Portal de Periódicos da CAPES

Complementation of the Mpg1 mutant phenotype in Magnaporthe grisea reveals functional relationships between fungal hydrophobins

1998; Springer Nature; Volume: 17; Issue: 14 Linguagem: Inglês

10.1093/emboj/17.14.3838

1460-2075

Michael J. Kershaw, Gavin E. Wakley, Nicholas J. Talbot,

Transgenic Plants and Applications

Article15 July 1998free access Complementation of the Mpg1 mutant phenotype in Magnaporthe grisea reveals functional relationships between fungal hydrophobins Michael J. Kershaw Michael J. Kershaw Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK Search for more papers by this author Gavin Wakley Gavin Wakley Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK Search for more papers by this author Nicholas J. Talbot Corresponding Author Nicholas J. Talbot Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK Search for more papers by this author Michael J. Kershaw Michael J. Kershaw Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK Search for more papers by this author Gavin Wakley Gavin Wakley Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK Search for more papers by this author Nicholas J. Talbot Corresponding Author Nicholas J. Talbot Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK Search for more papers by this author Author Information Michael J. Kershaw1, Gavin Wakley1 and Nicholas J. Talbot 1 1Department of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4QG UK *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:3838-3849https://doi.org/10.1093/emboj/17.14.3838 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info The functional relationship between fungal hydrophobins was studied by complementation analysis of an mpg1− gene disruption mutant in Magnaporthe grisea. MPG1 encodes a hydrophobin required for full pathogenicity of the fungus, efficient elaboration of its infection structures and conidial rodlet protein production. Seven heterologous hydrophobin genes were selected which play distinct roles in conidiogenesis, fruit body development, aerial hyphae formation and infection structure elaboration in diverse fungal species. Each hydrophobin was introduced into an mpg1− mutant by transformation. Only one hydrophobin gene, SC1 from Schizophyllum commune, was able partially to complement mpg1− mutant phenotypes when regulated by its own promoter. In contrast, six of the transformants expressing hydrophobin genes controlled by the MPG1 promoter (SC1 and SC4 from S.commune, rodA and dewA from Aspergillus nidulans, EAS from Neurospora crassa and ssgA from Metarhizium anisopliae) could partially complement each of the diverse functions of MPG1. Complementation was always associated with partial restoration of a rodlet protein layer, characteristic of the particular hydrophobin being expressed, and with hydrophobin surface assembly during infection structure formation. This provides the first genetic evidence that diverse hydrophobin-encoding genes encode functionally related proteins and suggests that, although very diverse in amino acid sequence, the hydrophobins constitute a closely related group of morphogenetic proteins. Introduction Fungal hydrophobins are small secreted, hydrophobic proteins which are fundamental to the developmental biology of fungi. More than 20 hydrophobin-encoding genes have now been recognized and they may prove to be ubiquitous in filamentous fungi (Wessels, 1997). In most cases, hydrophobin genes have been identified as mRNAs abundantly transcribed during particular developmental processes such as sporulation, fruit body formation or fungal infection of plants and animals. Genetic studies have now implicated hydrophobins in all of these morphogenetic processes. Hydrophobin genes have been mutated using targeted gene disruption and cause significant effects on development. In Aspergillus nidulans and Neurospora crassa, for example, null alleles at the rodA and EAS genes respectively cause spores to lose their surface hydrophobicity and clump together (Stringer et al., 1991; Bell-Pederson et al., 1992; Lauter et al., 1992). This 'easily wettable' phenotype is due to the absence of a hydrophobin-encoded protein layer on the surface of conidia (Templeton et al., 1995). In the mushroom Schizophyllum commune, four hydrophobin genes have been described—SC1, SC3, SC4 and SC6 (Wessels et al., 1991; Wessels, 1997)—and are differentially expressed