Phospholipase D (PLD) has been implicated in various cellular processes including membrane degradation, vesicular trafficking and signal transduction. Previously, we described a PLD gene family in tomato (Lycopersicon esculentum) and showed that expression of one of these genes, LePLDβ1, was induced by treatment with the fungal elicitor xylanase. To further investigate the function of this PLD, a gene-specific RNAi construct was used to knock down levels of LePLDβ1 transcript in suspension-cultured tomato cells. Silenced cells exhibited a strong decrease in xylanase-induced PLD activity and responded to xylanase treatment with a disproportionate oxidative burst. Furthermore, LePLDβ1-silenced cell-suspension cultures were found to have increased polyphenol oxidase activity, to secrete less of the β-d-xylosidase LeXYL2 and to secrete and express more of the xyloglucan-specific endoglucanase inhibitor protein XEGIP. Using an LePLDβ1–green fluorescent protein (GFP) fusion protein for confocal laser scanning microscopy-mediated localization studies, untreated cells displayed a cytosolic localization, whereas treatment with xylanase induced relocalization to punctuate structures within the cytosol. Possible functions for PLDβ in plant–pathogen interactions are discussed.
A plant's susceptibility or resistance to a certain pathogen often depends on its ability to recognize pathogen attack. Once the pathogen has been perceived, the plant can respond with a plethora of defensive strategies (Nimchuk et al., 2003). For example, plants can accumulate compounds and proteins that are toxic to the attacker, limit their nutritional value with respect to the pathogen or fortify their cell walls (Dixon, 2001; Schilmiller and Howe, 2005). Together, such adaptations constitute the plant defense response.
Phospholipid signaling has recently been implicated in the responses of plants to various environmental cues, both biotic and abiotic (Meijer and Munnik, 2003; Wang, 2004). Phosphatidic acid (PA)-mediated signal transduction, in particular, has been shown to play an important role in numerous processes (Munnik, 2001; Testerink and Munnik, 2005). The lipid second messenger PA can be generated by two distinct enzymatic routes, either directly by phospholipase D (PLD) or via the combined action of phospholipase C (PLC) and diacylglycerol kinase (DGK).
The Arabidopsis genome contains 12 PLDs that can be subdivided into several classes depending on their sequence similarity, substrate specificity and enzymatic requirements. Arabidopsis has two calcium-independent PLDs, AtPLDζ1 and AtPLDζ2, that are similar to mammalian and yeast PLDs with regard to their N-terminal lipid-binding domains, namely a pleckstrin homology domain and a PHOX homology domain (Qin and Wang, 2002). The remaining 10 PLDs in the Arabidopsis genome harbor an N-terminal C2 domain and are subdivided into the α, β, γ, δ and ɛ PLD classes (Elias et al., 2002; Wang, 2000). Whereas α-class PLDs require millimolar concentrations of calcium and low pH for in vitro activity, the β- and γ-class PLDs require phosphatidylinositol 4,5-bisphosphate (PIP2) and physiological calcium concentrations (low micromolar range) and pH (neutral) for in vitro activity (Wang, 2000). The δ-class PLD distinguishes itself in that its in vitro activity is stimulated by oleic acid (Wang and Wang, 2001). The in vitro activity of PLDɛ has not been studied to date.
Evidence of PA signaling as a part of the plant defense response has been observed in several independent studies (Laxalt and Munnik, 2002). An increase in expression of various Arabidopsis PLDs was demonstrated in response to both virulent and avirulent strains of Pseudomonas syringae (de Torres Zabela et al., 2002). The same is true for several PLDs in rice treated with Xanthomonas oryzae (Young et al., 1996). In addition, a change in membrane distribution of PLD upon microbial infiltration was observed (Young et al., 1996). Furthermore, PLC/DGK-generated PA was shown to be crucial for the oxidative burst response following the Avr4/Cf-4 interaction in transgenic tobacco cell-suspensions (de Jong et al., 2004). Recently, PLD activation in response to the elicitor chitosan has been ascertained in rice cell-suspension cultures and this activity was also suggested to be involved in the oxidative burst response (Yamaguchi et al., 2003, 2004, 2005).
