Research

Developmentally regulated conversion of surface-exposed chitin to chitosan in cell walls of plant pathogenic fungi
Blackwell Science, Ltd

Nour Eddine El Gueddari1, Una Rauchhaus2, Bruno M. Moerschbacher1 and Holger B. Deising2
1 2

Westflische-Wilhelms-Universitt Mnster, Institut fr Biochemie und Biotechnologie der Panzen, Hindenburgplatz 55, D-48143 Mnster, Germany; Martin-Luther-Universitt Halle-Wittenberg, Institut fr Panzenzchtung und Panzenschutz, Ludwig-Wucherer-Str. 2, D-06099 Halle (Saale), Germany

Summary
Authors for correspondence: Holger B. Deising Tel: +49 345 5522660 Fax: +49 345 5527120 Email: deising@landw.uni-halle.de Bruno M. Moerschbacher Tel: +49 251 8324794 Fax: +49 251 8328371 Email: moersch@uni-muenster.de Received: 17 June 2002 Accepted: 28 June 2002

Conversion of surface-exposed chitin to chitosan in cell walls of in vitro- and in vivo-differentiated infection structures of two rust fungi, the wheat stem rust fungus Puccinia graminis f. sp. tritici and the broad bean rust fungus Uromyces fabae, and of the causal agent of maize anthracnose, Colletotrichum graminicola, were studied. Epi-uorescence microscopy with the uorescence-labeled lectin wheat germ agglutinin (WGA) revealed that surfaces of infection structures formed on the plant cuticle expose chitin, whereas surfaces of structures formed after invading the host do not. To identify chitin modication by de-N-acetylation, we raised polyclonal antibodies specically recognizing de-N-acetylated chitosan. These antibodies labeled only those infection structures that differentiate inside the plant, indicating that chitosan is exposed on cell wall surfaces post penetration. Surface modication of the fungal cell walls by chitin de-N-acetylation is discussed as a fungal strategy to protect cell walls of pathogenic hyphae from enzymatic hydrolysis by host chitinases, and to avoid generation of an auto-catalytic defense response system in the invaded host tissue. Key words: biotrophic/hemibiotrophic pathogens, chitin, chitosan, fungal cell wall biogenesis, pathogenicity. New Phytologist (2002) 156: 103112

Introduction
The interaction between a plant pathogenic fungus infecting above-ground plant organs and a potential host begins with the establishment of intimate contact of the surfaces, that is fungal cell wall and plant cuticle, and mutual recognition events control disease development. Chemical and physical plant surface properties initiate fungal infection structure differentiation and invasion of the host (Podila et al., 1993; Deising et al., 1996; Mendgen et al., 1996), and molecular recognition of fungal cell wall components may subsequently induce active defense mechanisms in the invaded plant (Boller, 1995). The outermost layers of the fungal cell wall are the initial targets for antimicrobial hydrolases often synthesized by the plant in response to pathogen attack. Cell walls

may thus be lyzed leading to disintegration of the hyphae, and soluble fragments of structural cell wall polymers may act as elicitors of defence responses in the plant (Boller, 1995; Vander et al., 1998). Thus, especially in longlasting biotrophic interactions, adaptations in the surface properties of fungal structures that are in direct contact with plant cells that help to evade hyphal lysis and elicitor production can be anticipated. Lectin cytochemistry has frequently been used to visualize the distribution of chitin in infection structures of plant pathogenic fungi (e.g. Chong et al., 1986; Heath, 1989; Harder & Chong, 1991). Using uorescence-labeled lectins Freytag & Mendgen (1991a) quantied the surface carbohydrate composition of urediniospore- and basidiospore-derived infection structures of different rust fungi. One of the most prominent changes in surface properties observed during infection

New Phytologist (2002) 156: 103 112 www.newphytologist.com

103

104 Research

structure differentiation was revealed by labeling of chitin with wheat germ agglutinin ( WGA). Chitin was generally found on cell wall surfaces of germ tubes and appressoria, but structures that under natural conditions are differentiated within the leaf tissue, that is infection hyphae and haustorial mother cells, were not labeled. Accordingly, WGA bound to the surfaces of germ tubes and appressoria of the hemibiotrophic bean anthracnose fungus, Colletotrichum lindemuthianum, indicating the presence of chitin (OConnell et al., 1992). No labeling was detected in young infection vesicles, which represent the biotrophic growth stage of this fungus (OConnell & Ride, 1990). These studies suggest that chitin may be missing or masked in the infection hyphae and vesicles by additional surface layers of cell wall material. However, Deising & Siegrist (1995) demonstrated that synthesis of chitin de-Nacetylases during differentiation of infection structures of the broad bean rust fungus correlates with the lack of surfaceexposed chitin. As chitosan is a poor substrate for chitinases (Ride & Barber, 1990), conversion of chitin to chitosan during intercellular growth may protect hyphae of plant pathogenic fungi from being lyzed by extracellular plant chitinases. Although WGA-gold-labeling studies have been performed with ultrathin-sections of different rust haustoria (Harder & Chong, 1991, and literature therein), no conclusions have yet been drawn as to whether the conspicuous lack of WGA-labeling in biotrophic fungal infection structures differentiated in vitro or within the host tissues is due to the absence of chitin, to masking of chitin polymers by layers of other cell wall compounds, or to enzymatic modication of chitin. The demonstration of partial or complete de-Nacetylation of chitin to yield chitosan during infection structure differentiation would require a specic probe allowing detection of chitosan as the product of the enzymatic reaction. Since lectins with specicity to chitosan are not commercially available (Linart et al., 1991), alternative chitosan-specic probes are needed to test for chitosan on the surfaces of hyphae (Hadwiger & Line, 1981; Grenier et al., 1991). Here, we describe the production and characterisation of a specic antiserum to completely de-N-acetylated chitosan. Results from microscopical analyzes using uorescence-labeled WGA and antiserum to chitosan indicate that chitin is present on the surface of infection structures differentiated on the cuticle of the host, and that this polymer is converted to chitosan on cell wall surfaces of infection structures invading the host tissue.

