WO2004054366A2

WO2004054366A2 - Screening methods for membrane binding fungicides

Info

Publication number: WO2004054366A2
Application number: PCT/BE2003/000222
Authority: WO
Inventors: Bruno Cammue; Karin Thevissen; Isabelle FRANÇOIS; Martina Leipelt; Ernst Heinz; Dirk Warnecke; Kathelijne Ferket
Original assignee: K.U.Leuvent Research And Development
Priority date: 2002-12-16
Filing date: 2003-12-16
Publication date: 2004-07-01
Also published as: AU2003289772A1; AU2003289772A8; GB0229148D0; WO2004054366A3

Abstract

The invention clearly demonstrates that the anti-fungal activity of given plant and insect defensins is mediated via the interaction of said defensins with fungal membrane lipids, said lipids being selected from the group comprising steryl glucosides and sphingolipids, more particularly the glycosphingolipid man nose-(inositol-phosphate)2-ceramide (M(IP)2C) and the glucosylceramides. Based on these findings a method is provided for assessing the anti-fungal properties of a compound characterised in that the method evaluates the interaction of a compound with the steryl glucosides and sphingolipids in or derived from the fungal cell membrane.

Description

ANTI-FUNGAL SCREENING TOOL

FIELD OF THE INVENTION

The present invention provides new fungal-specific, membrane targets allowing the screening of compounds for anti-fungal activity. Said targets comprising fungal glycosphingolipids and steryl glucosides.

BACKGROUND OF THE INVENTION

In medical applications, there is a limited number of antifungal compounds available for the treatment of the extending spectrum of fungi causing severe diseases in immunocompromised patients. This is mainly due to the development of resistance to some of the currently used agents, their low fungicidal activity or severe side effects. Although mammals have diverged from yeasts and fungi over one billion years of evolution, recent studies underscore that much of the basic cellular machinery and signalling are remarkably highly conserved (1, 2). New chemical classes of antifungals, which selectively inhibit or kill fungi, must be identified. In the search for new drug candidates, different strategies can be distinguished. Traditionally, empirical screening examines the ability of different compounds to inhibit the growth of target organisms. The mode of action, selectivity, toxicity and efficacy of active samples can then be further identified and characterized. Consequently, the target sites blocked or influenced by the respective compounds are not predifined and are identified only after the biological activity has been found. This approach has the advantage of simplicity. However, one of the drawbacks is that the search can lead to compounds, which affect or disturb already known biochemical targets.

Plant defensins are small (45-54 amino acids), basic peptides that have a characteristic three-dimensional folding pattern comprising an α-helix and a triple-stranded β-sheet, stabilized by eight disulfide-linked cysteines (reviewed by (2)). They are structurally related to insect and mammalian defensins. Plant defensins can inhibit the growth of a broad range of fungi but are nontoxic to either mammalian or plant cells. Most cationic (positively charged) antimicrobial peptides, including insect and human defensins, induce membrane permeabilization after initial electrostatic binding to negatively charged (phospho)lipids on the target cell surface. In contrast, plant defensins induce membrane permeabilization through specific interaction with high-affinity binding sites on fungal cells and membrane fractions (3-5).

This invention describes the target sites for antifungal plant defensins in the fungal membrane, being sphingolipids (mannosylated inositol phosphoryl sphingolipids and glucosylceramides) and steryl glucosides, and the use of fungal deletion mutants lacking these target sites as screening tools for the discovery of new antifungal compounds with high selectivity. Furthermore, the present invention allows to set up of High Throughput-Screening (HTS) assays to discover inhibitors directed toward binding sites of antifungal plant defensins. In the in fungi and yeasts important sphingolipids are (mannosylated) inositol phosphoryl ceramides (namely IPC, MIPC and M(IP)₂C). There is growing evidence that fungi maintain two separate pools of ceramides to be used for the synthesis of different sphingolipids (8). Ceramide backbones with very long chain C₂₄ and C₂₆ fatty acids bound to the sphingobase 4-hydroxysphinganine are directed to the synthesis of the inositol-containing sphingolipids, whereas ceramide backbones with Cι₆ or Cι₈ fatty acids linked to the sphingobase 9-methyl-4,8-sphingadienine are exclusively used as precursors for biosynthesis of glucosylceramide (GlcCer). In many fungal and yeast species significant amounts of both inositol-containing sphingolipids and glucosylceramides are detected (9, 10). Steryl glucosides are another group of complex lipids present in eukaryotic membranes.

SUMMARY OF THE INVENTION

The present invention clearly demonstrates that the anti-fungal activity of given plant and insect defensins is mediated via the interaction of said defensins with fungal membrane lipids, said lipids being selected from the group comprising steryl glucosides and sphingolipids, more particularly the glycosphingolipid mannose-(inositol- phosphate)2-ceramide (M(IP)₂C) and the glucosylceramides. Therefore, a first object of the present invention is a method for assessing the anti-fungal properties of a compound characterised in that the method evaluates the interaction of a compound with the steryl glucosides and sphingolipids in or derived from the fungal cell membrane. In a first embodiment the interaction of a potential antifungal compound with said membrane lipids is tested in a fungal cell based assay, in a more preferred embodiment the fungal cells have an altered expression of said lipids as compared to their wild type counterparts. In a second embodiment the interaction of a potential antifungal compound with said membrane lipids is tested in a cell independent assay evaluating the binding of said compound to said membrane lipids.

The invention demonstrates for the first time that the anti-fungal activity of Dm-AMP1 , Ah-MAP1 and Ct-AMP1 is mediated via the binding of said plant defensins to fungal membrane sphingolipids and more specifically to mannose-(inositol-phosphate)₂- ceramide (M(IP)₂C) and not to structurally related fungal glucosylceramides, nor other membrane phospholipids. Furthermore, the invention shows that the anti-fungal activity of the radish plant defensin RsAFP2 is mediated via the binding of RsAFP2 to fungal membrane glucosylceramides. In addition it was found that RsAFP2 binds specifically to fungal glucosylceramides, while it had no affinity for animal glucosylceramides. Hence the binding sites for RsAFP2 or DmAMPI in fungal membranes, namely glucosylceramides and mannosylated phosphoryl inositol-containing sphingolipids, respectively, are preferred targets for antifungal compounds for the preventive or therapeutic treatment of animals and/or plants against fungal infections. Therefore, an important embodiment of the present invention is the use of fungal sphingolipids such as glucosylceramides and M(IP)₂C as targets for the screening of anti-fungal compounds.

In a first embodiment the screening for antifungal compounds with RsAFP2-like activity is done by comparing how said compounds affect the growth of a wild type fungal strain and that of a corresponding mutant strain having lower levels of glucosyl ceramides or structurally different glucosyl ceramides in its cell membrane. In a more preferred embodiment, the screening for Rs-AFP2-like activity of the compound is done in parallel with screening for DmAMP1-like activity of the compound by comparing how said compounds affect the growth of a wild type fungal strain and that of a corresponding mutant strain having lower levels of mannosylated inositol phosphoryl ceramides or structurally different mannosylated inositol phosphoryl ceramides in its cell membrane. In a particular embodiment yeast strains are used in the screening..

In a second embodiment the screening for anti-fungal Rs-AFP2-like activity is done by comparing how said compounds affects the permeabilization of the membrane of a wild type fungal strain and that of a corresponding mutant strain having lower levels of glucosyl ceramides or structurally different glucosyl ceramides in its cell membrane. In a more preferred embodiment, the screening for Rs-AFP2-like activity of the compound is done in parallel with screening for DmAMP1-like activity of the compound by comparing how said compounds affect the permeability of the membrane of a wild type fungal strain and that of a corresponding mutant strain having lower levels of mannosylated inositol phosphoryl ceramides or structurally different mannosylated inositol phosphoryl ceramides in its cell membrane. In a particular embodiment yeast strains are used in the screening. In a third embodiment the screening for the anti-fungal activity of compounds is done using an assay system, preferably an ELISA based system, wherein the compounds to be tested compete with a reference compound for binding on fungal membrane sphingolipids. In a preferred embodiment, the reference compound is a defensin or a compound mimicking the antifungal activity of a defensin. In a more preferred embodiment, the defensin is a plant defensin or an insect defensin. In a most preferred embodiment, the reference compound is selected from the group comprising plant defensins from radish (RsAFP2), dahlia (DmAMPI), Aesculus hippocastanum (AhAMPI), Clitorea terneata (CtAMPI) or the insect defensin-like peptide heliomicin.

In a fourth embodiment, the interaction of compounds with fungal, plant and animal sphingolipids is investigated in a BIACORE-based HTS-assay. In this way it will be possible to identify compounds that bind to fungal sphingolipids, but which have no affinity for plant and/or animal sphingolipids. Using such technique, no reference compound for binding has to be used. In a more preferred embodiment, the sphingolipids are glycosphingolipids, more particularly M(IP)₂C and the glucosylceramides.

