Necrosis- and Ethylene-Inducing Peptide from Fusarium oxysporum Induces a Complex Cascade of Transcripts Associated with Signal Transduction and Cell Death in Arabidopsis

Abstract

Treatment of Arabidopsis (Arabidopsis thaliana) with a necrosis- and ethylene-inducing peptide (Nep1) from Fusarium oxysporum inhibited both root and cotyledon growth and triggered cell death, thereby generating necrotic spots. Nep1-like proteins are produced by divergent microbes, many of which are plant pathogens. Nep1 in the plant was localized to the cell wall and cytosol based on immunolocalization results. The ratio of chlorophyll a fluorescence (F685 nm/F730 nm) significantly decreased after 75-min treatment with Nep1 in comparison to the control. This suggested that a short-term compensation of photosynthesis occurred in response to localized damage to cells. The concentrations of most water-soluble metabolites analyzed were reduced in Arabidopsis seedlings after 6 h of Nep1 treatment, indicating that the integrity of cellular membranes had failed. Microarray results showed that short-term treatment with Nep1 altered expression of numerous genes encoding proteins putatively localized to organelles, especially the chloroplast and mitochondria. Short-term treatment with Nep1 induced multiple classes of genes involved in reactive oxygen species production, signal transduction, ethylene biosynthesis, membrane modification, apoptosis, and stress. Quantitative PCR was used to confirm the induction of genes localized in the chloroplast, mitochondria, and plasma membrane, and genes responsive to calcium/calmodulin complexes, ethylene, jasmonate, ethylene biosynthesis, WRKY, and cell death. The majority of Nep1-induced genes has been associated with general stress responses but has not been critically linked to resistance to plant disease. These results are consistent with Nep1 facilitating cell death as a component of diseases caused by necrotrophic plant pathogens.

The fungal plant pathogen Fusarium oxysporum produces necrosis- and ethylene-inducing peptide (Nep1), a 24-kD necrosis and ethylene-inducing peptide (Bailey, 1995; Bailey et al., 1997). Nep1 and Nep1-like proteins (NLPs) from other microbes cause necrosis and induce ethylene production in dicots but are inactive in monocots (Bailey, 1995; Veit et al., 2001). NLPs have been identified in many different microorganisms including fungi and bacteria (for review, see Pemberton and Salmond, 2004). Bae et al. (2005a) identified complex multiple copies of NEP1 orthologs in five Phytophthora species. Phytopthera sojae necrosis-inducing protein was produced during the necrotrophic phase of infection in soybean (Glycine max; Qutob et al., 2002). The presence of multiple NLPs in various plant pathogenic microorganisms suggests a significant role for these genes in pathogenicity.

Plant responses to Nep1 include induction of pathogen-related (PR) genes, changes in K⁺ and H⁺ channel fluxes, callose apposition, accumulation of reactive oxygen species (ROS) and ethylene, altered cell respiration, the hypersensitive response, and localized cell death (Jennings et al., 2001; Fellbrich et al., 2002; Keates et al., 2003). In Arabidopsis (Arabidopsis thaliana), spotted knapweed (Centaurea maculosa), and dandelion (Taraxacum officinale), Nep1 caused the breakdown of the cuticle layer and internal chloroplast membrane structures 1 to 4 h after treatment began (Keates et al., 2003). Various genes involved in plant stress responses including wounding, drought, senescence, and disease resistance, were also induced (Keates et al., 2003). A related peptide, NPP1, induced rapid accumulation of the salicylic acid (SA)-dependent PR1 transcript in Arabidopsis (Fellbrich et al., 2002).

The objective of this study was to characterize the responses of Arabidopsis to Nep1 with respect to changes in plant growth, ultrastructural modifications, metabolite levels, and transcription profiles. Only a handful of genes that respond to Nep1 treatment have been identified to date, even though the effects of Nep1 on plants can be dramatic. In this investigation, high-throughput screening based on DNA microarray technology was used to identify Nep1 responsive genes in Arabidopsis. In addition, chlorophyll a fluorescence imaging was employed as a nondestructive method to assess plant vigor subsequent to Nep1 treatment. Molecular and biochemical analyses of the response of Arabidopsis to Nep1 should enhance our understanding of the signaling networks that are involved in plant responses to Nep1 and NLPs, a family of proteins of increasing importance in many different plant microbe interactions.

RESULTS AND DISCUSSION

Nep1 Inhibits Seedling Growth and Root Development

To assess the effects of Nep1 on plant growth, Arabidopsis seedlings were grown for 5 d on agarose plates containing 1× Murashige and Skoog (MS) basal salts supplemented with 1% Suc. When Arabidopsis seedlings were treated with Nep1, the growth of roots and cotyledons was inhibited in comparison to control plants (Figs. 1 and 2). Necrotic spots were visible on cotyledons of Nep1-treated seedlings (Fig. 1A, inset), particularly in association with stomata (Fig. 1C). The formation of necrotic spots was also observed on cotyledons of dark-grown seedlings (Fig. 1J). Nep1 treatment inhibited the formation of root hairs and damaged root tips both under dark and light conditions (Fig. 1, E, G, and L). White deposits were detected in roots of Nep1-treated seedlings, suggesting that callose formation had occurred (Fig. 1, E and L). Nep1 treatment altered the shape of cells on the root surfaces resulting in distorted root growth. When grown under light, cotyledon development and hypocotyl elongation were inhibited 79% and 55% by Nep1 treatment, respectively (Fig. 2). The inhibitory effects of Nep1 on root growth were severe (92% inhibition), and the inhibition was greater in light-grown than in dark-grown seedlings (Fig. 1, A and H). However, there was negligible inhibition of cotyledon development and hypocotyl elongation in response to Nep1 using dark-grown seedlings (Fig. 1H). Nep1 penetrates plant tissues via openings such as stomata, hydathodes, or wounds (Bailey et al., 2000; Jennings et al., 2001). Developing roots have many penetration points, including areas where lateral roots emerge and root tips that resemble wounded tissues (Scheres et al., 2001). Nep1 was able to damage the root epidermis and cortex. However, Nep1 was unable to penetrate the casparian strip and enter the xylem of unwounded roots. Nep1 was shown to be transported through the xylem if access to this tissue was available (Jennings et al., 2001). Etiolated Arabidopsis seedlings have unexpanded cotyledons with minimal mature stomata, thereby limiting penetration of Nep1 through these points (Nadeau and Sack, 2001). The movement of Nep1 from the root into the xylem and to the shoot has not been demonstrated. This would require significant movement across cell membranes and between cells through the casparian strip before entering the vascular tissues for transport to the shoot. As a result, Nep1 treatment caused less damage to the developing hypocotyls and cotyledons of etiolated compared to light-grown seedlings. Other factors may contribute to the reduced response to Nep1 in the dark. For example, in the absence of light, ROS production would be limited in the etioplasts due to the absence of chlorophyll and thylakoid membrane structures. The reduced production of ROS might cause less damage to plants.