were propagated on Vicia faba cv. Con Amore plants as previously described (Deising et al., 1991). To induce infection structure differentiation, spores were inoculated onto polyethylene membranes scratched with brass brushes; the mean distance between scratches was 24 m. Inoculated membranes were sprayed with distilled H2O and kept at 100% rh and 20C in the dark for 24 h (Deising et al., 1991). Urediniospores of the wheat stem rust fungus Puccinia graminis (Pers.) f.sp. tritici (Eriks. & E. Henn) (race 32) were propagated on fully susceptible, adult wheat plants (Triticum compactum cv. Little Club) growing in automatically regulated growth chambers (20C, 70% rh., 16-h-photoperiod at 70 W m2 in the visible range) in prefertilised soil. Differentiation of infection structures was induced by a mild heat shock (2 h, 30C) starting 2 h after sowing urediniospores (0.1 mg) into polystyrene Petri dishes ( 5 cm) and adding distilled water (5 ml), as previously described (Daniels, 1996). The maize anthracnose fungus Colletotrichum graminicola (Cesati) Wilson (teleomorph Glomerella graminicola (Politis)), isolate CgM2, was cultivated on oatmeal agar as previously described (Anderson & Nicholson, 1996), except that black light irradiation (TL-D36W/08 lamps, Phillips, Hamburg, Germany) was used instead of white uorescent light. As uorescence-labeling of appressoria may be hampered by the presence of melanin, the melanin-decient mutant Cg1.502 (Rasmussen & Hanau, 1989) was also used to study in vitrodifferentiated infection structures. Conidia (2030 mg) were suspended in 1.5 ml of distilled water and centrifuged (2 min, 6000 r.p.m). After repeated washings, the pellet was resuspended in 1 ml of distilled water, diluted 100-fold in DIFCO Potato Dextrose Broth (4.8 mg l1) to yield 710 106 conidia ml1. This suspension was used to inoculate susceptible Zea mays cv. Mutin plants or applied onto cover slips and incubated in sealed Petri dishes on moist lter paper at room temperature (Mercure et al., 1994). Thirty-six to 48-h-oldstructures that had developed appressoria were incubated in 3 M glycerol for up to 2 d to induce formation of infection vesicles in vitro. Immunisation and production of the antiserum Polyglucosamine (chitosan) with an average degree of polymerisation of 848 and a degree of N-acetylation of 0% was kindly provided by A. Domard, Lyon, France, and used as an antigen. Antibodies were raised in two male Chincilla-bastard rabbits provided by Dr K. Thomae, Biberach/Ri, Germany. At the rst antigen injection, the animals were 30 wk old. Within 13 wk, the animals received six antigen injections subcutaneously, with 200 g of the above chitosan preparation each. The antigen was dissolved in a total volume of 1 ml, containing 0.5 ml Ribi-Adjuvans-System (Sigma, Deisenhofen, Germany). Five days after antigen injections, blood was taken

Materials and Methods

Fungal material and induction of infection structure differentiation Urediniospores of the broad bean rust fungus Uromyces fabae (Pers.) Schroet. (isolate I2) were from the rust spore collection of K. Mendgen, University of Konstanz, Germany. Spores

www.newphytologist.com New Phytologist (2002) 156: 103 112

Research

105

and analyzed for antibody titer. Five days after the last boost, the animals were bled, and the antiserum was stored at 70C. Pre-immune serum was taken immediately before the rst antigen injection. Specicity of the antiserum In order to test the specicity of the antiserum, glycolchitosan (Sigma, Deisenhofen, Germany) and glycol-chitin, generated from glycol-chitosan (Trudel & Asselin, 1989) were covalently bound to discs ( 5 mm) excised from Immobilon-AV-membrane (Millipore, Eschborn, Germany) via free amino groups. The discs were incubated for 16 h at room temperature in 300 l of polysaccharide solutions (0.1, 1, 10, and 100 g ml1), washed with 300 l of phosphate buffered saline (PBS; 0.25 M NaCl in 0.01 M NaH2PO4/ Na2HPO4, pH 7.6) containing 0.05% (v/v) Tween 20, incubated with PBS containing 2% (w/v) BSA for 2 h at room temperature, and washed three times with PBS/Tween 20. If not otherwise noted, washing steps were performed for 15 min. The discs were subsequently incubated with the preimmune serum or antiserum (diluted 1 : 10 in 1% (w/v) BSA in (PBS) for 2 h at room temperature, washed three times with PBS/ Tween 20, incubated with alkaline phosphatase-conjugated goat-anti rabbit secondary antibody (diluted 1 : 32 000 with 1% BSA (w/v) in (PBS) for 1 h at room temperature, washed three times with PBS/Tween 20, and incubated with 0.1 M TrisHCl buffer, pH 9.6, containing 1 mM MgCl2, 0.005% (w/v) 5-bromo-4-chloro-3-indolyl-phosphate (BCIP) and 0.01% (w/v) nitro blue tetrazolium (NBT) for 510 min at room temperature and scanned at 600 dpi. In order to test for equal binding of glycol-chitin and glycol-chitosan to the Immobilon-AV-membrane, both polysaccharides were derivatised with a uorescent label at their reducing ends. To 300 l of a 100-g ml1 solution of glycol-chitin or glycol-chitosan was added 60 l of a 1-mg ml1 solution of BODIPY-FL hydrazide (Molecular Probes, Eugene, Oregon, USA). This mixture was allowed to react for 4 h in the dark at room temperature, 24 l of 1 mM NaCNBH3 were then added as reducing agent, and the mixture was incubated for 30 min. The reaction products were dialyzed against distilled water until the dialysate was free of uorescence. The labeled polysaccharides were bound to the Immobilon-AV-membrane as described above, and binding of glycol-chitin and glycol-chitosan was visualised under an Olympus BX40 epiuorescence microscope (Olympus, Hamburg, Germany) with lter combination U-MNB (BP 470 490; DM 500; BA 515) and digitalized using a video camera. Labelling of fungal infection structures Chitin staining Rust infection structures were incubated with 2% (w/v) BSA in PBS for 30 min at room temperature,

