Portal de Periódicos da CAPES

A permeable cuticle in Arabidopsis leads to a strong resistance to Botrytis cinerea

2007; Springer Nature; Volume: 26; Issue: 8 Linguagem: Inglês

10.1038/sj.emboj.7601658

1460-2075

Michael Bessire, Céline Chassot, Anne-Claude Jacquat, Matt Humphry, Sandra Borel, Jean MacDonald-Comber Petétot, Jean‐Pierre Métraux, Christiane Nawrath,

Plant-Microbe Interactions and Immunity

Article29 March 2007free access A permeable cuticle in Arabidopsis leads to a strong resistance to Botrytis cinerea Michael Bessire Michael Bessire Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Céline Chassot Céline Chassot Department of Biology, University of Fribourg, Fribourg, Switzerland Search for more papers by this author Anne-Claude Jacquat Anne-Claude Jacquat Department of Biology, University of Fribourg, Fribourg, Switzerland Search for more papers by this author Matt Humphry Matt Humphry Department of Plant–Microbe Interactions, Max Planck Institute for Plant Breeding Research, Köln, Germany Search for more papers by this author Sandra Borel Sandra Borel Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Jean MacDonald-Comber Petétot Jean MacDonald-Comber Petétot Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Jean-Pierre Métraux Jean-Pierre Métraux Department of Biology, University of Fribourg, Fribourg, Switzerland Search for more papers by this author Christiane Nawrath Corresponding Author Christiane Nawrath Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Michael Bessire Michael Bessire Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Céline Chassot Céline Chassot Department of Biology, University of Fribourg, Fribourg, Switzerland Search for more papers by this author Anne-Claude Jacquat Anne-Claude Jacquat Department of Biology, University of Fribourg, Fribourg, Switzerland Search for more papers by this author Matt Humphry Matt Humphry Department of Plant–Microbe Interactions, Max Planck Institute for Plant Breeding Research, Köln, Germany Search for more papers by this author Sandra Borel Sandra Borel Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Jean MacDonald-Comber Petétot Jean MacDonald-Comber Petétot Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Jean-Pierre Métraux Jean-Pierre Métraux Department of Biology, University of Fribourg, Fribourg, Switzerland Search for more papers by this author Christiane Nawrath Corresponding Author Christiane Nawrath Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland Search for more papers by this author Author Information Michael Bessire1, Céline Chassot2, Anne-Claude Jacquat2, Matt Humphry3, Sandra Borel1, Jean MacDonald-Comber Petétot1, Jean-Pierre Métraux2 and Christiane Nawrath 1 1Department of Plant Molecular Biology, University of Lausanne, Lausanne, Switzerland 2Department of Biology, University of Fribourg, Fribourg, Switzerland 3Department of Plant–Microbe Interactions, Max Planck Institute for Plant Breeding Research, Köln, Germany *Corresponding author. Department of Plant Molecular Biology, University of Lausanne, Biophore Building, Quarter UNIL/Sorge, CH-1015 Lausanne, Switzerland, Tel.: +41 21 692 4256; Fax: +41 21 692 4195. E-mail: [email protected] The EMBO Journal (2007)26:2158-2168https://doi.org/10.1038/sj.emboj.7601658 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The plant cuticle composed of cutin, a lipid-derived polyester, and cuticular waxes covers the aerial portions of plants and constitutes a hydrophobic extracellular matrix layer that protects plants against environmental stresses. The botrytis-resistant 1 (bre1) mutant of Arabidopsis reveals that a permeable cuticle does not facilitate the entry of fungal pathogens in general, but surprisingly causes an arrest of invasion by Botrytis. BRE1 was identified to be long-chain acyl-CoA synthetase2 (LACS2) that has previously been shown to be involved in cuticle development and was here found to be essential for cutin biosynthesis. bre1/lacs2 has a five-fold reduction in dicarboxylic acids, the typical monomers of Arabidopsis cutin. Comparison of bre1/lacs2 with the mutants lacerata and hothead revealed that an increased permeability of