Previously our laboratory demonstrated PLC and PLD activation in response to pathogenic elicitation in tomato cell-suspension cultures (Msk8; Lycopersicon esculentum Mill.) treated with xylanase (van der Luit et al., 2000). Subsequently, we showed that expression of one PLD gene, LePLDβ1, is specifically up-regulated after xylanase treatment (Laxalt et al., 2001). Xylanase (endo-1,4-β-xylanase), from the green mold Trichoderma viride, is a family 11 glucoside hydrolase that is characterized by its ability to break down various xylans to produce short-chain xylo-oligosaccharides. It is a potent elicitor of plant defense responses in tomato (Enkerli et al., 1999; Furman-Matarasso et al., 1999). Recently, two likely xylanase receptors were cloned from tomato, establishing the relationship between this elicitor and the plant defense response as a ligand–receptor interaction (Ron and Avni, 2004).
In this paper we investigate the role of LePLDβ1 in the Msk8 response to xylanase elicitation. A silencing strategy was employed to resolve whether diminished LePLDβ1 expression influences the xylanase-induced PLD activity, and the phenotypic consequences of silencing are explored. A green fluorescent protein (GFP) tagging approach was used to examine the influence of xylanase on the intracellular localization of LePLDβ1.
Rapid LePLDβ1expression and LePLDβ1 activation in response to xylanase treatment
In untreated cell-suspension cultures, the LePLDβ1 transcript was hardly detectable by Northern blot analysis. Basal expression levels could only be visualized after a long exposure (Figure 1a) but levels rapidly increased when cells were treated with 100 μg ml−1 xylanase (Figure 1b). Exposure to xylanase led to a readily detectable level of LePLDβ1 transcript within 3 h and this expression was still present after 24 h, albeit to a lesser extent. The induction of LePLDβ1 expression preceded the increase in the transcript level of pathogenesis-related protein 1 (PR1), which was detected only after 24 h of xylanase treatment (Figure 1b).
In order to study the function of LePLDβ1 in more detail, LePLDβ1 expression was knocked down in Msk8 cell-suspension cultures using a gene-specific RNAi construct (see Experimental procedures). Several silenced lines were obtained that had diminished basal levels of LePLDβ1 transcript compared with empty-vector control lines (Figure 1a). Upon xylanase treatment, the effect of the silencing construct on expression of LePLDβ1 was more obvious; a representative silenced line displaying a ±90% reduction in transcript level is shown in Figure 1(b).
The acquisition of LePLDβ1-silenced lines allowed us to test whether this PLD is responsible for the xylanase-induced PLD activity in Msk8 cell suspensions reported previously (Laxalt et al., 2001; van der Luit et al., 2000). In vivo changes in phospholipid composition can be monitored by pre-labeling cells with radioactive phosphate. After treatment, lipids can be extracted, separated by thin layer chromatography and quantified by phosphoimaging. The activity of PLD can be visualized using this enzyme's unique transphosphatidylation activity, which entails that in the presence of primary alcohols PLDs can transfer the lipid moiety of their substrate to the alcohol producing a phosphatidyl alcohol (Munnik, 2001; Munnik et al., 1995). In this case, cells were treated with xylanase in the presence of 1-butanol. The activity of PLD in Msk8 cells incubated with 100 or 200 μg ml−1 xylanase increased 2- and 2.5-fold, respectively, within 1 h as measured by the production of phosphatidylbutanol in xylanase-treated cultures compared with mock-treated cultures (Figure 2a,b). However, in LePLDβ1-silenced cell-suspension cultures this increase in activity was markedly reduced, reaching at most a 1.5-fold increase (Figure 2a,b). These results indicate that both expression of LePLDβ1 and activity of LePLDβ1 are increased upon treatment with xylanase.
Enhanced reactive oxygen species response in LePLDβ1-silenced cell-suspension cultures
The plant defense response consists of a wide variety of reactions, e.g. accumulation of phytoalexins, pathogenesis-related (PR) proteins, proteinase inhibitors and glucanase inhibitors and lignification and callose deposition. One of the earliest measurable responses to perception of elicitors is the production of reactive oxygen species (ROS). We therefore decided to study this phenomenon in Msk8 cell-suspension cultures exposed to xylanase and examine in particular whether silencing LePLDβ1 affected the response.