washed three times with PBS/Tween 20 and incubated with FITC-conjugated WGA (0.1 mg ml1 in PBS, 1% (w/v) (BSA) for 1 h at room temperature. After three washes with PBS/Tween 20, infection structures were mounted in distilled water and analyzed by epiuorescence microscopy, as described above. For chitin staining in C. graminicola infection structures, Texas-Red-conjugated WGA (0.1 mg ml1 in PBS, 5% (w/v) (BSA) was used. The specicity of the WGA staining procedure for chitin was tested by preincubation of WGA with either chitosan or chitin (1 mg ml1 PBS) for 1 h at room temperature. Subsequently, the preadsorbed probes were used for labeling of fungal infection structures as described above. WGA-labeled infection structures were analyzed with an Olympus BX40 epiuorescence microscope equipped with lter combination EX: 470490; DM: 500; BA: 515 or with a Nikon Eclips 600 microscope (Nikon, Dsseldorf, Germany), equipped with lter combination EX: 540589; DM: 595; BA: 600660 to visualize Texas-Red uorescence. Chitosan staining Rust infection structures were incubated with 2% (w/v) BSA in PBS at room temperature for 2 h, washed three times with PBS/Tween 20 and incubated with either preimmune serum or antiserum raised against chitosan, diluted 10-fold in PBS containing 0.5% (w/v) BSA. One hour incubation at room temperature was followed by three washes with PBS/Tween 20, and infection structures were then incubated with Texas-Red-conjugated goat antibodies to rabbit IgG (Dianova, Hamburg, Germany), 100-fold diluted with PBS containing 1% (w/v) BSA, for 1 h. After another three washes in PBS/Tween 20, the infection structures were mounted in distilled water and analyzed with an Olympus BX40 epiuorescence microscope equipped with lter combination U-MWG (BP 510550; DM 570; BA > 590). To stain chitosan in C. graminicola, antiserum against chitosan, diluted 1000-fold in PBS containing 5% (w/v) BSA and an FITC-conjugated secondary goat antibody to rabbit IgG (Sigma, Deisenhofen, Germany) diluted 200-fold in PBS containing 5% (w/v) BSA were used. Fluorescence microscopy was done using a Nikon Eclips 600 uorescence microscope (Nikon, Dsseldorf, Germany), equipped with lter combination EX: 465495; DM: 505; BA: 515555. The specicity of immuno-labeling of chitosan was tested by preincubation of the antiserum with either chitosan or chitin (1 mg ml1 PBS) for 1 h at room temperature. Subsequently, the preadsorbed probes were used for labeling of fungal infection structures as described above. To obtain labeling patterns with the same color codes, Texas-Red- and FITC-uorescence was electronically adapted in micrographs of C. graminicola. Double staining for chitin and chitosan Rust infection structures were incubated with 2% (w/v) BSA in PBS for 2 h at room temperature, washed with PBS/Tween 20 three times

New Phytologist (2002) 156: 103 112 www.newphytologist.com

106 Research

and incubated with either preimmune serum or antiserum raised against chitosan, diluted 10-fold in PBS containing 0.1 mg ml1 FITC-conjugated WGA and 1% (w/v) BSA. One hour incubation at room temperature was followed by three washes with PBS/Tween 20, and infection structures were then incubated with Texas Red-conjugated goat antibodies to rabbit IgG (Dianova, Hamburg, Germany), 100-fold diluted with PBS containing 1% (w/v) BSA, for 1 h. After another three washes in PBS/Tween 20, infection structures were mounted in distilled water and analyzed with an Olympus BX40 epiuorescence microscope equipped with lter combination U-MNB. Staining of in planta differentiated infection structures For in planta-staining of rust infection structures, P. graminisinfected wheat leaves and U. fabae-infected broad bean leaves were taken as soon as uredosori had formed, and thin sections were made by hand. Combined WGA- and antibody-staining was performed as described above. For staining of in planta-differentiated infection structures of C. graminicola, infection sites of maize leafs were excised 2 5 d after inoculation and xed in 3% (w/v) paraformaldehyde in PBS supplemented with 0.05% (v/v) Triton X-100. After xation, the samples were rinsed, dehydrated, embedded in polyethylene glycol (PEG) and sectioned according to Van Lammeren et al. (1985). Microtome (HM 335 E, Microm, Walldorf, Germany) sections of 3 m were applied to polylysine-coated microscope slips. A 10-min wash in PBS was followed by 5 min incubation in 0.1 M NH4Cl in PBS and after blocking in 5% (w/v) BSA in PBS for 30 min, sections were stained for chitin or chitosan as described above.

Fig. 1 Specicity of the rabbit antiserum to chitosan. Water-soluble glycol-chitin and glycol-chitosan were immobilized via free amino groups to carboxy groups on an afnity membrane (Immobilon-AV). Equal binding of glycol-chitin and glycol-chitosan to the membrane was evident under the uorescence microscope when polymers labeled with a uorescence tag (Bodipy-FL) at their reducing ends were used (a). Specic binding of the chitosan-antiserum to glycolchitosan, but not to glycol-chitin, was visualized using alkaline phosphatase-conjugated secondary antibodies. Pre-immune serum did not label glycol-chitosan (b).