Figure 1.

Effect of F. oxysporum Nep1 on Arabidopsis growth. Nep1 treatment caused necrotic lesions and inhibited cotyledon and root growth. Sterilized Arabidopsis seeds were mixed with Nep1 (20 μg mL⁻¹), 0.001% (v/v) Silwet-L77, 0.3% agarose, 1% Suc in 1× MS media, and plated onto petri dishes. Controls were treated with 0.001% (v/v) Silwet-L77. Seeds were germinated and grown in growth cabinets at 22°C for 5 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ PAR with 16 h light or under dark condition. A, Control seedling (left, no Nep1) and Nep1-treated seedling (right). Inset shows necrotic spots on cotyledons. B, Normal stomata. C, Necrotic region on Nep1-treated cotyledon. D, Normal roots with root hairs. E, Damaged root without root hairs from Nep1-treated seedling. F, Normal root tip. G, Abnormal root tip from Nep1-treated seedling. H, Control (left) and Nep1-treated (right), etiolated seedlings. I, Normal cotyledon of etiolated seedling. J, Cotyledon with necrotic regions in Nep1-treated, etiolated seedling. K, Normal root hairs from etiolated control seedling. L, Abnormal root without root hairs in Nep1-treated etiolated seedling. Growth conditions, A to G, 16 h light; H to L, dark.

Figure 2.

Inhibitory effect of F. oxysporum Nep1 on Arabidopsis growth. Sterilized Arabidopsis seeds were mixed with Nep1 (20 μg mL⁻¹), 0.001% (v/v) Silwet-L77, 0.3% agarose, 1% Suc in 1× MS media, and plated onto petri dishes. Controls were treated with 0.001% (v/v) Silwet-L77. Seeds were germinated and grown in growth cabinets at 22°C for 5 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ PAR with 16 h light. Nep1 treatment reduced the length of hypocotyl and root, 55% and 93%, respectively. The area of cotyledon was reduced by 79%. Values are the means ± se from 10 independent replicates, with at least six seedlings per replicate.

Nep1 Treatment Damages Membranes and Chloroplast Ultrastructure

Ultrastructural changes in response to Nep1 treatment were examined in 5-d-old seedlings using transmission electron microscopy. While control plants developed normal chloroplasts (Fig. 3A) chloroplasts in Nep1-treated cotyledons were filled with plastoglobuli and unstacked thylakoid membrane structures (Fig. 3B). Nep1 damage to other organelles was not observed. Keates et al. (2003) reported similar alterations to chloroplasts in Arabidopsis and in two invasive weed species within 4 h of Nep1 treatment. Changes to chloroplasts and membrane structures also have been observed in association with the hypersensitive response and during leaf senescence (Goodman et al., 1986; Meyer and Heath, 1988). Based on the immunolocalization of Nep1 in treated Arabidopsis tissues, electron-dense gold particles were mainly associated with the cell wall and cytosol (Fig. 3D). Gold particles were not observed to be associated with membrane structures, such as chloroplast envelope, mitochondria, vacuole, and endoplasmic reticulum. This result indicates that Nep1 can penetrate through plant cells but may be not able to penetrate into organelles.

Figure 3.

Ultrastructure of Arabidopsis chloroplasts and immunolocalization of Nep1. Sterilized Arabidopsis seeds were mixed with Nep1 (20 μg mL⁻¹), 0.001% (v/v) Silwet-L77, 0.3% agarose, 1% Suc in 1× MS media, and plated onto petri dishes. Controls were treated with 0.001% (v/v) Silwet-L77. Seeds were germinated and grown in growth cabinets at 22°C for 5 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ PAR with 16 h light prior to sampling for transmission electron microscopy and immunolocalization of Nep1. A, Chloroplasts from a control true leaf. B, Chloroplasts from a Nep1-treated true leaf. Nep1 treatment causes breakdown of chloroplast internal membrane structures and generates lipophilic bodies. C, In control plant, gold particles were not detected. D, Numerous gold particles are present over the plant cell wall/membrane and cytosol. Scale bar = 500 nm. CH, Chloroplast; CW, cell wall; CY, cytosol.

Nep1 Causes Large-Scale Leakage of Cell Metabolites

Changes of metabolites in Arabidopsis seedlings were analyzed with gas chromatography coupled to a mass selective detector. For these measurements, 7-d-old seedlings were grown in liquid culture and Nep1 treatments were for 0, 1, and 6 h. As shown in Table I, most metabolite concentrations were markedly decreased in the 6-h Nep1-treated samples compared to the controls. Conversely, Arabidopsis metabolites were little affected by Nep1 following 1 h of treatment. These results suggest that Nep1 treatment caused a general leakage of the metabolites after 6 h of treatment, which may be due to membrane damage (i.e. depolarization). The treatment of tobacco (Nicotiana tabacum) cell suspensions with an ethylene biosynthesis-inducing endoxylanase also caused plasma membrane depolarization resulting in the leakage of 1-aminocyclopropane-1-carboxylate (ACC) into the growing medium (Bailey et al., 1992). Cryptogein, a 10-kD elicitor of Phytophthora cryptogea, caused the acidification of the cytosol, alkalization of the medium, and the depolarization of plasma membrane in tobacco (Pugin et al., 1997). Val uptake was also inhibited in cryptogein-treated tobacco cells, which might be due to the decrease of proton motive force (Bourque et al., 2002). Because amino acid transporters are also proton symporters (Li and Bush, 1990), the same explanation might also impact the decrease of amino acids in 6-h Nep1 treatment.