Hydrogen peroxide (H2O2) is a typical ROS produced by plant cells in response to pathogenic elicitation. An assay adapted from Felix et al. (1999) was used to study production of H2O2 by making use of the oxidation of 5-aminosalicylic acid (ASA) by H2O2 peroxidases, which leads to the formation of a dark stain. Control cells and LePLDβ1-silenced cells were spread as flat lawns in Petri dishes and test solutions containing different amounts of xylanase were applied locally. In the presence of ASA, a dark stain was visible after 24 h around the areas where cells had responded with production of H2O2; 50 ng of xylanase barely induced macroscopic staining while 5 μg induced massive staining around the area of application (Figure 3). The LePLDβ1-silenced cell-suspension cultures responded to the same range of elicitor concentrations; however, the staining was more intense and typically gave higher backgrounds (Figure 3).
These results prompted us to look at the production of H2O2 in more detail, focusing on the immediate ROS response following elicitation, termed the oxidative burst. Production of H2O2 was monitored by measuring the decrease in fluorescence of the H2O2-sensitive probe pyranin in a microtiter-plate assay. This reporter loses its fluorescence at 512 nm when oxidized by H2O2 peroxidases (Apostol et al., 1989; de Jong et al., 2004). The oxidative burst response during the first 30 min following elicitation with 50 ng ml−1, 500 ng ml−1 and 5 μg ml−1 xylanase was examined. Control Msk8 cells clearly responded to 500 ng ml−1 and 5 μg ml−1 xylanase within the assay period, displaying a biphasic response to 5 μg ml−1 xylanase, with an initial strong burst between 5 and 10 min followed by a secondary weaker burst after 15 min (Figure 4a). As with the ASA assay (Figure 3), the LePLDβ1-silenced cell-suspension cultures responded to the same range of elicitor concentrations; however, the biphasic response seen in control cell-suspension cultures was not discernible in silenced lines. Instead, the initial strong burst continued throughout the assay period (Figure 4b). Together, these results show that silencing LePLDβ1 affects the ROS response to xylanase in tomato cells.
Elevated polyphenol oxidase activity in LePLDβ1-silenced cell suspensions
Polyphenol oxidases (PPOs) oxidize phenols to form quinones: highly reactive molecules that can covalently link encountered nucleophiles resulting in the formation of melanin-like black or brown condensation polymers. Polyphenol oxidase enzymes are present in the thylakoid lumen and their phenolic substrates in the vacuole. Upon loss of cell integrity, enzyme and substrate converge and quinones are produced (Koussevitzky et al., 2004). Polyphenol oxidase has been linked to several plant–pathogen interactions. Its activity and gene expression is inducible by wounding and pathogens and over-expression of PPO in tomato has been shown to increase resistance to the bacterial pathogen P. syringae (Howe and Ryan, 1999; Li and Steffens, 2002; Thipyapong and Steffens, 1997).
When being shaken in the incubator there was no visible difference between the control and LePLDβ1-silenced cell suspensions. We noticed that the culture media of the silenced lines developed an orange/brown coloration when the old culture flasks were left standing on the laboratory bench after subculturing. To further investigate this color change, 3 ml of 1-week-old control and LePLDβ1-silenced cell-suspension cultures was transferred to test tubes and photographed after 0, 24 and 48 h. Whereas control cell-suspension cultures showed no or hardly any coloration, even after 48 h, the silenced lines displayed a strong coloration of the medium within 24 h and the formation of a dark precipitate within 48 h (Figure 5a). This coloration was derived from the plant cells, as cell-free medium from 1-week-old cultures displayed no change in color (data not shown). Staining with fluorescein diacetate of cell suspensions that had been left standing showed that both the control and silenced cell suspensions had started to lose cell integrity within 24 h and that, in both cases, all cells had perished within 48 h (data not shown). Coloration of the medium could be abated by co-incubation with 10 mm of the antioxidant ascorbic acid (data not shown). The discharged compound was most likely an oxidated phenolic. We tried to identify it by mass spectrometry but definitive identification of the compound(s) was not possible due to its resistance to fragmentation (data not shown).