Results
Production and characterisation of a rabbit antiserum to chitosan The weak antigenicity of chitosan makes this polymer ideally suited for many medical applications (Felse & Panda, 1999), but renders antiserum production difcult (Hadwiger & Line, 1981; Spindler-Barth & Buss, 1997). The specicity of the antiserum obtained was tested using glycol-chitin and glycol-chitosan as target molecules. The two polysaccharides differ greatly in their physical properties, chitosan being positively charged at physiological pH, due to the free amino groups, and chitin being hydrophobic, due to the acetyl groups. To ensure equivalent loading of the membranes with chitin and chitosan, the polymers were bound covalently to the membrane via their free amino groups. As chitin, in contrast to chitosan, contains very few free amino groups, equal binding of the polymers was checked by labeling their reducing ends with the uorescent BODIPY-FL label prior to membrane binding. Comparable intensities of uorescence of

the chitin- and chitosan-coated membranes indicated equivalent binding of the polymers (Fig. 1a). Comparable densities of chitin and chitosan on the Immobilon-AV membrane allowed analysis of the specicity of the antibody for chitosan. Fig. 1(b) clearly demonstrates that the antiserum specically labeled chitosan, and that binding to chitin did not occur. Pre-immune serum did not cross-react with chitosan. This antibody probe thus allows discrimination between chitin and chitosan on fungal infection structures. Chitin and chitosan staining of fungal infection structures Both rust fungi used in this study, that is P. graminis f. sp. tritici and U. fabae, differentiate complex infection structures after perception of (a) signal(s) provided by the host leaf or an articial inductive signal. To induce infection structure differentiation in P. graminis, a heat shock signal has been applied, while infection structures of U. fabae were induced by surface ridges. Under natural conditions, germ tubes and appressoria are formed on the plant cuticle, while substomatal

www.newphytologist.com New Phytologist (2002) 156: 103 112

Research

107

Fig. 2 Distribution of chitin and chitosan on the surface of in vitro differentiated infection structures. Double-staining of Puccinia graminis f. sp. tritici (a) and Uromyces fabae (b) showed that infection structures normally differentiated on the cuticle, that is germ tube (gt) and appressorium (ap) are strongly labeled by FITC-WGA (green uorescence), indicative of surface-exposed chitin. Structures formed in the intercellular space, that is substomatal vesicle (sv) and infection hypha (ih) expose chitosan, as shown by red uorescence of the Texas-Red-labeled antibody. Urediniospores (sp) were not labeled. Bright eld microscopy (c,e) showed that conidia (co) of Colletotrichum graminicola form a short germ tube (gt), a melanized appressorium (ap) and an infection vesicle (iv) in vitro. While germ tubes show green uorescence indicative of chitin, uorescence can not be seen in wild-type appressoria, due to melanization (d). Appressoria of a melanin-decient mutant (insert in c,d) are clearly labeled by WGA. Some faint staining of chitin can also be observed at the tip of the infection vesicle (d, arrowhead). Chitosan on the surface of infection vesicles (iv) is specically stained by antibodies (f). Bars are 30 m (a,b) or 10 m (cf).

vesicles, infection hyphae and haustorial mother cells develop in the intercellular space. In vitro -induced rust infection structures (Fig. 2a,b) did not differ in morphology (appearance of septa, number of nuclei, Calcouor-staining of cell walls) from those formed in vivo to invade the plant (Collins et al., 2001). Likewise, C. graminicola formed short germ tubes and appressoria on hard surfaces such as glass or

plastic, and after some days structures resembling infection vesicles were found (Fig. 2c,e). To test for possible changes of the chitin/chitosan composition on the surface of fungal infection structures, antibodies raised to chitosan and the chitin-specic lectin WGA were used as probes in uorescence microscopy studies. Double staining of in vitro differentiated infection structures of both rusts and WGA-staining of C. graminicola infection structures showed that chitin was readily accessible on the cell wall surfaces of germ tubes and appressoria (Fig. 2a,b,d). While WGA staining of melanized appressoria of wild-type C. graminicola often failed (Fig. 2d), appressoria of the melanindecient mutant Cg1.502 showed clear WGA labeling (Fig. 2d, insert), indicative of chitin on the appressorial cell wall surface. Germ tubes and appressoria that are normally formed on the leaf surface were not stained by the chitosanspecic antiserum (Fig. 2a,b,f ). By contrast, all subsequently formed structures, that is substomatal vesicles and infection hyphae of rusts (Fig. 2a,b) and infection vesicles of C. graminicola (Fig. 2f ) formed in vitro were labeled by the antiserum raised to chitosan, but not by WGA (Fig. 2a,b,d). Only at the tips of infection hyphae of the wheat stem rust fungus (Fig. 2a) and infection vesicles of C. graminicola (Fig. 2d), some WGA-labeling was observed. Labeling of the infection structures thus appeared to be strictly complementary. All structures labeled by WGA did not cross-react with the antiserum to chitosan, and vice versa. This is particularly evident in infection structures double-stained by WGA and antibodies to chitosan (Fig. 2a,b). To see whether or not this labeling pattern can also be observed during plant infection, thin sections obtained from rust-infected wheat and broad bean leaves and from C. graminicola-infected maize leaves were incubated with WGA or with chitosan-specic antibodies (Fig. 3). To obtain an overview of the distribution of chitin and chitosan of in planta growing rusts, thin sections were subjected to double staining and were observed at low magnication. Hyphae that had invaded wheat (Fig. 3a) or broad bean leaves (Fig. 3b) and developed in the intercellular space were labeled by the antiserum to chitosan and thus appear red. Only uredinia formed on the upper and lower surface of a host leaf (Fig. 3b), showed green WGA labeling. Labeling of xylem elements (Fig. 3a) by antibodies were nonspecic, as indicated by controls with preimmune serum (data not shown). Infection structures of C. graminicola formed within a maize epidermal cell (Fig. 3c,d), that is infection vesicle and primary hyphae, were strongly labeled by antibodies to chitosan. However, some antibody labeling was also associated with the appressorial base (Fig. 3d). The upper part of the appressorium shows some background labeling comparable with nonspecic uorescence of the plant cell walls. Interestingly, cross sections of in planta growing Colletotrichum infection hyphae were faintly labeled with WGA (Fig. 3e,f ). This labeling pattern is consistent with the hypothesis that only the outermost cell wall

