Effect of cold-induced changes in physical and chemical leaf properties on the resistance of winter triticale (×Triticosecale) to the fungal pathogen Microdochium nivale

Authors


E-mail: szechynska@ifr-pan.krakow.pl

Abstract

This study showed that several mechanisms of the basal resistance of winter triticale to Microdochium nivale are cultivar-dependent and can be induced specifically during plant hardening. Experiments and microscopic observations were conducted on triticale cvs Hewo (able to develop resistance after cold treatment) and Magnat (susceptible to infection despite hardening). In cv. Hewo, cold hardening altered the physical and chemical properties of the leaf surface and prevented both adhesion of M. nivale hyphae to the leaves and direct penetration of the epidermis. Cold-induced submicron- and micron-scale roughness on the leaf epidermis resulted in superhydrophobicity, restricting fungal adhesion and growth, while the lower permeability and altered chemical composition of the host cell wall protected against tissue digestion by the fungus. The fungal strategy to access the nutrient resources of resistant hosts is the penetration of leaf tissues through stomata, followed by biotrophic intercellular growth of individual hyphae and the formation of haustoria-like structures within mesophyll cells. In contrast, a destructive necrotrophic fungal lifestyle occurs in susceptible seedlings, despite cold hardening of the plants, with the host epidermis, mesophyll and vascular tissues being digested and becoming disorganized as a result of the low chemical and mechanical stability of the cell wall matrix. This work indicates that specific genetically encoded physical and mechanical properties of the cell wall and leaf tissues that depend on cold hardening are factors that can determine plant resistance against fungal diseases.

Introduction

The plant cell wall is the matrix where a pathogen and a plant initiate their interaction, and in muro events have a substantial impact on the outcome (Cantu et al., 2008). For successful colonization, fungal pathogens must penetrate plant tissues through natural openings, such as stomata or wounds. Simultaneously, or successively, pathogens gain access to both the plant cell interior and cellular nutrients through an enzymatic weakening and breaching of the host cell wall and/or by the high turgor pressure created inside invasion structures. Conversely, plants respond to fungal action by reinforcing the cell wall, inhibiting cell-wall-degrading enzymes and killing potential intruders via antimicrobial means (Hückelhoven, 2007).

The chemical and mechanical properties of the plant cell wall are factors that can determine the plant's basal resistance to pathogens. Cell walls of a typical grass species contain 25% cellulose, 10% pectin and 55% hemicellulose (on a dry weight basis; Vorwerk et al., 2004). Unbranched and unsubstituted cellulose microfibrils consist of thousands of (1, 4)-β-d-glucose monomers and form insoluble crystalline complexes through extensive inter- and intramolecular hydrogen bonding and hydrophobic interactions. Pectins build the highly hydrated gel-like matrix of the wall and interact with associated proteins, phenolic compounds and ions. Hemicelluloses (high-molecular-weight polymers) are hydrogen-bonded to the surface of the cellulose microfibrils and embedded in a pectin matrix. Pectins and hemicelluloses show considerable heterogeneity and dependence on the cell's specific functions, its stage of development and environmental conditions (Vorwerk et al., 2004; Burton et al., 2010).

Such wall structure provides mechanical support to the cell, but its inherent porosity allows water and small molecules to diffuse through the cell wall. Molecules that weigh about 50 kDa diffuse through a typical wall in less than 1 h, and molecules below 17 kDa would be expected to diffuse in a few minutes (Vorwerk et al., 2004). Thus, the hydrolytic enzymes released by pathogens, which are typically 30–40 kDa, can probably diffuse from the cell wall surface to the plasma membrane before a physical contact between the pathogen cell and the host plasma membrane is established (Vorwerk et al., 2004). Relatively small fungal pectinases would be able to start the process of cell wall degeneration, and the loss of pectins would increase the ability of other degrading enzymes (cellulases, polygalacturonases, xylanases and proteinases) to reach other components of the cell wall (Hückelhoven, 2007; Cantu et al., 2008).

Plants can actively protect their cell wall integrity and restrict the passage of small molecules by structural and chemical modifications of cell wall compounds (Hückelhoven, 2007). Such modifications can also defend against the compressive forces of fungal hyphae and prevent pathogens from using the products of plant metabolism (Bechinger et al., 1999). The general plant strategy includes the production of the cuticle, cutin and soluble waxes embedded in the matrix and deposited on the external surface, although some pathogens are able to breach the surface of the reinforced host cells by releasing enzymes, such as cutinases and lipases (Kikot et al., 2008). Furthermore, internal cellulose microfibrils and the matrix of the plant cell wall become encrusted with lignin, a large randomly cross-linked polymer consisting of p-coumaryl, coniferyl and sinapyl alcohols, that can act as a functional permeability barrier and mechanical reinforcement of the cell wall (Burton et al., 2010). The expression of lignin biosynthetic genes and the accumulation of soluble and cell-wall-bound phenolics are induced in response to infection with various pathogens, e.g. in Arabidopsis thaliana challenged with Xanthomonas campestris, in flax inoculated with Botrytis cinerea and in barley during powdery mildew infection (Von Röpenack et al., 1998; Lauvergeat et al., 2001; Hano et al., 2006). It has also been shown that inhibitors of the phenylpropanoid pathway, such as the phenylalanine ammonium lyase inhibitors α-aminooxy-β-phenylpropionic acid or 2-aminoindan-2-phoshonic acid, and the cinnamic acid dehydrogenase inhibitor 2-hydroxyphenylaminosulphinyl acetic acid (1, 1-dimethyl ester), weaken the resistance of barley to Blumeria graminis or that of A. thaliana to Hyaloperonospora parasitica (Hückelhoven, 2007).

Similar changes in cell wall composition and structure have also been observed in plants exposed to different abiotic stresses. Cold, drought or high light, as well as mechanical injuries, lead to reduction in cellulose biosynthesis and cause enhanced lignin production in a large number of plant species, such as poplar, rice, pine, A. thaliana and soyabean (Hernández-Blanco et al., 2007; Moura et al., 2010). In this way, acclimation of the plant cell to abiotic stress may simultaneously increase the basal resistance to pathogens (biotic stress) as a result of cross-tolerance (Karpinski & Szechynska-Hebda, 2010; Szechynska-Hebda et al., 2010).

The present study correlated changes in the biochemical and mechanical properties of the plant cell wall during cold hardening with the induction of resistance to the fungal pathogen Microdochium nivale (pink snow mould). This fungus is a psychrophilic pathogen able to invade winter cereals, forage, amenity and sports turf grasses. It infects both crowns and leaves, either under the snow or during rainy winter weather. The leaves in the prostrate winter rosette stage are killed, forming pink or straw-coloured mats, which later dry to papery films (Nakajima & Abe, 1996). Conversely, the conditioning of plant seedlings by low temperature promotes genotype-dependent resistance to M. nivale infection (Golebiowska & Wedzony, 2009; Szechynska-Hebda et al., 2011) and the mechanism of cold-induced resistance is dependent on the up-regulation of a wide range of stress-response genes, comprising those related to defence and low-temperature stress (Gaudet et al., 2011; Szechynska-Hebda et al., 2011). Microdochium nivale behaves as a biotroph (living at the expense of viable host cells) in the resistant triticale cv. Hewo and as a necrotroph (feeding on dead biomass) in susceptible tissue of cv. Magnat. Although extensive plant tissue maceration is typically a feature of necrotrophic infections, both biotrophs and necrotrophs secrete diverse cell-wall-degrading proteins (Cantu et al., 2008). Therefore, the plant genetic potential for cell-wall reinforcement can be crucial in defence reactions. The work presented here investigated the roles of adhesiveness and wettability of the leaf surface, and cell wall reinforcement, stability and permeability, in defence against the destructive action of M. nivale.

Materials and methods

Plant growth conditions

As the expression of cold-induced resistance to the fungal pathogen M. nivale is dependent on plant genotype, two previously selected (Golebiowska & Wedzony, 2009) cultivars of hexaploid winter triticale (×Triticosecale), namely Hewo (able to develop resistance after cold treatment) and Magnat (susceptible to fungal infection despite plant hardening) were used for the experiments. Seeds were surface-sterilized in 0·05% (v/v) NaOCl for 15 min, rinsed with sterile water and germinated on moistened filter paper. A commercial potting mix of peat:soil:sand (1:1:1) was sterilized at 120°C and 0·1 MPa and germinated seeds (c. 48 h) were planted in pots. Seedlings were grown in a climatic chamber under light of 100 ± 10 μmol (quantum) m−2 s−1 PAR, 8/16 h (day/night), at 20°C/17°C and 60–67% relative humidity (RH), for 7 days. Starting on the 8th day after potting, the plants were subjected to a hardening period consisting of 7, 14, 21, 28, 48 or 98 days at 4/4°C, 8/16 h (day/night). The control, unhardened plants were grown at 20/17°C, 8/16 h (day/night) until reaching the same developmental stage as the fully cold-hardened plants.

Plant inoculation

Mycelium was derived from a monosporal isolate of M. nivale with a high virulence (38z/5a/01) collected from rye by Dr Maria Pronczuk from the Institute of Plant Breeding and Acclimatization, Radzikow, Poland. The fungus was grown for 10 days in darkness at 20°C on potato dextrose agar (PDA) medium.

To determine the effect of cold-hardening duration on the expression of snow mould resistance a seedling regrowth test was used. Seedlings hardened at 4°C for 7, 14, 21, 28, 42 or 98 days and nonhardened seedlings (control) were inoculated by spreading a cube of PDA with mycelium, c. 0·5 × 0·5 × 0·5 cm, onto the soil surrounding each seedling. The pots were then covered with moistened blotting paper and black plastic foil to imitate the conditions occurring under snow cover and kept for a period of 3 weeks at 4°C in darkness. The paper and foil were then removed and the plants cut 4 cm above soil level and allowed to regrow for 10 days under light at 100 ± 10 μmol (quantum) m−2 s−1 PAR, 8/16 h (day/night), at 20/17°C. The frequency of plants that had regenerated shoots on the 10th day after transferring to the optimal growth conditions was used as the indicator of resistance/susceptibility. The values (%) were expressed as a mean from five independent experiments (100 plants).

For microscopic analysis of the degree of fungal growth and spread through leaf tissues, a cube of PDA with mycelium (0·5 × 0·5 × 0·5 cm) was introduced directly onto the upper surface of the second fully expanded leaf of intact triticale seedlings. In order to ensure optimal conditions for infection development, each seedling was covered with a plastic bag and kept in darkness at 4°C for 7 or 10 days. Control leaves were treated with pieces of PDA without mycelium and processed in the same way as the infected plants. A minimum of 20 leaves per treatment was collected for microscopic analysis.

To study cell wall permeability and leaf-adhesion properties, an extract was prepared by homogenizing M. nivale mycelium with redistilled water, then filtrating through syringe filters (0·2 μm). A drop of extract (100 μL) was introduced into a cotton ball, then fastened to the upper surface of the second fully expanded leaf with a plastic strip. Redistilled water replaced the fungal extract in the control. Seedlings were incubated for 24 h in darkness at 4°C and 100% humidity. Again, a minimum of 20 leaves per treatment was collected for microscopic analysis.

Light, fluorescence and scanning electron microscopy

Cell walls, chloroplasts and the accumulation of phenolic compounds were imaged by autofluorescence in living tissue. Microdochium nivale hyphae were visualized with calcofluor white (0·01%) and bromophenol blue (0·01%) in both living tissue and after tissue dehydration in a graded ethanol series: 10 min each in 25% (v/v), 50 and 75%. Nuclei were stained with DAPI (0·01%) after dehydration in the graded ethanol series. The tissues were analysed under UV fluorescence (excitation filter 365 nm, dichroic mirror 395 nm, barrier filter 420 nm) and white light with the Nikon Eclipse E-600 microscope equipped with a digital camera DXM 1200F.

For scanning electron microscopy (SEM), samples were fixed overnight at 4°C with primary fixative containing 2% glutaraldehyde, 2% paraformaldehyde, 0·13 m sucrose and 0·01 m 2-mercaptoethanol in 0·1 m cacodylate buffer (pH 7·4), and then washed four times (30 min each) in 0·1 m cacodylate buffer. Secondary fixation of specimens took place in 2% osmium tetroxide in the same buffer as used in the primary fixative. Samples were washed three times in distilled water for 15 min and dehydrated in a graded ethanol series: 30 min each in 25% (v/v), 50, 70 and 95% and three times (30, 45 min and 1 h) in 100%. Samples were then critical-point-dried and sputter-coated with 15 nm gold in a Cressington 108 auto sputter coater. Samples were observed with a JEOL JSM-50A scanning electron microscope operating at 15 kV.

Adhesion properties of leaf epidermis

Freshly detached leaves of the hardened and unhardened seedlings of both tested cultivars were used for experiments. Additionally, leaves were preincubated with fungal extract and water (as a control) as described earlier. Leaves were mounted on double-sided tape to the edge of a ruler (Fig. S2) and droplets (7 μL) of both redistilled water and fungal extract were placed on the leaf surface. At least five replicate droplets were measured on five leaves taken from different plants and treatments. An optical microscope (Motic SMZ-168) and camera (Moticam 2000) were used for imaging the leaf surfaces and droplets. Both the static contact angle and width-to-height ratio, indicating leaf surface wettability, were calculated from the projected image of the droplet with motic images plus 2.0 ml software. A low contact angle (<100°) of the liquid droplet to the surface is indicative of easy wetting and shows that fluid can spread over a large area of the surface, making the width-to-height ratio >1·3 (hydrophilic surface). Angles over 120° and width-to-height ratios below 1·28 indicate that the surface is hydrophobic and very difficult to wet. A surface is superhydrophobic if it has a water contact angle around 150° (Gaskin et al., 2005; Bhushan & Jung, 2008).

Physicochemical properties of cell wall

The stability of tissue and individual cell wall components was studied by simultaneous differential scanning calorimetry (DSC) and thermogravimetry (TG) combined with quadruple mass spectrometry (QMS) (Netzsch STA409 CD). Twenty-milligram samples of dry powdered leaves were introduced into sealed aluminium pans with lids. The samples were heated at 5°C min−1 to 400°C. Thermograms were recorded against an empty pan placed in the reference vessel. The final thermogram of the sample was obtained after the baseline was subtracted. An ion current was detected at a temperature corresponding to changes of the sample weight. Control peaks corresponding to the degradation of individual cell wall components were determined against the (Sigma–Aldrich) standards: phenols mix (EPA 8270), cellulose fibres (C6288) and lignin (471003) (e.g. Yao et al., 2008). Each experiment was performed for three biological repeats and two technical replicates.

Results

Mycelium growth pattern among genotypes and treatments

Cold-induced resistance to M. nivale was evaluated based on the regrowth ability of unhardened and hardened (7–98 days) seedlings after 3 weeks' incubation with mycelium (Fig. 1, Fig. S1). Unhardened Hewo and Magnat plants suffered 100% mortality after inoculation. The Magnat plants inoculated with M. nivale mycelium were very susceptible and exhibited a lower seedling survival than Hewo plants. Although 42 days of cold pretreatment induced partial resistance of Magnat seedlings (35·0%), 98 days of hardening before inoculation were still not sufficient to result in the regrowth of all seedlings (40·7%) in optimal conditions. The exposure of the Hewo seedlings to cold for longer than 14 days promoted their resistance to infection. Increasing the duration of hardening before inoculation enhanced plant survival (Fig. 1) and green biomass production (Fig. S1) almost linearly. These (Fig. 1) and earlier results (e.g. Golebiowska & Wedzony, 2009) reveal that a period of 28 days is sufficient to distinguish a resistance level between both cultivars. This period enabled Hewo plants to successfully survive infection (89% regrowth), while Magnat plants were still highly susceptible (12% regrowth). Therefore, further experiments were performed for plants hardened for 28 days.

Figure 1.

Effect of cultivar and duration of plant hardening on triticale seedling resistance to Microdochium nivale. Two cultivars of triticale, Magnat (triangles) and Hewo (squares), were subjected to hardening periods of 0, 14, 21, 28, 42 or 98 days at 4°C. Plants were then inoculated with M. nivale mycelium and incubated at 4°C for 21 days. Control (c) noninfected plants were hardened for 28 days and incubated in similar conditions to the infected plants. After incubation plants were cut at 4 cm above soil level and kept in optimal conditions. Plant regrowth data (percentage of plants regenerating shoots) are means of 100 plants from five independent experiments ± SE.

The success of fungi in plant tissue colonization can largely be attributed to their ability to form hyphae. Mycelium easily grew on leaf surfaces of both triticale cultivars, forming numerous thread-like hyphae (Fig. 2). However, the dynamics of fungal growth varied significantly between the cultivars and temperature treatments. The ability of hyphae to rapidly form a random network of mycelium was characteristic of the fungal invasion of cv. Magnat tissue. On the seventh day post-infection (dpi), dense hyphal networks were found on leaf surfaces of both unhardened and hardened Magnat seedlings (Fig. 2a,c). By contrast, although unhardened Hewo epidermis was also covered by dense mycelium (Fig. 2b), only a few fine hyphae were detectable on hardened Hewo leaf surfaces (Fig. 2d). In contrast to the random hyphal pattern of mycelium on the Magnat leaves (Fig. 2a,c), linear hyphal growth between the epidermal cells was favoured on the surface of hardened Hewo leaves (Fig. 2d), which implies that individual fungal hyphae exhibit apical dominance (Harris, 2008).

Figure 2.

Hyphal growth and invasion pattern of Microdochium nivale on triticale leaf surfaces. Two cultivars of triticale, Magnat (a,c,e,g) and Hewo (b,d,f,h), were subjected to a hardening period of 28 days at 4°C (c–h). Control plants were grown at 20°C (a,b). Plant leaves were then inoculated with mycelium, covered with plastic bags and incubated at 4°C, for 7 (a–f) or 10 days (g,h). Analyses were performed with a light microscope (a–d) and under UV light (e–h). For visualization of M. nivale hyphae, bromophenol blue (a–d,f) and calcofluor white (g,h) were used. Images show representative examples of leaves selected from 20 biological repeats. Bars, 20 μm.

Abundant hyphae of M. nivale gained access to plant tissues by direct penetration of plant stomata (Fig. 2a–d). On the other hand, when the infected leaves were viewed with fluorescence microscopy, hardened leaves of cv. Magnat exhibited changes in the surface fluorescence close to the infected area (Fig. 2e), indicating that the hyphae affected cell wall composition very early (7 dpi), i.e. before visible symptoms of tissue degradation occurred. Under UV excitation, leaves normally emit red fluorescence from chlorophyll and blue-green fluorescence from cell wall components. However, the microscopic imaging of Magnat leaves showed that the leaf surface emitted dark-yellow fluorescence (Fig. 2e). The dark-yellow fluorescence next to the fungal hyphae was not detected on Hewo leaf surfaces at the same time (Fig. 2f). When the plants were incubated with mycelium for longer than 1 week, Magnat leaf tissue developed deep lesions (dark areas without the red autofluorescence of the chloroplasts; Fig. 2g). The lesions were adjacent to the tissue areas that had little or no apparent damage. In the resistant, hardened Hewo plants, tested at 10 dpi, the polarized growth of hyphal bundles was accompanied by minor spots of yellow fluorescence in the infection sites (mainly between epidermal cells) and the infected tissue appeared to be morphologically unaffected (the red autofluorescence of the chloroplasts was still detected even below the thick layer of mycelium; Fig. 2h). Through SEM analysis, the three-dimensional aspects of the lesions were more obvious (Fig. 3a,c). Ten days of hyphal growth on the surface of cv. Magnat seedlings was followed by digestion of the outer leaf layer and the loss of epidermal integrity (Fig. 3a). Microdochium nivale hyphae directly penetrated the epidermis (Fig. 3a). By contrast, the surface of Hewo leaves invaded by the fungus seemed to be intact, and the mycelium could easily be separated from the leaf surface, even where a thick layer of mycelium developed (Fig. 3b).

Figure 3.

Growth of Microdochium nivale hyphae in leaf tissues of two cultivars of triticale, Magnat (a,c,f) and Hewo (b,d,e). After a hardening period (28 days at 4°C), leaves were inoculated with mycelium and plants were covered with plastic bags and incubated at 4°C for 10 days (a–e) or 3 days (f). Microscopic analyses were performed with SEM. Images show representative examples of leaves selected from 20 biological repeats. Bars, 10 μm for (a–d).

By 10 dpi, the hyphae progressively penetrated Magnat leaf tissues. The fungus spread through the mesophyll both extra- and intracellularly, causing degeneration of cell walls and protoplasts (Fig. 3c). The host cells became disorganized and eventually collapsed. Penetration of hyphae proceeded in the vascular tissue. A bundle of parallel-oriented hyphae used the vascular tissue as a route for long-distance penetration of the tissue. The vascular elements, including the lignified elements, also became degraded (Fig. 3c). Despite dense mycelium developing on Hewo leaf surfaces (Fig. 3b), hyphae were rarely observed in the apoplast or intracellularly within mesophyll and vascular tissue (Fig. 3d,e). Individual infection hyphae of M. nivale penetrating into the host tissues formed swollen vesicle-like structures resembling haustoria (haustoria-like structures) within the mesophyll cells (Fig. 3e).

First line of defence – physical properties of host epidermal cells

The surface properties of sensitive and resistant leaves can be the reason for altered interactions between the fungus and the host plant at the initial contact area (Figs 2g,h & 3a,b). Submicron- and micron-scale roughness of a leaf determined the formation of a specific mycelium pattern on its surface, differing between susceptible and resistant leaves. On the micron scale, the leaf surface of cv. Hewo revealed a rigid three-dimensional network (Fig. 3b), while the leaf surface of cv. Magnat was relatively smooth (Fig. 3f). On the submicron scale, fine structures were detected on the surface of epidermal cells of both unhardened cultivars (Fig. 4a,b,e,f) and the number of these structures increased greatly when leaves were treated for 28 days with cold (Fig. 4c,d,g,h). The combined micron- (Fig. 3b,f) and submicron- (Fig. 4c,d,g,h) scale roughness can modify the hydrophobic properties of the leaf surface and affect hyphal adhesion (Fig. 3a,b). Hydrophobicity of a surface (wettability) can be characterized by the static contact angle between a water droplet deposited on a leaf and the surface of the leaf (e.g. Bhushan & Jung, 2008). Therefore, a simple quantitative analysis of the contact angle and width-to-height ratio for both redistilled water droplets and fungal extract droplets allowed this study to distinguish the cultivar-dependent hydrophobic properties of hardened and unhardened leaves (Table 1; Fig. S2). Although differences in the adhesion of redistilled water droplets to fresh detached leaves of unhardened cvs Magnat and Hewo were not so obvious (contact angle around 125° and width-to-height ratio around 1·25), the adhesion of droplets containing the fungal homogenate was greatly enhanced on the Magnat leaf surface. Extract droplet contact angles of 97·60° and 122·65° and width-to-height ratios of 1·46 and 1·28 were measured on the unhardened Magnat and Hewo leaf surfaces, respectively (Table 1). After hardening of cv. Hewo, an extremely hydrophobic leaf surface was observed. Both water and fungal extract droplets sat on those leaves in an approximately spherical shape (Table 1; contact angles >140°, width-to-height ratio <1·17), leaving air trapped in the roughness (Fig. S2). It is worth noting that droplet placement and deposition were particularly difficult on hardened Hewo plants because of the droplets rolling off leaf surfaces. The incubation of leaves with fungal extract for a period of 24 h induced cultivar-dependent changes in leaf surface hydrophobicity and droplet adhesion. The contact angles between water droplets and extract-treated leaf surface decreased greatly for unhardened and hardened Magnat, as well as for unhardened Hewo plants. The control incubation for 24 h with water did not change the contact angles significantly. By contrast, hardened Hewo leaf surfaces were still very difficult to wet with water droplets and no variation in the contact angle was present either in the leaves pretreated with fungal extract for 24 h or with water as a control, or in the leaves not incubated at all.

Figure 4.

Effect of a hardening period (28 days at 4°C; c,d,g,h) on the micron- and submicron-scale pattern of the leaf surface of two cultivars of triticale, Magnat (a,c,e,g) and Hewo (b,d,f,h). Control plants were grown at 20°C (a,b,e,f). Plant leaves were then inoculated with mycelium, plants covered with plastic bags and incubated at 4°C, for 7 days. Microscopic analyses were performed with SEM. (e–h) Fragments corresponding to white squares on (a)–(d), respectively. Images show representative examples of leaves selected from 20 biological repeats. Bars, 10 μm for (a–d); 5 μm for (e–h).

Table 1. Genotype-dependent changes in adhesion properties of the triticale leaf surface after cold treatment
PreincubationDropletsM20H20M4H4
  1. Seedlings of two triticale cultivars, Magnat (M) and Hewo (H), were subjected to a hardening period (28 days at 4°C, 8/16 h day/night) (treatments M4 and H4).

  2. Control plants were grown at 20°C (M20 and H20). Leaf surfaces were untreated, or preincubated for 24 h with freshly prepared fungal homogenate or for 24 h with redistilled water (control). Droplets (7 μL) of redistilled water or fungal homogenate were applied to the leaf surface. The contact angle between droplet and leaf surface, as well as droplet width-to-height ratio, are expressed as mean (n = 25) ± SE.

 Contact angle
NoneWater121·99 ± 1·47129·35 ± 9·02130·52 ± 8·16151·28 ± 11·95
NoneFungal extract97·60 ± 4·47122·65 ± 9·18114·68 ± 9·52143·47 ± 10·89
Control (water)Water113·49 ± 3·15130·54 ± 1·11129·28 ± 2·43140·15 ± 4·45
Fungal extractWater99·36 ± 4·16119·68 ± 8·28117·17 ± 8·38140·85 ± 4·65
 Width/height
NoneWater1·29 ± 0·021·20 ± 0·011·20 ± 0·041·14 ± 0·04
NoneFungal extract1·46 ± 0·031·28 ± 0·051·31 ± 0·031·17 ± 0·04
Control (water)Water1·32 ± 0·041·20 ± 0·011·20 ± 0·011·18 ± 0·02
Fungal extractWater1·41 ± 0·031·30 ± 0·031·30 ± 0·021·19 ± 0·02

Fungal products can also influence host cell wall permeability. To test this hypothesis, the kinetics of accumulation of the fluorescent dye, DAPI, were determined in the nuclei of plants which were hardened (0–98 days) and treated with fungal extract (Fig. 5; Fig. S3). DAPI is a stain which binds very easily to AT-rich regions in the DNA and gives strong fluorescence. Because the plant tissues were kept in 75% ethanol until analysis, the properties of the cell wall, rather than plasma membrane permeability, determined the differences in DAPI influx to the cell interior of cvs Hewo and Magnat. Time-course experiments revealed that 10 min after DAPI staining the number of fluorescent nuclei in control leaves (unhardened, not treated with fungal extract) was higher for Magnat tissue (51·2%) than Hewo tissue (18·6%), whereas 1 h of incubation with DAPI resulted in staining of all nuclei in both cultivars (Fig. 5; Fig. S3). DAPI entered much more rapidly into Magnat and Hewo cells when the unhardened leaves had earlier been incubated with the fungal homogenate (92 and 68·6% of cells stained after 10 min, respectively, Fig. 5). A short hardening period (of up to 14 days) did not prevent the enhancement of cell wall permeability by the fungal homogenate in both cultivars and the number of stained nuclei after 10 min was similar with unhardened tissue. However, an almost linear decrease in DAPI diffusion was observed with the prolongation of the plant hardening period. Moreover, 60 min of staining revealed that Hewo cell walls were less permeable to the dye than Magnat cell walls.

Figure 5.

Effect of cold hardening on cell wall permeability of two triticale cultivars, Hewo (squares) and Magnat (triangles). Plants were hardened for 0–98 days at 4°C and then incubated for 24 h with Microdochium nivale fungal extract. Unhardened control plants (c) were incubated with redistilled water. Leaf cell wall permeability to DAPI (percentage of stained cells) was measured 10 min (dashed lines, open symbols) and 60 min (solid lines, filled symbols) after staining. Data are means of 20 biological repeats ± SE.

Effect of chemical composition on cell wall stability

Because of the differences in the cell wall properties between hardened Hewo and Magnat leaves, different patterns of accumulation of secondary cell wall components were expected. Therefore, to determine the cell wall composition, the leaves were examined using thermal analysis under nonisothermal conditions. The analysis was based on an evaluation of the thermal behaviour of individual cell wall compounds in order to obtain data about their stability depending on state transitions during processing (e.g. Yao et al., 2008). Data from the DSC/TA/QMS analyses are presented as: temperature of maximum peaks (°C; indicating the temperature of maximal thermal decomposition of individual cell wall components; Fig. 6a), mass loss (%; TG curve; mass loss during thermal decomposition; Fig. 6b) and total ion current (A; determined for atomic mass 18, corresponding to H2O released from breaking of bonds; Fig. 6c). At linear heating rates under a dynamic, inert gas atmosphere (helium), the DSC signal clearly indicated a two-step degradation between 180°C and 400°C (Fig. 6a). Great differences in cell wall stability were found between cold-hardened and control plants. In the hardened leaves, the first exothermal step occurred at 200–270°C and was related to the release of many organic compounds, which increased the mechanical resistance of the cell wall, e.g. fats, waxes, alkaloids, terpenes, glycosides and phenols. Because the maximum peak of temperature describes the point of highest velocity of the reaction, and the maximum peak of the Hewo cell wall was shifted to higher temperature, these substances were degraded more slowly and had higher stability. The resulting mass changes at a temperature corresponding to the end of the DSC peaks were around 21% for both cultivars (Fig. 6b). Peaks at these temperatures were absent from the DSC curves recorded for unhardened cell walls (Fig. 6a). It indicates that cell walls of hardened plants showed higher amounts of encrusting substances. The presence of organic components, which increased the mechanical resistance of hardened cell walls, was also clearly confirmed by the ion current peak at 230–240°C. Moreover, microscopic analysis showed the localized accumulation of phenolic compounds in the cell wall, as detected by autofluorescence under blue light, and a higher phenolic content was determined spectrophotometrically (Fig. S4). Higher amounts of phenolic substances of hardened cell walls were found in cv. Hewo, in particular.

Figure 6.

Stability and composition of cell wall components of triticale seedlings studied by simultaneous differential scanning calorimetry (a) and thermogravimetry (b) combined with quadruple mass spectrometry (c) for hardened (28 days at 4°C; solid lines) and unhardened (20°C; dashed lines) of cv. Hewo (black lines) and Magnat (grey lines). The ion current corresponds to atomic mass 18 (H2O). Each curve is a mean of three biological repeats and two technical replicates.

The second peak of thermal degradation occurred in the range of 275–350°C for hardened leaves (Fig. 6a). At a similar range of temperature (250–350°C), two peaks on the DSC curve were recorded for unhardened cell walls. This step of the thermal process was accomplished at about 24 and 42% of hardened and unhardened tissue initial mass, respectively (Fig. 6b). The different thermal behaviour of the hardened and unhardened plants pointed to various chemical compositions of the main cell wall polymers. The second peak for the hardened plants (Fig. 6a) corresponded to exothermic cellulose degradation. Because the decomposition process of the Hewo cell wall was shifted to a slightly higher temperature (Fig. 6a), and a lower ion current was derived for gaseous H2O released from bonds breaking during cellulose decomposition to glucose monomers (Fig. 6c), it was concluded that the higher cellulose stability of hardened plants was the result of longer chains of polymers and more regular crystalline-like complexes.

The two-step degradation of unhardened cell walls suggests separate reactions for hemicelluloses and cellulose (Fig. 6a). Hemicelluloses exhibited lower molecular weights than cellulose and were decomposed at temperatures of around 305°C. The lower exothermic effect at the higher temperature during decomposition of hemicelluloses in the Hewo cell wall (Fig. 6a) indicated their different chemical composition and lower amount of easily degraded molecules, when compared to cv. Magnat. Consistent with these results, a higher level of the third DSC peak, corresponding to cellulose, was recorded for Hewo (Fig. 6a). Moreover, the lower mass loss and ion current of total two-step degradation (Fig. 6b,c, respectively) provided evidence that the components of the Hewo cell wall were more thermally stable.

Discussion

The ability of rapidly growing hyphae to generate a mycelium network on the plant surface represents one of the most important, yet least understood aspects of fungal interactions with host organisms (Harris, 2008). There are fungi for which apical growth seems to be a programmed feature associated with rapid hyphal extension and mycelium formation. In most cases hyphal tip cells are engaged in nutrient acquisition and sensing of the local environment, whereas subapical cells generate new hyphae by lateral branching (Harris, 2008). The experiments in the present study indicate that the pattern of mycelium formation by the pathogen M. nivale on the triticale leaf surface can be dependent on the three-dimensional structure of the leaf surface and its physicochemical properties. Random mycelium grew on the whole leaf surface of Magnat seedlings, both unhardened and hardened. When the hardened leaves of the resistant cv. Hewo were infected, the fungus formed unusually straight hyphae, mainly parallel to the long axis of the epidermal cells. The cold-induced rough topography, production of surface contours and higher hydrophobicity of the Hewo leaf surface required the fungus to navigate and coordinate its growth until hyphae reached the appropriate location, i.e. the stomata, whereupon they underwent extensive branching. The formation of intimate nonpathogenic associations between epidermal cells of the Hewo leaf resulted in reduction of total mycelium area and success of colonization. Conversely, the preference of hyphal growth along the host epidermal cells facilitates the reaching of stomata by hyphae, because the stomata are located between epidermal cells, which are themselves aligned in parallel rows. All together, these data demonstrate that specific properties of the resistant leaf surface may act as a trigger for specific hyphal behaviour, known also as thigmotropic growth and thigmodifferentiation (Brand & Gow, 2009).

Previously, it was shown that filamentous fungi grew slowly and exhibited apical dominance in the absence of a prospective host (Akiyama et al., 2005). This pattern changed dramatically in the presence of exudates derived from the host (Harris, 2008). A similar mechanism can be considered for M. nivale. The dense network of hyphae developed on the surface of susceptible Magnat seedlings enhanced nutrient assimilation, whereas the nutrient assimilation could be strongly restricted by specific properties of the Hewo leaf surface forcing the fungus to apical growth. In fact, it was demonstrated in this study that the composition of the leaf surface could determine the availability of nutrients. Despite seedling hardening, the composition of the Magnat cell wall resulted in easy loss of tissue, cells and cell wall integrity under hyphal action, leading to an efficient digestion of the leaf surface by the fungus. Under the same conditions, the Hewo tissue remained intact.

The mechanical action of fungal hyphae on the Magnat leaf surface was aided by fungal chemical products. The hydrophilic properties of the leaf surface and cell wall permeability were originally higher in the susceptible Magnat seedlings and were greatly enhanced when the leaves were treated with fungal homogenate for 24 h. The surfaces of the most hardened Hewo leaves were extremely difficult to wet, nonpermeable and appeared unaffected after treatment with fungal extract. In hardened plants, organic substances affecting the hydrophobic and electric properties of leaf surface, such as fats, waxes, cutins, alkaloids, terpenes, glycosides and phenols, were detected by thermal analysis, autofluorescence and spectrophotometry. These compounds are waterproof, increasing surface hydrophobicity, and helping to block the entry of pathogenic fungi (Gaskin et al., 2005; Bhushan & Jung, 2008; Teisala et al., 2011). Hence, data indicating a higher amount and/or stability of hydrophobic molecules in the hardened cv. Hewo can be linked to greater resistance of the leaf surface against penetration by fungal hyphae and fungal substances. Furthermore, submicron- and micron-scale roughness of leaf surfaces can increase their hydrophobicity significantly (Bhushan & Jung, 2008). On the micron scale, Hewo leaf surfaces showed a highly rigid three-dimensional network, while those of Magnat were flatter. On the submicron scale, lower adhesion to hardened Hewo leaves may arise as a result of fine structures on the surface. Rough surfaces that possess micron-scale irregularities combined with submicron-scale fine structures entrap air so that water cannot fill the space between the structures. This causes superhydrophobicity (Bhushan & Jung, 2008) and results in the lower adhesion of both water droplets and fungal hyphae to these surfaces (Teisala et al., 2011). Moreover, superhydrophobicity efficiently prevents fungal growth and development because M. nivale is extremely sensitive to a lack of moisture (Xu et al., 2008).

When the fungal pathogen enters the plant tissue, it almost always occupies the extracellular niches at the beginning of the colonization process. Despite this fact, the nutrients that enable pathogen growth must be derived from the host cell interior and the progress of M. nivale infection can proceed by necrotrophic invasion (Golebiowska & Wedzony, 2009). Lethal fungal invasion in cv. Magnat leaves resulted from the fungus switching to necrotrophic growth in less than 3 dpi. The penetration of the hyphae into the mesophyll and vascular tissue was executed via total disintegration of the mesophyll cells and partial digestion of the vessels. Hyphae can also grow within xylem and sieve elements, causing embolism of the water transport system, and, finally resulting in rapid plant death (Perfect & Green, 2001). However, the present study provides cytological evidence that M. nivale adopts a uniquely biotrophic infection strategy in resistant plants (cv. Hewo) for a prolonged period and spreads without causing loss of host cell viability. Despite very dense mycelium on the leaf surface, only individual infection hyphae of M. nivale penetrating into the host tissues via the stomata and swollen, vesicle-like structures resembling haustoria within the mesophyll cells were visualized. Haustoria can only develop during direct plant–pathogen interactions and are formed from both fungal and host components (Mendgen & Hahn, 2002); therefore further confirmation of the nature of the demonstrated structures is needed. The introduction of an individual hypha between mesophyll cells and its use to transport nutrients seems to be a simple and energy-saving mechanism for the mycelium to survive on the surface of a resistant seedling.

The involvement of enzymes weakening the cell wall, e.g. cellulases, xylanases and pectinases during infection and colonization of wheat spikes by Fusarium culmorum, F. graminearum, F. avenaceum and M. nivale, has been reported earlier (Kang & Buchenauer, 2000; Wanyoike et al., 2002; Kang et al., 2005). This paper presents new findings concerning cell wall resistance to fungal digestion within different triticale cultivars. The effectiveness of the intracellular penetration of fungal hyphae appeared to be related to the structure and physicochemical reinforcement of cells located in the deeper layers of leaf tissue. In Hewo seedlings, cold treatment resulted in the accumulation of phenolic compounds, which are known to inhibit disease development through mechanisms such as inhibition of extracellular fungal enzymes and antioxidant activity controlling the negative effect of ROS during pathogen-generated mechanical injuries (Dixon & Paiva, 1995). Moreover, some phenols are used as substrates for the production of lignin and suberin, which, when incorporated into a cell wall, lower its permeability and make it more resistant to both the enzymatic digestion and mechanical pressure applied during penetration by fungal hyphae.

Other mechanisms can be involved in the protection of cell wall integrity as well as mechanical and chemical stability. Spatially and temporally controlled heterogeneity in the chain length of cellulose and its degree of crystallinity can modify cell wall resistance of different plants, organs, tissues and individual cells (Burton et al., 2010). It was also shown that different cells have specific patterns of cellulose orientation (Verbelen et al., 2001), determining its mechanical properties (Kerstens et al., 2001). However, definitive proof of the importance of structural and chemical strengthening of the cellulose network in penetration resistance is rare in the literature. This study provides evidence for this assumption, showing the reduced thermal degradation of celluloses in the hardened Hewo cell wall, expressed as lower exothermal peaks of mass loss, lower energetic effects and a lower ion current for gaseous H2O. The higher thermal stability of cellulose indicates that the Hewo cell wall consists of tightly compacted aggregates of long cellulose chains, stabilized through the formation of intermolecular hydrogen bonding and hydrophobic interactions between the sugar rings. This will reduce their solubility and enhance their ability to form a crystal-like configuration. Such a crystalline structure resists thermal decomposition better than the structure of the Magnat cell wall, which is abundant in hemicelluloses but has less cellulose. Until now, the resistance of A. thaliana to fungal infection was attributed to activity of the enzyme CESA3, an essential enzyme for primary cell wall cellulose synthesis and celllose quantity (Ellis et al., 2002). However, cell wall compaction can determine host resistance during the pathogen invasion. Reduced turgor caused by pathogen-induced damage to the wall might activate a cell wall integrity-sensing system and trigger resistance (Vorwerk et al., 2004).

In conclusion, cold can induce genotype-dependent alterations in the cell walls of the leaf epidermis, mesophyll and vascular tissue, including enhancement of the content of small lipid-like substances, stabilization of the cell wall cellulose/hemicellulose matrix and establishment of specific physical and structural properties in disease resistant cultivars. These factors determine the strategy of fungal invasion and, in consequence, the success of colonization.

Acknowledgements

This work was supported by the project 595/N-COST/2009/0 and partly by the Welcome/2008/1 Program operated within the framework of the Foundation for Polish Science, co-financed by the European Regional Development Fund and by project NCN N N310 778640.

Ancillary