These observations prompted us to examine whether there was more PPO activity in the silenced lines. Assaying for PPO activity in protein extracts of untreated cell cultures and cultures that had been left standing for 24 h showed that silenced lines indeed had more PPO activity (Figure 5b). Purple coloration caused by oxidation of p-phenylenediamine, indicative of PPO activity, was more intense in protein extracts of silenced lines than in protein extracts from control culture, irrespective of whether cell suspensions had been left standing (Figure 5b). These results indicate that knocking down LePLDβ1 expression affects the phenolic metabolism of Msk8 cell-suspension cultures.
Anomalous protein secretion in culture medium of LePLDβ1-silenced cell suspensions
The plant cell wall is a sturdy casing that keeps the cell from bursting, protects it against dehydration and also serves as a physical barrier to invasive attackers. It is composed of polysaccharides (cellulose, hemicelluloses and pectins), proteins and aromatic compounds such as lignins (Goujon et al., 2003). During development, the cell wall undergoes several modifications to allow for expansion and differentiation of the cell. In order to execute these changes, plant cells secrete cell-wall-modifying enzymes that can hydrolyze and/or reattach cell-wall components. Plant pathogens also secrete cell-wall-modifying enzymes; in their case, to breach the plant's protective barrier or to make oligo- and monosaccharides accessible for nutrition (Igawa et al., 2004).
The altered phenolic composition of the silenced cell-suspension culture media led us to investigate the protein composition of the media. Accordingly, medium proteins were precipitated, separated by SDS-PAGE and stained with Coomassie. Two protein bands were observed, with an apparent mass of 51 and 67 kDa (Figure 6a), that clearly differed in abundance when the medium of silenced lines was compared with that of control cell suspensions. Both bands were excised, digested with trypsin and analyzed by mass spectrometry. Strikingly, the 67 kDa band that was less abundant in the LePLDβ1-silenced cell-suspension culture medium could be identified by the tandem mass spectroscopy (MS/MS) of six peptides as LeXYL2, a β-d-xylosidase [total Mascot score 249 (for more information see http://www.matrixscience.com/); accession number AB041812; Itai et al., 2003]. Plant β-d-xylosidases have been suggested to be involved in remodeling of cell walls (Goujon et al., 2003). Surprisingly, the 51 kDa band that was more abundant in silenced cell-suspension culture medium was identified by the MS/MS of four peptides as the xyloglucan-specific endoglucanase inhibitor protein XEGIP (total Mascot score 113; accession number AY155579; Qin et al., 2003). This protein has been shown to form a complex with the xyloglucan-specific endoglucanase from Aspergillus aculeatus, inhibiting its glucanase activity. For both identified proteins the theoretical mass corresponds well with the observed mass on SDS-PAGE.
Expression of XEGIP in control and silenced lines was subsequently analyzed (Figure 6b). Whereas untreated control cell-suspension cultures exhibited no detectable XEGIP transcript, elicitation with 100 μg ml−1 xylanase induced clear expression within 3 h. In contrast, untreated LePLDβ1-silenced cell-suspension cultures already exhibited XEGIP expression and this transcript level was increased considerably beyond control cell-suspension culture levels after xylanase treatment (Figure 6b). These results suggest that diminished LePLDβ1 expression leads to an altered expression and protein secretion pattern in tomato cell-suspension cultures, especially of proteins involved in xyloglucan hydrolysis and/or the plant defense response.
Relocalization of LePLDβ1– GFP after xylanase treatment
Localization studies concerning PLDβs have, up to now, focused on the distribution of PLD activity. These studies have implied that PIP2-dependent, low-micromolar [Ca2+] requiring PLD activity is present both in the cytosol and on various membranes (Pappan et al., 1997). However, such activity assays do not distinguish between PLDs of the β and γ class, let alone between members of the same class.