Table I.

Changes of metabolite levels in 7-d-old Arabidopsis seedlings in response to Nep1 treatment (10 μg mL⁻¹)

Sample means are in mg g⁻¹ FW and are for n = 8 to 10. Significant differences were for P ≤ 0.01 (^**) and P > 0.05, respectively. FW, Fresh weight.

Compound	Treatment	Time			% Change in 6 h
				0 h		1 h	6 h
			μg g−1 FW
Lactic acid	Control	65.62 ± 11.66	56.47 ± 5.99	47.05 ± 6.08	+0.9
	Nep1		38.36 ± 9.97	47.48 ± 7.46
Val	Control	6.31 ± 1.56	5.04 ± 1.00	5.71 ± 1.20	−93.5
	Nep1		4.99 ± 0.86	0.37 ± 0.26**
Leu	Control	2.24 ± 0.30	2.20 ± 0.65	2.55 ± 0.59	−95.2
	Nep1		2.84 ± 0.59	0.12 ± 0.12**
Gly	Control	15.91 ± 3.85	5.18 ± 0.47	6.75 ± 1.52	−44.3
	Nep1		11.49 ± 2.18**	3.76 ± 0.94
Ser	Control	70.87 ± 8.46	51.69 ± 6.13	59.82 ± 6.18	−81.4
	Nep1		47.25 ± 6.31	11.14 ± 1.37**
Thr	Control	3.87 ± 0.29	4.86 ± 0.27	4.02 ± 0.28	−44.4
	Nep1		4.68 ± 0.57	2.24 ± 0.26**
Maleic acid	Control	4.34 ± 1.05	4.02 ± 0.33	4.13 ± 0.32	−32.9
	Nep1		4.96 ± 0.39	2.77 ± 0.20**
Asp	Control	8.77 ± 1.63	4.60 ± 0.34	5.86 ± 0.46	−78.6
	Nep1		6.36 ± 1.10	1.25 ± 0.55**
Glu	Control	226.90 ± 20.85	248.46 ± 34.79	257.18 ± 19.26	−66.2
	Nep1		212.01 ± 24.73	86.96 ± 9.61**
Orn	Control	126.62 ± 9.77	104.98 ± 11.32	110.39 ± 13.05	−87.2
	Nep1		119.10 ± 9.64	4.15 ± 3.78**
Lys	Control	24.39 ± 6.57	18.40 ± 2.94	17.80 ± 2.53	−94.5
	Nep1		16.09 ± 3.04	0.98 ± 0.39**
Myoinositol	Control	14.41 ± 1.42	11.26 ± 1.47	11.11 ± 0.57	+58.9
	Nep1		13.82 ± 1.20	27.07 ± 3.20**

Although we were looking for increases in specific metabolites to help us identify metabolic pathways for future studies related to Nep1 action, the vast majority of metabolites decreased in concentration after Nep1 treatment. The concentration of myoinositol increased in response to 6 h of Nep1 treatment (59% increase). This was likely due to uptake from the liquid media, because myoinositol is present in the MS basal salts mixture used here. There was no evidence of transcriptional induction involved in inositol biosynthesis in the microarray results. Alternatively, it is possible inositol is synthesized in response to Nep1 treatment as biosynthesis of membrane and cell wall components collapse. Inositol has important roles in membrane biogenesis, signaling, and plant growth (for review, see Stevenson et al., 2000). Myoinositol is also used for stress-induced accumulation of a methylinositol, d-ononitol using myoinositol O-methyltranserase (OMT), and inositol-containing lipids are components of membranes. Lactic acid was not altered in response to Nep1 treatment, possibly because of its localization in roots and hypocotyls that are less affected by Nep1. Another possibility is that Nep1 treatment induced the synthesis of lactic acid compensating for any leakage. Lactic acid may accumulate in the cytosol, lowering the cytoplasmic pH and thereby contributing to cell death.

After 48 h of exposure to Nep1 in liquid culture, the treated Arabidopsis seedlings were transferred to a petri dish containing moist filter paper that was sealed, placed in the dark for 1 d, and subsequently transferred to light as described in “Materials and Methods.” Although necrotic spots were observed after 1 d in the light, the Nep1-treated seedlings completely recovered in the absence of further treatment. In comparison, isolated necrotic spots were not detected in seedlings continuously exposed to Nep1 in liquid culture, and all seedlings died after 1 week of exposure to Nep1.

Chlorophyll a Fluorescence Imaging Suggests a Compensatory Response to Nep1 Treatment

Images of chlorophyll a fluorescence emissions at 685 and 730 nm were measured from the adaxial surfaces of 7-d-old Arabidopsis cotyledons grown in liquid culture treated with either Nep1 or 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU; Sigma). Fluorescence emission in the red and far-red regions of the spectrum (F685 and F730) in plants emanates from protein-bound chlorophyll a molecules in the photosynthetic apparatus (Bailey-Serres and Chang, 2005). PSII generates both F685 and F730 fluorescence, while PSI generates a small fraction of F730 fluorescence. The F685/F730 fluorescence ratio has been used as an indicator of plant vigor, and it has an inverse relationship with photosynthetic activity (Stober et al., 1994; Kim et al., 2001b). Representative false-color, fluorescence images of the control, Nep1, and DCMU-treated samples are presented in Figure 4. The F685/F730 ratio decreased approximately 4% within 75 min after treatment with Nep1 when compared to the control. In comparison, DCMU treatment increased the F685/F730 ratio 10% compared to the control within the first 15 min of Nep1 treatment. After 75 min, the F685/F730 ratios were 2.47 ± 0.05, 2.37 ± 0.02, and 2.71 ± 0.04 for the control, Nep1, and DCMU treatments, respectively. DCMU blocks electron transport from PSII to PSI, and this results in a rapid increase of chlorophyll a fluorescence (Hodges and Barber, 1986). The F685/F730 ratios for Nep1-treated samples continued to decrease until the end of the 3-h measurement period when the fluorescence ratios were 2.55 ± 0.04 and 2.12 ± 0.03 for the control and Nep1-treated samples, respectively. A decrease in F685/F730 ratio suggests more active photosynthetic activity in the Nep1-treated samples compared to either the control or DCMU-treated samples. This may be a compensatory response of healthy cells/chloroplasts for the Nep1-induced damage to cells/chloroplasts.

