous fluorescent light of 150 μmol m⁻² s⁻¹. F_v/F_m value was measured using CF-1000 chlorophyll fluorescence measurement system (Morgan Scientific) as previously described (Artus et al., 1996; Jang et al., 2003).

Effects of Catalase and Ascorbic Acid on the Hydrogen Peroxide Accumulation Induced by Glufosinate Ammonium

Approximately 4-week-old transgenic rice was sprayed with mock (250 μg mL⁻¹ Tween 20) or 100 μg mL⁻¹ glufosinate ammonium. Twenty-four hours after treatment, mock (distilled water), 5,000 units mL⁻¹ catalase, or 10 mm ascorbic acid were treated via vacuum infiltrated at 710 mmHg for 10 min after complete washing with distilled water. After complete washing, infiltrated rice plants were inoculated 3 h later with M. grisea strain KJ201 or C. miyabeanus strain HIH-1. Inoculation and estimation of disease development were performed as described above.

RNA Preparation and Transcription Analyses

Rice leaves were harvested for RNA isolation at 0, 1, 2, 3, and 4 d after fungal inoculation or 1 d after glufosinate ammonium spray. Total RNA was also extracted from 4-week-old transgenic Arabidopsis plants. To determine whether the effects of glufosinate ammonium on PR1/PDF1.2 transcriptions could be translocated systemically, glufosinate ammonium was sprayed on rosette leaves of the 35S:bar-transgenic Arabidopsis and rosette and cauline leaves were harvested. In addition, salicylic acid- (500 μm) or jasmonic acid (100 μm)-treated rosette leaves were harvested from transgenic Col-0 24 h after treatment. Harvested plant materials were preserved at −70°C. Total RNA was extracted using a lithium chloride-precipitation method (Davis and Ausubel, 1989). For hybridization analyses, 5 μg of total RNA per lane was separated electrophoretically in denaturing formaldehyde-agarose gel (8% formaldehyde, 0.5× MOPS, 1.5% agarose) and blotted to Hybond-N⁺ membrane (Amersham Pharmacia Biotech) by capillary transfer. Uniform sample loading was confirmed by staining ribosomal RNA bands in the gel with ethidium bromide. DNA probes were labeled with [α-³²P] dCTP using a random primer labeling kit (Boeringer-Mannheim). The PR1, PBZ1, and POX22.3 probes were derived from cDNA clones (Kim et al., 2001a; Ahn et al., 2005a). The 379-bp eGFP probe was prepared by PCR using forward primer 5′-cacatgaagcagcacgactt-3′ and reverse primer 5′-tgctcaggtagtggttgtcg-3′. Analyses of Arabidopsis PR1 and PDF1.2 gene transcriptions were performed using the reverse transcription-PCR as described (Pieterse et al., 1998) with some modifications. First-strand cDNA was synthesized from 50 ng of total RNA of the leaves using the Reverse-iT first-strand synthesis kit and anchored oligo(dT) as indicated by the manufacturer's instructions (AB gene). Independent PCR using equal aliquots (0.5 μL) of cDNA samples was performed using PR1/PDF1.2-specific primers as described (Vieira Dos Santos et al., 2003). The TUBULIN gene was amplified as a quantitative control (Lee et al., 2000).