Polyclonal antibodies directed against LePLDβ1 were generated but, unfortunately, immunolocalization studies failed due to a lack of specificity of the antibodies (data not shown). We therefore employed a GFP tagging strategy in combination with confocal laser scanning microscopy to examine the intracellular distribution of LePLDβ1 in Msk8 cells upon xylanase exposure. In untreated cells, the LePLDβ1–GFP signal was cytosolic (Figure 7b). Intriguingly, after xylanase treatment the LePLDβ1–GFP signal relocalized to punctuate structures in the cytosol (Figure 7d). These punctuate structures were 1–2 μm in diameter, mobile and mainly localized close to the plasma membrane (Figure S1a,b). In contrast, the LePLDα1–GFP signal in Msk8 cells did not relocalize in response to xylanase but remained cytosolic (Figure 7a,c). These results suggest that LePLDβ1 localization in Msk8 cells is specifically altered in response to xylanase.
LePLDβ1 activation in response to xylanase
The plant PLD family is a diverse group that can be subdivided into several classes: the α, β, γ, δ, ɛ and ζ classes (Elias et al., 2002; Wang, 2000). The PLDζs resemble mammalian and yeast PLDs, possessing an N-terminal PHOX homology and pleckstrin homology domain, while the remaining classes all have C2 domains in their N-terminal region. The fact that this multitude of PLDs can be found in all higher plants examined so far suggests that they have specific functions and are not redundant. The occurrence of so many PLDs in plants makes it difficult to determine which one is activated in response to a certain stimulus.
Here, we have shown that we are able to silence LePLDβ1 expression in Msk8 cell-suspension cultures using an RNAi construct (Figure 1a,b). Basal LePLDβ1 transcript levels were absent in cultures expressing an LePLDβ1 silencing construct (Figure 1a). Xylanase treatment of cell suspensions led to a prolific induction of LePLDβ1 expression that was markedly reduced in silenced lines, although some induction (±10%) was still detectable (Figure 1b).
Xylanase also induces an increase in PLD activity, as can be measured by the increased formation of phosphatidylbutanol in xylanase-treated cell suspensions compared with mock-treated cell suspensions. LePLDβ1-silenced cell-suspension cultures reached only a third of the increase in PLD activity seen in control cultures (Figure 2a,b), indicating that LePLDβ1 is responsible for the xylanase-induced PLD activity. The remaining increase in activity observed in silenced lines can be explained by incomplete gene silencing or by the activity of another PLD. The fact that we can still detect residual expression in the silenced lines argues for the former explanation, although the latter cannot be excluded.
To our knowledge, no localization studies have been done to date with respect to PLDs of the β class specifically. Assays for PIP2-dependent, low-micromolar [Ca2+] requiring PLD activity in protein isolates have demonstrated that this PLD activity can be found both in cytosolic and membrane fractions (Pappan et al., 1997). However, these enzymatic requirements are also present in PLDs of the γ class (Fan et al., 1999; Pappan et al., 1998). Consequently, going by these studies it is unclear whether PLDβ is present at all in untreated plants, and, if so, whether it is soluble or on the membrane. Fan et al. (1999) tried to detect Arabidopsis PLDβ in different tissues using antibodies but were unable to show the presence of this isoform, although they could detect PLDβ mRNA. Our expression studies show that LePLDβ1 transcript is detectable in untreated Msk8 cell-suspension cultures (Figure 1a).
Expression of an LePLDβ1–GFP fusion protein confers a cytosolic fluorescent signal in untreated cell-suspension cultures (Figure 7b). Upon treatment with xylanase, the LePLDβ1–GFP signal relocalized to punctuate structures in the cytosol, whereas LePLDα1–GFP remained diffuse throughout the cytosol (Figure 7c,d). This relocalization is another strong indication that LePLDβ1 is activated by xylanase treatment.
Involvement of LePLDβ1 in the plant defense response
LePLDβ1-silenced cell-suspension cultures responded differently to xylanase than did control cultures. The oxidative burst, as measured by the pyranin assay, was much more intense in silenced lines (Figure 4). It might seem that LePLDβ1 is a negative regulator of the oxidative burst response to xylanase. Yet, this is difficult to tie in with the fact that the PA produced by PLC is a positive regulator of the oxidative burst (de Jong et al., 2004). This interpretation is also in conflict with the findings of Yamaguchi et al. (2005), who suggest that the PLD activity observed after treatment of rice cell-suspension cultures with chitosan positively regulates ROS production, defense-related gene expression and accumulation of phytoalexin.