In a fifth embodiment, the compounds to be tested are screened using a lipid bilayer setup in which a mixture of phosphoglycerolipids, with or without purified sphingolipids, is pasted over the hole separating two buffer chambers. Said lipid bilayer setup allows identifying compounds, which increase the permeability of the sphingolipid- supplemented lipid bilayer at particular voltages.

The invention further demonstrates the involvement of steryl glucosides in the mediation of the antifungal effect of HsAFP Therefore, another object of the invention is a method for assessing the antifungal properties of a compound said method evaluating the interaction of said compound with steryl glucosides. In a preferred embodiment said method comprises an evaluation of the sensitivity of a fungal strain having decreased levels of steryl glucosides towards said antifungal compound and comparing said sensitivity to that of a corresponding wild type fungal strain. In another preferred embodiment said method comprises an evaluation of the sensitivity of a fungal strain having increased levels of steryl glucosides towards said antifungal compound and comparing the sensitivity to that of a corresponding wild type fungal strain

The use of the above described screening methods is particularly valuable in the identification of antifungal peptides, said peptides having a specific affinity for fungal glucosylceramides, (mannosylated) inositol phosphoryl-containing sphingolipids such as M(IP)₂C and steryl glucosides. In a more preferred embodiment said peptides belong to or are structurally related to the family of the defensin peptides.

DETAILED DESCRIPTION OF THE INVENTION

Legends to the figures

Figure 1. Binding of DmAMPI on mannosylated inositolphosphoryl-containing sphingolipids. Dose-response curves are presented for DmAMPI -binding on mannosylated inositolphosphoryl-containing sphingolipids (black squares), fungal glucosylceramides (open triangles) and monogalactosyldiacylglycerols from soybean (open squares). DmAMPI was used at a concentration of 200 ng/ml. Binding on glucosylceramides was determined immunologically by measuring the absorbance at 405 nm. Data are means +- SE of triplicates

Figure 2. Binding of RsAFP2 and heliomicin on glucosylceramides. Dose- response curves are presented for A, RsAFP2- and S, heliomicin-binding on glucosylceramides from P. pastoris (squares), human spleen (triangles) and monogalactosyldiacylglycerols from soybean (circles). RsAFP2 and heliomicin were used at a concentration of 200 ng/ml. Binding on glucosylceramides was determined immunologically by measuring the absorbance at 405 nm. Data are means +- SE of triplicates.

Figure 3. Binding of an inactive variant of RsAFP2 to fungal glucosylceramides. Binding of 250 ng/ml RsAFP2 (open bars) or the biologically inactive variant RsAFP2(Y38G) (dotted bars) to glucosylceramides from P. pastoris and human spleen is presented. Each glycolipid was used at a concentration of 2 μg/ml. Data are means +- SE of triplicates.

Figure 4. RsAFP2-induced membrane permeabilization of P. pastoris parental strain but not of corresponding gcs-deletion mutant. RsAFP2-induced membrane permeabilization was measured by SYTOX Green uptake of parental strain of P. pastoris (black square) and g/cs-deletion mutant (open square). Pretreated yeast cells suspended in SMF1 , were incubated in the presence of different concentrations of RsAFP2 for 2h, after which fluorescence was measured. Data are means +- SE of duplicates Figure 5. Membrane permeabilization and growth inhibition induced by different plant defensins in N. crassa WT, MUT16 and MUT24. Dose-response curves of membrane permeabilization measured by SYTOX Green fluorescence and growth inhibition of N. crassa WT (triangles), N. crassa MUT16 (squares) and N. crassa MUT24 (circles) suspended in SMF1 are shown. Mycelium fragments were treated with DmAMPI , Hs-AFP1 and RsAFP2, and fluorescence was measured at different time points, only the values at 360' are shown here.

Figure 6. HPTLC profile of GIPCs from N. crassa wild-type (lane 2), MUT16 (lane 3) and MUT24 (lane 4).

Figure 7. HPTLC profile of neutral lipids from N. crassa wild-type (lane 1), MUT 16 (lane 2) and MUT24 (lane 3).

Mode of action of plant defensins suggests therapeutic potential

Every living organism, whether it's a mammal, a plant or even a microbe, is confronted with a constant threat of attack by various kinds of pathogens. Despite this threat, illness is the exception. This illustrates that all these organisms must have evolved mechanisms to defend themselves against pathogen attack. The most sophisticated form of pathogen defense is elaborated only in mammals and is based on the specific recognition and subsequent elimination of potential pathogens: the adaptive immune response (12). However, other organisms cannot rely on such a highly developed immune system. Their immune system is based largely on the production of compounds with general antimicrobial properties: the innate immune response (13). This strategy involves, among other responses, the production of cationic antimicrobial peptides (AMPs) that generally have a broad activity spectrum. Until now, only one class of AMPs was found to be conserved between plants, invertebrates and vertebrates, namely defensins. Plant defensins

Defensins are small, highly basic cysteine-rich peptides that share a common three- dimensional structure, (reviewed in (2)). The global fold of plant defensins comprises a cysteine-stabilized αβ motif (CSαβ motif) consisting of an α-helix and a triple-stranded β-sheet, organised in a βαββ architecture and stabilized by four disulfide bridges (2). The plant defensin family is quite diverse regarding amino acid composition: the sequence conservation is restricted to eight structurally important cysteines. Some plant defensins were not found to display any antimicrobial activity, while others were found to have antifungal or antibacterial activity in vitro. Plant defensins appear not to be toxic to either mammalian or plant cells. Insect defensins combine an α-helix and a double-stranded β-sheet stabilized by three disulfide-bridges organised in a CSαβ motif, as for plant defensins (14, 15). An even higher homology to plant defensins is found for the insect defensin-like peptide heliomicin which carries a triple-stranded β- sheet in a βαββ fold forming a CSαβ motif (16). In addition to similarities in their global folds, heliomicin and the radish plant defensin RsAFP2 exhibit very similar distributions of hydrophobic residues (16) and display similar biological activities: both peptides are antifungal rather than antibacterial (17, 18). The similarity in structures of defensins from plants, mammals and insects supports the ancient origin of defensins.

Mode of antimicrobial activity of plant defensins

Most cationic (positively charged) antimicrobial peptides, including insect and human defensins, induce membrane permeabilization after initial electrostatic binding to negatively charged (phospho)lipids on the target cell surface. There are currently two models describing the mode of activity of such cationic peptides (reviewed by (19, 20)). One model postulates the formation of multimeric pores within microbial membranes. After initial electrostatic binding of these positively charged peptides to negatively charged (phospho)lipids on the target cell surface, they insert into the energized cell membrane and most likely form multimeric ion-permeable channels in a voltage- dependent manner. The second model postulates an electrostatic charge-based mechanism of membrane permeabilization, as in case of magainins, cecropins and HBD-2, antimicrobial peptides isolated from frog skin, insect hemolymph and human skin, respectively. They are thought to bind as monomers onto anionic lipid headgroups of the membrane, thereby covering the membrane in a 'carpet-like' manner. The subsequent neutralization of the anionic lipid headgroups disrupts the integrity of the lipid bilayer, causing transient gaps and allowing ions and larger molecules to cross the membrane.

In contrast to mammalian and insect defensins, plant defensins displaying antifungal effects have never been shown to induce ion-permeable pores in artificial membranes composed of phospholipids, nor change the electrical properties of artificial lipid bilayers. This demonstrates that a direct interaction between plant defensins and plasma membrane phospholipids is unlikely (21). Using radiolabeled Heuchera sanguinea plant defensin Hs-AFP1 and dahlia plant defensin Dm-AMP1 , we demonstrated the existence of high-affinity binding sites for these plant defensins on fungal cells and membrane fractions (3, 5). Binding of Hs-AFP1 and Dm-AMP1 to these sites was found to be partially irreversible, yet highly specific, as binding could only be competed for by highly homologous but not by more distantly related plant defensins. Furthermore, plant defensins were found to permeabilize membranes of susceptible yeast and fungal cells (4). In conclusion, plant defensins induce membrane permeabilization through specific interaction with high-affinity binding sites on fungal cells and membrane fractions (3-5).

Recently, the theory that many cationic peptides exert their antimicrobial activity not only through permeabilization of the membrane but in addition have cytoplasmic targets gains support. Binding studies with human α-defensin HNP-1 revealed binding of HNP-1 on both the bacterial plasma membrane/cell wall and the cytosol. In addition, HNP-1 was found to inhibit DNA biosynthesis (22), indicating that HNP-1 has DNA as the secondary intracellular target for antibacterial action. Furthermore, γ- as well as ω- hordothionin, plant defensins isolated from barley endosperm, were found to inhibit protein synthesis in eukaryotic as well as prokaryotic cell-free systems (23). However, it is currently not clear whether these defensins actually exhibit antifungal effects. This and other evidence has led to the suggestion that membrane disruption by itself is not the primal cause of antimicrobial activity of possibly many cationic peptides, but rather inhibition of DNA, RNA or protein synthesis.