the cuticle facilitates perception of putative elicitors in potato dextrose broth, leading to the presence of antifungal compound(s) at the surface of Arabidopsis plants that confer resistance to Botrytis and Sclerotinia. Arabidopsis plants with a permeable cuticle have thus an altered perception of their environment and change their physiology accordingly. Introduction The cuticle covers the epidermal cell wall of aerial tissues forming a boundary between the plant and its environment. It represents a primary barrier minimizing water and solute loss and protecting the plants against various abiotic and biotic stresses, but may also act in a number of developmental processes, such as delimitation of organs as well as trichome and stomata formation (Goodwin and Jenks, 2005; Nawrath, 2006; Riederer, 2006). The cuticle is composed of epicuticular and intracuticular waxes and a structural polymer, cutin, that are laid down in a well-organized manner (Jeffree, 2006). Waxes are a mixture of very long-chain fatty acids and their derivatives, and in some species, triterpenes and β-diketones (Kunst and Samuels, 2003; Jetter et al, 2006). Cutin is a polyester that is, in many plant species, formed of C16 and C18 ω-hydroxylated fatty acids that may carry additional hydroxy or/and epoxy groups in the mid-chain position (Kolattukudy, 2001). However, the cutin of Arabidopsis has an unusual composition as it consists of high amounts of C16 and C18 dicarboxylic acids (Bonaventure et al, 2004; Xiao et al, 2004; Franke et al, 2005). The large amounts of monomers only carrying carboxyl groups in the Arabidopsis cutin require other monomers rich in hydroxy groups in order to form polyesters. Glycerol might be such a monomer as it has recently been found to be a cutin monomer in several plant species (Graça et al, 2002). Although a number of mutants potentially affected in the deposition of the cuticular polyesters have been characterized in recent years (Yephremov and Schreiber, 2005; Nawrath, 2006), changes in the amount and composition of cuticular polyesters have been determined only for a few mutants. The Arabidopsis mutant attenuation of type three genes (att1) having only 30% of the wild-type (WT) cuticular polyester encodes the cytochrome P450-dependent monooxygenase CYP86A2 (Xiao et al, 2004). The Arabidopsis mutant lacerata (lcr) carries a mutation in the homologue CYP86A8 and has a pleiotropic phenotype, including a permeable cuticle and a number of developmental disorders, such as organ fusions (Wellesen et al, 2001). Both mutants are most likely affected in cutin monomer biosynthesis as cytochrome P450s of the CYP86A type have been shown to have fatty acid hydroxylase activity (Benveniste et al, 1998). The organ fusion mutant hothead (hth) that exhibits a 30% reduction in dicarboxylic acids in its polyesters indicates the involvement of an oxidoreductase-dependent pathway for the formation of dicarboxylic acids in Arabidopsis (Krolikowski et al, 2003; Kurdyukov et al, 2006a). The extracellular α/β hydrolase BODYGUARD (BDG) may be involved in the formation of the polyester itself (Kurdyukov et al, 2006b). The bdg mutant accumulates more ester-bound cutin monomers than WT in the outer extracellular matrix surface, but is unable to form a functional continuous cuticular membrane. A potential function of bdg in both hydrolysis and synthesis of the polyester has been hypothesized as a number of phenotypes of bdg, including structural aspects of the polyester, resemble those observed in transgenic Arabidopsis plants expressing a fungal cutinase (Sieber et al, 2000; Kurdyukov et al, 2006b). WAX2 and LONG-CHAIN FATTY ACID SYTHETASE 2 (LACS2) are also enzymes potentially involved in cutin biosynthesis. LACS2 is an epidermis-specific long-chain acyl-CoA synthetase, whose absence, as that of WAX2, leads to altered properties and ultrastructure of the cuticular membrane (Schnurr et al, 2004). The LACS2 protein expressed in Escherichia coli has a preference for hydroxylated fatty acids, in accordance with the idea that LACS2 plays an important role in cutin biosynthesis (Schnurr et al, 2004). The proposed function of the cuticle as a