However, untreated LePLDβ1-silenced cell-suspension cultures were already aberrant. They displayed higher background staining in the ASA assay (Figure 3), and showed pyranin oxidation in the absence of xylanase (Figure 4b). Furthermore, silenced lines had higher PPO activity (Figure 5b), secreted less LeXYL2 (Figure 6a) and expressed and secreted more XEGIP (Figure 6). Increased PPO activity and secretion of proteins that inhibit pathogenic cell-wall-degrading enzymes are clear indicators of the plant defense response (York et al., 2004; Thipyapong et al., 2004). A decreased secretion of endogenous cell-wall-modifying enzymes has, to our knowledge, not been reported as a response to pathogenic elicitation but it is comprehensible that a plant would suspend developmental cell-wall-modification while it deals with a pathogen. Hence, the lack of basal PLDβ1 levels already has an effect on plant defense. It appears that LePLDβ1-silenced cell-suspension cultures are responding without elicitation; or that they are primed to respond and that the slightest elicitation sets them off. However, it seems that only a subset of defensive strategies are primed, as PR1 expression is not affected by the silencing (Figure 1b).
The function of LePLDβ1 in the plant defense response remains undefined. The silencing strategy employed here has shed some light but there is still much to be discovered. LePLDβ1–GFP in Msk8 cells relocalized to what appear to be vesicles after xylanase treatment (Figure 7b and Figure S1). One could speculate that these are endocytic vesicles; this fits with the fact that the recently cloned tomato xylanase receptors have an endocytosis signal (Ron and Avni, 2004). Alternatively, these punctuate structures could be secretory vesicles delivering proteins or other compounds to the cell surface in response to elicitation. Arabidopsis PLDβ1 has been reported to bind actin in vitro (Kusner et al., 2003), maybe PLDβ1 can form a tether between vesicular membranes and the actin cytoskeleton, as has been proposed for the plasma membrane and cortical microtubules (Dhonukshe et al., 2003). Co-localization studies and studies with cytoskeleton-affecting drugs could yield enlightening results in the future.
Plant cell-suspension cultures
Suspension-cultured cells (L. esculentum Mill.; line Msk8; Felix et al., 1991) were grown at 24°C in the dark at 125 r.p.m. in MS medium supplemented with 3% (w/v) sucrose, 5.4 μm NAA, 1 μm 6-benzyladenine and vitamins (pH was adjusted to 5.7 with 1 m KOH) as described by Felix et al. (1991) and used 4–6 days after weekly subculturing.
Construction of silencing and GFP-tagging plasmids
For the silencing plasmid, an inverted repeat (RNAi) construct specific for LePLDβ1 was generated targeting the gene's 3′ untranslated region (UTR). Polymerase chain reaction amplification of the LePLDβ1 cDNA, cloned previously by Laxalt et al. (2001), was performed with the following oligonucleotides: 1_5′-CGGGATCCCCATCGATGATCACTTCACTCATAAAGAC-3′ (reverse) with a BamHI and a ClaI restriction site, 2_5′-CCGGAATTCCGTTCTCTGGAAGGCACCAGAGATACT-3′ (forward) with an EcoRI restriction site and 3_5′-CCGGAATTCGGCCAGAGGGTAATCCGACCAG-3′ (forward) with an EcoRI restriction site. The PCR products resulting from primer combinations 1–2 and 1–3 were ligated in a 1–2/3–1 orientation into the pGreen1K binary vector which was modified to contain the 35S-Tnos cassette from pMON999 (Verdonk et al., 2005).