New Phytologist (2002) 156: 103 112 www.newphytologist.com

108 Research

Fig. 3 Intercellular infection hyphae (ih) of Puccinia graminis f. sp. tritici (a) and Uromyces fabae (b) show intensive red uorescence, indicating chitosan. Only uredinia with urediniospores (sp) show signicant chitin labeling (green uorescence). Labelling of xylem (xy) by antibodies is nonspecic, as indicated by controls with preimmune serum (data not shown). Bright eld micrograph of a microtome section of an infection site of Colletotrichum graminicola (c) showing an appressorium (ap) on the cuticle and an infection vesicle (iv) and initials of primary hyphae (ph) in an epidermal maize cell. The inner cell wall of the base of the appressorium and the walls of the infection vesicle (iv) and primary hyphae (ph) are strongly labeled by the antiserum to chitosan (d). Older infection sites are intensively colonized by infection hyphae (ih), as shown by bright eld microscopy (e). When cross sections of these hyphae were stained by WGA (f) chitin of the inner cell wall layers of infection vesicle (iv) and hyphae (ih) was labeled. Wild-type appressoria (ap) were not labeled by the probes used. In (cf) bars are 10 m.

layer is de-N-acetylated by extracellular fungal chitin de-Nacetylases. Surprisingly, cross-sections of melanized appressorial walls only occasionally showed faint WGA-labeling. In all three fungal systems the labeling pattern was similar, irrespective of whether they had formed on articial surfaces or during plant infection. As chitosan synthesis is only known to occur by the tandem action of chitin synthetase and chitin de-N-acetylase (Davis & Bartnicki-Garcia, 1984), the results presented clearly indicate that chitin is converted to chitosan when the pathogens penetrate the plant leaf tissue (Figs. 2 and 3). To further test the specicity of the probes, blocking experiments and labeling with preimmune serum were performed. To demonstrate that the labeling pattern observed with WGA and the antiserum was the result of hapten/lectin- and antigen/antibody-specic reactions, both probes were separately preincubated with glycol-chitin and glycol-chitosan prior to probing fungal structures (Fig. 4). While WGA-staining of chitin on infection structures of the wheat stem rust fungus was unaffected by preincubation of the lectin with glycolchitosan (Fig. 4a), it was signicantly reduced by preincubation

of the lectin with glycol-chitin (Fig. 4b). The fact that labeling was not completely blocked may indicate a minor degree of nonspecicity. Conversely, antiserum-staining of chitosan was unaffected by preincubation of the antiserum with glycol-chitin (Fig. 4c), but was completely abolished by preincubation of the antiserum with glycol-chitosan (Fig. 4d). By contrast to the antiserum against chitosan, preimmune serum did not cross-react with cell wall surface epitopes of any of the infection structures analyzed. For example, hyphae of C. graminicola were strongly labeled by antibodies to chitosan (Fig. 3d), but were not labeled by the preimmune serum (Fig. 4e,f ). These data clearly demonstrate that the probes used are specic and allow to discriminate between chitin and its de-N-acetylated counterpart, chitosan.

Discussion
Fungal plant pathogens, when colonising their host plant tissues, encounter an elaborate defence system consisting of chemical and physical preformed resistance factors and of induced resistance reactions ( Jabs & Slusarenko, 2000;

www.newphytologist.com New Phytologist (2002) 156: 103 112

Research

109

Fig. 4 Test for specicity of WGA-staining of chitin and antibody-staining of chitosan by preincubation with glycol-chitin and -chitosan. WGA-FITC-staining of chitin on the surface of in vitro-differentiated infection structures of the wheat stem rust fungus was unaffected by preincubation of the lectin with glycol-chitosan (a), but was greatly reduced by preincubation of the lectin with glycol-chitin (b). Conversely, antiserum-Texas-Red staining of chitosan on the fungal surface was unaffected by preincubation of the antiserum with glycol-chitin (c), but was completely abolished by preincubation of the antiserum with glycol-chitosan (d). Maize leaves infected with C. graminicola showed extensive colonization with infection hyphae, as revealed by bright-eld microscopy (e). Fluorescence microscopy showed no labeling on sections treated with preimmune serum (f). ap, appressorium; gt, germ tube; ih, infection hypha; sp, spore; sv, substomatal vesicle. Bars are 30 m (ad) and 10 m (ef).

Manseld, 2000; Moerschbacher & Mendgen, 2000). Hydrolases, such as chitinases and -1,3-glucanases, represent standard antifungal enzymes found in most plants (Broekaert, Terras & Cammue, 2000). Typically, some chitinase and glucanase isoenzymes are constitutively expressed in the plant tissue and secreted into the cell wall or the vacuole. These enzymes are thought to lyze cell walls of penetrating fungal hyphae (Mauch & Staehelin, 1989). In addition to interfering with hyphal growth directly (Mauch et al., 1988; Toyoda et al., 1991; Arlorio et al., 1992), antifungal hydrolases may generate soluble fungal cell wall fragments, which act as elicitors of active defense responses (Felix et al., 1993; Ryan & Farmer, 1991; Vander et al., 1998). As the induced defense mechanisms include further synthesis of chitinases and -1,3glucanases, an autocatalytically accelerating defense system is generated (Benhamou et al., 1990; Boller, 1995).