diffusion barrier has been intensively studied. Intercuticular waxes seem to play an important role in sealing the cuticle and are thought to be densely packed in the spaces of the cutin polyester that serves as a scaffold (Goodwin and Jenks, 2005). This model would explain why some mutants affected in cutin deposition are strongly disturbed in their permeability barrier (Sieber et al, 2000; Kurdyukov et al, 2006b). Although the cuticle forms a diffusion barrier, molecules can diffuse through the cuticle. Lipophilic non-ionic molecules diffuse through the hydrophobic cuticle and their diffusion kinetics are different from those of charged molecules that travel in an apolar path, while H2O diffuses in both routes (Schreiber, 2005; Riederer and Friedmann, 2006; Burghardt and Riederer, 2006). Phytopathogenic fungi often secrete cutinases during their initial contact with a plant that liberates cutin monomers serving as signals for the induction of cutinases necessary for penetration and for a number of differentiation processes in fungi. For instance, germination and appressorium formation in Magnaporte grisea and formation of the appressorial tube in Erysiphe graminis are induced by cutin monomers (Kolattukudy et al, 1995; Francis et al, 1996; Gilbert et al, 1996). However, a crucial role of cutinases in the invasion of a plant was not always found. For example, deletion of the cutinases and cutinolytic lipases in the necrotrophic fungus Botrytis cinerea did not hinder the pathogen to enter intact plant tissues (Reis et al, 2005). Very little is known about the significance of cutin for plant defence (Yephremov and Schreiber, 2005; Nawrath, 2006). The Arabidopsis mutant att1 was identified by its attenuating effect on virulence gene induction in phytopathogenic bacteria and its increased disease susceptibility toward bacterial infection (Xiao et al, 2004). In contrast, cutinase-expressing Arabidopsis plants and the bdg mutant were found to be strongly resistant to the necrotrophic fungus B. cinerea because of a multifactorial defence mechanism (Chassot et al, 2007). This resistance phenomenon has been seen in relation to earlier findings that cutin monomers may induce defence responses in plants (reviewed in Chassot and Métraux, 2005). A forward genetic screen in Arabidopsis made it possible to dissect the potential pathway linking cutin or cutin monomers to resistance to B. cinerea. A mutant, botrytis resistant1 (bre1), displaying strong resistance to infection by B. cinerea, was also found to be defective in its cuticular membrane. bre1 carries a mutation in the LACS2 gene that was previously shown to be involved in cuticle development (Schnurr et al, 2004), and was here found to be important for cutin biosynthesis. The study of bre1/lacs2 and a number of other Arabidopsis mutants impaired in cutin monomer biosynthesis demonstrated that the increase in permeability of the cuticular membrane directly correlates with the amount of antifungal compounds released to the plant surface and Botrytis resistance. An interrupted cuticular membrane allowed the diffusion of signals and effector molecules across the cuticle resulting in an arrest of infection by Botrytis and Scleotiorum Sclerotinia. Results Isolation and characterization of the bre1 mutant A total of 13 000 plants of an M2 population of EMS-mutagenized Arabidopsis plants were screened in parallel for two phenotypes: resistance to B. cinerea and increased cuticular permeability. One mutant of nearly normal size was strongly resistant to infection with B. cinerea strain (BMM) and was called bre1. Infection sites on bre1 leaves remained usually symptom free, in contrast to those on Col leaves (less than 2% of outgrowing lesions) (Figure 1A–C). In addition, bre1 displays increased cuticle permeability compared with WT as determined by Calcofluor staining. Calcofluor white fluoresces when it binds to β-glucans, such as cellulose in the cell wall. bre1 showed an extensive staining in all parts of the plant. In particular, the staining of cotyledons, the second rosette leaf pair, as well as most parts of the flower was distinctly different in bre1 compared with WT