Figure 4.

Chlorophyll a fluorescence image detection after treatment of Nep1. Chlorophyll a fluorescence images were measured from the adaxial surfaces of Arabidopsis cotyledons treated with Nep1. DCMU (200 mm) was used as an electron blocking agent in an electron transport chain of photosystem. Seedlings were grown at 22°C for 7 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ PAR with 24 h light. Seven-day-old Arabidopsis seedlings grown in the liquid culture were treated with Nep1 (10 μg mL⁻¹) without Silwet. Control seedlings were treated with an equivalent volume of sterile distilled water. Fluorescence was measured for up to 180 min after treatment. Chlorophyll a fluorescence images were from ratio images of F685/F730, and the graph (bottom) was drawn based on the ratio of F685/F730. Relative fluorescence intensity is given in vertical color scales on the left.

Total chlorophyll (a + b) and carotenoid concentrations (x + c) were not altered by Nep1 treatment (data not shown). However, the chlorophyll a/b ratio was lower in the 6-h Nep1-treated seedlings (2.79 ± 0.12) than in the controls (3.59 ± 0.15). Ninety percent of chlorophyll b is bound to the antennae protein light-harvesting complex II (LHCII), while 90% of chlorophyll a is bound to chlorophyll-protein complexes other than LHCII, such as the PSII core complex (Kura-Hotta et al., 1986). In Nep1-treated tissues, PSII core complexes might be degraded faster than LHCII during the rearrangement of chloroplast internal membrane structures. A reduced chlorophyll a/b ratio (22% reduction) also occurred during leaf senescence in rice (Tang et al., 2005).

Nep1 Rapidly Induces Large-Scale Changes in Gene Expression

Changes in transcript levels in Arabidopsis seedlings in response to Nep1 treatment were measured by DNA microarray analysis using Affymetrix ATH1 chips. Thirty minutes after treatment with Nep1, the steady-state level of 458 transcripts was up-regulated, and 26 transcripts were down-regulated by more than 2-fold. Transcripts showing more than 2-fold induction and all of the repressed transcripts are listed in Supplemental Table I. Genes showing altered expression levels were divided into eight functional categories for the up-regulated transcripts, and six for the down-regulated categories (Fig. 5). Gene ontologies were from the microarray elements with additional information from http://www.arabidopsis.org. Among the transcripts up-regulated by Nep1 treatment, the most abundant categories consisted of genes involved in catalytic activities (32%) and genes involved in DNA-/RNA-binding activities (31%). Nep1 treatment induced large groups of transcripts encoding proteins associated with the endomembrane system (91 transcripts), plastids (72 transcripts), mitochondria (45 transcripts), and nucleus (24 transcripts). The largest categories of down-regulated genes were transcription regulators (27%) and genes with unknown functions (27%).

Figure 5.

Pie charts showing the number of genes identified on the Arabidopsis microarray (ATH1, Affymatrix) showing a 2-fold or greater induction (A) or repression (B) in the expression of the transcripts from Arabidopsis seedlings after 30-min treatment of 10 μg mL⁻¹ Nep1. Genes were categorized into functional groups based on the annotation from The Arabidopsis Information Resource and MIPS.

Nep1 Induces Genes Involved in the Production of ROS and Signal Transduction

Nep1 treatment induces the production of ROS (Jennings et al., 2001) and, in this study, many genes involved in ROS and signal transduction were induced by Nep1 treatment. ROS accumulate in many plants during microbial pathogen attack and can function in establishing inaccessibility (resistance) or accessibility (susceptibility) to disease (Hückelhoven and Kogel, 2003). ROS, originating primarily from chloroplasts, mitochondria, and microbodies, function in plants as both toxic by-products and key regulators of plant biological processes such as growth, cell death, hormone signaling, and biotic and abiotic stress (for review, see Mittler et al., 2004). A plasma membrane-located NADPH oxidase is responsible for the respiratory oxidative burst (Torres et al., 2005). In the Arabidopsis genome, there are 10 homologs of the mammalian gp91^phox respiratory burst NADPH-oxidase subunit (AtrbohA–AtrbohJ; Torres et al., 2005). In our study, the transcript encoding NADPH oxidase (At5g47910, AtrbohD) was up-regulated by Nep1 treatment (7.2-fold). The ROS generated by AtrbohD can suppress the SA-activated cell death process that surrounds the NADPH oxidase-activated sites (Torres et al., 2005). This could be one of the mechanisms that generate isolated necrotic spots after Nep1 treatment. However, we did not detect a large induction of genes responsive to SA or that were involved in SA biosynthesis after a 30-min treatment with Nep1. This finding suggests that SA is not associated with the early responses of Arabidopsis to Nep1.

Short-term treatment with Nep1 induced transcripts for ROS sensing and signaling processes associated with stress responses. These included the calcium-binding elongation factor (EF) hand family proteins, (putative) calmodulin (CAM), CAM-binding and (putative) CAM-related proteins, cytochrome P450 family protein, WRKY family transcription factors, and Ser/Thr protein kinase (PK19). There are at least three putative ROS sensing mechanisms: (1) receptor proteins, (2) redox-sensitive transcription factors, and (3) direct inhibition of phosphatases by ROS (for review, see Mittler et al., 2004). The downstream signaling processes involved in ROS sensing utilize Ca²⁺ and Ca²⁺-binding proteins (i.e. CAM) and involve the activation of both G proteins and phospholipid signaling.

In plants, the change of [Ca²⁺]_cyt levels are monitored by at least four major families: (1) CAM, (2) CaM-like proteins, (3) Ca²⁺-dependent protein kinases, and (4) other Ca²⁺-binding proteins without EF hands (helix-loop-helix structure). In our study, 18 transcripts involved in calcium sensing were up-regulated by Nep1 treatment: calcium-binding EF hand family proteins, (putative) CAM, CAM-binding, (putative) CAM-related proteins, and Ca²⁺-ATPases. The induction of a transcript encoding CAM-related protein (CmCAL-1) was also reported in spotted knapweed and dandelion within 15 min in response to Nep1 (Keates et al., 2003). Calcium has a role in mediating responses to diverse signals, such as biotic and abiotic stresses (for review, see White and Broadley, 2003).