Hydrolases attacking fungal cell walls have been used in a number of transgenic approaches to generate crop plants with increased resistance to phytopathogenic fungi. Broglie et al. (1991) showed that transgenic tobacco (Nicotiana tabaccum cv. Xanthi) and canola (Brassica napus) plants expressing a bean chitinase exhibit elevated resistance to Rhizoctonia solani. As chitinases act synergistically with -1,3-glucanases (Mauch et al., 1988), coexpression of chitinase and glucanase genes in transgenic plants increased resistance in some cases (SelaBuurlage et al., 1993; Jach et al., 1995). However, failure of this approach in other pathosystems gave rise to speculations that a different set of cell wall hydrolysing enzymes may be required to disintegrate the highly complex fungal cell walls. Thus, analysing the polymer composition of fungal cell wall surfaces and their alterations during pathogenesis is crucial for the understanding of plantfungus interactions and for

New Phytologist (2002) 156: 103 112 www.newphytologist.com

110 Research

constructing transgenic plants with elevated levels of resistance to plant pathogenic fungi. Electron microscopic studies of the cell walls of rust infection structures have revealed striking changes in cell wall composition upon the differentiation steps from appressorium to substomatal vesicle, and from haustorial mother cell to haustorium. In both cases, the outer layers of the more distal infection structure is continuous with an inner layer of the more proximal infection structure (Littleeld & Heath, 1979). Several reports have shown that chitin is present only in inner cell wall layers of intercellular rust hyphae (Chong et al., 1985, 1986) but is not exposed on surfaces of fungal infection structures growing within the host tissue (Mendgen et al., 1988; OConnell & Ride, 1990; Freytag & Mendgen, 1991b; Deising et al., 1996; Deising, Werner & Wernitz, 2000). These observations have led to the hypothesis that restriction of fungal chitin to subsurface areas avoids presentation of the substrate to antifungal apoplastic plant chitinases (Freytag & Mendgen, 1991a; Freytag & Mendgen, 1991b; Schoffelmeer et al., 1999). However, the above studies did not allow discrimination between masking by complete covering and masking by enzymatic modication of chitin. Based on WGA-labeling and staining with antibodies raised to chitosan, we conclude that surface-exposed chitin polymers of cell walls of the three fungi analyzed are de-N-acetylated to chitosan during penetration and invasive growth within the host tissue. These data corroborate previous suggestions based on the developmentally regulated induction of chitin de-Nacetylases in the broad bean rust fungus after appressorium formation (Deising & Siegrist, 1995). As all cytochemical staining methods, lectin-based staining has to be interpreted with some caution, as lectin specicity is never absolute (for a detailed discussion concerning WGA, see Kaminskyj & Heath, 1983; Heath, 1989; OConnell, 1991). The complementary patterns of lectin staining for chitin and immunological staining for chitosan shown in this study, and the concomitant appearance of the chitin de-N-acetylase (Deising & Siegrist, 1995), strongly argue in favor of the specicity of both staining techniques. However, a nal decision will have to await detailed chemical studies of the degree of Nacetylation of the chitin/chitosan present in the different fungal infection structures. Conversion of chitin into chitosan by de-N-acetylation may protect fungal infection structures from hydrolytic attack by chitinases present in the host tissue. Ride & Barber (1990) have shown that decreasing degrees of N-acetylation reduce the utilization of this substrate by plant chitinases. Freytag & Mendgen (1991b) found that a mixture of -1,3-glucanase and -1,3-glucanase containing some chitinase activity completely dissolved only those rust infection structures that expose chitin on their surface. By contrast, the enzymes did not affect infection hyphae and haustorial mother cells, which, as shown by immunocytochemistry in this study, expose chitosan on their surface. Likewise, yeast (Saccharomyces

cerevisiae) mutants that are decient in genes encoding chitin de-N-acetylase show reduced ascospore wall rigidity and increased susceptibility to lytic enzymes (Pammer et al., 1992; Christodoulidou et al., 1996). Concomitantly, as elicitoractive chitin oligomers can not be liberated in the absence of chitinase substrate, triggering of resistance reactions may also be delayed or avoided ( Vander et al., 1998). De-N-acetylation of chitin may thus represent a crucial precondition for successful infection and host tissue colonisation. If, in further studies, chitin de-N-acteylases should prove to be essential in fungal pathogenesis, these enzymes may be promising new candidates for fungicide targets (Struck et al., 1998). Our nding that chitosan, but not chitin, is present on the surface of the cell walls of fungal infection structures growing in planta suggests that chitinases, although effectively degrading cell walls of vegetative fungal hyphae, may not be sufcient for the digestion of cell walls of infection hyphae. Possibly, chitosanases might represent more efcient defense enzymes. Approaches to generate disease-resistant transgenic crop plants over-expressing chitosanases alone or in combination with -1,3-glucanases and /or chitinases are therefore underway.

Acknowledgements
The chitosan used as an immunogen was kindly provided by Dr Alain Domard, Lyon, France. We thank Dr I. Kuhlmann and Dr D. Schopper, University of Konstanz, for help with the production of antibodies. We gratefully acknowledge preliminary experiments by D. Lasso-Agredo, C. Buddenborg and T. Conze. We also thank U. Beike for help with the in vitro-induction of fungal infection structures, and Dr G. Hause, Biozentrum of Martin-Luther-University HalleWittenberg, for help with analyses of in planta-growing infection structures. This project was supported nancially by the Deutsche Forschungsgemeinschaft (DFG; De 403/51/52; Mo 476/23).