plants (Figure 1F, I, and L). All phenotypes were inherited in a recessive manner. Figure 1.Phenotypes of bre1/lacs2-2 and lacs2-3 mutant isolates. (A) Morphology of the rosette of Col-0, lacs2-2, and lacs2-3. (B–D) Macroscopic evaluation of symptoms 3 days after inoculation with B. cinerea of Col-0, bre1/lacs2-2, and lacs2-3. Scale bar: 1 cm. (E–M) Calcofluor white-stained organs of Col-0, bre1/lacs2-2, and lacs2-3, viewed under UV light. Col-0 pictures are shown with the corresponding pictures in bright-field microscopy as inset. Leaves (E–G), young seedlings (H–J), and the tip of a flowering inflorescence (K–M) are shown. Download figure Download PowerPoint BRE1 has a mutated LACS2 gene BRE1 was mapped to the middle of chromosome 1 close to LACS2 (Lukowitz et al, 2000), which has been previously identified to be important for cuticle development (Schnurr et al, 2004). Arabidopsis plants (accession Col-0) carrying a T-DNA insertion in the LACS2 gene were obtained that were also resistant to B. cinerea (quantitative data shown in Figure 6B), stained strongly with Calcofluor, and had no other obvious morphological phenotypes (Figure 1A, D, J, and M) (Rosso et al, 2003). Complementation analysis between bre1 and the LACS2 T-DNA insertion line revealed that BRE1 is indeed LACS2, as both mutants did not complement each other in the F1 generation (Supplementary Figure 1). We therefore named bre1 lacs2-2 and the T-DNA insertion line lacs2-3. Figure 2.Ultrastructure of the epidermal extracellular matrix of rosette leaves. Lower epidermal extracellular matrix of Col-0 (A) and lacs2-3 (B, C) leaves. Extracellular matrix at the point of fusion between the lower epidermis of one leaf and the upper epidermal cell layer of another leaf in lacs2-3 (D). Scale bar: 500 nm (A) and 800 nm in (B–D). Download figure Download PowerPoint Sequence analysis revealed mutations in LACS2-2 leading to the expression of a truncated cDNA (Supplementary Figure 2A and C). Mutations accompanied by insertion of the T-DNA in LACS2-3 lead to an RNA-null allele (Supplementary Figure 2B and C). As lacs2-2 and lacs2-3 had very similar phenotypes during the initial characterization, only data for lacs2-3 are presented in the following studies. The cuticle of the lacs2 mutant The ultrastructure of the cuticular membrane of the lacs2-3 mutant was analyzed by transmission electron microscopy (TEM). The distinct osmium-dense cuticular membrane representing insoluble lipid-derived polymers such as cutin, visible at the epidermal outer extracellular matrix of WT plants, could not be seen at the leaf epidermis of lacs2-3 (Figure 2A and B). Interestingly, the remaining outer epidermal extracellular matrix sometimes had a distinct laminated structure (Figure 2C). Growing plants on a large scale revealed that lacs2-3 plants very rarely formed organ fusions (one organ fusion in less than 1% of plants). Ultrastructural analysis of such organ fusions revealed that both epidermal cell walls are directly fused without any osmium-dense layer in between them (Figure 2D). Figure 3.Composition of aliphatic monomers of leaf polyesters. Large rosette leaves of 5- to 6-week-old plants of Col-0 and lacs2-3 were analyzed. (A) Hydroxylated fatty acids and their derivatives and (B) fatty acids and fatty alcohols. DiOH, dihydroxy carboxylic acid; DiCA, dicarboxylic acid (n=3; ±s.e.; the experiment was repeated once with similar results). Download figure Download PowerPoint The cuticular membrane of lacs2-3 was not visible as an electron-opaque layer, indicating that the osmium-dense, insoluble lipids might be largely missing. Thus, the amount and composition of ester-bound lipids were analyzed from plant material after complete elimination of soluble lipids. Analysis of residue-bound lipids of lacs2-3 in comparison with Col-0 plants revealed that a number of ester-bound monomers were reduced in the lacs2-3 mutant (Figure 3A and B). The total amount of ω-hydroxylated fatty acids and their derivatives was reduced 4- to 5-fold in lacs2-3 compared with Col-0 (Figure 3A). The impact of LACS2 mutation on the amount of dicarboxylic acids, characteristic of Arabidopsis cutin, was particularly