Green fluorescent protein fusion constructs were generated by cloning LePLDα1 and LePLDβ1 into pEZR(K)-LN. The pEZR(K)-LN plasmid was created by ligating the expression cassette from pEZS-LN (David Ehrhardt, Stanford University; more information is available at http://deepgreen.stanford.edu/) into pCambia 2300. The following oligonucleotides were used to add restriction sites to previously cloned cDNAs: LePLDα1 5′-CCGTCGACATGGCTCAGATTCAGCTTCATGG-3′ (forward) with a SalI restriction site and 5′-GGGGATCCGTAGTGAGGTTGGGAGGAAGGTAG-3′ (reverse) with a BamHI restriction site; LePLDβ1 5′-GCGAATTCTAATGGCTCATTTCTCTTATTC-3′ (forward) with an EcoRI restriction site and 5′-GGAGATCTATGGTGAGATTTTCTTGAACACCAGTGA-3′ (reverse) with a BglII restriction site.
All constructs were transferred to Agrobacterium tumefaciens strain EHA105 carrying the pJIC.SaRep plasmid. Transfection was achieved by co-cultivating 8 ml of 4-day-old Msk8 cells for 3 days at 23°C in a Petri dish with 200 μl of an overnight A. tumefaciens culture carrying the appropriate vector, in the presence of 0.2 mm acetosyringon. The cells were then plated on filters in Petri dishes with supplemented MS/agar containing 250 μg ml−1 carbenicilin and 40 μg ml−1 kanamycin. Every 3 days, cells were transferred to a new Petri dish containing carbenicilin and increasing amounts of kanamycin (twice on 40 μg ml−1, twice on 100 μg ml−1 and twice on 200 μg ml−1). After 3 weeks, developing calli were transferred to individual Petri dishes with 250 μg ml−1 carbenicilin and 200 μg ml−1 kanamycin. Pieces of independently transformed calli were transferred to liquid MS medium containing the same antibiotic concentrations and cultured as described above.
In vivo PLD measurements
To assay PLD activity in living cells, the production of phosphatidylbutanol was measured (Munnik et al., 1995). In brief, cells were pre-labeled with 32Pi for 3 h and subsequently treated with cell-free medium with or without xylanase (Trichoderma viride, Fluka BioChemika) in the presence of 0.5% (v/v) 1-butanol. Incubations were stopped and lipids extracted as described before (van der Luit et al., 2000). The 32P-labeled phosphatidylbutanol was separated from the rest of the phospholipids on heat-activated TLC plates using the organic upper phase of an ethyl acetate mixture: ethyl acetate/iso-octane/formic acid/water (12:2:3:10; Munnik et al., 1998). Lipids were visualized by autoradiography and quantified by phosphoimaging (Molecular Dynamics, Sunnyvale, CA, USA).
Northern blot analysis
Total RNA from tomato cell-suspension cultures was isolated using the TRIzol-LS reagent method (Gibco, Gaithersburg, MD, USA). Ten micrograms of RNA was separated by denaturing 1.4% formaldehyde-agarose gel electrophoresis, transferred onto Hybond-XL nylon membranes (Amersham Pharmacia, Buckinghamshire, UK) and hybridized with 32P-labeled probes in modified Church solution at 65°C. Membranes were washed three times for 15 min with wash buffer [1× SSC, 0.1% (w/v) SDS] and the probe signal was visualized by autoradiography.
Hydrogen peroxide measurements
The ASA assay was performed as described by Felix et al. (1999). Briefly, 10 ml 5-day-old Msk8 cell suspensions with or without 10 mm ASA were placed in 6 cm Petri dishes. Most of the medium was removed using a narrow-tipped pipette, leaving a layer of wet cells. Cell-free medium in a volume of 1 μl containing 0 ng μl−1, 5 ng μl−1, 50 ng μl−1, 500 ng μl−1 or 5 μg μl−1 xylanase was applied locally onto the lawn of cells and staining was visualized after 24 h using a flat-bed scanner.
The pyranin oxidative burst assay was performed as described previously (de Jong et al., 2004). In short, Msk8 cells were washed and resuspended to a 100 mg ml−1 cell density in assay buffer (5 mmMES/NaOH, pH 5.7, 175 mm mannitol, 0.5 mm K2SO4, 0.5 mm CaCl2). Cells were dispensed in 250 μl aliquots in 24-well microtiter plates and allowed to equilibrate overnight in the dark at 20°C. To measure the elicitor-induced oxidative burst, 250 μl assay buffer, supplemented with the elicitor and the fluorescent probe pyranin (8-hydroxy-1,3,6-pyrenetrisulfonic acid; Aldrich Chemical Company, Milwaukee, WI, USA) was added to the cells to produce the desired final concentration of elicitor and 10 μg ml−1 pyranin. The quenching of pyranin fluorescence due to the production of H2O2 was recorded at 1-min intervals using an excitation wavelength of 405 nm and emission wavelength of 512 nm in a fluorescence spectrophotometer (Perkin-Elmer, Boston, MA, USA).
Polyphenol oxidase assay
Msk8 cells were ground in liquid nitrogen and taken up in extraction buffer [0.1 m Tris-HCl pH 7.0, 0.1 m KCl, 1% (v/v) Triton, 1 mm EDTA pH 8.0 and 5% (w/v) polyvinylpolypyrrolidone; Li and Steffens, 2002]. This suspension was centrifuged at 4°C for 20 min at 10 000 g and the supernatant was transferred to new tubes and assayed for protein content according to Bradford (1976). Protein extracts were spotted onto Hybond-ECL nitrocellulose membrane (Amersham Pharmacia, Buckinghamshire, UK) in a 1 μl volume containing 10 mg ml−1 protein. The membrane was then placed onto Whatman 3MM paper soaked in PPO staining solution (0.2 m sodium phosphate, 0.1 m citric acid, 15 mm catechol and 0.05% (w/v) p-phenylenediamine; Howe and Ryan, 1999) for 1 h, rinsed with water and imaged with a flat-bed scanner.
Identification of culture medium proteins
Protein from 3 ml of culture medium was precipitated overnight at −20°C with four volumes of precipitation buffer [0.1 m sodium acetate, 95% (v/v) ethanol]. The protein was spun down by centrifugation for 10 min at 15,500 g and the pellet was resuspended in 100 μl loading buffer [8% (w/v) SDS, 40% (v/v) glycerol, 20% (v/v) β-mercapto-ethanol, 240 mm Tris-HCl pH 6.8 and 0.08% (w/v) Bromophenol Blue]. Samples were separated using an 8% SDS-PAGE gel and stained with Brilliant Blue G-colloidal Coomassie (Sigma-Aldrich, Steinheim, Germany). Selected bands were identified by mass spectrometry as described by Testerink et al. (2004).
Confocal scanning laser fluorescence microscopy
Microscopic imaging was performed as described previously (Dhonukshe et al., 2003). Briefly, 200 μl 4–6-day-old cell-suspension culture samples were prepared in eight-well NUNC chambers (Nunc, Naperville, IL, USA). Images were acquired using confocal laser scanning microscopy based on the Zeiss LSM 510 system with a ×40 water-immersion objective (Zeiss, Jena, Germany). Images were captured using LSM 510 image-acquisition software (Zeiss).
We thank Dr Gert-Jan de Boer (Enza Zaden, Enkhuizen, The Netherlands) for kindly providing the pEZR(K)-LN vector, Dr Rob Schuurink (Swammerdam Institute for Life Sciences, Amsterdam, The Netherlands) for advice concerning the PPO assay, Dr Alex Mithöfer (Max-Planck-Institute for Chemical Ecology, Jena, Germany) for his efforts towards the identification of the phenolic compound in the culture medium, Dr Titia Sijen (Hubrecht Laboratory, Netherlands Institute of Developmental Biology, Utrecht, The Netherlands) for assistance in designing the silencing construct, Dr Phil Mullineaux (University of Essex, Colchester, UK) and Dr Roger Hellens (Biotechnology and Biological Sciences Research Council) for kindly providing the pGreen1K vector. Research in TM's lab was funded by the Netherlands Organization for Scientific Research (NWO: 805-33-343 and 805-48-005), the Royal Dutch Academy of Arts and Sciences (KNAW) and the European Commission (HPRN-CT-2000-00093). AL was funded by the Fundación Antorchas and the Consejo Nacional de Investigaciones Científicas y Técnicas.