Calcium/CAM complexes are involved in senescence that in turn is mediated by ethylene, free radical formation, lipoxygenase, and phospholipase A2 (for review, see Mittler et al., 2004). Calcium/CAM complexes also activate protein kinases that are involved in signal transduction and phytoalexin synthesis by cell wall elicitors (Vogeli et al., 1992). In addition, [Ca²⁺]_cyt can regulate the activity of phospholipase D (PLD), an enzyme that cleaves membrane phospholipids into a soluble head group and phosphatidic acid. PLD is also involved in the cellular responses to ethylene, pathogen attack, leaf senescence, and drought (Wang et al., 2000; Ritchie et al., 2002). In tomato (Lycopersicon esculentum) and the resurrection plant (Craterostigma plantagineum), the transcript for PLD was rapidly induced by a fungal elicitor xylanase, dehydration, and abscisic acid (ABA; Frank et al., 2000; Laxalt et al., 2001). In our study, a transcript (At4g29780) encoding a protein that has putative calcium ion binding and phospholipase A2 activity was induced (23-fold) by Nep1 treatment.

Receptor proteins have been shown to detect ROS, which results in the influx of Ca²⁺ and the activation of phospholipase C/D (PLC/PLD) activity leading to the release of phosphatidic acid, possibly leading to the activation of the protein kinase OXI1 (Rentel et al., 2004). Although OXI1 induction was not observed in this study, we found the induction of many different transcripts encoding signal transduction proteins included receptor-like kinases (11 transcripts), small GTP-binding proteins (two transcripts), MYB transcription factors (six transcripts), and protein kinases (18 transcripts).

Nine WRKY genes were induced by Nep1 treatment, while WRKY65 was repressed moderately (55% reduction). Keates et al. (2003) also detected the induction of WRKY18 within 15 min after treatment with Nep1. WRKY proteins are transcription factors that recognize the W-box elements that exist in the promoters of many plant PR genes (for review, see Ulker and Somssich, 2004). In the Arabidopsis genome, there are 74 WRKY genes that are important in various physiological processes including disease resistance, responses to abiotic stress, senescence, and development. ROS also activate the expression of transcription factors, such as WRKY, Zat, RAV, GRAS, and Myb families (for review, see Mittler et al., 2004).

Nep1 treatment induced four transcripts encoding cytochrome P450 proteins. Keates et al. (2003) also found that a putative cytochrome P450 (ToCYP-1) transcript was highly induced within 15 min of Nep1 treatment in dandelion. The cytochrome P450 family consists of 246 members in Arabidopsis (Yamanaka et al., 2003). These catalyze various oxidative reactions and play an important role in the biosynthesis of lipophilic compounds and in detoxification of herbicides in plant tissues (Narusaka et al., 2004). CYP transcripts, including the two induced by Nep1 (At5g45340, CYP707A3, 8.6-fold induction and At3g48520, and CYP94B3, 3.2-fold induction) are induced by various biotic and abiotic stresses (Narusaka et al., 2004; Saito et al., 2004). CYP707A3 encodes the key enzyme (+)-ABA 8′-hydroxylase that hydroxylates the 8′-methyl group of ABA resulting in the inactivation of ABA (Saito et al., 2004).

Nep1 Induces Genes Involved in Ethylene Synthesis, Senescence, and Cell Death

As shown previously (Bailey, 1995; Veit et al., 2001), Nep1 treatment induces ethylene production in dicotyledonous plants. In this investigation, Nep1 induced the expression of 31 transcripts related to ethylene. The affected transcripts were involved in ethylene biosynthesis (ACC synthase [ACS] 4, 6, and 7) or were APETALA2 domain-containing transcription factors and ethylene-responsive element binding factors (ERF 1, 2, 4, 5, 8, 11, and putative). During ethylene biosynthesis, ACS converts S-adenosyl-l-Met to ACC in a rate-limiting process (for review, see Bleecker and Kende, 2000). Ethylene has important roles in plant growth and development, such as fruit ripening, senescence, cell death, abscission, seed germination, and flowering (for review, see Bleecker and Kende, 2000).

We also observed that Nep1 treatment induced genes expressed during senescence or that have been shown to respond to jasmonic acid (JA). These included lipoxygenases (plastidic LOX3, At1g17420, 58-fold induction; putative plastidic LOX, At1g72520, 22-fold induction), 12-oxophytodienoate reductase (At1g76690, 2.4-fold induction), and allene oxide cyclase (AOC2, At3g25780, 6.3-fold induction). In Arabidopsis, JA is produced during leaf senescence, and the genes mentioned above were strongly induced either during leaf senescence (He et al., 2002) or by the exogenous application of JA (Maucher et al., 2000). The induction of JA synthesis by various elicitors and wounding has been reported previously (Hildebrand et al., 1998; Wang et al., 2000). This report shows that Nep1 rapidly induces genes involved in JA biosynthesis.

Nep1 treatment also induced genes involved in protein turnover. For example, four transcripts of an ATPase associated with diverse cellular activity (AAA)-type proteins were highly up-regulated by Nep1 (up to 37-fold induction). The AAA-type ATPases, which are subunits of the 26S proteasome (Fu et al., 1999) might be associated with protein degradation under stress conditions. This finding suggests that protein turnover may be accelerated during cell death in response to Nep1 treatment.

Nep1 Induces Transcripts Involved in Stress Response

Nep1 treatment induced three known and 20 putative transcripts encoding disease resistance associated proteins (up to 26-fold induction). Recognition of pathogens by plant resistance (R) genes can trigger resistance responses in plants. The majority of R genes contain a nucleotide-binding site and a Leu-rich repeat domain (for review, see Gregory et al., 2003).

Two genes of a τ class glutathione S-transferase (GST) were induced by Nep1: GSTU11 (At1g69930, 2.7-fold induction) and GSTU12 (At1g69920, 12.5-fold induction). In Arabidopsis, there are 47 GST genes that are divided into four classes (φ, τ, ζ, and θ; Wagner et al., 2002). Plant GSTs have roles in the detoxification and tolerance to herbicides, but GSTs are differentially regulated in response to various abiotic and biotic stresses (Wagner et al., 2002).