strong, and the amount was reduced to 15–20% of WT amount. ω-Hydroxylated fatty acids and their dihydroxy fatty acids were reduced to less than 50%. Thus, lacs2 has a strongly reduced content of cuticular polyester. Figure 4.Cuticular transpiration. Loss of water from 5- to 6-week-old rosettes of Col-0 and lacs2-3 plants was measured at the indicated time points (n=4–6; ±s.e.; experiment was repeated twice with similar results). Download figure Download PowerPoint Permeability of the cuticle of lacs2 and other cuticle mutants An intact cuticular membrane is important to reduce water loss from the plant under drought conditions (Goodwin and Jenks, 2005; Burghardt and Riederer, 2006). Water loss was taken as a measure of cuticular permeability and was determined in fully expanded but not yet bolting rosettes excised at the level of the hypocotyls. As shown in Figure 4, the rate of water loss from lacs2-3 is significantly higher than that of Col-0 plants. As the stomata of both Col-0 and lacs2-3 plants were well closed during the experiment (Supplementary Figure 3), lacs2-3 is likely to have a higher cuticular transpiration than Col-0. Figure 5.Permeability of leaf cuticles of mutants having an altered cuticular polyester. (A) Droplets of a toluidine blue solution were incubated on the leaves for 2 h and then washed with water. An overview of typically stained leaves from each genotype is given in the top panel. A microscopic view of the typical staining underneath a droplet that was incubated for 2 h on leaves from each genotype is given below its overview. (B) Toluidine blue droplets were incubated on lacs2-3 leaves for the time points indicated below each leaf and then briefly rinsed with water. Typically stained leaves are presented. A colour version of this figure is available at the EMBO Journal online. Download figure Download PowerPoint Sensitivity to xenobiotics such as herbicides also reflects the permeability of the cuticular membrane. Permeability of the upper surface of Arabidopsis leaves was specifically determined to avoid the problem of molecule exchange via stomata, as these openings are mostly positioned on the abaxial side of leaves. Besides WT and lacs2-3 plants, two other mutants, hth and lcr, reported to have a permeable cuticle were included in this experiment (Lolle et al, 1998; Wellesen et al, 2001). lacs2-3 as well as lcr showed severe damage at a lower concentration of BASTA® compared with hth and WT Col-0 plants (Table I). Table 1. Sensitivity to BASTA® of hth, lcr, and lacs in comparison with WT Col-0 Basta (nl/ml) Col hth lcr lacs2 1 +++++ +++++ +++++ +++++ 10 +++++ +++++ ++++ ++++ 20 ++++ ++++ +++ +++ 40 ++++ ++++ ++ ++ 60 +++ +++ +/− +/− +++++: fully green; ++++: chlorotic tips or spots; +++: whole leaves chlorotic, some dying; ++: all expanded leaves died, but vigorous regrowth; +: poor or late regrowth; +/−: dead or regrowth after 10 days only. Permeability of the cuticle of lacs2 and other mutants was further assessed with toluidine blue applied in the form of droplets on the adaxial sites of the leaves (Tanaka et al, 2004; Kurdyukov et al, 2006b). As toluidine blue can be directly applied to plant tissues without solvent treatment, permeability of a functional cuticular membrane can be investigated by this staining method, in contrast to Calcoflour staining. Two hours after the application of toluidine blue solution, Col-0 leaves showed a small stained spot covering less than 10% of the area on which the droplet was applied, whereas lacs2 leaves showed dark blue staining on 75–100% of the area. Interestingly, the lcr leaves showed staining on approximately 25–50% of the droplet area, whereas hth had a staining pattern similar to that of Col-0 (Figure 5A). A shorter incubation time, that is, 30 min, resulted in a similar pattern, but the staining was weaker (data not shown). Leaves of lacs2-3 were already stained after only 5 min and the whole area underneath droplets was stained after 30 min (Figure 5B). Figure 6.Resistance to Botrytis of mutants having differences in their cuticular polyesters. (A) Macroscopic symptoms 3 days after inoculation with B. cinerea (BMM) on leaves