References
Anderson DW, Nicholson RL. 1996. Characterization of a laccase in the conidial mucilage of Colletotrichum graminicola. Mycologia 88: 996 1002. Arlorio M, Ludwig A, Boller T, Bonfante P. 1992. Inhibition of fungal growth by plant chitinases and -1,3-glucanases. A morphological study. Protoplasma 171: 3443. Benhamou N, Joosten MH, De Wit PJGM. 1990. Subcellular localisation of chitinase and its potential substrate in tomato root tissues by Fusarium oxysporum f. sp. radicislycopersici.. Plant Physiology 92: 1108 1120. Boller T. 1995. Chemoreception of microbial signals in plant cells. Annual Review of Plant Physiology and Plant Molecular Biology 46: 189 214. Broekaert WF, Terras FRG, Cammue BPA. 2000. Induced and preformed antimicrobial proteins. In: Slusarenko A, Fraser RSS, Loon LC, eds. Mechnisms of resistance to plant diseases. Dordrecht, The Netherlands: Kluwer Academic Publishers, 371477. Broglie K, Chet I, Holliday M, Cressman R, Biddle P, Knowlton S, Mauvais CJ, Broglie R. 1991. Transgenic plants with enhanced resistance to the fungal pathogen Rhizoctonia solani. Science 254: 1194 1197.

www.newphytologist.com New Phytologist (2002) 156: 103 112

Research
Chong J, Harder DE, Rohringer R. 1985. Cytochemical studies on Puccinia graminis f. sp. tritici in a compatible wheat host. I. Walls of intercellular hyphal cells and haustorium mother cells. Canadian Journal of Botany 63: 1713 1724. Chong J, Harder DE, Rohringer R. 1986. Cytochemical studies on Puccinia graminis f. sp. tritici in a compatible wheat host. II. Haustorium mother cell walls at the host cell penetration site, haustorial walls, and the extrahaustorial matrix. Canadian Journal of Botany 64: 25612575. Christodoulidou A, Bouriotis V, Thireos G. 1996. Two sporulation-specic chitin deacetylase-encoding genes are required for the ascospore wall rigidity of Saccharomyces cerevisiae. Journal of Biological Chemistry 271: 31420 31425. Collins TJ, Moerschbacher BM, Read NR. 2001. Synergistic induction of wheat stem rust appressoria by chemical and topographical signals. Physiological and Molecular Plant Pathology 58: 259266. Daniels U. 1996. Die Flssigkultur von Puccinia graminis f. sp. tritici, Rasse 32, als Modellsystem fr die Interaktion eines obligaten Parasiten mit seinem Wirt. PhD thesis. Aachen, Germany: University of Aachen, Germany. Davis LL, Bartnicki-Garcia S. 1984. Chitosan synthesis by the tandem action of chitin synthetase and chitin deacetylase from Mucor rouxii. Biochemistry 23: 1065 1073. Deising H, Heiler S, Rauscher M, Xu H, Mendgen K. 1996. Cellular aspects of rust infection structure differentiation: Spore adhesion and fungal morphogenesis. In: Nicole M, Gianinazzi-Pearson V, eds. Histology, ultrastructure and molecular cytology of plantmicroorganism interactions. Dordrecht, The Netherlands: Kluwer Academic Publishers, 135156. Deising H, Jungblut PR, Mendgen K. 1991. Differentiation-related proteins of the broad bean rust fungus Uromyces viciae-fabae, as revealed by high resolution two-dimensional polyacrylamide gel electrophoresis. Archives of Microbiology 155: 191198. Deising H, Siegrist J. 1995. Chitin deacetylase activity of the rust Uromyces viciae-fabae is controlled by fungal morphogenesis. FEMS Microbiology Letters 127: 207212. Deising HB, Werner S, Wernitz M. 2000. The role of fungal appressoria in plant infection. Microbes and Infection 2: 16311641. Felix G, Regenass M, Boller T. 1993. Specic perception of subnanomolar concentrations of chitin fragments by tomato cells: Induction of extracellular alkalinization, changes in protein phosphorylation, and establishment of a refractory state. Plant Journal 4: 307316. Felse AP, Panda T. 1999. Studies on application of chitin and its derivatives. Bioprocess Eigineering 20: 505 512. Freytag S, Mendgen K. 1991a. Carbohydrates on the surface of urediniospore- and basidiospore-derived infection structures of heteroecious and autoecious rust fungi. New Phytologist 119: 527534. Freytag S, Mendgen K. 1991b. Surface carbohydrates and cell wall structure of in vitro-induced uredospore infection structures of Uromyces viciaefabae before and after treatment with enzymes and alkali. Protoplasma 161: 94 103. Grenier J, Benhamou N, Asselin A. 1991. Colloidal gold-complexed chitosanase: a new probe for ultrastructural localization of chitosan in fungi. Journal of General Microbiology 137: 20072015. Hadwiger LA, Line RF. 1981. Hexosamine accumulations are associated with the terminated growth of Puccinia striiformis on wheat isolines. Physiological Plant Pathology 19: 249 255. Harder DE, Chong J. 1991. Rust haustoria. In: Mendgen K, Lesemann D-E, eds. Electron microscopy of plant pathogens. Berlin, Germany: Springer-Verlag, 235 250. Heath MC. 1989. In vitro formation of haustoria of the cowpea rust fungus, Uromyces vignae, in the absence of a living plant cell. I. Light microscopy. Physiological and Molecular Plant Pathology 35: 357366. Jabs T, Slusarenko AJ. 2000. The hypersensitive response. In: Slusarenko A, Fraser RSS, Loon LC, eds. Mechnisms of resistance to plant diseases. Dordrecht, The Netherlands: Kluwer Academic Publishers, 279 323. Jach G, Grnhardt B, Mundy J, Logemann J, Pinsdorf E, Leah R, Schell J, Maas C. 1995. Enhanced quantitative resistance against fungal disease by combinatorial expression of different barley antifungal proteins in transgenic tobacco. Plant Journal 8: 97109. Kaminskyj SGW, Heath MC. 1983. Histological responses of infection structures and intercellular mycelium of Uromyces phaseoli var. typica and U. phaseoli var. vignae to the HNO2-MBTH-FeCl3 and the IkI-H2SO4 tests. Physiological Plant Pathology 22: 173179. Linart Y, Gautier C, Domard A. 1991. Isolation from Rubus cellsuspension cultures of a lectin specic for glucosamine oligomers. Planta 184: 813. Littleeld LJ, Heath MC. 1979. Ultrastructure of rust fungi. New York, USA: Academic Press. Manseld JW. 2000. Antimicrobial compounds and resistance, the role of phytoalexins and phytoanticipins. In: Slusarenko A, Fraser RSS, Loon LC, eds. Mechnisms of resistance to plant diseases. Dordrecht, The Netherlands: Kluwer Academic Publishers, 325370. Mauch F, Mauch-Mani B, Boller T. 1988. Antifungal hydrolases in pea tissue. II. Inhibition of fungal growth by combinations of chitinase and -1,3-glucanase. Plant Physiology 88: 936942. Mauch F, Staehelin A. 1989. Functional implications of the subcellular localization of ethylene-induced chitinase and -1,3-glucanase in bean leaves. Plant Cell 1: 447457. Mendgen K, Hahn M, Deising H. 1996. Mechanisms and morphogenesis of penetration by plant pathogenic fungi. Annual Review of Phytopathology 34: 367386. Mendgen K, Schneider A, Sterk M, Fink W. 1988. The differentiation of infection structures as a result of recognition events between some biotrophic parasites and their hosts. Journal of Phytopathology 123: 259272. Mercure EW, Kunoh H, Nicholson RL. 1994. Adhesion of Colletotrichum graminicola conidia to corn leaves: a requirement for disease development. Physiological and Molecular Plant Pathology 45: 407 420. Moerschbacher B, Mendgen K. 2000. Structural aspects of defense. In: Slusarenko A, Fraser RSS, Loon LC, eds. Mechnisms of resistance to plant diseases. Dordrecht, The Netherlands: Kluwer Academic Publishers, 231277. OConnell RJ. 1991. Cytochemical analysis of infection structuress of Colletotrichum lindemuthainum using uorochrome-labelled lectins. Physiological and Molecular Plant Pathology 39: 189 200. OConnell RJ, Nash J, Bailey JA. 1992. Lectin cytochemistry: A new approach to understanding cell differentiation, pathogenesis and taxonomy in Colletotrichum. In: Bailey JA, Jeger MJ, eds. Colletotrichum: biology, pathology and control. Wallingford, UK: CAB International, 6787. OConnell RJ, Ride JP. 1990. Chemical detection and ultrastructural localization of chitin in cell walls of Colletotrichum lindemuthianum. Physiological and Molecular Plant Pathology 37: 39 53. Pammer M, Briza P, Ellinger A, Schuster T, Stucka R, Feldmann H, Breitenbach M. 1992. DIT101 (CSD2, CAL1), a cell cycle-regulated yeast gene required for synthesis of chitin in cell walls and chitosan in spore walls. Yeast 8: 10891099. Podila GK, Rogers LM, Kolattukudy PE. 1993. Chemical signals from avocado surface wax trigger germination and appressorium formation in C. gloeosporioides. Plant Physiology 103: 267272. Rasmussen JB, Hanau RM. 1989. Exogenous scytalone restores appressorial melanization and pathogenicity in albino mutants of Colletotrichum graminicola. Canadian Journal of Plant Pathology 11: 349 352. Ride JP, Barber MS. 1990. Purication and characterization of multiple forms of endochitinase from wheat leaves. Plant Science 71: 185 197. Ryan C, Farmer E. 1991. Oligosaccharide signals in plants: a current assessment. Annual Review of Plant Physiology and Plant Molecular Biology 42: 651674. Schoffelmeer EAM, Klis FM, Sietsma JH, Cornelissen BJC. 1999. The cell wall of Fusarium oxysporum. Fungal Genetics and Biology 27: 275 282.