One transcript encoding an immediate-early fungal elicitor family protein (At3g02840) was highly up-regulated by Nep1 treatment (64-fold induction). This protein in Arabidopsis shares 40% identity with a parsley (Petroselinum crispum) immediate-early fungal elicitor protein CMPG1 (AAK69402), which was identified as a fast-responding gene to pathogen-derived elicitor, Pep25 oligopeptide elicitor from P. sojae (Kirsch et al., 1997). The elicitor Pep25 binds specifically to the plasma membrane leading to ion fluxes, hydrogen peroxide formation, activation of numerous defense-related genes, and induction of the plant defense response.

Three transcripts encoding putative OMTs (At1g21100, At1g21110, and At1g21120) that catalyze the transfer of a methyl group to the oxygen atom of an acceptor molecule were highly induced by Nep1 (6.5-, 16-, and 13.9-fold induction, respectively). These sequences share 42% and 46% amino acid identity with tobacco COMT I and II (caffeic acid OMT). Isoforms of tobacco COMT are divided into two distinct classes (Jaeck et al., 1996). Class I COMT was constitutively expressed in lignified tissues (Jaeck et al., 1996), while Class II COMT was strongly induced by various biotic and abiotic elicitors (Toquin et al., 2003). A COMT II gene was expressed strongly in cells surrounding necrotic lesions, indicating the implication of COMT II in the synthesis of phenylpropanoid compounds that accumulate in a ring of living cells surrounding necrotic tissues.

Nep1 Alters Gene Expression Targeting Organelles and Signal Transduction

The ATH1 array results subsequently were verified with QPCR using representative genes from Figure 5 and in Supplemental Table II: CAM (1–3), mitochondria (4–6), chloroplast (7–10), membrane (11–14), apoptosis (15–17), jasmonate (18 and 19), SA (20–22), ethylene (23–25), and WRKY (26 and 27). Transcripts of CLH1 encoding chlorophyllase 1 (At1g19670), which did not respond to Nep1 treatments, also were quantified using QPCR. This result agreed with the observation that chlorophyll content was unchanged in Arabidopsis after Nep1 treatment.

In each instance, transcription induction ratios from the QPCR data verified the microarray results (Fig. 6). However, the mean QPCR response was 4.0-fold greater than that of the ATH1 microarray. This finding is typical of hybridization-based methods versus QPCR (Holland, 2002). The involvement of calcium/CAM was reconfirmed in QPCR (Fig. 6, sections 1–3). We also tested the genes encoding proteins localized into chloroplast, mitochondria, and endomembrane system (Fig. 6, sections 4–14). The high induction of the transcripts indicated that Nep1 treatment significantly affected the transcripts for proteins targeted to these organelles. Three transcripts encoding putative disease resistance proteins (Fig. 6, sections 15–17) that are involved in apoptosis were also highly up-regulated. Three genes were examined for JA response (Fig. 6, sections 7, 18, and 19): lipoxygenase 3 (LOX3) and putative AOC for JA biosynthesis, and putative GST for JA response. LOX3 and putative AOC were highly induced (greater than 21-fold at 30 min), verifying results of the microarray, while putative GST was not significantly up-regulated by Nep1. In comparison, the three transcripts selected to examine the involvement of SA in the response to Nep1 (Fig. 6, sections 20–22) showed that induction was less than 6-fold at 30 min. As expected, transcripts related to ethylene response to Nep1 treatment were highly induced (Fig. 6, sections 23–25). Induction of four transcripts was much higher at 30 min than 3 h, suggesting an early, rapid response to Nep1. Induction of three WRKY transcripts was also confirmed by QPCR.

Figure 6.

Confirmation of microarray data by QPCR. Seedlings were grown at 22°C for 7 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ PAR with 24 h light. Seven-day-old Arabidopsis seedlings grown in the liquid culture were treated with Nep1 (10 μg mL⁻¹) without Silwet and harvested 30 and 180 min after treatment. Control seedlings were treated with an equivalent volume of sterile distilled water and harvested 0, 30, and 180 min after treatment. In total, 27 genes were selected from each of the expression profiles and functional categories. Genes were tested with cDNA templates from three different biological samples. Numbers inside the graph near the values for 30 min are fold induction from the microarray result. No number means less than 2-fold change in the microarray experiment. y axis, Relative expression; x axis, time after treatment (minutes); white circle, Nep1-treated sample; black circle, control.

Nep1 Induces Cell Death

Using microarray analysis, we detected the induction of genes involved in ROS production and signal transduction, ethylene biosynthesis, membrane modification, stress, and cell death. In addition, genes whose products are targeted to specific organelles and membrane structures were rapidly induced by Nep1 treatment. The majority of genes that responded to Nep1 treatment were associated with general stress responses but were not critically linked to resistance to plant disease. These results are consistent with Nep1 facilitating cell death as a component of diseases caused by necrotrophic plant pathogens.

Results presented here characterize the broad array of plant genes that respond to Nep1 treatment. Plant recognition of microbial elicitors and the resulting signal transduction involves the ion channels (Ca²⁺, anionic, and K⁺), NADPH oxidase, phospholipases, and many yet unidentified proteins. The activation of many of these proteins triggers secondary messengers including [Ca²⁺]_cyt accumulation, cytosolic pH decrease, ROS accumulation, plasma membrane depolarization, and changes in metabolism, which ultimately lead to plant defense and/or cell death. The involvement of all of these systems in the Nep1 response is indicated based on the altered gene expression observed here. Multiple genes encoding ACS, a control point for ethylene biosynthesis, along with genes involving JA biosynthesis are rapidly induced in response to Nep1 treatment. As plant hormones, ethylene and JA may serve as secondary signals inducing signal transduction cascades (Bleecker and Kende, 2000; He et al., 2002) involved in the response to Nep1. The NLPs are widespread among microbes in nature including many plant pathogens. Evidence continues to accumulate supporting the hypothesis that NLPs are virulence factors in many different plant/pathogen interactions (Qutob et al., 2002; Mattinen et al., 2004; Wang et al., 2004; Bae et al., 2005a). The continued characterization of Nep1-mediated cell death signaling in plants will enhance our understanding of biological processes leading to susceptibility in plant-pathogen interactions and improve our ability to control plant diseases.