of different genotypes. White arrowheads indicate examples of symptom-free inoculation sites. Scale bar: 1 cm. (B) Number of outgrowing lesions developed 3 days after inoculation with B. cinerea (BMM) in different genotypes (n=4; ±s.e.; different letters indicate significance at P<0.005 using ANOVA). The experiment was repeated twice with similar results. Download figure Download PowerPoint Taken together, our data demonstrate that the lacs2-3 has a highly permeable cuticle that is even more permeable than that of lcr, in comparison with WT Col-0, whereas an increased cuticular permeability of hth leaves could not be confirmed in our experimental settings. Resistance to B. cinerea of mutants having altered cuticular permeability Resistance of lacs2 plants to Botrytis was compared with other mutants affected in cuticle formation. The visual phenotypes of the leaves 3 days after inoculation are depicted in Figure 6A. The number of lesions that were larger than the original infection site (outgrowing lesions) is given as a quantitative measure of the resistance level (Figure 6B). The lacs2-3 mutant exhibited a strong resistance to infection with B. cinerea strain BMM, as described above, with less than 5% outgrowing lesions in comparison with WT Col-0 (75% outgrowing lesions). A significant increase in the resistance to B. cinerea was also observed in lcr (45% outgrowing lesions), whereas no increase in resistance was observed in hth. In addition, infection sites could be observed in lcr that did not develop any disease symptoms, similar to the symptom-free infection sites in the lacs2-3 mutant (Figure 6A). The frequency of symptom-free lesions in lcr was significantly lower than that in the lacs2-3 mutant (30 and 95%, respectively). Similar results were obtained with B. cinerea strain BO5.10. Analysis of leaf diffusates for antifungal compounds The permeability of the cuticle correlated with the resistance to B. cinerea (Figures 5A and 6) but not with PDF1.2 and camalexin production that have often been correlated to Botrytis resistance (Penninckx et al, 1998; Berrocal-Lobo et al, 2002; Coego et al, 2005; Kliebenstein et al, 2005) (Supplementary Figure 4). In the search for antimicrobial compounds at the surface of the plant that might arrest the infection at an early stage, droplets of 1/4 strength potato dextrose broth (PDB) were incubated on the leaves so that antifungal substances could diffuse into them. The activity of these diffusates was analyzed in two assays: (1) spores of B. cinerea were germinated in vitro in the presence or absence of diffusates and the length of the hyphae was measured (in vitro assay). (2) Botrytis spores were also tested for their ability to infect WT plants in the presence or absence of diffusates (in vivo assay). The results of in vitro assay are presented in Figure 7A. The 1/4 PDB diffusate from lacs2-3 collected after 18 h of incubation inhibited the development of Botrytis hyphae, whereas the diffusates from Col-0, lcr, and hth did not. After 44 h, the 1/4 PDB diffusates collected from lcr and lacs2-3 were active, whereas these from hth and Col-0 were not. The growth-inhibiting effect of diffusate from lacs2-3 was much stronger than that from lcr, whereas the 1/4 PDB diffusates from hth and Col-0 were not active. Results of the in vivo test are presented in Figure 7B. Botrytis spores that were mixed with 1/4 PDB diffusates, collected after 18 or 44 h from lacs2-3 plants, were unable to infect Col-0 plants (0% outgrowing lesions). Spores incubated with 1/4 PDB diffusates from lcr collected after 44 h could not infect Col-0 plants (0% outgrowing lesions), but when collected after 18 h could still infect Col-0 normally (85% outgrowing lesions). Spores incubated with 1/4 PDB diffusates from hth or Col-0 had no significant reduction in infection capability (70–85% outgrowing lesions). Figure 7.Analysis of PDB diffusates for the presence of growth-inhibiting activities against B. cinerea. (A) Germination and growth of B. cinerea (BMM) in vitro in the presence of leaf diffusates collected at 18 h (grey bars) and 44 h (black bars) from plants of different genotypes. Three independent experiments were undertaken and six representative pictures were evaluated ±s.d. (B) B. cinerea (BMM) spores were incubated on WT Col-0 leaves in the presence of 1/4 PDB diffusates collected from leaves of different genotypes at indicated time points. Typical symptoms 3 days after inoculation are presented. Download figure Download PowerPoint Both tests clearly indicate that the surface of plants with an increased cuticle permeability releases antifungal compounds against Botrytis and that the activity depends on the time of diffusion. Comparison of the in vitro and in vivo tests suggests that a reduction of 50% and more in the growth of the hyphae in vitro leads to an arrest in infection in vivo. Characterization of antifungal compounds and their induction The antifungal compounds present in the 1/4 PBD diffusate obtained from lacs2-3 plants after an incubation time of 42 h were characterized by testing their in vitro activity on Botrytis spores after different treatments: the active compound(s) were resistant to lyophilization and heat treatment and could not be extracted into ethyl acetate, hexane, or chloroform, indicating that they are of polar nature. The antifungal activity of the lacs2 diffusate could be diminished by treatment with a lipase from Mucor miehei. The germlings of Botrytis were four to eight times longer than without lipase treatment of the diffusate (Figure 8 and Supplementary Figure 5). Depending on the PDB diffusate preparation, 12–60% of growth of the germlings in 1/4 PDB treated with lipase could be restored. Similar results were obtained with a lipase from Candida antarctica that had higher specific activity (data not shown). Proteinase K treatment has only a small effect on the activity of PDB diffusates (Supplementary Figures 5 and 8). Germlings were only 2–3 times longer than when incubated in untreated PDB diffusate. Between 5 and 18% of the germling length grown in 1/4 PDB treated with proteinase K could be restored (Figure 8). Figure 8.Analysis of the activity of different PDB diffusates from lacs2 after enzymatic digestions. Germination and growth of B. cinerea (BMM) in the presence of different 1/4 PDB diffusates (PDB DIF) from lacs2 that had undergone no treatment (control), treatment with lipase from M. miehei at 32 U/μl (lipase), or proteinase K (protease) are shown. Lengths of the germlings grown in treated 1/4 PDB diffusates from lacs2 were normalized to those grown in treated 1/4 PDB (100%). Three to four representative pictures were evaluated ±s.d. Download figure Download PowerPoint These results indicate that the lacs2 diffusate most likely contains several components. Differences in the efficiency of the treatments may reflect differences in the composition of diffusates that may result from slightly different induction kinetics of the compounds. One of the prominent active compounds of the PDB diffusate from lacs2 plants may be a lipid or a compound that contains a linkage to an acyl group that is important for its activity. This lipase-sensitive compound is of relatively low molecular weight, as it could be filtrated through a membrane with a molecular weight cutoff of 3000 Da (data not shown). Interestingly, only PDB itself—even when adjusted to different pH—induced the antifungal compounds in lacs2-3 plants and not the major components of PDB (sucrose and starch), different pH, different ion concentrations, and osmotic potential (Table II). The inducer is, therefore, most likely a minor component of PDB that acts as an elicitor to induce antifungal compounds inhibiting the growth of the hyphae of B. cinerea and S. sclerotiorum, but not of Alternaria brassicicola and Plectosphaerella cucumerina (Supplementary Figure 6). In accordance with this observation, lacs2-3 also showed increased resistance to infection by Sclerotinia (Figure 9). Interestingly, the resistance of lacs2 to infection by Sclerotinia could be further increased by a 20-h preincubation period of the future inoculation sites with 1/4 strength PDB, again demonstrating the importance of PDB-mediated induction for the defence response observed in permeable cuticle mutants (Figure 9). Figure 9.Influence o