111

New Phytologist (2002) 156: 103 112 www.newphytologist.com

112 Research
Sela-Buurlage MB, Ponstein AS, Vloemans SA, Melchers LS, Van den Elzen PJM, Cornelissen BJC. 1993. Only specic tobacco chitinases and -1,3glucanases exhibit antifungal activity. Plant Physiology 101: 857863. Spindler-Barth M, Buss U. 1997. ELISA for Determination of Chitin and Chitosan. In: Muzzareli RAA, Peter MG, eds. Chitin handbook. Grottammare, Italy: European Chitin Society, 913. Struck C, Hahn M, Mendgen K. 1998. Infection structures of plant pathogenic fungi potential targets for plant disease control. Journal of Plant Diseases and Protection 105: 581589. Toyoda H, Matsuda Y, Yamaga T, Ikeda S, Morita M, Tamai T, Ouchi S. 1991. Suppression of the powdery mildew pathogen by chitinase microinjected into barley coleoptile epidermal cells. Plant Cell Reports 10: 217220. Trudel J, Asselin A. 1989. Detection of chitinase activity after polyacrylamide gel electrophoresis. Analytical Biochemistry 178: 362366. Van Lammeren AAM, Keijzer CD, Willemse MTM, Kieft H. 1985. Structure and function of the microtubular cytosceleton during pollen development in Gasteria verrucosa (Mill.) H. Duval. Planta 165: 111. Vander P, Varum KM, Domard A, El Gueddari NE, Moerschbacher BM. 1998. Comparison of the ability of partially N-acetylated chitosans and chitoologosaccharides to elicit resistance reactions in wheat leaves. Plant Physiology 118: 13531359.

About New Phytologist

New Phytologist is owned by a non-prot-making charitable trust dedicated to the promotion of plant science. Regular papers, Letters, Research reviews, Rapid reports and Methods papers are encouraged. Complete information is available at www.newphytologist.com All the following are free essential colour costs, 100 offprints for each article, online summaries and ToC alerts (go to the website and click on Synergy) You can take out a personal subscription to the journal for a fraction of the institutional price. Rates start at 83 in Europe/$133 in the USA & Canada for the online edition (go to the website and click on Subscriptions) If you have any questions, do get in touch with Central Ofce ( newphytol@lancaster.ac.uk; tel + 44 1524 594691) or, for a local contact in North America, the USA Ofce (newphytol@ornl.gov; tel 865 576 5251)

www.newphytologist.com New Phytologist (2002) 156: 103 112