MATERIALS AND METHODS

Plant Materials: Agarose-Grown Seedlings

To study the effects of Nep1 on plant growth, sterilized Arabidopsis (Arabidopsis thaliana; L. Henyh [Columbia ecotype]) seeds were mixed with Nep1 (20 μg mL⁻¹ plus 0.001% [v/v] Silwet-L77) from Fusarium oxysporium in 0.3% agarose, 1% Suc, 1× MS media, 1× vitamin B5, and 0.1 mm MES. Nep1 was isolated and purified from the culture filtrates of F. oxysporum f. sp. erythroxyli grown for 6 d in Czapek-Dox broth with 1% (w/v) casamino acids (Bailey, 1995). The purified Nep1 was stored in 20 mm MES and 300 mm KCl, pH 5.0, at −20°C until used.

Silwet (0.001% [v/v]) was used only when seedlings were grown in the media containing agarose. We developed the agarose plate system that reduced required amounts of Silwet to minimize the potential for Silwet-induced responses in the developing seedlings. In general, less surfactant required more Nep1 to achieve the same response due to reduced penetration of the tissues. We used a relatively high amount of Nep1 (20 μg mL⁻¹) to ensure that strong responses were generated during the treatment. Control samples were similar except that the Nep1 solution was omitted. The final solution was adjusted to pH 5.7 with NaOH and plated onto empty petri dishes (4 mL for 100-mm plates). Seeds were surface sterilized with one-third strength commercial bleach for 10 min. The sterilized seeds were rinsed five times with sterile, deionized water prior to plating. Seedlings were grown in growth chambers at 22°C for 5 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ photosynthetically active radiation (PAR) with 16 h light or under dark condition. Growth parameters were measured by flattening 5-d-old seedlings with a coverslip and tracing the outline with a Nikon SMZ 1500 microscope and a Nikon Digital camera DXM1200. Area and length were calculated with a 1-mm² standard with ASSESS (Image Analysis Software for Plant Disease Quantification; The American Phytopathological Society).

Transmission Electron Microscopy and Immunogold Location of Nep1

Arabidopsis seedlings were harvested and rinsed after treatment with Nep1 as described above. Fixing, embedding, sectioning, and staining were performed as described in Bae et al. (2005b). For immunogold labeling, freshly dissected leaf samples were fixed for 2 h at 4°C in a freshly prepared solution containing 4% (v/v) paraformaldehyde and 0.2% (v/v) glutaraldehyde in 0.1 m cacodylate buffer, pH 7.4. The specimens were washed in the same buffer and dehydrated with an ascending ethanol series and were embedded in LR White (London Resin). Leaf sections of control and Nep1-treated seedlings were prepared for immunolocalization as described in Kim and Koh (1997). Ultrathin sections (80 nm) were mounted on uncoated nickel grids (300 mesh) that were treated initially with a polyclonal antibody raised against Nep1. The sites of antibody binding were labeled with protein A (10 nm gold particle diluted 1:50 in bovine serum albumin-phosphate-buffered saline buffer). After immunogold labeling, the replicas were washed three times with phosphate-buffered saline. After counterstaining with 4% (v/v) uranyl acetate, the sections were examined with a JEOL 1200 EX or a PHILIPS CM12 transmission electron microscope. Control samples were treated as above without the use of the primary antibody.

Plant Material-Liquid Culture-Grown Seedlings

For metabolite analysis, fluorescence imaging, and the DNA microarray and QPCR analyses, approximately 100 sterilized seeds were transferred to sterile 125-mL Erlenmeyer flasks containing 20 mL of 1× MS basal salts supplemented with 1× vitamin B5 in 0.1 mm MES-NaOH, pH 5.7, buffer. Seedlings were grown in controlled environment chambers as described above except that the flasks were continuously agitated with a rotary shaker for 7 d. Seedlings were grown at 22°C for 7 d under fluorescent lights providing 150 ± 10 μmol m⁻² s⁻¹ PAR with 24-h days. One-week-old seedlings were exposed to Nep1 (10 μg mL⁻¹) without Silwet. We applied 10 μg mL⁻¹ of Nep1 instead of 20 μg mL⁻¹, because Nep1 has better accessibility to seedlings in liquid culture compared to agarose media to the point that Silwet was not required, which allowed us to reduce the required amount of Nep1. Seedlings were harvested after treatment, filtered through Whatman number 1 filter paper (Whatman), rinsed thoroughly with deionized water, and were immediately frozen in liquid nitrogen.

Metabolite Analyses

Soluble metabolites from Arabidopsis seedlings were extracted and measured by gas chromatography coupled to a mass selective detector essentially as described by Roessner et al. (2000). We examined various metabolites indicative for carbon and nitrogen metabolism, including nine amino acids, two organic acids (lactic acid and maleic acid), and one sugar alcohol (myoinositol). They were chosen because they are abundant and give easily identifiable, discrete peaks on gas chromatography/mass spectrometry. One-week-old liquid-cultured seedlings were exposed to Nep1 (10 μg mL⁻¹) for 0, 1, and 6 h. Seedlings were harvested as described above. Frozen plant material was ground to a fine powder under liquid N₂ and 0.05 g (fresh weight) tissue was extracted at 0°C with 1.4 mL ice-cold methanol in a ground glass tissue homogenizer. The extracts were heated to 70°C for 15 min, diluted with an equal volume of deionized water, and centrifuged at room temperature for 15 min at 12,000g. An aliquot of the supernatant was evaporated under N₂ gas at 37°C. Dried samples were derivatized with 80 μL of freshly prepared methoxyamine (20 mg mL⁻¹ in pyridine) for 90 min at 30°C and then with 80 μL of N-methyl-N-(trimethylsilyl) fluoracetamide for 30 min at 37°C. Metabolites were chromatographed on a Hewlett-Packard model 6890 gas chromatography system using a 0.25-mm by 30-m SPB-50 column (Supelco). Retention times, mass spectra, and quantification were obtained via four-point curves with known chemical standards. Ribitol was added to each extract prior to homogenization and was used as an internal standard.