during development of fruit bodies. SC3, a hydrophobin which is produced during hyphal growth, is required for production of upwardly projecting aerial hyphae as sc3− gene disruption mutants are unable to form hydrophobic aerial structures (van Wetter et al., 1996). SC3 was the first hydrophobin to be purified, and this showed that hydrophobins can respond to interfaces between water and air, or between water and solid surfaces (Wösten et al., 1993, 1994b, 1995). Exposure to such an interface caused spontaneous aggregation of SC3 monomers resulting in an amphipathic protein polymer. The hydrophobic side of aggregated SC3 hydrophobin possessed a rodlet architecture identical to that found on aerial hyphae of S.commune (Wösten et al., 1993), suggesting that SC3 is secreted from fungal cells and forms the hydrophobic coating of hyphae after interfacial self-assembly (Wösten et al., 1994a). Significantly, SC3 self-assembly also occurred in response to hydrophobic surfaces, and sc3− mutants were deficient in their attachment to hydrophobic substrates (Wösten et al., 1994b, 1995). In addition to development, hydrophobins play roles in fungal pathogenicity. Targeted gene replacement of MPG1 in the rice blast fungus Magnaporthe grisea produced mutants with reduced pathogenicity (Talbot et al., 1993). This was correlated with a reduced ability to form infection structures called appressoria. Magnaporthe grisea appressoria are dome-shaped cells produced from the tips of fungal hyphae in response to the hydrophobic rice leaf surface (Hamer et al., 1988; Talbot, 1995). Their role is to generate turgor and mechanically breach the rice cuticle (Howard and Valent, 1996; de Jong et al., 1997). Purification of MPG1 indicated that hydrophobin self-assembly occurs on the rice leaf surface and the resulting amphipathic hydrophobin layer then acts as a conformational cue for appressorium development (Talbot et al., 1996). However, MPG1 also encodes a spore wall rodlet protein (like EAS and rodA) and is required for efficient production of conidia. In M.grisea, therefore, it appears that a hydrophobin functions during both conidiogenesis and appressorium development. The diversity of hydrophobins described so far suggests that these morphogenetic proteins serve a large number of specific functions in filamentous fungi (Talbot, 1997; Wessels, 1997). Hydrophobins are indeed very diverse proteins, as the 20 hydrophobins described to date share only 4.3% amino acid identity (Wessels, 1997). They do, however, share a common hydropathy profile, and all hydrophobins have eight cysteine residues characteristically spaced in their amino acid sequences. Two possibilities exist regarding the functional relatedness of hydrophobins. The different functions of hydrophobins may, for example, simply reflect the particular developmental stage at which they are produced. In this case, any hydrophobin gene would be able to complement the mutant phenotype of another, as long as it is expressed at the appropriate developmental stage. Conversely, it may be that the considerable divergence in hydrophobin sequence reflects real functional diversity. If this is the case, then complementation of hydrophobin mutant phenotypes would perhaps be restricted to only hydrophobins fulfilling the same function in different fungi. We decided to address these questions by complementation of the mpg1− mutant phenotypes in the rice blast fungus M.grisea. In this report, we describe the introduction of seven hydrophobin-encoding genes into M.grisea mpg1− mutants. We show that six of the seven hydrophobins we tested could substitute, at least partially, for MPG1 during fungal pathogenicity when expressed as MPG1 promoter fusions. Complementation of MPG1 mutant phenotypes consistently was associated with formation of a rodlet protein layer, indicating that hydrophobin self-assembly is fundamental to the function of MPG1. This provides the first evidence that hydrophobins are functionally related and suggests that although very diverse in amino acid sequence, they form a closely related group of morphogenetic proteins. Results In order to test functional relationships between the fungal hydrophobins, we selected diverse heterologous hydrophobin-encoding genes and introduced these individually into an M.grisea mpg1− null mutant. Complementation experiments were designed such that each hydrophobin-encoding gene was expressed in M.grisea, either under the control of its own endogenous promoter or as a gene fusion to the MPG1 promoter. In this way, the relative contribution of the regulated expression of MPG1, or the characteristics of its hydrophobin gene product to biological function, could be assessed. Identification of a functional MPG1 promoter A functional MPG1 promoter initially was isolated for expression of alternative hydrophobin genes and to establish the precise pattern of expression of MPG1 in relation to its known functions (Talbot et al., 1993, 1996). A 1.28 kb promoter fragment from the 5′ end of the MPG1 gene was isolated and fused to a synthetic allele of the GFP reporter gene, encoding green fluorescent protein (Prasher et al., 1992; Stearns, 1995; Chiu et al., 1996), providing a non-invasive reporter of MPG1 expression. The sGFP gene was subcloned in-frame with the translation initiation codon at the 3′ end of the MPG1 promoter and introduced into an M.grisea transformation vector conferring hygromycin B resistance. In a control experiment, sGFP was fused to the constitutively expressed A.nidulans gpd promoter and transformed into a wild-type M.grisea strain, Guy-11 (see Materials and methods). Transformants were selected and analysed by Southern blot to ensure that they contained single insertions. A total of eight M.grisea transformants containing single copy integrations of the plasmids were selected for further analysis, four containing the MPG1(p)::sGFP::Hph construct, and four the gpd(p)::sGFP::hph construct. Conidia from MPG1(p)::sGFP::Hph transformants were incubated on rice leaves, or sterilized onion epidermis, for 24 h and infection structures were allowed to form. GFP expression could be observed clearly in appressoria of all MPG1(p)::sGFP::Hph transformants, but fluorescence did not migrate into infection pegs or hyphae which entered the plant epidermis (Figure 1A and B). This strongly indicates that MPG1 expression is restricted to early infection-related development, prior to plant penetration. Rice leaf infections were incubated for a further 72 h in conditions of high humidity, allowing disease lesions to form and produce aerially borne conidia (Figure 1C). Conidia from MPG1(p)::sGFP::Hph showed extensive GFP fluorescence, indicating that high level MPG1 expression also occurs during conidiogenesis (Figure 1D). Very little variation was observed between the four MPG1(p)::sGFP::Hph transformants examined. In control experiments, gpd(p)::sGFP::Hph showed GFP fluorescence at all stages of development (not shown). Figure 1.Spatial regulation of MPG1 expression during plant infection by M.grisea. Micrographs on the left are bright field images viewed with Hoffman modulation optics (Nikon). GFP fluorescence is shown on the right. Conidia of an MPG1(p)::sGFP::Hph transformant were incubated on sterile onion epidermis, and allowed to form appressoria and penetrate the epidermis. (A) Appressoria and penetration pegs after 24 h. (B) Infection hyphae after 48 h. Bar = 10 μm. (C and D) Aerially borne conidia from an MPG1(p)::sGFP::Hph transformant emerging from a rice blast disease lesion, 96 h after plant inoculation. Bar = 120 μm Download figure Download PowerPoint Expression of diverse hydrophobin-encoding genes in an mpg1::Hph null mutant of M.grisea Having isolated a functional MPG1 promoter, seven alternative hydrophobin-encoding genes were selected for complementation tests. These were SC1, SC3 and SC4 from S.commune, the rodA and dewA genes from A.nidulans, EAS from N.crassa and the ssgA from Metarhizium anisopliae. Heterologous hydrophobin genes were selected to encompass the diversity of functions so far attributed to hydrophobins: SC1 and SC4 represent dikaryon-specific fruit body hydrophobins; rodA, dewA and EAS are all spore wall rodlet proteins; and ssgA has been implicated in appressorium development by the insect pathogen M.anisopliae, suggesting a function similar to MPG1. The origins and functions of all hydrophobins used in this study are given in Table I. Table 1. Fungal hydrophobins used in this study Name Taxonomic class Organism Mutant phenotype determined Biological function Reference MPG1 Ascomycetes Magnaporthe grisea + Conidial spore wall protein. Involved in conidium and appressorium formation. Talbot et al. (1993), 1996 Required for full pathogenicity SC1a Basidiomycetes Schizophyllum commune − Unknown. