Dynamics of cell wall components of Magnaporthe grisea during infectious structure development

Authors


*E-mail marie@affrc.go.jp; Tel. (+81) 29 838 8461; Fax (+81) 29 838 7408.

Summary

Oligosaccharides derived from cell wall of fungal pathogens induce host primary immune responses. To understand fungal strategies circumventing the host plant immune responses, cell wall polysaccharide localization was investigated using fluorescent labels during infectious structure differentiation in the rice blast fungus Magnaporthe grisea. α-1,3-glucan was labelled only on appressoria developing on plastic surfaces, whereas it was detected on both germ tubes and appressoria on plant surfaces. Chitin, chitosan and β-1,3-glucan were detected on germ tubes and appressoria regardless of the substrate. Major polysaccharides labelled at accessible surface of infectious hyphae were α-1,3-glucan and chitosan, but after enzymatic digestion of α-1,3-glucan, β-1,3-glucan and chitin became detectable. Immunoelectron microscopic analysis showed α-1,3-glucan and β-1,3-glucan intermixed in the cell wall of infectious hyphae; however, α-1,3-glucan tended to be distributed farther from the fungal cell membrane. The fungal cell wall became more tolerant to chitinase digestion upon accumulation of α-1,3-glucan. Accumulation of α-1,3-glucan was dependent on the Mps1 MAP kinase pathway, which was activated by a plant wax derivative, 1,16-hexadecanediol. Taken together, α-1,3-glucan spatially and functionally masks β-1,3-glucan and chitin in the cell wall of infectious hyphae. Thus, a dynamic change of composition of cell wall polysaccharides occurs during plant infection in M. grisea.

Introduction

Magnaporthe grisea is a model plant pathogen for the study of plant–microbe interactions. This fungus is also the causal agent of rice blast disease, leading to enormous economic damage in rice production. Infection with M. grisea starts when conidia attach to plant surfaces. Upon recognition of the plant surface, the fungus germinates and develops an infection-specific structure called an appressorium from the tip of the germ tube (Howard and Valent, 1996). Previous studies have revealed that appressorium formation is induced by thigmotropic or chemical cues (e.g. plant wax derivatives) from the attachment surface (Uchiyama et al., 1979; Lee and Dean, 1994; Gilbert et al., 1996). Thigmotropic cue recognition signals are transmitted by heterotrimeric G-proteins, consisting of α, β and γ subunits, to intracellular effectors such as adenylate cyclase (Nishimura et al., 2003; Liu et al., 2007). cAMP signalling is essential for appressorium formation, and accumulation of intracellular cAMP or exogenous cAMP triggers appressorium differentiation (Lee and Dean, 1993; Choi and Dean, 1997; Nishimura et al., 2003; Liu et al., 2007). A penetration peg, developing from the appressorium, expands to form a primary hypha, which differentiates into a bulbous secondary infectious hypha in susceptible plant epidermal cells (Kankanala et al., 2007).

The cell wall is the outermost layer exposed to surrounding environments and is essential to maintain cellular structure and integrity in fungi. The major components of fungal cell walls are polysaccharides: β-glucans (β-1,3-glucan and β-1,6-glucan), α-glucans (α-1,3-glucan and α-1,4-glucan), chitin and mannan (Perez and Ribas, 2004; Latge, 2007). The importance of cell wall components in host–parasite interactions has been studied in both mammalian and plant pathogenic fungi. Cell walls are major sources of microbe-associated molecular patterns (MAMPs) that are recognized by host mammalian and plant cells. Fungal MAMPs include β-glucans, mannans and chitins, which induce innate immune responses in mammalian host cells (Poulain and Jouault, 2004; Theis and Stahl, 2004; Brown and Gordon, 2005; Reese et al., 2007). In mammalian immune systems, pattern-recognition receptors specific for β-glucans and mannan have been identified on phagocytic cells (Marodi et al., 1993; Brown and Gordon, 2005). Fungal cell wall β-glucans and chitins have also been reported as MAMPs that evoke primary immune responses in plant cells (Ryan and Farmer, 1991; Felix et al., 1993; Zhang et al., 2002; Altenbach and Robatzek, 2007; de Wit, 2007). β-glucan binding proteins have been identified as components of a proposed receptor complex that detects branched β-glucan derivatives from oomycetes in soybean (Glycine max), French bean (Phaseolus vulgaris) and barrel medic (Medicago truncatula) (Mithofer et al., 1996; 1999; Leclercq et al., 2008). Recently, a pattern-recognition receptor for chitin oligosaccharides was identified in rice (Oryza sativa) (Kaku et al., 2006).

It has been assumed that pathogenic fungi have developed mechanisms to avoid host recognition or to protect themselves from host immune attacks. Recently, Rappleye et al. (2007) has presented evidence that α-1,3-glucan masks β-glucan on the cell wall of Histoplasma capsulatum to block host recognition of the fungal invasion. A correlation between reduction in α-1,3-glucan and virulence has also been observed in other dimorphic fungal pathogens, such as Paracoccidioides brasiliensis and Blastomyces dermatitidis (San-Blas et al., 1977; Hogan and Klein, 1994; Rappleye and Goldman, 2006). Similarly to the case of H. capsulatum, it has been assumed that the fungal α-1,3-glucan conceals cell wall MAMPs on the fungal cell surface to circumvent recognition by hosts (Hogan et al., 1996; Rappleye and Goldman, 2006). In the plant pathogenic fungi Puccinia graminis f. sp. tritici, Uromyces fabae and Colletotrichum graminicola, chitosan is exposed on cell wall surfaces of infectious structures that differentiate in planta (El Gueddari et al., 2002). Conversion of chitin to chitosan by de-N-acetylation is postulated to be a fungal strategy to protect the cell wall from enzymatic hydrolysis by plant chitinases (El Gueddari et al., 2002).

In this study, we investigated localization of major cell wall polysaccharides during development of infectious structures in M. grisea. Our cytological analysis using fluorescent labels demonstrated infectious structure-specific localization of cell wall polysaccharides. Major polysaccharides detected at label-accessible surface of infectious hyphae were α-1,3-glucan and chitosan; β-1,3-glucan and chitin became detectable after enzymatic digestion of α-1,3-glucan. Immunoelectron microscopic observation of infectious hyphae revealed that α-1,3-glucan was localized more outwardly in the cell wall compared with β-1,3-glucan. The accumulated α-1,3-glucan interfered with the digestion of chitin in the cell wall by chitinase. Taken together, α-1,3-glucan masks β-1,3-glucan and chitin in the cell wall of infectious hyphae, suggesting that α-1,3-glucan protects the fungal cell wall from digestive enzymes produced by plant cells during infection. A plant wax derivative induced α-1,3-glucan accumulation via activation of the Mps1 signal transduction pathway. Our results demonstrate that M. grisea dynamically changes the composition of cell wall polysaccharides upon recognition of a plant cue. We discuss fungal mechanisms to circumvent host plant recognition by dynamically changing its cell wall components.

Results

Phylogenetic analyses of α-1,3-glucan synthases, β-1,3-glucan synthases and chitin synthases from fungi

We identified a single gene encoding α-1,3-glucan synthase in the M. grisea genome, and designated it as MgAGS1. The results of phylogenetic analysis using the ClustalW program showed that MgAgs1p is more closely related to α-1,3-glucan synthases in Neurospora crassa than those in Schizosaccharomyces pombe or Aspergillus species (Fig. S1A). A gene encoding a putative β-1,3-glucan synthase in M. grisea, MgFKS1, is also a single gene in the fungal genome. Phylogenetic analysis showed that MgFks1p is closely related to β-1,3-glucan synthases in N. crassa (Gls1p) and Cryptococcus neoformans (Fig. S1B). Seven chitin synthase genes were identified in the M. grisea genome (MgCHS1, MgCHS2, MgCHS3, MgCHS4, MgCHS6, MgCHS7 and MgCSM1). As shown in Fig. S1C, phylogenetic analysis showed that these genes can be classified into the seven known classes of fungal chitin synthases (Roncero, 2002; Martin-Urdiroz et al., 2008).

Localization of cell wall polysaccharides at label-accessible surface of germ tubes and appressoria developing on cover glass

Appressorium formation of M. grisea is induced on some artificial substrates such as plastic coverslips and cover glass as well as on plant surfaces. We observed localization of major cell wall components in infectious structures (germ tubes, appressoria) developing on cover glass (inductive surfaces) 24 h after the start of incubation for wild-type M. grisea. We used fluorophore-labelled antibodies to detect α-1,3-glucan and β-1,3-glucan, fluorophore-labelled lectins to detect chitin and mannan, and a fluorescent dye to detect chitosan. As shown in Fig. 1A, α-1,3-glucan was label-accessible only on appressoria, while β-1,3-glucan, chitin, chitosan and mannan were clearly labelled on both appressoria and germ tubes (Fig. 1B–E), showing infectious structure-specific localization of each cell wall polysaccharide. We also confirmed that the infectious structure-specific localization of cell wall polysaccharides is not altered on cover glass with prolonged incubation up to 48 h (data not shown).

Figure 1.

Localization of cell wall polysaccharides on infectious structures developed on cover glass. Localization of cell wall polysaccharides on infectious structures developing on cover glass in the absence/presence of exogenous plant wax derivative (1,16-hexadecanediol) was observed at 24 h post incubation. α-1,3-glucan and β-1,3-glucan were detected by fluorophore-labelled antibodies, chitin and mannan by fluorophore-labelled lectins, and chitosan by the fluorescent dye eosin. Images were taken using bright-field optics (left panels of each pair) and epifluorescent optics (right panels of each pair); (A) and (F) α-1,3-glucan; (B) and (G) β-1,3-glucan; (C) and (H) chitin; (D) and (I) chitosan and (E) and (J) mannan. Scale bar = 20 μm. More than 200 samples were observed for each polysaccharide and typical images are presented.

Transcription of cell wall polysaccharide synthase genes during appressorium development on plastic surfaces

We examined the expression patterns of genes related to synthesis of cell wall polysaccharides: α-1,3-glucan synthase gene MgAGS1, β-1,3-glucan synthase gene MgFKS1 and seven chitin synthase genes (MgCHS1, MgCHS2, MgCHS3, MgCHS4, MgCHS6, MgCHS7 and MgCSM1). MgAGS1 and MgFKS1 are single genes, whereas chitin synthase genes are redundant in the M. grisea genome. We analysed the transcriptional levels of these genes during appressorium development on plastic surfaces by quantitative reverse-transcription PCR (qRT-PCR) analysis. On plastic surfaces, germination from conidia was observed < 2 h after the start of incubation, germ tube extension at < 4 h, formation of small incipient appressoria at approximately 7 h and formation of appressoria at > 10 h (Fig. S2A); appressoria were mature and fully melanized at 24 h after the start of incubation (data not shown). MgAGS1 expression levels were temporally elevated at 7 h after the start of incubation and declined by 24 h to levels similar at 0 h (Fig. S2B). Thus, MgAGS1 expression was induced, especially in the early stage of appressorium development. In contrast, MgFKS1 expression levels were almost constant throughout appressorium development (Fig. S2C). The expression of seven chitin synthase genes was temporally overlapped during appressorium development (Fig. S2D–J); however, the levels were very low compared with those of MgAGS1 and MgFKS1. The gene expression profiles obtained from the qRT-PCR analyses (Figs. S2B–J) were consistent with our above cytological observations, which indicated non-specific localization of chitin and β-1,3-glucan in the appressoria and germ tubes, and appressorium-specific localization of α-1,3-glucan (Fig. 1A–C).

Localization of cell wall polysaccharides at label-accessible surface of infectious structures developing on plant surfaces and in planta

When M. grisea invades plant cells, a penetration peg develops from an appressorium and differentiates to infectious hyphae (Howard and Valent, 1996). To investigate localization of cell wall polysaccharides on the infectious structures developing in plant cells, rice sheath cells from the susceptible cultivar LTH were inoculated with fungal conidia for up to 24 h. We first observed infectious hyphae developing in susceptible rice cells at 24 h post inoculation (hpi) without fixation. α-1,3-glucan was label-accessible on the infectious hyphae, but β-1,3-glucan or chitin were not (Fig. S3). As this result could have been due to insufficient staining of the fungal cell wall in the rice cells, we fixed the samples with formaldehyde for further observation. We observed germ tubes and appressoria at 16 hpi and infectious hyphae at 24 hpi. α-1,3-glucan was antibody-accessible on germ tubes and appressoria (Fig. 2A), as well as on infectious hyphae (Fig. 2B). β-1,3-glucan was antibody-accessible on appressoria (Fig. 2A), but not on germ tubes or infectious hyphae (Fig. 2B). Chitin was clearly labelled on germ tubes and appressoria (Fig. 2A), but hardly on infectious hyphae (Fig. 2B). Chitosan was detected on germ tubes, appressoria and infectious hyphae (Fig. 2C). Our cytological observations indicate that the composition or the label accessibility of cell wall polysaccharides of each fungal infectious structure (germ tube, appressorium and infectious hypha) is significantly different when grown on plant surfaces and cover glass.

Figure 2.

Localization of cell wall polysaccharides on infectious structures developing on rice leaf sheath cells. Leaf sheaths of susceptible of rice cultivar LTH were inoculated with wild-type M. grisea. The fungal infectious structures were stained with fluorophore-labelled antibodies to detect α-1,3-glucan and β-1,3-glucan, with fluorophore-labelled lectins to detect chitin and mannan, and with the fluorescent dye eosin to detect chitosan. Samples were prefixed by formaldehyde before observation. More than 300 samples were observed for each time stage and representative images are presented.
A. Localization of cell wall polysaccharides at 16 hpi.
B. Localization of cell wall polysaccharides at 24 hpi.
C. Localization of chitosan at 24 hpi.
D. Detection of cell wall polysaccharides on infectious hyphae after enzymatic digestion of α-1,3-glucan. The formaldehyde-fixed samples were treated with α-1,3-glucanase. Images were taken at 24 hpi. More than 50 samples were observed and representative images are presented. BF, bright-field optics; Ap, appressorium; Co, conidium; Gt, germ tube; Ih, infectious hypha. Scale bar = 20 μm.

Transcription of cell wall synthase genes during infectious growth in rice cells

We analysed the transcriptional levels of cell wall polysaccharide synthase genes during infectious growth in planta by qRT-PCR. Total RNA for qRT-PCR analysis was extracted from rice sheath cells at 24 or 48 hpi of M. grisea conidia. We confirmed that infectious hyphae had developed in the rice sheath cells at 24 hpi, and had vigorously extended at 48 hpi (data not shown). The expression level of MgAGS1 was significantly higher at 48 hpi than at 24 hpi (P < 0.001), indicating that expression of MgAGS1 increases during infectious growth (Fig. 3A). Interestingly, in contrast to MgAGS1, the transcription level of MgFKS1 at 48 hpi was markedly lower than at 24 hpi (P < 0.01), indicating that expression of MgFKS1 decreases sharply over time (Fig. 3B). The transcriptional levels of four of the seven chitin synthase genes, MgCHS1, MgCHS4, MgCHS7 and MgCSM1, were very low compared with those of MgFKS1 and MgAGS1, while the chitin synthase gene, MgCHS2, was expressed continuously during infectious growth (P <  0.05, Fig. S4). No transcription of these fungal genes was observed in total RNA extracted from uninoculated rice sheath cells (at time 0 in Fig. 3A and B and Fig. S4).

Figure 3.

Expression of genes encoding synthases for α-1,3-glucan and β-1,3-glucan during development of infectious hyphae in rice cells. Leaf sheaths of susceptible rice cultivar LTH were inoculated with wild-type M. grisea. Expression levels of synthase genes for α-1,3-glucan and β-1,3-glucan were analysed by qRT-PCR at 0, 24 and 48 hpi, and normalized to the actin mRNA.
A. Expression of the α-1,3-glucan synthase gene, MgAGS1.
B. Expression of the β-1,3-glucan synthase gene MgFKS1. Asterisks on bars indicate significant differences in the gene expression levels compared with those at 0 h; *< 0.001, **P < 0.01.

Induction of α-1,3-glucan by exogenous plant wax derivative during development of infectious structures

In M. grisea, 1,16-hexadecanediol, which is derived from plant wax esters by hydrolysis, induces appressorium formation under non-inductive conditions (Gilbert et al., 1996). Hence, plant wax is regarded as one of the cues by which M. grisea recognizes plant surfaces (Talbot, 2003). As α-1,3-glucan was detected on the germ tubes developing on plant surfaces but not on cover glass (Figs 1A and 2A), we hypothesized that plant wax induces α-1,3-glucan accumulation at the accessible surface of cell wall of the germ tubes. As expected, accumulation of α-1,3-glucan was evident on both appressoria and germ tubes in the presence of exogenous 1,16-hexadecanediol (Fig. 1F). Importantly, this localization pattern of α-1,3-glucan was similar to that observed on plant surfaces at 16 hpi (Fig. 2A). Note that 1,16-hexadecanediol induced no effect on the infectious structure-specific localization of other cell wall polysaccharides, such as β-1,3-glucan, chitin, chitosan and mannan (Fig. 1G–J).

Localization of α-1,3-glucan in a deletion mutant of heterotrimeric G-protein β subunit

In M. grisea, heterotrimeric G-protein signalling is involved in surface recognition, which induces appressorium formation via activation of cAMP signalling (Liu and Dean, 1997; Fang and Dean, 2000; Nishimura et al., 2003; Liu et al., 2007). A disruption mutant of a heterotrimeric G-protein β subunit, mgb1, is deficient in appressorium formation on inductive (plastic or plant) surfaces and in invasive growth (Nishimura et al., 2003). As exogenous 1,16-hexadecanediol fails to differentiate appressoria from mgb1 conidia on inductive surfaces (M. Nishimura, unpubl. data), we investigated whether the 1,16-hexadecanediol-dependent accumulation of α-1,3-glucan on germ tubes requires the G-protein signal transduction pathway. In mgb1, accumulation of α-1,3-glucan was induced on germ tubes on cover glass in the presence, but not in the absence, of exogenous 1,16-hexadecanediol (Fig. 4D and E), as in the wild type (Fig. 4A and B). Consistently, exogenous 1,16-hexadecanediol did not alter the infectious structure-specific localization pattern of β-1,3-glucan, chitin, chitosan and mannan on cover glass in mgb1 (Fig. S5). In addition, exogenous cAMP, which is known to induce appressorium formation on non-inductive surfaces (Lee and Dean, 1993), showed no induction of α-1,3-glucan accumulation on germ tubes (Fig. 4C) or effect on the structure-specific localization pattern of β-1,3-glucan, chitin, chitosan and mannan (Fig. S5) of the wild type on cover glass. Over all, G-protein signalling does not appear to be involved in the 1,16-hexadecanediol-dependent accumulation of α-1,3-glucan.

Figure 4.

Accumulation of α-1,3-glucan in a deletion mutant of heterotrimeric G-protein β subunit (mgb1) and mutants related to Mps1 MAP kinase pathway. Localization of α-1,3-glucan in germ tubes and appressoria developing on cover glass were detected by a fluorophore-labelled antibody. Left panels of each pair are images taken using bright-field optics, and right panels are images taken using epifluorescent optics. More than 200 samples were observed for each mutant and typical images are presented.
A. Wild-type conidia incubated on cover glass for 24 h.
B. Wild-type conidia incubated on cover glass for 24 h in the presence of 50 μM 1,16-hexadecanediol.
C. Wild-type germ tube developed on cover glass in the presence of 10 mM exogenous cAMP.
D. mgb1 germ tube developing on cover glass.
E. mgb1 germ tube developing on cover glass in presence of 50 μM exogenous 1,16-hexadecanediol.
F. mps1 incubated on cover glass for 24 h.
G. mps1 incubated on cover glass in presence of 50 μM 1,16-hexadecanediol.
H. mkk1DD, in which MAPKK for the Mps1-mediated pathway is constitutively activated, incubated on cover glass for 24 h. Scale bar = 20 μm.

Accumulation of α-1,3-glucan is dependent on Mps1 MAP kinase

The Mps1 MAP kinase in M. grisea is an orthologue of MpkAp in Aspergillus nidulans and Slt2p/Mpk1p in Saccharomyces cerevisiae, which are key enzymes for cell wall integrity (CWI) (Xu et al., 1998; Levin, 2005; Fujioka et al., 2007). In the mps1 deletion mutant, a drastic reduction in aerial hyphal growth and hypersensitivity to fungal cell wall-digesting enzymes has been observed (Xu et al., 1998). In addition, mps1 fails to differentiate infectious hyphae in live plant cells, although development of infectious hyphae has been observed in heat-killed rice sheath cells (Park et al., 2004; Mehrabi et al., 2008). As Mps1p may play a role the maintenance of CWI, we investigated whether cell wall polysaccharide accumulation is Mps1p-dependent. In the mps1 mutant, no α-1,3-glucan was detected on the accessible-surface of germ tubes and appressoria (Fig. 4F). Consistently, suppression of MgAGS1 transcription was confirmed in mps1 (data not shown). Even in the presence of exogenous 1,16-hexadecanediol, mps1 failed to accumulate α-1,3-glucan on the infectious structures (Fig. 4G). On the other hand, β-1,3-glucan, chitin, chitosan and mannan were detected regardless of the presence of 1,16-hexadecanediol (Fig. S5).

Saccharomyces cerevisiae and S. pombe possess two MAPKKs that are redundant in function and are involved in CWI signalling, while M. grisea possesses a single MAPKK (Mkk1p) that is involved in the Mps1 MAP kinase cascade. Skh1p is the MAPKK involved in the CWI signal transduction pathway of S. pombe. Mutations in Skh1p (S234D and T238D) constitutively activate the downstream Spm1 MAP kinase (Loewith et al., 2000). As Ser234 and Thr238 in Skh1p correspond to Thr369 and Thr375 in Mkk1p, we generated a mutant (mkk1DD) by replacing the endogenous MKK1 in the wild-type strain with MKK1T369D,T375D to form a constitutively active Mps1p (Fig. 5A and B); phosphorylation of Mps1p was confirmed by Western blot analysis. As shown in Fig. 5C, the intensity of the signal corresponding to phosphorylated Mps1p, detected by binding of an antiphospho-p44/42 antibody, was increased in mkk1DD compared with that in wild type. When the same membrane was stripped and then incubated with an anti-p44/42 antibody, the intensity of the signal corresponding to total Mps1p in mkk1DD was comparable to that in wild type. No Mps1p was detected in mps1 (Fig. 5C). Therefore, we concluded Mps1p is constitutively activated in mkk1DD.

Figure 5.

Construction of mkk1DD mutant.
A. Mutation sites in Mkk1p. White arrow indicates the MKK1 gene and the three black boxes indicate exons. Amino acid and DNA sequences of the mutated sites (the underlined text) in MKK1 are indicated at the bottom. The threonine residues at 369 and 375 in Mkk1p were mutated to aspartic acid in MKK1DD. Primers used in this study are indicated by arrows and arrowheads.
B. Replacement of MKK1 with MKK1DD. The replacement of MKK1 with MKK1DD was confirmed by Southern hybridization analysis using probes 1 and 2 described in Experimental procedures. Genomic DNA extracted from mycelia of wild-type (Guy11) and mkk1DD strains was digested with HindIII. Probe 1, amplified by Int-F/Mkk1-R2 from pGeMKK1, hybridized to both MKK1 and MKK1DD (left panel). Probe 2, amplified with Int-F/DD-R from pGeMKK1DD, showed specific hybridization to MKK1DD (right panel).
C. Phosphorylation of Mps1p in the wild-type (Guy11), mkk1DD and mps1 strains. Proteins were extracted from submerged cultures in YG medium. Protein (50 μg) was subjected to SDS-PAGE and blotted onto PVDF membranes for Western blot analysis. Mps1p was detected by a p44/42 MAP kinase antibody (indicated by p44). Phosphorylated Mps1p was detected by a phospho-p44/42 MAP kinase (Thr202/Tyr204) antibody (indicated by Phospho-p44). Note that Mps1p in mkk1DD was constitutively phosphorylated.
D. Phosphorylation of Mps1p in the wild-type strain Guy11 in the presence of a plant wax derivative. Protein was prepared from Guy11 mycelia cultured in YG media with 50 μM 1,16-hexadecanediol for 0, 20 and 40 min. Signal intensity of phosphorylated Mps1p increased over time (< 40 min).

We found that the accumulation of α-1,3-glucan was observed not only on appressoria, but also on germ tubes developing on cover glass in mkk1DD (Fig. 4H). In addition, we confirmed the phosphorylation of Mps1p in the presence of 1,16-hexadecanediol (Fig. 5D). Taken together, our results demonstrate that accumulation of α-1,3-glucan in the fungal cell wall is regulated by the Mps1 MAP kinase pathway, which can be activated by 1,16-hexadecanediol.

Accumulation of α-1,3-glucan blocks enzymatic digestion of chitin in the fungal cell wall

During infection, the fungal cell wall is attacked by cell wall-digesting enzymes, e.g. β-1,3-glucanases and chitinases, which are produced by plant cells as an immune response (van Loon et al., 2006). As α-1,3-glucan was specifically induced during plant infection, we tested whether the accumulated α-1,3-glucan contributes in protection of β-1,3-glucan or chitin in the cell wall from digestive enzymes during infectious structure development in M. grisea. When mycelia were digested with chitinase alone for 24 h, more spheroplasts (1.36 ± 0.50 × 106 ml−1) were generated from fresh mycelia of mps1 compared with wild-type mycelia (0.53 ± 0.18 × 106 ml−1). In contrast, only a few spheroplasts (0.13 ± 0.05 × 106 ml−1) were generated from mkk1DD mycelia (Fig. 6A). When mycelia were codigested with chitinase and α-1,3-glucanase, the number of spheroplasts generated from mycelia of the wild type or mkk1DD was significantly increased compared with those after digestion with chitinase alone; the increase was more significant in mkk1DD (Fig. 6A, right and left panels). In mps1, the number of spheroplasts generated after codigestion with chitinase and α-1,3-glucanase (1.79 ±  0.23 × 106 ml−1) was comparable to that generated after digestion with chitinase alone (Fig. 6A, centre panel). These results suggest that lack of α-1,3-glucan increased sensitivity of the cell wall of mycelia to chitinase. In contrast, accumulation of α-1,3-glucan appears to increased tolerance of the fungal cell wall to chitinase. We next examined whether accumulation of α-1,3-glucan on germ tubes increases tolerance of the cell wall to chitinase digestion. In the wild-type strain, approximately 30% of germ tubes were lysed after chitinase digestion, while > 70% were lysed after codigestion with chitinase and α-1,3-glucanase (Fig. 6B, WT). Importantly, in the presence of 1,16-hexadecanediol, which induces α-1,3-glucan accumulation, the fungal cell wall was more tolerant to chitinase digestion; > 90% of the germ tubes were not lysed by chitinase alone (Fig. 6B, WT + wax). Even after the codigestion with chitinase and α-1,3-glucanase, approximately 55% of the germ tubes were still non-lysed (Fig. 6B, WT + wax). Similarly, mkk1DD germ tubes were more tolerant to chitinase digestion, even in the presence of α-1,3-glucanase, than the wild-type strain (Fig. 6B, WT, mkk1DD). In contrast, mps1 germ tubes, which lack α-1,3-glucan, were very susceptible to chitinase digestion compared with the wild-type strain. Moreover, the addition of exogenous 1,16-hexadecanediol showed no increase in tolerance of mps1 germ tubes against chitinase digestion (Fig. 6B, mps1, mps1 + wax). Taken together, the accumulation of α-1,3-glucan correlates with the increase in tolerance of cell-wall chitin to chitinase digestion, suggesting its role in protection of the fungal cell wall against digestive enzymes secreted by plants during infection.

Figure 6.

Cell wall susceptibility assay against chitinase digestion in wild-type, mps1 and mkk1DD strains.
A. Effectiveness of α-1,3-glucanase on chitinase treatment of fungal mycelia. Four-day-old mycelia (1 μg) was digested in reaction buffer containing 15 μg chitinase alone or both 10 μg α-1,3-glucanase and 15 μg chitinase. After 0, 4 and 24 h incubation at 30°C, the number of spheroplasts was determined by a hemocytometer. Control samples were incubated in reaction buffer containing no cell wall-digesting enzymes. Results represent the mean ± SD of triplicate experiments.
B. Effect of α-1,3-glucan accumulation on the cell walls of germ tubes on their tolerance against chitinase digestion. Conidia of the wild-type (Guy11), mps1 and mkk1DD strains were incubated on plastic coverslips with or without 50 mM 1,16-hexadecandiol for 16 h. Appressorium-forming conidia were incubated with reaction buffer containing 15 μg chitinase for 4 h. To examine effectiveness of adding α-1,3-glucanase on chitinase treatment, appressorium-forming conidia were coincubated with 10 μg α-1,3-glucanase and 15 μg chitinase. Control samples were incubated in reaction buffer containing no enzymes for 4 h. Results are presented as the percentage of germ tubes that were digested after 4 h incubation. Results represent the mean ± SD of three experiments.

Spatial localization of α-1,3-glucan in the cell wall of infectious hyphae

It has been reported that α-1,3-glucan masks the surface of the cell wall of H. capsulatum and blocks host recognition of fungal invasion by MAMP β-1,3-glucan (Rappleye et al., 2007). In M. grisea, accumulation of α-1,3-glucan increases tolerance of the fungal cell wall to chitinase digestion (Fig. 6A and B). In addition, α-1,3-glucan was label-accessible on infectious hyphae, while chitin and β-1,3-glucan were not (Fig. 2B). Therefore, we presumed that α-1,3-glucan locates at surface of the cell wall of infectious hyphae and masks other polysaccharides in the cell wall. To test our hypothesis, we treated formaldehyde-fixed infectious hyphae with α-1,3-glucanase to digest α-1,3-glucan. Chitin and β-1,3-glucan, but not α-1,3-glucan, were detected on infectious hyphae after enzymatic digestion (Fig. 2D). These results indicate that α-1,3-glucan interferes with binding of the fluorescently labelled lectin or antibody to chitin and β-1,3-glucan. We further analysed localization of α-1,3-glucan and β-1,3-glucan in the cell wall of infectious hyphae (at 48 hpi) by immunoelectron microscopy using the same primary antibodies used for fluorescent microscopy. At this infection stage in susceptible rice cells, fungal colonization was observed to extend to several cells both laterally and vertically from the originally infected cells, and host membranous structures were not observed in the colonized cells. These infectious hyphae had single-layered, electron-lucent cell walls (Fig. 7A). Localization of β-1,3-glucan, detected by gold particles, was observed throughout from the vicinity of plasma membrane to the cell wall surface, and double labelling showed that β-1,3-glucan intermixed with α-1,3-glucan (Fig. 7A). We quantitatively analysed the localization of α-1,3-glucan and β-1,3-glucan in infectious hyphae. Relative distances of α-1,3-glucan and β-1,3-glucan were standardized for the cell wall thickness; that is, the average distance of each gold particle from the plasma membrane was calculated as a percentage of the average thickness of the fungal cell wall in each image. As expectedly, analysis of the relative distances revealed that α-1,3-glucan was distributed more distantly from the fungal plasma membrane compared with β-1,3-glucan (n = 16, paired t-test, P < 0.01, Fig. 7B). Thus, α-1,3-glucan exists in and at the surface of the cell wall of infectious hyphae and masks β-1,3-glucan in the cell wall.

Figure 7.

Immunoelectron microscopy of α-1,3-glucan and β-1,3-glucan in infectious hyphae. Leaf sheaths of susceptible rice cultivar LTH were inoculated with wild-type M. grisea, incubated for 48 h, and prepared for electron microscopy.
A. Double labelling of α-1,3-glucan with 18 nm colloidal gold and β-1,3-glucan with 12 nm colloidal gold α-1,3-glucan was intermixed with β-1,3-glucan in electron-lucent fungal cell walls of infectious hyphae. PCW, plant cell wall; FCW, fungal cell wall; FM, fungal mitochondrion; FV, fungal vacuole. Scale bar = 500 nm.
B. Relative distances of α-1,3-glucan and β-1,3-glucan from the fungal plasma membrane; 0% = at fungal plasma membrane; 100% = at the most surface of the cell wall of infectious hyphae. Although distribution of α-1,3-glucan and β-1,3-glucan within the cell wall overlapped, α-1,3-glucan was located farther from the plasma membrane than β-1,3-glucan on average (n = 16, P < 0.01).

Discussion

Fluorescence labelling and imaging techniques revealed specific localization of the major cell wall polysaccharides during infectious structure development in M. grisea. The cell wall polysaccharides β-1,3-glucan, chitin, chitosan and mannan were label-accessible on appressoria and germ tubes, while α-1,3-glucan was exclusively detected on appressoria on cover glass. qRT-PCR analysis in addition to cytological observations showed that the genes encoding β-1,3-glucan synthase (MgFKS1) and chitin synthases were expressed throughout the process of appressorium development, whereas the gene encoding α-1,3-glucan synthase (MgAGS1) was specifically expressed when incipient appressoria were formed.

Although no accumulation of α-1,3-glucan on germ tubes was observed on cover glass, it was clearly detected on cover glass in the presence of exogenous 1,16-hexadecanediol or on plant surfaces. These results provide evidence that α-1,3-glucan accumulation on germ tubes is induced by a plant cue. In a deletion mutant of the Mps1p MAP kinase, which is deficient in infectious hypha formation in planta, α-1,3-glucan did not accumulate on either germ tubes or appressoria, regardless of the presence of exogenous 1,16-hexadecanediol. On the other hand, Mps1p was phosphorylated in the presence of 1,16-hexadecanediol. Thus, Mps1p is required for α-1,3-glucan accumulation in response to exogenous 1,16-hexadecanediol. 1,16-hexadecanediol also induces appressorium formation, which requires heterotrimeric G-protein β subunit Mgb1p, a component of the cAMP signalling pathway. However, 1,16-hexadecanediol induced accumulation of α-1,3-glucan on the germ tubes of mgb1, although no appressorium formation was observed (Fig. 4D and E). Consistently, exogenous cAMP did not induce α-1,3-glucan accumulation on germ tubes on cover glass (Fig. 4C). Therefore, Mgb1p or cAMP signalling is indispensable for appressorium formation, but dispensable for α-1,3-glucan accumulation. Taken together, the recognition signal for 1,16-hexadecanediol is most likely transmitted by at least two signalling pathways: the G-protein signal transduction pathway, which induces appressorium formation via activation of cAMP signalling, and the Mps1 signal transduction pathway, which induces accumulation of α-1,3-glucan (Fig. 8). It should be noted that α-1,3-glucan accumulates on appressoria without exogenous 1,16-hexadecanediol. Therefore, environmental signals such as thigmotropic cues that induce appressorium formation on cover glass may induce α-1,3-glucan accumulation via direct or indirect activation of the Mps1 signal transduction pathway (Fig. 8).

Figure 8.

Model for dual regulation of appressorium formation and α-1,3-glucan accumulation by plant cues in M. grisea. A plant cue (plant wax derivative) recognition signal is transmitted by heterotrimeric G-protein signalling and triggers appressorium formation. At the same time, the plant cue is also recognized by a sensor protein that is involved in the activation of Mps1 MAP kinase. Activated Mps1p then induces in accumulation of α-1,3-glucan in infectious structures. A thigmotropic cue induces appressorium formation via G-protein signal transduction pathway and may be involved in α-1,3-glucan accumulation in appressoria via activation of the Mps1 MAP kinase signalling cascade (see details in Discussion).

In S. cerevisiae, the CWI signalling pathway involving Slt2p/Mpk1p has been well studied (Levin, 2005). Cell wall stresses stimulate the CWI signalling cascade, and activated Slt2p/Mpk1p induces expression of genes encoding integral cell wall proteins and proteins related to cell wall biogenesis via phosphorylation and activation of MADS-box transcription factor Rlm1p (Jung and Levin, 1999). Transcription of a β-1,3-glucan synthase (FKS1) and a chitin synthase (CHS3) is fully dependent on Rlm1p, while a second β-1,3-glucan synthase (FKS2) is Rlm1p-independent but Slt2p/Mpk1p-dependent (Jung and Levin, 1999). Genetic studies have demonstrated that MpkAp, an orthologue of Slt2p/Mpk1p, is also involved in the CWI signalling pathway in A. nidulans, Aspergillus fumigatus and Aspergillus oryzae (Mizutani et al., 2004; Fujioka et al., 2007; Valiante et al., 2008). In A. oryzae, disruption of KexB, a subtilisin-like processing enzyme, results in constitutive activation of MpkAp and subsequently in an increase in expression of some cell wall-related genes (Mizutani et al., 2004). In Aspergillus niger, cell wall stress increases deposition of chitin and transcription of genes encoding a glutamine : fructose-6-phosphate amidotransferase (GFAA) and an α-1,3-glucan synthase (AGSA) (Ram et al., 2004; Damveld et al. 2005a,b). GfaAp is responsible for a step in the metabolic pathway of UDP-N-acetylglucosamine, a sugar donor in chitin biosynthesis (Ram et al., 2004). In a deletion mutant of RLMA (an orthologue of Rlm1p in A. niger), expression of AGSA is abolished and that of GFAA is decreased. Moreover, expression of AGSA is fully dependent on Rlm1p binding sites in its promoter region (Damveld et al. 2005b). In A. nidulans, among the genes related to cell wall biogenesis that are induced in response to cell wall stress, only α-1,3-glucan synthases (AGSA and AGSB) are expressed in a MpkA-RlmA-dependent manner (Fujioka et al., 2007). Transcription of GFAA is partly dependent on MpkA-RlmA signalling but that of β-1,3-glucan synthase and other chitin synthase-related genes is MpkAp-independent (Fujioka et al., 2007). Nevertheless, RlmAp in A. nidulans functionally complements an RLM1 deletion mutant in S. cerevisiae (Fujioka et al., 2007). In this study, we have shown that α-1,3-glucan but not β-1,3-glucan, chitin, mannan and chitosan were undetectable in the cell wall of mps1, the deletion mutant of the orthologue of Slt2p/Mpk1p in M. grisea. This result suggests that expression of α-1,3-glucan synthase fully depends on Mps1p, while that of other cell wall biogenesis-related enzymes, such as β-1,3-glucan synthase, chitin synthases, mannan transferases and chitin deacetylases, is Mps1p-independent. We have further demonstrated that deposition of α-1,3-glucan occurs in an Mps1p-dependent manner in response to addition of the plant wax derivative. Therefore, regulation of α-1,3-glucan synthesis by the CWI signalling pathway is most likely conserved in filamentous fungi, while activation of the CWI pathway by plant cues may be a phenomenon specific to plant pathogens.

In M. grisea, mps1 has severe defects in aerial hyphal development and conidiation, and is hypersensitive to cell wall-digesting enzymes (Xu et al., 1998). Autolysis has also been observed in mps1 when grown on plate media (Xu et al., 1998). This mutant has lost pathogenicity due to its inability to form infectious hyphae after penetration of live plant cells; however, it can form infectious hyphal-like structures in heat-killed plant cells (Xu et al., 1998; Park et al., 2004; Mehrabi et al., 2008). Like Mps1p, Mig1p, a homologue of Rlm1p and RlmAp, is also required for infection of live rice cells (Mehrabi et al., 2008). The mig1 mutant penetrates live rice cells and differentiates primary infectious hyphae but fails to develop secondary infectious hyphae, even though hyphal-like structures develop from mig1 appressoria in heat-killed rice cells (Mehrabi et al., 2008). Vegetative growth and conidiation are reduced in mig1, but the reduction is not as severe as observed for mps1. Moreover, the mutant does not exhibit autolysis on plate media and is not hypersensitive to cell wall digesting enzymes (Mehrabi et al., 2008). Mig1p might be one of the transcription factors regulated by Mps1 signalling; however, no Rlm1p/Mig1p binding site was found in the MgAGS1 promoter region (Mehrabi et al., 2008). In this study, we have showed that MgAgs1p expression is Mps1p-dependent. Identifying a transcriptional factor for MgAGS1 in the Mps1 signal transduction pathway is an objective of future research.

By immunofluorescent microscopy, α-1,3-glucan and chitosan were the major cell wall polysaccharides detected at the label-accessible surface of the infectious hyphae (Fig. 2B and C); β-1,3-glucan and chitin were hardly detectable (Fig. 2B). However, treatment with α-1,3-glucanase allowed for β-1,3-glucan and chitin detection (Fig. 2D), suggesting that α-1,3-glucan interferes with the binding of antibody and lectin to β-1,3-glucan and chitin in the cell wall of infectious hyphae. Further analysis also provides an evidence that the accumulated α-1,3-glucan contributes in protection of chitin in the cell wall from chitinase digestion (Fig. 6A and B). In addition, immunoelectron microscopy revealed that α-1,3-glucan and β-1,3-glucan existed as mosaic in the cell wall of infectious hyphae, but α-1,3-glucan was localized more outwardly compared with β-1,3-glucan (Fig. 7A and B). Taken together, α-1,3-glucan masks β-1,3-glucan and chitin and interferes with the binding of antibodies, lectins and digestive enzymes to β-1,3-glucan and chitin in the cell wall of infectious hyphae. However, in the cell wall of appressoria, α-1,3-glucan did not interfere with the binding of antibody to β-1,3-glucan or lectin to chitin. Further studies are needed to fully understand this interference mechanisms by α-1,3-glucan.

Plants are known to induce secretion of various enzymes that degrade fungal cell walls, e.g. β-1,3-glucanase and chitinase, in response to fungal infection as a defence mechanism (van Loon et al., 2006). These enzymes prevent fungal infection not only by blocking development of infectious hyphae (Mauch et al., 1988; Mauch and Staehelin, 1989; Toyoda et al., 1991), but also by releasing MAMPs from hyphae (Ryan and Farmer, 1991; Felix et al., 1993; Vander et al., 1998). It has been assumed that chitin is deacetylated to chitosan in cell walls of infectious hyphae to avoid generation of MAMPs in some plant pathogenic fungi, e.g. P. graminis f. sp. tritici (a wheat pathogen), U. fabae (a broad bean pathogen) and C. graminicola (a maize pathogen) (El Gueddari et al., 2002). In fact, N-acetylglucosamine oligomers generated from chitin function as a MAMP in wheat, whereas N-glucosamine oligomers generated from chitosan do not (Vander et al., 1998). Thus, cell wall composition of fungal pathogens plays an important role in plant–pathogen interactions. Our results demonstrate a dynamic change in composition of cell wall polysaccharides during plant infection in M. grisea. Especially, α-1,3-glucan displayed a specific transcription and accumulation patterns during infection (Figs 2B and 3A). Rice cells produce various enzymes that hydrolyse fungal cell walls in response to fungal infection (van Loon et al., 2006); importantly, genes encoding chitosan and α-1,3-glucan degrading enzymes are not found in the rice genome. Interestingly, α-1,3-glucan physically and functionally masks other cell wall polysaccharides in the infectious hyphae; indeed, the accumulation of α-1,3-glucan contributes in protection of cell wall from digestive enzymes (Figs 6A and B, and 7B). M. grisea may protect its cell wall from the polysaccharide-hydrolytic enzymes secreted from rice cells by masking its cell wall surface with α-1,3-glucan during infection. In support of this, we have preliminary observations that a deletion mutant in the α-1,3-glucan synthase gene (MgAGS1) is non-pathogenic to live rice plants (T. Fujikawa and M. Nishimura, in preparation). Further study is needed to clarify contribution of chitosan in protection of the fungal cell wall during infection. Recently, a novel role of α-1,3-glucan has been reported in a human pathogen, H. capsulatum, in which the cell wall β-1,3-glucan is masked by α-1,3-glucan to block host recognition of β-1,3-glucan via the dectin-1 receptor (Rappleye et al., 2007). Concealing cell wall MAMPs by α-1,3-glucan may be important for successful infection in some fungal pathogens. In conclusion, both plant and mammalian pathogenic fungi appear to have similar strategies to circumvent host recognition by dynamically changing their cell wall components.

Experimental procedures

Strains and growth media

Wild-type M. grisea strain Guy11 (Leung et al., 1988), a deletion mutant of MGB1, the gene encoding heterotrimeric G-protein β subunit, which was generated from Guy11 (nw98; Nishimura et al., 2003), and mps1, a deletion mutant of the Mps1 MAP kinase of Guy11 (Xu et al., 1998), were used in this study. For conidiation, strains were cultured on oatmeal medium (3% ground oatmeal, 0.5% glucose, 1.6% agar) at 25°C under constant fluorescence for 10 days. For submerged culture, the fungus was grown in YG medium (yeast extract 0.5%, glucose 2.0%) for 5 days at 27°C with gentle shaking (90 r.p.m.). Storage of these fungal strains was as previously described (Nishimura et al., 2003).

Observation of fungal infectious structures

A conidial suspension (30 μl of 1 × 105 conidia ml−1 in distilled water) was dropped on cover glass (Matsunami, Osaka, Japan) and incubated in a moist chamber at room temperature for up to 48 h to observe germination and appressorium formation. cAMP (Invitrogen, Carlsbad, CA, USA) or the plant wax derivative (1,16-hexadecanediol: Aldrich Chemical Co., Milwaukee, WI, USA) was added to the conidial suspension at a final concentration of 10 mM and 50 μM respectively. To observe invasive growth of the fungi in plant cells, 50 μl conidial suspension (1 × 106 ml−1 conidia in distilled water) was placed on sheath cells from rice cultivar LTH (susceptible to Guy11) or on onion epidermal cells, which were incubated at room temperature for up to 48 h. Infectious hyphae developing in plant cells were observed at 16 and 48 h after inoculation.

Immunofluorescent labelling of fungal cell wall polysaccharides

Fungal conidia incubated on inductive surfaces (cover glass or rice leaf sheath cells) were gently rinsed with distilled water before fixation. For observation of unfixed fungal samples, the water-rinsed samples were directly fluorescent-labelled. Fungal cells on cover glass were fixed by 3% (v/v) formaldehyde solution in distilled water at 65°C for 30 min. Fungal cells on rice leaf sheath cells were fixed by 3% (v/v) formaldehyde in 90% (v/v) ethanol until the rice cells were decolorized. The fixed fungal cells were washed three times in PBS buffer (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HPO4, 1.5 mM KH2PO4, pH 7.4) before being infiltrated by 1% (v/v) Tween 20 in PBS buffer (PBS-T). For enzymatic digestion of α-1,3-glucan, the fixed fungal cells were incubated with 30 μl of 5 μg μl−1 purified α-1,3-glucanase (from Bacillus circulans KA-304; Yano et al., 2006) in PBS buffer for 6 h at room temperature and then washed with PBS-T buffer three times before immunofluorescent labelling. To observe localization of α-1,3-glucan or β-1,3-glucan on the cell wall of M. grisea, we used fluorescently labelled antibodies as previously described (Wheeler and Fink, 2006; Rappleye et al., 2007). Briefly, the fixed samples were incubated with primary antibodies overnight at room temperature before incubation with secondary antibodies for 2–24 h in the dark. The antibody-labelled cells were rinsed with PBS buffer before observation using an epifluorescent microscope. To detect α-1,3-glucan in the cell wall, 20 μl mouse IgMγ MOPC-104E (Sigma, St Louis, MO, USA) (0.1 mg ml−1 in PBS buffer) was used as the primary antibody and 20 μl Alexa Fluor 488 goat anti-mouse IgM (μ chain) antibody (Invitrogen) (0.1 mg ml−1 in PBS buffer) as the secondary antibody. The MOPC-104E antibody specifically recognizes α-1,3-linked glycosyl polysaccharide (Marion et al., 2006; Rappleye et al., 2007). Cell wall β-1,3-glucan was detected using 20 μl monoclonal β-1,3-glucan-specific antibody (Biosupplies, Parkville, Australia) (0.1 mg ml−1 in PBS buffer) as the primary antibody and 20 μl Alexa Fluor 594 goat anti-mouse IgG (H + l) antibody (Invitrogen) (0.1 mg ml−1 in PBS buffer) as the secondary antibody (Hohl et al., 2005; Wheeler and Fink, 2006). Chitin, chitosan and α-1,6-mannan were detected as described by Ramonell et al. (2005), Baker et al. (2007) and Briza et al. (1988) respectively. Briefly, the fixed samples were incubated overnight in the dark with 20 μl wheat germ agglutinin Alexa Fluor 350 conjugate (Invitrogen) 10 μg ml−1 in PBS buffer to detect chitin, 20 μl eosin Y (Sigma) (0.05% (v/v) in PBS buffer) to detect chitosan and 20 μl of 0.1 mg ml−1 FITC conjugated concanavalin A (Sigma) in PBS buffer to detect α-1,6-mannan. The stained samples were rinsed with PBS before microscopic observation with a Leica DR system (Wetzlar, Germany). For observation of α-1,3-glucan and mannan, we used a GFP filter cube (excitation filter BP 470/40 nm, 500 nm dichromatic mirror, suppression filter BP 525/50 nm), for β-1,3-glucan and chitosan, a Y3 filter cube (excitation filter BP 545/30 nm, 565 nm dichromatic mirror, suppression filter BP 610/75 nm), and for chitin, a A4 filter cube (excitation filter BP 360/40 nm, 400 nm dichromatic mirror, suppression filter BP 470/40 nm).

Immunogold labelling of α-1,3-glucans and β-1,3-glucans

Conidia of M. grisea were inoculated onto leaf sheaths of a susceptible rice cultivar and then incubated in a humid chamber for 48 h. The samples were fixed with 2% glutaraldehyde in 50 mM PIPES buffer (pH 6.9) for 2 h at room temperature before rinsing in the same buffer, and then cut into small pieces with a razor blade. The fixed samples were then post-fixed with 1% osmium tetroxide in the buffer for 1 h at room temperature and rinsed in distilled water. Next, the samples were dehydrated using a graded series of ethanol and propylene oxide before embedding in Spurr's resin. Ultrathin sections were cut with a diamond knife and directly picked up on nickel grids. The sections were pre-incubated with 1% normal goat serum (NGS) in 10 mM Sörensen's phosphate buffer (pH 7.0) containing 150 mM NaCl (PBS) and incubated with primary antibodies to α-1,3-glucan or β-1,3-glucan for immunofluorescent microscopy. The antibodies for α-1,3-glucan and β-1,3-glucan were diluted 500- and 2000-fold, respectively, with 1% NGS in PBS (NGS/PBS). The sections were rinsed with NGS/PBS and then treated with 18 and 12 nm colloidal gold-conjugated secondary antibodies (Jackson ImmunoResearch, 1:10 and 1:40 dilution in NGS/PBS respectively). Double labelling was conducted on one side for detection of α-1,3-glucan and the other side for β-1,3-glucan. The labelled sections were observed without staining of electron-dense material using a transmission electron microscope (JEM-1200EX, JEOL, Tokyo, Japan). Spatial localization of α-1,3-glucan and β-1,3-glucan in the fungal cell wall was determined using double-labelled ultrathin sections in which the distance of each gold-labelled polysaccharide from the hyphal plasma membrane was measured. Sixteen images containing one or two hyphal cross sections taken at 15 000 or 12 000 times magnification were imported into Image-Pro Express software (MediaCybernetics, Bethesda, MD, USA). The perpendicular distance of the gold particles corresponding to α-1,3-glucan (labelled with 18 nm colloidal gold, a total of 251 gold particles in all 16 images, equal to 10–20 gold particles/image) and β-1,3-glucan (labelled with 12 nm colloidal gold, a total of 1141 gold particles in 16 images, equal to >50 particles/image) from the hyphal cell membrane was determined. Cell wall thickness of the hyphal sections in each image was also measured at about 10 positions per hypha. In each image, an average distance (in nm) of the position of each gold particle from the plasma membrane was then calculated as a percentage of the average thickness of the cell wall. The standardized distances (relative distance) were used for comparison of positions of α-1,3-glucan and β-1,3-glucan in the cell wall (paired t-test, n = 16).

Quantitative qRT-PCR

Total RNA was isolated from germinating conidia or appressorium-forming conidia from the inductive hydrophobic surface of GelBond film (Takara, Shiga, Japan) using a Qiagen Plant mini easy kit (Tokyo, Japan). The fungal cells developing on the GelBond surface were collected with silicon scrapers (Toray, Tokyo, Japan) and resuspended in RNAlater (Ambion, Austin, TX, USA). Total RNA from rice leaf sheaths inoculated with fungal conidia was isolated using a Qiagen Plant mini easy kit. cDNAs were synthesized from the total RNA samples using ExScript RT reagent kit (Takara) with oligo-dT primers and were used as templates for quantitative qRT-PCR analyses. SYBR Premix ExTaq kit (Takara) was used for labelling and amplification of template cDNA for qRT-PCR. The gene-specific primers were designed to amplify approximately 300 bp of unique sequence from corresponding genes (Table 1). qRT-PCR was performed using a Stratagene Mx3000p system (Stratagene, Tokyo, Japan) following the manufacturer's instructions. Transcription levels of these genes were quantified by delta-Ct methods (Livak and Schmittgen, 2001). Statistical analyses were done according to Welch's t-test, and significance was defined as a P-value of < 0.05.

Table 1.  Primers used in this study.
Target genePrimer nameSequence (5′ to 3′)
MgAGS1RTAGS1-FCCTTTGTCGCGCCGTTTG
RTAGS1-RCCGTCCTTGGTCGTAGTGAG
MgFKS1RTFKS1-FTGGCATACAATTTCGCAGCCGG
RTFKS1-RTTGTTATGGCCTTTGGTGG
MgCHS1RTCHS1-FTCGACCCCTGGCACATG
RTCHS1-RGGATAAAACGGAGGTAGTAGT
MgCHS2RTCHS2-FTATTCTACTCGCCTTGATCGCC
RTCHS2-RTCGTCGTTGAAAGGTGCCAGGG
MgCHS3RTCHS3-FGCGGTCTCTCGGTTCCC
RTCHS3-RCACCTCTGTAGTAGTCCAAT
MgCHS4RTCHS4-FGCTGGATCAACTCGACTATTCACA
RTCHS4-RGTCTTCTCACCCGCCGTCTTGC
MgCHS6RTCHS6-FGTGCCCCGATGAACCCGCTC
RTCHS6-RCCCTTGTCGGCAGTACGCAG
MgCHS7RTCHS7-FACACTGGCCGCGTTCCT
RTCHS7-RACGCTGAAGACCCAAGACTAAC
MgCSM1RTCSM1-FCATGAGCCCGTTGGAGC
RTCSM1-RCAGAAGCGTGGTCAGGTAA
MgACTINRTActin-FCAACTCGATCATGAAGTGCGATGT
RTActin-RGCTCTCGTCGTACTCCTGCTT

Sequence analysis

We screened the M. grisea genome database (http://www.broadinstitute.org/annotation/genome/magnaporthe_grisea/MultiHome.html) for proteins that were highly homologous to α-1,3-glucan synthases, β-1,3-glucan synthases or chitin synthases in A. nidulans, A. niger and A. fumigatus. The homology search was performed by the blastp program (Altschul and Lipman, 1990). Phylogenetic analysis of these genes using amino acid sequences was done with the ClustalW program (Thompson et al., 1994).

Construction of mkk1DD mutant

The primers used for generation of the mkk1DD mutant are listed in Table 2. A 2.5 kb DNA fragment containing the MKK1 gene (GenBank XM_369967) together with 450 bp upstream and 350 bp downstream of MKK1 was amplified from Guy11 genomic DNA by ExTaq polymerase (Takara) with primers Mkk1-F and Mkk1-R (Fig. 5A). The amplified fragment was cloned into the pGEM-T vector (Promega, Madison, WI, USA) and named pGeMKK1. To replace Thr 375 with Asp in Mkk1p, the MKK1 gene in pGeMKK1 was mutated with primer Mkk1-T375D using the GeneEditor in vitro Site-Directed Mutagenesis System (Promega). The plasmid carrying the MKK1T375D allele was named pGeMKK1T375D. To replace Thr 369 with Asp in Mkk1p, 1.6 and 0.9 kb DNA fragments amplified from pGeMKK1T375D with primer sets Mkk1-F/Mkk1-G368-EcoRV-R and Mkk1-G371-DraI-F/Mkk1-R were ligated after respective enzymatic digestion with EcoRV and DraI. The ligated fragment was amplified with primers Mkk1-F/Mkk1-R and cloned into the pGEM-T vector (Promega). The resulting plasmid, which carried MKK1T369D,T375D (Fig. 5A, see underlined text), was named pGeMKK1DD. After confirming the sequence, MKK1DD (MKK1T369D,T375D) amplified with Mkk1-F/Mkk1-R from pGeMKK1DD was cotransformed with pAhd5 (a kind gift from Dr Jin-Rong Xu) into Guy11 by the protoplast PEG transformation method as described by Kito et al. (2008). The replacement of MKK1 by MKK1DD in Guy11 was first screened by PCR with primers Int-F/DD-R and confirmed by Southern hybridization analysis with probes 1 and 2 (Fig. 5B). Probe 1 was amplified with primers Int-F/Mkk1-R2 from pGeMKK1 to detect wild-type MKK1 and probe 2 was amplified with primers Int-F/DD-R from pGeMKK1DD. Replacement of MKK1 with MKK1DD in the mkk1DD mutant was also confirmed by sequencing. DNA sequences of all constructs and DNA fragments used in this study were confirmed using an ABI 4100 sequencer (Applied Biosystems).

Table 2. MKK1DD construction primers used in this study.
Primer nameSequence (5′ to 3′)
Mkk1-FGCATGCCCAGGTCCGGA, SphI site is underlined
Mkk1-RGTCGACGCTAGGTCCCAAAA, SalI site is underlined
Mkk1-T375DGGCGAGGCAAACGACTTCATTGGAACG
Mkk1-G368-EcoRV-RGATATCACCAAAGTCACCCGACACTCC, EcoRV site is underlined
Mkk1-G371-DraI-FTTTAAAGGCGAGGCAAACGACTTC, DraI site is underlined
Int-FCAGATACTACGGCGCGTTTGAG
WT-RGTGTTTGCCTCGCCCTTGGT
DD-RTCGTTTGCCTCGCCTTTATC
Mkk1-R2TCACTCTGCAGGCTTGGCTTCAC

Western blot analysis

Total protein was extracted following the procedures described by Yoshimi et al. (2005). Briefly, M. grisea strains mps1, mkk1DD and Guy11 were grown in YG medium for 5 days at 27°C. To determine the effect of a plant wax derivative on phosphorylation of Mps1p, Guy11 was grown in YG medium supplemented with 50 μM 1,16-hexadecanediol. The fungal mycelia were collected through Miracloth (Calbiochem, La Jolla, CA, USA) and ground in liquid nitrogen. Ice-chilled extraction buffer containing protease inhibitors and phosphatase inhibitors [50 mM Tris-HCl, pH 7.5, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, 50 mM sodium fluoride, 5 mM sodium pyrophosphate, 0.1 mM sodium vanadate, protease inhibitor cocktail set (Roche, Basel, Switzerland)] (one tablet per 50 ml) was added to the ground mycelia and the suspension was mixed vigorously. After incubation for 30 min at 60°C, the suspension was centrifuged at 8000 g for 6 min at 4°C. The concentration of total protein in the supernatant was determined using the BCA protein assay reagent (Pierce, Rockford, IL, USA). Protein samples (50 μg) were subjected to SDS-PAGE in a 10% polyacrylamide gel and then the gel was blotted onto a PVDF membrane for Western blot analysis. Phosphorylation of Mps1p in M. grisea was examined by Western blot analysis using ECL Plus Western blotting detection reagent (GE Healthcare, Little Chalfont, Buckinghamshire, UK) with a phospho-p44/42 MAP kinase (Thr202/Tyr204) antibody (Cell Signalling Technology, Beverly, MA, USA) as the primary antibody. The presence of Mps1p was detected by a p44/42 MAP kinase antibody (Cell Signalling Technology). A horseradish peroxidase conjugate was used as the secondary antibody.

Cell wall chitinase digestion susceptibility assay

Susceptibility of the fungal cell wall to chitinase digestion was determined according to methods described in Mizuno et al. (1997). Washed 4-day-old mycelia (1 μg fresh weight) grown in YG medium at 25°C were resuspended in 1 ml of 50 mM potassium phosphate buffer (pH 6.5, containing 0.2 mg ml−1 sodium azide, 0.5 M mannitol). To determine susceptibility of the fungal cell wall to chitinase, 15 μg purified chitinase A from Streptomyces cyaneus SP-27 (Yano et al., 2008) was added to 1 ml potassium phosphate buffer. To determine the effect of additional α-1,3-glucanase on lysis of the fungal cell wall by chitinase, 10 μg purified α-1,3-glucanase from B. circulans KA-304 (Yano et al., 2006) was added to the buffer containing 15 μg ml−1 purified chitinase A. After 0, 4 and 24 h incubation at 30°C, the number of spheroplasts generated from mycelia was counted by a hemocytometer. Mycelia incubated in the potassium phosphate buffer containing no enzyme were used as controls. To observe susceptibility of germ tubes to cell wall-digesting enzymes, 30 μl conidial suspension (1 × 105 ml−1 conidia in distilled water) was incubated on cover glass for 16 h at 25°C. To induce accumulation of α-1,3-glucan, 50 μM 1,16-hexadecandiol was added to the conidial suspension. Appressorium-forming conidia were covered with 100 μl potassium phosphate buffer containing 15 μg ml−1 purified chitinase A from S. cyaneus SP-27 and incubated for 4 h at 25°C. To observe the effect of additional α-1,3-glucanase on spheroplast generation by chitinase, purified α-1,3-glucanase from B. circulans KA-304 was added to the chitinase A/potassium phosphate buffer at a final concentration of 10 μg ml−1. Samples incubated in potassium phosphate buffer containing no enzymes for 4 h were used as controls.

Nomenclature

In this report, we used the following nomenclature for all organisms to avoid confusion: mutant: all small letters in italics, GENE: gene names in capital letters in italics, e.g. AGS1 or MgAGS1 (M. grisea AGS1) and proteins: first letter capitalized and the letter p at the end.

Acknowledgements

We are grateful to Dr Jin-Rong Xu for providing the mps1 mutant, Drs Akihiro Moriwaki, Yoko Nishizawa, Kenichiro Saito and Eiichi Minami for their fruitful discussions, and Toshiaki Harashima for critical reading of this manuscript. This work was supported by in part by grants from the Research and Development Program for New Bio-industry Initiatives (BRAIN) and from the Ministry of Agriculture, Forestry and Fisheries of Japan (Molecular analysis of rice–microbe interactions, PMI-0009) to M.N.

Ancillary