Fluorescence Image Detection

Chlorophyll a fluorescence images of sample materials at 685 and 730 nm along with 685/730-nm ratio images were obtained using a hyperspectral fluorescence imaging system (Kim et al., 2001a). Samples consisted of 7-d-old Arabidopsis seedlings grown in liquid culture as described above. Using a 96-well, nonfluorescent microtiter plate (Fisher Scientific) as a sample holder, 30 seedlings were placed individually in a 0.5-mL well with the adaxial surface up in a 10 × 3 (row × column) arrangement. The wells were filled with liquid culture media (control). Immediately following the initial fluorescence measurement, the Nep1 (Column 2) and DCMU (Sigma; column 3) treatments were applied, and subsequent images were acquired at approximately 15-min intervals for a 3-h period. The final concentrations of Nep1 and DCMU were 10 μg mL⁻¹ and 200 mm, respectively. DCMU was used as a blocking agent of electron transport from PSII to PSI, resulting in a rapid increase of chlorophyll a fluorescence (Hodges and Barber, 1986).

DNA Microarray Analyses

To study the Nep1 responsive genes in Arabidopsis, experiments were performed with the Affymetrix GeneChip Arabidopsis ATH1 genome array (www//affymetrix.com;catalogno.900385), which contains more than 22,500 probe sets representing approximately 24,000 genes. Probe arrays are spotted onto a glass array, and the glass substrate was coated with linkers containing photolabile protecting groups, which were removed by illumination. By repeating the masking and illumination, a specific set of oligonucleotide probes is synthesized in a selected location.

Two independent biological replications were conducted using two independent samples and two microarrays. For these experiments whole Arabidopsis seedlings were grown in liquid culture for 7 d as described above. Samples were harvested 30 min after treatment with Nep1 (10 μg mL⁻¹) or sterilized, deionized water that was used as a control. The harvested samples were ground to a fine powder under liquid nitrogen using a mortar and pestle. Total RNA was extracted using an RNeasy Mini kit (Qiagen) as described by the manufacturer and then purified with a RNeasy MinElute Cleanup kit (Qiagen). Samples for hybridization were labeled according to protocols developed by the vendor (http://www.affymetrix.com/support/technical/index.affx). Labeling, hybridization of biotinylated cRNA, and imaging procedures were performed using standard Affymetrix protocols (Affymetrix GeneChip Expression Analysis technical manual) in the University of Maryland DNA Microarray Core Facility located in Rockville, Maryland (http://www.umbi.umd.edu/∼cbr/macore/macorestart.htm). Briefly, double-stranded cDNA was synthesized from 5 μg of total RNA, and an in vitro transcription reaction was used to generate biotin-labeled cRNA. Exogenous positive poly(A) spike-in controls (four Bacillus subtilis genes) were used to monitor the entire eukaryotic target-labeling process. Fragmented cRNA was hybridized to the array for 16 h, and washing and staining were performed on the fluidics station. After hybridization, the probe array was scanned using GeneArray scanner with 570-nm wavelength and 3-μm pixel value, which was controlled by GeneChip Operating software (GCOS 1.2). The data image file was analyzed for probe intensities. The output of the GCOS was further analyzed using Microsoft Excel. Differences in transcript abundance (fold change) were calculated and fold change was accepted only if the corresponding change call showed a significant change (I = increase, D = decrease), which was generated by GCOS. Transcripts were selected only when they showed a significant change more than 2-fold in both replicates (Supplemental Table I). Genes were categorized into functional groups based on the annotation from The Arabidopsis Information Resource and Munich Information Center for Protein Sequences (MIPS; http://mips.gsf.de). All GeneChip datasets were deposited for open access (accession no. GSE4638; http://www.ncbi.nlm.nih.gov/projects/geo/).

QPCR

Selected Nep1 responsive Arabidopsis transcripts identified by DNA microarray analysis were further examined using QPCR. For these experiments, 7-d-old seedlings were grown in liquid culture and treated with Nep1 or the appropriate control for 0, 30, and 180 min. Total RNA was extracted with the RNeasy Mini kit according to the manufacturer's recommendation except that an extra DNase I treatment was used. One microgram of each RNA sample was used to generate first-strand cDNA using SuperScript III RNase H⁻ reverse transcriptase (Invitrogen) with an oligo(dT)₂₀ primer. The synthesized first-strand cDNA was diluted 10-fold and used as a template for QPCR. QPCR analysis with selected primer sets was performed using an Mx3000P QPCR system and Brilliant SYBR Green QPCR Master mix (Stratagene). Primers for selected genes were designed to be 23 to 27 mers with a melting temperature of 60°C ± 3°C using the Primer 3 program of Biology WorkBench (http://workbench.sdsc.edu/). Most products generated were 200 to 250 bp, although a few were shorter. Primer sequences are available in Supplemental Table II. A dissociation (melting) curve was run for each gene at the end of the amplification reaction to determine whether genes other than the gene of interest were amplified in the PCR reaction.

Reactions contained 12.5 μL of 2× Brilliant SYBR Green QPCR Master mix, 5 μL of 10-fold diluted cDNA, and 2,500 nm of each gene-specific primer and diluted reference dye (final concentration = 300 nm) in a final volume of 25 μL. A master mix of cDNA, 2× SYBR Green QPCR Master mix, and reference dye was prepared to reduce pipetting errors and to ensure the same amount of reagent in each well. A threshold of 0.1 was manually defined to obtain a threshold cycle (C_T) value, which is the cycle number that is required for the SYBR Green fluorescent signal (ΔRn) to cross the threshold value. Averages and ses for C_T values were calculated for each gene of interest based on three replications with three different biological samples. The following default thermal profile was used: 95°C for 10 min, 40 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 30 s.

Translation elongation factor 1-α (EF-1-α, At5g60390), a constitutively expressed gene, was used as an expression control. PCR efficiencies (E) of all primers were calculated using dilution curves with five dilution points, a 3-fold dilution, and the equation E = [10^(−1/slope)] − 1 as described previously by Pfaffl (2001). To compare data from different PCR reactions and cDNA template, C_T values for all genes of interest (C_TGOI) were normalized to the C_T values of EF-1-α (C_TTEF) for each replication. The ratios of treatment/control gene expression were then obtained from the equation [(E_GOI)_TGOI^ΔC/(E_TEF)_TTEF^ΔC], where ΔC_T = C_TC − C_TT, C_TT is the C_T for treatment, and C_TC is the C_T for the corresponding control.