Application of plant defensins

Plant defensins can be used as lead molecules for the development of antifungal medicines. Control of eukaryotic pathogens constitutes one of the most challenging problems in medicine. This problem has deepened during the last decades due to the increase of a number of immunocompromised patients, because of human immunodeficiency virus, cancer, and organ transplantation (24). A new threat has emerged due to the appearance of fungal strains with multi-drug resistance and a battery of new species, both yeast and filamentous fungi, which are increasingly recognised as opportunistic pathogens (25). To date, a limited number of antifungal compounds is available for the treatment of an extending spectrum of pathogenic fungi. Currently available antifungals can be grouped into three classes based on their targets, namely azoles, polyenes and fluorinated pyrimidines (reviewed by (19)). Azoles inhibit the synthesis of ergosterol (the main fungal sterol), and cause accumulation of aberrant sterols, hereby impairing membrane functions. Polyenes physicochemically interact with fungal membrane sterols, thereby disrupting membrane integrity and causing leakage of cytoplasmic contents and cell death. Fluorinated pyrimidines affect nucleic acid metabolism. However, due to the development of resistance, there is an urgent need for the development of new chemical classes of antifungals aimed at totally new fungal targets. Toxicity has been a major hurdle in the development of antifungal agents because mammalian cells, in contrast to bacterial cells, share many structures and metabolic pathways with fungal cells. Therefore, the discovery of antifungal agents that possess selective toxicity against the eukaryotic fungal cell remains an important scientific challenge. Antifungal compounds should target molecules that are ubiquitous in fungal cells, but are rarely present in mammalian cells. In the search for new antifungal agents that selectively inhibit or kill fungi, components of the cell membrane or cell wall and virulence factors are currently exploited as putative antifungal targets (24). Moreover, fungal cell walls and membranes are attractive targets for development of antifungal agents because developing resistance would mean that microbes have to change very fundamental elements of their cells.

Membrane components in the fungal plasma membrane that present an attractive new target are sphingolipids. Several inhibitors of sphingolipid biosynthesis have been discovered in recent years, some of which act at a biosynthetic step unique to fungi and have potent and selective antifungal activity, such as aurebasidin, rustmicin, khafrefungin and australifungin (reviewed by (19)). Therapeutic potential of plant defensins

Various insect and mammalian defensins are currently being tested in clinical trials to examine their use in combating bacterial and fungal infections (27). The company Micrologix Biotech Inc. developed different defensin variants which are currently evaluated in a late-stage clinical trial to prevent catheter-related infections caused by bacteria such as Pseudomonas aeruginosa and Staphylococcus aureus, and for their potential to cure severe acne and to prevent acute S. aureaus infections. The French company EntoMed is conducting preclinical studies with a defensin-like antifungal insect peptide to combat Candida albicans and Aspergillus fumigatus, which often cause fatal infections in immunocompromised patients (26, 27)]. The present invention is based on the observation that plant defensins, in contrast to insect and mammalian defensins, interact with specific structures in the fungal membrane, such as phosphorylinositol-containing sphingolipids or glucosylceramides. This finding points towards a high selectivity and hence, interesting perspectives for treatment of fungal infections. Importantly, various plant defensins such as Dm-AMP1 , Hs-AFP1 and Rs- AFP2 are active against the human pathogen Candida albicans at micromolar concentrations (Table 3) supporting the potential of plant defensins for the development of therapeutics. Since the mode of action of only a few plant defensin members has been studied in detail yet, this family of antimicrobial peptides might represent a pool of membrane-targetting compounds with a great potential to be used in the development of future therapeutic antifungals.

Example 1: Materials and Methods

Materials and Microorganisms

The antifungal peptides DmAMPI , HsAFP , AcAMPI , AceAMPI , lbAMP4, RsAFP2 and RsAFP2(Y38G) were isolated as described previously (11, 18, 29-32). Anti-

RsAFP2 and -DmAMPI serum from rabbit was purified as described previously (18).

Anti-glucosylceramide serum from rabbit was kindly provided by Dr. L. Brade

(Research Center Borstel, Center for Medicine and Biosciences, Borstel, Germany).

Heliomicin and anti-heliomicin rabbit serum were kindly provided by Dr. J.L. Dimarcq (Entomed, Strasbourg, France) and Dr. P. Bulet (Institut de Biologie Moleculaire et

Cellulaire, CNRS, Strasbourg, France). Phosphatase-coupled goat anti-rabbit immunoglobulins and glucosylceramides isolated from human spleen were purchased from Sigma (St. Louis, MO). SYTOX Green was obtained from Molecular Probes (Eugene, Oreg.). Ethyl methanesulfonate (EMS) was obtained from Sigma. SYTOX Green was purchased from Molecular Probes (Eugene, Oreg.). Yeast strains used in this study are Saccharomyces cerevisiae strain BY4741 (Invitrogen), Pichia pastoris strain GS115 (Invitrogen) and the corresponding P. pastoris ρ/cs-deletion strain GS115zlgcs (8), Candida albicans strain SC5314 CAI4 (33) and the corresponding C. albicans gcs-deletion strain (8). Neurospora crassa (strain 74-OR23-1A, FGSC collection number 987, termed wild-type) and its mutants were grown on half strength potato dextrose broth agar (PDA, 12 g/l PDB, Difco, Detroit, Mich.; 15 g/l agar, Difco).

Antifungal Activity Assay

Antifungal activity of protein samples against yeast strains was assayed by microspectrophotometry of liquid cultures grown in microtiter plates as described previously (4-6). Growth media used were either Y_ PDB (12 g/l Potato Dextrose Broth, Sigma), YPD (10 g/l Yeast Extract, Difco; 20 g/l Peptone, Difco; 20 g/l glucose), PDB (24 g/l Potato Dextrose Broth, Sigma) or PDB/YPD (24 g/l PDB; 2 g/l Yeast Extract, Difco; 4 g/l Peptone, Difco; 4 g/l glucose), supplemented with 50 mM HEPES, pH 7.0., unless otherwise stated.

Antifungal activity of protein samples against N. crassa wild-type and N. crassa mutants was assayed as described previously (34), except that suspensions of mycelium fragments were used instead of spore suspensions. Briefly, fragments from the edges of the mycelium lawns were transferred to 50 ml half-strength PDB and incubated for 24-48 h at 25°C with continuous shaking. 500 μl aliquots of these cultures were transferred to 2 ml polypropylene microcentrifuge tubes with o-ringed screw caps each containing 5 glass beads (1mm diameter). The mycelium was fragmented by high speed reciprocal shaking using a Phastprep apparatus (Bio 101/Savant, Farmingdale). The obtained mycelium fragment suspensions were 100-fold diluted for use in antifungal activity assays. Growth medium was either 1/2 PDB or Synthetic Medium for Fungi 1 (SMF1 , (4)).

SYTOX Green Uptake

Membrane permeabilization of the yeast P. pastoris was measured as described previously by (4), with minor modifications. Briefly, cells of an overnight culture of P. pastoris in PDB/YPD were incubated with RsAFP2 or 0.5% SDS in PDB/YPD for 2 h. After incubation, cells were washed with SMF1 (4) supplemented with 0.25 μM SYTOX Green and 50 mM HEPES, pH 7.0. Hundred μl-aliquots of this cell suspension were incubated with RsAFP2 in white 96-well microplates (PE white; Perkin-Elmer, Norwalk, Conn.) for 2 h at RT, whereafter fluorescence was measured. Membrane permeabilization induced by 0.5 % SDS corresponds to 100 % membrane permeabilization.

Membrane permeabilization of the fungus N. crassa was measured as described previously (4) except for the preparation of the suspension of mycelium fragments. N. crassa WT (~10⁵ conidiospores/ml) and mutants (fragments from the mycelium on the agar plates) were incubated in 50 ml 1/2 PDB for 20 h and 36 h, respectively. Fragmentation of the mycelium in SMF1 was achieved by vortexing vigorously. Aliquots of this mycelium suspension were used in the SYTOX Green uptake assay.

Purification, Quantification and Analysis of Glucosylceramides

Purification of glucosylceramide from P. pastoris was performed as previously descibed (9). Purification of monogalactosyldiacylglycerols from soybean (Glycine max) and quantification of glucosylceramides from P. pastoris, human spleen and monogalactosyldiacylglycerols was performed as described (35).

Microtiter Plate Binding Assay (ELISA)

Binding of antifungal peptides with glucosylceramides was evaluated by using an ELISA-based assay as described previously (36, 37). Stock solutions of all glycolipids were prepared in methanol:chlorophorm:water (16:16:5; vol:vol:vol) at a concentration of 500 μM. Lipids were applied in 75-μl aliquots to the wells of microtiter plates and allowed to dry overnight at RT. All subsequent handling steps were performed at 37°C. Blocking buffer was 3% (w/v) BSA (Sigma) in PBS and washing buffer was 10% blocking buffer. Anti-RsAFP2, anti-DmAMP1 , anti-heliomicin rabbit antiserum and phosphatase-coupled goat anti-rabbit immunoglobulin were 1000-fold diluted in washing buffer. Plotted values are means of triplicates adjusted for the plate background. Plate background values are the absorbance readings of methanol:chlorophorm:water (16:16:5; vol:vol:vol)-coated wells incubated with peptides, and antisera. Chemical mutagenesis of N. crassa and isolation of RsAFP2-resistant N. crassa mutants

Conidiospores of the N. crassa wild-type were mutagenised by treatment with ethylmethanesulfonate (EMS). To this end, 100 μl aliquots of conidiospores (~10⁷ spores/ml in 20 % (v/v) glycerol) were resuspended in 100 μl 100 mM sodium phosphate (pH 7.5). After adding 1 μl EMS solution, the spore suspension was incubated at 25°C with continuous shaking (300 rpm). After 45', EMS was inactivated by addition of 200 μl 5% thiosulfate. To calculate the rate of spontaneous mutagenesis, an additional aliquot of conidiospores was treated similarly, except that EMS was omitted. Spore suspensions were diluted in 70 ml half-strength PDB containing 4 μM RsAFP2. The suspensions were then divided in 100 μl subcultures in 96-well flat- bottom microtiterplates (Greiner) which were subsequently incubated at room temperature for 5 days and scored for growth on a regular base. Putative mutants were transferred to PDA plates.

Extraction, purification of lipid fractions and High performance thin layer chromatography (HPTLC)

Extraction, purification of lipid fractions from fungal mycelium and HPTLC were performed as described previously (49-52).

EXAMPLE 2: Mode of action of plant defensins isolated from dahlia (Dm-AMP1), Aesculus hippocastanum (Ah-AMPI) and Clitoria te neata fCt-AMP1)

Via a genetic complementation approach, IPT1 was identified as a gene determining sensitivity of Saccharomyces cerevisiae towards Dm-AMP1 (6). This gene encodes inositol phosphotransferase, an enzyme involved in the last step of the synthesis of the sphingolipid mannose-(inositol-phosphate)₂.ceramide (M(IP)₂C. Strains with a nonfunctional IPT1 allele lacked M(IP)₂C in their membranes, bound significantly less DmAMPI compared to parental S. cerevisiae strain, and were highly resistant to DmAMPI -induced membrane permeabilization (6). It was shown previously that lowered levels of M(IP)₂C or altered levels of other sphingolipids may act to inhibit or stimulate some ABC transporters involved in pleiotropic drug resistance, like Pdrδp and Yorlp respectively, resulting in different resistance patterns (38). Since elimination of IPT1 does not lead to enhanced resistance to other plant defensins, like Hs-AFP1 (results not shown), Dm-AMP1 resistance in S. cerevisae Aiptl is most likely not the result of inhibition and/or stimulation of ABC-transporters.

Sphingolipids are one of the three major types of lipids found in eukaryotic membranes, along with sterols and phosphoglycerolipids. They associate with sterols in the plasma membrane to form patches or rafts that are highly enriched in glycosyl-phosphatidyl- inositol (GPI)-anchored membrane proteins (39). Possibly, membrane patches containing sphingolipids act as binding sites for dahlia plant defensin. The interaction facilitates the insertion of DmAMPI in the plasma membrane, leading to alterations in membrane permeability and finally to fungal growth arrest. Alternatively, GPI-anchored proteins present in sphingolipid rafts could act as docking sites for these defensins and facilitate their insertion into the fungal plasma membrane, leading to alterations in membrane permeability (6).

Interestingly, binding of Dm-AMP1 could be competed for by the highly homologous defensins Ah-AMP1 and Ct-AMP1 , isolated from Aesculus hippocastanum and Clitoria ternatea, respectively (5). This suggests that these defensins share the same binding site. Indeed, elimination of IPT1 not only leads to Dm-AMP1 resistance but also to Ah- AMP1 and Ct-AMP1 resistance. Distantly related plant defensins, like Heuchera sanguinea Hs-AFP1 , are not able to compete for the Dm-AMP1 binding site and elimination of IPT1 does not lead to enhanced resistance to Hs-AFP1 (results not shown). This demonstrates that Hs-AFP1 and Rs-AFP2 probably have distinct binding sites on the plasmamembrane than Ah-AMP1 , Ct-AMP1 and Dm-AMP1.

Results and Discussion

In order to determine the identity of the DmAMPI -binding sites on the fungal membrane, several experiments were performed. First, we observed that S. cerevisiae

mutant, grown in 34 PDB medium, still produces M(IP)₂C, possibly via a rescue pathway (results not shown). In V_ PDB, /pf7-deletion mutant is as sensitive towards DmAMPI as wild type strain (Table 1). Apparently, DmAMPI -sensitivity is not linked with presence of functional /PT7-encoding protein (Iptlp) but with the presence of M(IP)₂C in the fungal membrane. Resistance to Dm-AMP1 seems to depend on modifications in sphingolipid composition rather than on alterations in Iptlp itself. Therefore, the membrane sphingolipid M(IP)₂C or proteins stabilized by M(IP)₂C are involved in constituting the binding site for Dm-AMP1. Sphingolipids associate with sterols in the plasma membrane to form patches (also called rafts) that are highly enriched in so called glycosyl-phosphatidyl-inositol (GPI)- anchored membrane proteins (39). GPI-anchored proteins are found in all eukaryotic cells. In yeast, GPI-proteins are found at the cell surface, either attached to the plasma- membrane or as an intrinsic part of the cell wall. Use of the Von Heijne algorithm and homology searches allowed the identification of 58 ORFs encoding putative GPI- anchored proteins. Families of plasma-membrane and cell wall proteins were assigned, revealing 20 potential plasma-membrane and 38 potential cell wall proteins (40). To study the possible involvement of GPI-anchored proteins in DmAMPI sensitivity, yeast disruptants in (non-essential) GPI-anchored proteins were tested for increased resistance towards DmAMPI (see Table 2). None of these yeast deletion mutants showed an increased level of resistance against DmAMPI compared to wild-type yeast. We also constructed various double and triple deletion mutants in GPI-CWPs (see Table 3). However, all tested deletion mutants were as sensitive to DmAMPI as wild type, indicating that none of the tested GPI-anchored proteins is necessary for the antifungal activity of DmAMPI .

To investigate whether the membrane sphingolipid M(IP)₂C interacts directly with DmAMPI , we used an ELISA-based binding assay in which glycolipids are coated to the wells of microtiter plates and interacting peptides are detected immunologically. Different (glyco)lipids were used in the assay: S. cerevisiae (mannosylated) inositol phosphoryl-containing sphingolipids (mixture of IPC, MIPC and M(IP)₂C), P. pastoris glucosylceramides, and monogalactosyldiacylglycerols from soybean.

Using the ELISA-based binding assay, DmAMPI was found to bind in a dose- dependent manner to (mannosylated) inositol phosphoryl-containing sphingolipids from S. cerevisiae. (Fig. 1). No significant binding of DmAMPI to different concentrations of fungal glucosylceramides nor to soybean monogalactosyldiacylglycerols could be detected, indicating that binding of DmAMPI to mannosylated) inositol phosphoryl- containing sphingolipids from S. cerevisiae is specific.

In conclusion, the presented results demonstrate for the first time a direct interaction of DmAMPI with specific sphingolipid structures in fungal membrane, being the mannosylated phosphoryl inositol-containing sphingolipids, possibly M(IP)₂C. EXAMPLE 3: Identification of the mode of action of radish plant defensin Rs- AFP2 as an anti-fungal compound and development of a screening method

P. pastoris and C. albicans mutant strains lacking glucosylceramide are resistant to RsAFP2 To investigate a possible role of glucosylceramides in RsAFP2-mediated growth inhibition, mutants of P. pastoris and C. albicans that are completely devoid of glucosylceramides were used (8). The gcs-deletion mutants and their corresponding parental strains were tested for sensitivity to growth inhibition by RsAFP2. C. albicans and P. pastoris are sensitive to RsAFP2 at concentrations of 1-2 μM and higher, whereas the corresponding gcs-deletion strains are at least 20-fold more resistant to RsAFP2 (Table 4). As a control, DmAMPI-sensitivity of all yeast strains was assessed. DmAMPI is equally active on both gcs-deletion mutants and the corresponding parental strains. S. cerevisiae and Schizosaccharomyces pombe, two yeast species that lack glucosylceramides in their membranes (9, 10), are fully resistant to RsAFP2 (Table 4).

RsAFP2 binds specifically to fungal glucosylceramide

To investigate whether the membrane lipid glucosylceramide interacts directly with RsAFP2, we used an ELISA-based binding assay in which glycolipids are coated on the wells of microtiter plates and interacting peptides are detected immunologically. Different (glyco)lipids were used in the assay: P. pastoris glucosylceramides, glucosylceramides from human spleen and soybean monogalactosyldiacylglycerols. The P. pastoris glucosylceramide contains a glucose moiety linked to a ceramide backbone, which consists of an unsaturated, methyl-branched sphingobase and a long chain α-hydroxy fatty acid (8, 9). The mammalian glucosylceramides used in this study consist of long chain non-hydroxy fatty acids linked to glucosylated 4-sphingenine base.

Using the ELISA-based binding assay, RsAFP2 was found to bind in a dose-dependent manner to purified glucosylceramide from P. pastoris (Fig. 2A). No significant binding of RsAFP2 to different concentrations of human glucosylceramides or soybean monogalactosyldiacylglycerols could be detected, indicating that binding between RsAFP2 and P. pastoris glucosylceramides is specific. Link between the antifungal activity of RsAFP2 and its interaction with fungal glucosylceramides

To address the link between the antifungal activity of RsAFP2 and its interaction with fungal glucosylceramides, the interaction of an RsAFP2 variant that is devoid of antifungal activity (RsAFP2(Y38G)) (32) with fungal glucosylceramides was assessed. Similarly to RsAFP2, RsAFP2(Y38G) interacts with glucosylceramides from P. pastoris, but not with glucosylceramides from human origin (Fig. 3). These results indicate that RsAFP2(Y38G) is competent for glucosylceramide binding, but is impaired in the induction of fungal growth inhibition. Hence, interaction of RsAFP2 with fungal glucosylceramides is not sufficient to induce fungal growth arrest. Based on this observation, it could be anticipated that P. pastoris cells pretreated with RsAFP2(Y38G) should be more resistant to subsequent treatment with RsAFP2. Therefore, we investigated a putative antagonism between RsAFP2(Y38G) and RsAFP2 by preincubating P. pastoris cells with either 40 μM RsAFP2(Y38G) or water and assessing the sensitivity of the preincubated P. pastoris cells to RsAFP2. The antifungal activity of RsAFP2 was not affected by the presence of RsAFP2(Y38G): the IC₅₀ value of RsAFP2 was the same in the absence or presence of RsAFP2(Y38G) (IC50 = 2 μM). Hence, no antagonism between RsAFP2(Y38G) and RsAFP2 was observed. This can be explained by the finding that some plant defensins reversibly bind to fungal and yeast cells (5). In this respect, it is possible that RsAFP2(Y38G), which does not display antifungal activity and therefore probably can not reach and/or affect the fungal target, can be displaced from its primary binding site by RsAFP2. Hence, this approach is not appropriate to demonstrate the link between the antifungal activity of RsAFP2 and its interaction with fungal glucosylceramides. Therefore, we investigated whether anti-glucosylceramide antibodies can inhibit the antifungal action of RsAFP2 by binding to glucosylceramides and hence, blocking the RsAFP2-target on the fungal plasma membrane. To this end, we have preincubated a P. pastoris cell suspension with either a rabbit antiserum against glucosylceramides (1/20 diluted), a non-immune rabbit serum (1/20 diluted) as a control, or water. The antifungal activity of RsAFP2 was found to be 4-fold reduced in the presence of the rabbit antiserum against glucosylceramides (IC₅₀ of RsAFP2 = 8 μM) as compared to the appropriate controls (IC_S0 of RsAFP2 = 2 μM). From these data, it can be concluded that the antifungal action of RsAFP2 can be reduced by blocking its target, namely fungal glucosylceramides, and that the observed interaction between RsAFP2 and fungal glucosylceramides in the ELISA-based binding assay is relevant for its antifungal activity.

RsAFP2 induces permeabilization of P. pastoris cells but not of the corresponding gcs-deletion mutant

In order to determine the role of glucosylceramides and GCS in RsAFP2-mediated membrane permeabilization of yeast cells, an assay based on the uptake of SYTOX green was used (4). SYTOX green is an organic compound that only penetrates cells with compromised plasma membranes and fluoresces upon interaction with nucleic acids (41). RsAFP2-induced SYTOX green uptake was assessed in P. pastoris parental strain and the corresponding gcs-deletion mutant. SYTOX green uptake in P. pastoris parental strain increases significantly upon treatment with RsAFP2 at concentrations higher than 2 μM (Fig. 4). No RsAFP2-induced membrane permeabilization can be detected in the gcs-deletion mutant.

Conclusion

We present for the first time evidence that the radish antifungal defensin RsAFP2 interacts directly with fungal glucosylceramides. Interaction of RsAFP2 to fungal glucosylceramides was found to be very specific since these petides did not interact with human glucosylceramides nor glycolipids from soybean. This is possibly due to differences in ceramide structure between fungal and human glucosylceramides. Glucosylceramides from various fungal and yeast species, such as Pichia pastoris, Candida albicans (9), Cryptococcus neoformans (42), Aspergillus fumigatus (43) and various Fusarium species (44), were characterized and their ceramide structure was found to carry the characteristic methyl-branched fungal sphingobase 9-methyl-4,8- sphingadienine. Since animals contain a large variety of different glucosylceramide structures but never possess a methyl group at C9 of the sphingobase, this methyl group might be an important determinant for binding specificity.

EXAMPLE 4: Use of the GCS-deletion strains in the elucidation of the mode of action of the insect defensin heliomicin

The three-dimensional structure of heliomicin, an insect defensin-like peptide from the lepidopteran Heliothis virescens, is highly homologous to radish plant defensin RsAFP2

(2, 46). In addition, both peptides display similar biological acitivities. Possible involvement of fungal glucosylceramides in heliomicin-induced antifungal activity was assessed. As for RsAFP2, growth of C. albicans and P. pastoris is inhibited by heliomicin starting from 2 μM, whereas gcs-deletion mutants of both yeast strains are at least 20-fold more resistant (Table 3). Similarly to RsAFP2, binding of heliomicin to fungal glucosylceramides was found to be very specific (Fig 2B): heliomicin binds in a dose-dependent manner to P. pastoris glucosylceramides and not to human glucosylceramides or soybean monogalactosyldiacylglycerols.

In addition, the ability of RsAFP2 to compete with heliomicin for binding on glucosylceramides from P. pastoris was tested. Binding of heliomicin to Pichia glucosylceramides could not be competed for by a 50-fold excess of RsAFP2 nor DmAMPI and vice versa, indicating that heliomicin and RsAFP2 are interacting with different domains of fungal glucosylceramide.

Conclusion

We have previously discussed the evolutionary conservation of defensins between eukaryotes and noticed that there is a close relationship between plant defensins, insect defensins and mammalian defensins which led to the hypothesis that all these types have evolved from a single precursor (2).

The presented data further substantiates this hypothesis, as the binding site of a specific plant defensin and an insect defensin-like peptide on susceptible fungi seems to be conserved, namely glucosylceramides. It is for the first time that it has clearly been demonstrated that related defense components from organisms belonging to different kingdoms target the same microbial structure.

Interestingly, the three-dimensional structure of plant defensins, such as DmAMPI and RsAFP2, is stabilized by 4 disulfide bridges, whereas the three-dimensional structure of the insect peptide heliomicin is stabilized by 3 disulfide bridges. Structurally, based on the number of disulfide bridges, DmAMPI is more related to RsAFP2 than heliomicin. However, functionally, heliomicin seems to interact with the same fungal membrane structures as RsAFP2, in contrast to DmAMPI

The observation that binding of heliomicin could not be competed for by RsAFP2 and vice versa indicates that heliomicin and RsAFP2 bind to different domains of the fungal glycosylceramides. EXAMPLE 5: Importance of membrane lipids in the resistance of fungi against plant defensins

We screened several Neurospora crassa mutants, obtained by chemical mutagenesis, for increased resistance to various plant defensins. In addition, these N. crassa mutants were tested for their cross-resistance towards other families of structurally different antimicrobial plant proteins. Based on the previously demonstrated key-role of inositolphosphoryl-containing sphingolipids and GlcCer in the mechanism of fungal growth inhibition by certain defensins (6, 35), two N. crassa mutants displaying defensin-resistance were characterized with respect to defensin-induced membrane permeabilization and complex lipid composition.

Chemical mutagenesis of N. crassa and isolation of RsAFP2-resistant N.crassa mutants

N. crassa mutants were generated by chemical mutagenesis of N. crassa wild-type using ethylmethanesulphonate (EMS) and selected for resistance towards 4 μM of the radish plant defensin RsAFP2. The frequency of appearance of RsAFP2-resistant mutants was 8 times higher in chemically mutagenised populations compared to non- mutagenised populations. Twenty-five N. crassa mutants were isolated (MUT1 to MUT25) and characterized further.

Antifungal activity assay and cross resistance ofN. crassa mutants

N. crassa wild-type grows superficially on PDA plates and is pink-orange coloured. In contrast to wild-type, all RsAFP2-resistant N. crassa mutants have a white and very compact appearance on solid medium. They tend to grow into the agar and grow at least 10-fold slower than the N. crassa wild-type. In addition, the ability to form ascospores seems to be lost for most of these mutants. Therefore, antifungal activity assays were performed on outgrowing mycelium fragments of these mutants. Resistance and cross-resistance of these mutants towards different types of antifungal peptides was evaluated. These included the plant defensins RsAFP2, Hs-AFP1 and DmAMPI , isolated from radish (Raphanus sativus), Heuchera sanguinea and dahlia (Dahlia merckii) respectively, and three structurally unrelated antimicrobial plant proteins, namely AceAMPI , extracted from onion seeds and shares structural analogies to plant nonspecific lipid transfer proteins (30), AcAMPI , isolated from Amaranthus caudatus seeds and shares sequence homology to the cysteine/glycine- rich domain of chitin-binding proteins (18) and IbAMPI , a highly basic cysteine-rich peptide from seeds of Impatiens balsamina (20 amino acid residues) (31).

As summarized in Table 5, mutants can be divided in three groups based on sensitivity to different antimicrobial proteins. The largest group of mutants shows resistance to all antifungal peptides tested (group I). A second group (II) of mutants was found to be resistant to all plant defensins, but remains sensitive towards the structurally unrelated antifungal peptides. A last group (group III) consists of mutants with various resistance patterns towards the different antimicrobial compounds tested. Because of their specific resistance to plant defensins, mutants of the second group (MUT 16 and MUT24) were chosen for further study. Both mutants appeared to be at least 16-fold more resistant to the plant defensins tested as compared to N. crassa wild-type. The specific resistance pattern to plant defensins could be extended to structurally related peptides of non-plant origin, such as the pathogen-inducible antifungal insect peptides heliomicin and termicin. Heliomicin is an antifungal protein produced by the lepidopteran Heliothis virescens, that displays a relatively high sequence similarity with antifungal plant defensins (16). Termicin, an antifungal peptide isolated from the termite Pseudacanthotermes spiniger, is stabilized by six cysteines that are arranged in a similar way as in insect defensins. N. crassa mutants MUT16 and MUT24 are 16-fold more resistant to termicin than N. crassa wild-type (Table 6). For heliomicin, this difference in sensitivity is even larger: MUT16 and MUT24 are at least 500-fold more resistant to heliomicin as compared to the N. crassa wild-type.

SYTOX Green uptake assay

Possible membrane permeabilization induced by RsAFP2-, Hs-AFP1- and DmAMPI in these mutants was tested using a SYTOX Green uptake assay. SYTOX Green is a high affinity nucleic acid stain that fluoresces upon nucleic acid binding. It can only penetrate cells with compromised plasma membranes (41 , 47, 48). Previously, we showed that DmAMPI , RsAFP2 and Hs-AFP1 induce membrane permeabilization on intact hyphae of N. crassa wild-type (4). Similarly, these plant defensins also induce membrane permeabilization on mycelium fragments of N. crassa wild-type (Fig. 5). In case of N. crassa mutants MUT16 and MUT24, no significant permeabilization was detected upon incubation of mycelium fragments in the presence of RsAFP2, Hs-AFP1 and DmAMPI (Fig. 5). . HPTLC profile of neutral and acidic lipid fractions of N. crassa WT, MUT16 and MUT24

Previously, we showed that the sphingolipid M(IP)₂C, present in membranes of S. cerevisae (6), plays an important role in the mode of action of dahlia plant defensin DmAMP Since N. crassa mutants MUT16 and MUT24 are resistant towards DmAMPI , we performed an High Performance Thin Layer Chromatography (HPTLC)- analysis of the acidic fraction of lipids representing mainly glucosylinositol phosphorylceramides (GIPCs) of N. crassa WT, MUT16 and MUT24. Bands are visualised upon orcinol staining, i.e. violet staining indicating the presence of hexose. In the HPTLC of Figure 6, the corresponding orcinol⁺ acidic lipid GIPC-profiles of N. crassa WT, MUT16 and MUT24 are compared. The WT profile appears to be relatively simple, exhibiting ~4 bands with Rf values higher than those in the mutant profiles. These may represent mono- or diglycosyl-IPCs, possibly segregated into two components each based on differences in hydroxylation of the ceramide fatty-Λ/-acyl moiety. By contrast, the mutant GIPC profiles appear to be distinct from the WT, but similar to each other. A multiplicity of lower Rf (more polar) components are visible in the mutant GIPC profiles.

Unlike S. cerevisiae, most fungi synthesize not only GIPCs but also neutral monohexosylceramides, such as glucosylceramides (49). Such glucosylceramides play a key role in the sensitivity of the yeast P. pastoris to RsAFP2. Therefore, we also analyzed neutral lipid fractions comprising mainly cerebrosides and steryl glucosides of N. crassa WT, MUT16 and MUT24. In Figure 7, an HPTLC of neutral lipid fractions extracted from WT, MUT16, and MUT 24 strains of N. crassa (Lanes 1-3, respectively) is shown. Three orcinol⁺ bands can be observed for each strain. Of these, components having Rf values characteristic for steryl β-glucoside (GlcSte) and β-glucosylceramide (GlcCer) were identified by comparison with standard compounds (see bands marked in Figure 7). The third orcinol⁺ band, having the lowest Rf value, did not comigrate with any standard compound. Preliminary analysis of the unknown low Rf component indicates that it is not a glycolipid, but rather a small oligosaccharide fragment. Interestingly, although the amounts of GlcCer are similar in all three stains, both mutants express considerably more GlcSte than the N. crassa WT. However, since GlcCer are binding sites for radish plant defensin RsAFP2 in fungal membranes, the structure of putative GlcCer components of wild-type and mutant N. crassa strains was analysed by ¹H-NMR. The GlcCer components in these strains were identified as 2- hydroxy-fatty-/V-acyl-(4E,8£ -9-methyl-4,8-sphingadienines. 9-Methyl-4,8- sphingadienine is the sphingobase present in glucosylceramides from various fungal and yeast species. GlcCer of N. crassa WT and mutants MUT16 and MUT24 differ remarkably in the expression of one key structural feature — while the fatty-Λ/-acyl moiety in the WT GlcCer is essentially 100% saturated, those of the mutant GlcCer express a high level (>95%) of fatty-N-acyl £-Δ³-unsaturation, as measured by the amplitude of characteristic E-vinyl proton signals in the NMR spectra (43, 49-51). Among the higher Rf components (low polarity, orcinol/hexose-) observed in the HPTLC, an abundant brownish stained band appears in the WT profile, but is nearly absent in the mutant profiles. The reciprocal expression of this band with respect to GlcSte suggests that it may represent free or O-acylated sterol.

Discussion

Twentyfive N. crassa mutants were generated by EMS mutagenesis and selected for resistance towards Rs-AFP1. The N. crassa mutants were divided into three major groups with respect to their sensitivity towards different antimicrobial proteins. The first group consists of N. crassa mutants showing resistance to all antifungal peptides tested. Such broad resistance towards structurally unrelated peptides might be due to an overall reduction in cell wall and/or membrane permeability. The N. crassa mutants of the second group, are resistant to all plant defensins tested, but they remain sensitive towards antifungal peptides which are not structurally related to plant defensins. A more specific mechanism must be the basis for this multi-resistance to members of the plant defensin family. By further characterization of two representatives of this second group, namely MUT16 and MUT24, in comparison with N. crassa wild-type, differences in plant defensin induced membrane permeablization were demonstrated. Additionally clear differences in the lipid profiles of N. crassa wild- type and mutants were shown. Analysis of the acidic lipid fraction revealed a novel GIPC structure in the N. crassa mutants MUT16 and MUT24 in comparison with the N. crassa wild-type phosphatidylinositol compounds. In addition, cerebrosides of N. crassa wild-type and N. crassa mutants appear to be structurally different: 90-95% of GlcCer from wild-type have delta-3 unsaturation in their fatty acid moiety, whereas the fatty acid moiety of N. crassa mutants MUT16 and MUT24 GlcCer is fully saturated. Furthermore, there is a clear accumulation of steryl glucosides in MUT16 and MUT24 as compared to N. crassa wild-type. These results indicate that the changes in membrane lipid composition in N. crassa mutants MUT16 and MUT24 are correlated with decreased plant defensin induced membrane permeabilization and resistance towards different plant defensins. This finding is in line with the data presented above showing that sphingolipids and glucocylceramides are sensitivity determinants for DmAMPI and Rs-AFP-2, respectively. Information on the mode of action of Dm-AMP1 and Rs-AFP2 indicates that the specific structure of glycosphingolipids in the fungal plasmamembrane, namely M(IP)₂C and GlcCer respectively, determines permeability, binding and resistance towards these plant defensins. In case of Hs-AFP1 , none of these glycolipids seems to be involved in the mode of action. Elimination of IPT1 in S. cerevisiae does not lead to enhanced resistance to Hs-AFP1 and also no increased resistance to Hs-AFP1 can be detected in the P. pastoris Ages deletion mutant as compared to P. pastoris wild-type. In contrast, the higher level of steryl glucosides in N. crassa mutants MUT16 and MUT24 might explain the resistance to Hs-AFP1. In the yeast P. pastoris, Ugt51p catalyses the biosynthesis of steryl glucosides; a lack of this enzyme leads to reduced levels of steryl glucosides (52). A P. pastoris Δugt51 strain, lacking the enzyme UDP- glucose:sterol glucosyltransferase, is a factor 4 more sensitive towards HsAFPI as compared to P. pastoris wild-type (results not shown), whereas this mutant is as sensitive towards DmAMPI and HsAFPI as wild-type. High amounts of sterol glucosides in the membrane seem to be linked with increased resistance towards HsAFPI. The precise role of steryl glucosides in the path leading to growth inhibition by HsAFPI remains to be elucidated.

In conclusion, membranes of the N. crassa mutants with broad resistance towards different members of the plant defensin family characterized in this study contain structurally different cerebrosides, novel glycosphingolipids and an altered level of steryl glucosides, underlining the important role of these membrane lipids in the mediation of the anti-fungal effect of plant defensins.

REFERENCE LIST

1. Cardenas, M. E., Cruz, . C, Del Poeta, M., Chung, N., Perfect, J. R. & Heitman, J. (1999) Clin. Microbiol. Rev. 12, 583-611. 2. Thomma, B. P. H. J., Cammue, B. P. A. & Thevissen, K. (2002) Planta (DO1 10.

1007/S00425-002-0902-6).

3. Thevissen, K., Osborn, R. W., Acland, D. P. & Broekaert, W. F. (1997) J. Biol. Chem. 272, 32176-32181.

4. Thevissen, K., Terras, F. R. & Broekaert, W. F. (1999) Appl. Environ. Microbiol. 65, 5451-5458.

5. Thevissen, K., Osborn, R. W., Acland, D. P. & Broekaert, W. F. (2000) Mol. Plant Microbe Interact. 13, 54-61.

6. Thevissen, K., Cammue, B. P., Lemaire, K., Winderickx, J., Dickson, R. C, Lester, R. L, Ferket, K. K., Van Even, F., Parret, A. H. & Broekaert, W. F. (2000) Proc. Natl. Acad. Sci. U. S. A 97, 9531-9536.

7. Dickson, R. C, Nagiec, E. E., Wells, G. B., Nagiec, M. M. & Lester, R. L (1997) J. Biol. Chem. 272, 29620-29625.

8. Leipelt, M., Warnecke, D., Zahringer, U., Ott, C, Muller, F., Hube, B. & Heinz, E. (2001) J. Biol. Chem. 276, 33621-33629. 9. Sakaki, T., Zahringer, U., Warnecke, D. C, Fahl, A., Knogge, W. & Heinz, E. (2001)

Yeast 18, 679-695.

10. Dickson, R. C. & Lester, R. L. (1999) Biochim. Biophys. Acta 1426, 347-357.

11. Osborn, R. W., De Samblanx, G. W., Thevissen, K., Goderis, I., Torrekens, S., Van Leuven, F., Attenborough, S., Rees, S. B. & Broekaert, W. F. (1995) FEBS Lett. 368, 257-262.

12. Matsunaga, T. & Rahman, A. (1998) Immunol. Rev. 166, 177-186.

13. Boman, H. G. (2000) Immunol. Rev. 2000. Feb. ;173. :5-16. 173, 5-16.

14. Bonmatin, J. M., Bonnat, J. L, Gallet, X., Vovelle, F., Ptak, M., Reichhart, J. M. , Hoffmann, J. A., Keppi, E., Legrain, M. & Achstetter, T. (1992) J. Biomol. NMR 2, 235- 256.

15. Cornet, B., Bonmatin, J. M., Hetru, C, Hoffmann, J. A., Ptak, M. & Vovelle, F. (1995) Structure. 3, 435-448.

16. Lamberty, M., Caille, A., Landon, C, Tassin-Moindrot, S., Hetru, C, Bulet, P. & Vovelle, F. (2001) Biochem. 40, 11995-12003. 17. Lamberty, M., Ades, S., Uttenweiler-Joseph, S., Brookhart, G., Bushey, D., Hoffmann,

J. A. & Bulet, P. (1999) J. Biol. Chem. 274, 9320-9326. 18. Terras, F. R., Schoofs, H. M., De Bolle, M. F., Van Leuven, F., Rees, S. B.,

Vanderleyden, J., Cammue, B. P. & Broekaert, W. F. (1992) J. Biol. Chem. 267, 15301-

15309. 19. Thomma, B. P. H. J., Cammue, B. P. A. & Thevissen, K. (2003) Current Drug Targets -

Infectious Disorders (In Press).

20. Oren, Z., Lerman, J. C, Gudmundsson, G. H., Agerberth, B. & Shai, Y. (1999) Biochem. J. 341 ( Pt 3), 501-513.

21. Thevissen, K., Ghazi, A., De Samblanx, G. W., Brownlee, O, Osborn, R. W. & Broekaert, W. F. (1996) J. Biol. Chem. 271 , 15018-15025.

22. Sharma, S. & Khuller, G. (2001) Curr. Microbiol. 2001. Jul. ;43. (1):74. -6. 43, 74-76.

23. Mendez, E., Moreno, A., Colilla, F., Pelaez, F., Limas, G. G., Mendez, R., Soriano, F., Salinas, M. & de Haro, C. (1990) Eur. J. Biochem. 194, 533-539.

24. Walsh, T. J., Viviani, M. A., Arathoon, E., Chiou, C, Ghannoum, M., Groll, A. H. & Odds, F. C. (2000) Med. Mycol. 2000. ;38. Suppl 1:335. -47. 38 Suppl 1 , 335-347.

25. Bolard, J. (1986) Biochim. Biophys Acta 864, 257-304.

26. Shai, Y. (2002) Curr. Pharm. Des 2002. ;8. (9. ):715. -25. 8, 715-725.

27. Gura, T. (2001) Science 291 , 2068-2071.

28. Zasloff, M. (2002) Nature 415, 389-395. 29. Broekaert, W. F., Terras, F. R., Cammue, B. P. & Osborn, R. W. (1995) Plant Physiol

108, 1353-1358. 30. Cammue, B. P., Thevissen, K., Hendriks, M., Eggermont, K., Goderis, I. J., Proost, P. , Van Damme, J., Osborn, R. W, Guerbette, F., Kader, J. C. et al. (1995) Plant Physiol 109, 445-455.

31. TAILOR, R. H., Acland, D. P., Attenborough, S„ Cammue, B. P. A., EVANS, I. J. R., Osborn, R. W., RAY, J. A., Rees, S. B. & Broekaert, W. F. (1997) Journal Of Biological

Chemistry 272, -24487.

32. De Samblanx, G. W., Goderis, I. J., Thevissen, K., Raemaekers, R., Fant, F., Borremans, F., Acland, D. P., Osborn, R. W., Patel, S. & Broekaert, W. F. (1997) J. Biol. Chem. 272, 1171-1179. 33. Fonzi, W. A. & Irwin, M. Y. (1993) Genetics 134, 717-728.

34. Broekaert, W. F., Terras, F. R. G., Cammue, B. P. A. & Vanderleyden, J. (1990) FEMS Microbiol. Lett. 69, 55-60.

35. Thevissen, K., Warnecke, D., Francois, I. E. J. A., Leipelt, M., Heinz, E., Ott, C, Thomma, B. P. H. J., Ferket, K. K. A. & Cammue, B. P. A. (2003) Proc. Natl. Acad. Sci. U. S. A. (Submitted).

36. Mamelak, D. & Lingwood, C. (2001) J. Biol. Chem. 276, 449-456.

37. Mylvaganam, M., Binnington, B., Hansen, H. C, Magnusson, G., Nyholm, P. G. & Lingwood, C. A. (2002) Biochem. J. (In Press).

38. Hallstrom, T. C, Lambert, L., Schorling, S., Balzi, E., Goffeau, A. & Moye-Rowley, W. S. (2001) J. Biol. Chem. 2001. Jun. 29. ;276. (26. ):23674. -80. 276, 23674-23680.

39. Bagnat, M., Keranen, S., Shevchenko, A., Shevchenko, A. & Simons, K. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 3254-3259.

40. Caro, L. H., Tettelin, H., Vossen, J. H., Ram, A. F., van den, E. H. & Klis, F. M. (1997) Yeast 13, 1477-1489. 41. Roth, B. L, Poot, M., Yue, S. T. & Millard, P. J. (1997) Appl. Environ. Microbiol. 63,

2421-2431.

42. Rodrigues, M. L., Travassos, L. R., Miranda, K. R., Franzen, A. J., Rozental, S., de Souza, W., Alviano, C. S. & Barreto-Bergter, E. (2000) Infect. Immun. 68, 7049-7060.

43. Toledo, M. S., Levery, S. B., Straus, A. H., Suzuki, E., Momany, M., Glushka, J., Moulton, J. M. & Takahashi, H. K. (1999) Biochem. 38, 7294-7306.

44. Duarte, R. S., Polycarpo, C. R., Wait, R., Hartmann, R. & Bergter, E. B. (1998) Biochim. Biophys. Acta 1390, 186-196.

45. Xu, X., Bittman, R., Duportail, G., Heissler, D., Vilcheze, C. & London, E. (2001) J. Biol. Chem. 276, 33540-33546. 46. Hoover, D. M., Rajashankar, K. R., Blumenthal, R., Puri, A., Oppenheim, J. J., Chertov,

O. & Lubkowski, J. (2000) J. Biol. Chem. 275, 32911-32918.

47. Matsuzaki, K., Yoneyama, S., Fujii, N., Miyajima, K., Yamada, K., Kirino, Y. & Anzai, K. (1997) Biochemistry 36, 9799-9806.

48. Schmitt, M. J. & Compain, P. (1995) Arch. Microbiol. 164, 435-443. 49. Levery, S. B., Momany, M., Lindsey, R., Toledo, M. S., Shayman, J. A., Fuller, M.,

Brooks, K., Doong, R. L, Straus, A. H. & Takahashi, H. K. (2002) FEBS Lett. 525, 59- 64. 50. Toledo, M. S., Levery, S. B., Straus, A. H. & Takahashi, H. K. (2000) J. Lipid Res. 41 , 797-806. 51. Toledo, M. S., Levery, S. B., Suzuki, E., Straus, A. H. & Takahashi, H. K. (2001)

Glycobiology 11 , 113-124. 52. Warnecke, D., Erdmann, R., Fahl, A., Hube, B., Muller, F., Zank, T., Zahringer, U. & Heinz, E. (1999) J. Biol. Chem. 274, 13048-13059. TABLES

Table 1. Antifungal activity of DmAMPI on Saccharomyces cerevisiae WT and ipt- deletion mutant

IC50 values (μM)

YPD V_ PDB

S. cerevisiae BY4741 4 2 /pt7-deletion mutant > 100 2

yeast disruptants was determined. ^bND = not determined Table 3. DmAMPI sensitivity of double and triple yeast deletion mutants in GPI- CWPs.

ORFs ^aDmAMP1 S/R

YLR110C YJR150C S

YKL096w YKL096w-a S

YKL046c YKL096w-a S

YKL096W YHR126C S

YKL046w YHR126C S

YLR110C YJR150C YLR391w-a S

DmAMPI sensitivity (S; sensitive towards concentrations of 2 μM DmAMPI) or resistance (R) of yeast disruptants was determined.

Table 5. Activity of antifungal peptides against Neurospora crassa strains

ICso (μM)*

Strain RsAFP2 Hs-AFPI DmAMPI AceAMPI lbAMP2 AcAMP2 wild-type 1.25 0.4 0.4 2 2.5 0.3

Groupl

MUT1 >20 >20 >20 16.7 >20 >20

MUT4 >20 >20 >20 >20 >20 >20

MUT5 >20 >20 >20 >20 >20 >20

MUT6 >20 >20 >20 >20 >20 20

MUT7 >20 >20 >20 >20 >20 20

MUT8 >20 >20 >20 >20 >20 >20

MUT9 >20 >20 >20 >20 >20 20

MUT10 >20 >20 >20 >20 >20 >20

MUT15 >20 >20 >20 >20 >20 20

MUT17 >20 >20 >20 >20 >20 >20

MUT20 >20 >20 >20 >20 >20 15

MUT21 >20 >20 >20 >20 >20 >20

Groupll

MUT2 >20 >20 >20 16.7 20 10

MUT12 >20 20 >20 4.2 10 5

MUT16 >20 20 20 <0.1 2.5 0.3

MUT19 >20 20 >20 4.2 10 5

MUT24 >20 20 >20 1.7 12 5

MUT25 >20 >20 >20 2 >20 10

Grouplll

MUT3 >20 >20 >20 >20 20 20

MUT11 >20 >20 >20 >20 10 5

MUT13 >20 >20 >20 8.3 >20 10

MUT14 >20 >20 >20 >20 20 10

MUT18 >20 1.25 20 8.3 5 2.5

MUT22 4 >20 >20 4.2 10 5

MUT23 >20 >20 >20 8.3 >20 15

Concentration required for 50% inhibition of fungal growth as defined under 'Materials and Methods'

Table 6. Activity of defensin-like peptides from insects against Neurospora crassa strains

ICso* (μM) strain heliomicin termicin

Wild-type 0.04 0.16

MUT16 >20 2.5

MUT24 >20 2.5

Concentration required for 50% inhibition of fungal growth as defined under 'Materials and

Methods'

Tests were performed in half strength PDB, supplemented with 50 mM HEPES, pH 7.0

Claims

1. A method for assessing the anti-fungal properties of a compound characterised in that the method evaluates the interaction of a compound with fungal cell membrane lipids said lipids being selected from the group consisting of sphingolipids and steryl glucosides.

2. A method according to claim 1 wherein the sphingolipids are glycosphingolipids.

3. A method according to claim 1 or 2 wherein the sphingolipid is a mannosylated inositol phosphoryl ceramide

4. A method according to claim 3 wherein the sphingolipid is M(IP)₂C.

5. A method according to claim 1 or 2 wherein the sphingolipid is glucosylceramide.

6. A method according to claim 1 comprising the steps of:

(a) providing a compound;

(b) incubating a wild type fungal strain in the presence of said compound;

(c) incubating a corresponding mutant fungal strain in the presence of said compound, said mutant having lower levels of glucosylceramide or structurally different glucosylceramide in its cell membrane as compared to the wild type strain;

(d) comparing the growth of the wild type and the mutant strain in the presence of said compound

7. A method according to claim 1 comprising the steps of:

(a) providing a compound;

(b) incubating a wild type fungal strain in the presence of said compound;

(c) incubating a corresponding mutant fungal strain in the presence of said compound, said mutant having higher or lower levels of steryl glucosides in its cell membrane as compared to the wild type strain;

8. A method according to claim 6 or 7 comprising the additional steps of:

(e) incubating a corresponding mutant fungal strain in the presence of said compound, said mutant having lower levels of mannosylated inositol phosphoryl ceramides or structurally different mannosylated inositol phosphoryl ceramides in its cell membrane as compared to the wild type strain;

(f) comparing the growth of the wild type and the mutant strain in the presence of said compound

9. A method according to claim 1 comprising the steps of:

(a) providing a compound;

(b) incubating a wild type fungal strain in the presence of said compound;

(d) comparing permeabilization of the membrane of the wild type and the mutant strain in the presence of said compound

10. A method according to claim 1 comprising the steps of:

(a) providing a compound;

(b) incubating a wild type fungal strain in the presence of said compound;

(d) comparing the permeabilization of the membrane of the wild type and the mutant strain in the presence of said compound

11. A method according to claim 6 or 7 comprising the additional steps of:

(e) incubating a corresponding mutant fungal strain in the presence of said compound, said mutant having lower levels of mannosylated inositol phosphoryl ceramides or structurally different mannosylated inositol phosphoryl ceramides in its cell membrane as compared to the wild type strain; (f) comparing the permeabilization of the membrane of the wild type and the mutant strain in the presence of said compound

12. A method according to any of claims 9 to 11 wherein the permeabilization of the membrane is determined using a fluorescent marker

13. A method according to claim 12 wherein the fluorescent marker is SYTOX-green.

14. A method according to any of the claims 6 to 13 wherein the fungal strain is a yeast strain.

15. A method according to any of the claims 1 to 5 wherein a compound is screened for specific binding affinity with the selected fungal cell membrane lipids, or a part thereof.

16. A method according to claim 15 comprising the steps of:

(a) providing a compound

(b) combining the selected cell membrane lipid with said compound for a sufficient time to allow binding under suitable conditions

(c) detecting the binding of the fungal glucosylceramides to the compound.

17. Use of any of the methods of claims 1 to 16 for the identification of peptides having an antifungal action, said antifungal action being mediated by the interaction of said peptides with the selecte plasma membrane lipids.

18. Use of any of the methods of claims 1 to 16 according to claim 18 wherein the peptides belong or are structurally related to the family of the defensin peptide family.

19. A method for increasing the permeability of yeast cells by bringing the yeast cells into contact with a chemical compound characterized in that the compound binds to the selected cell membrane lipids.

20. A method according to claim 19 wherein the compound is a peptide.

21. A method according to claim 20 wherein the peptide belongs to or is structurally related to the defensin peptide family.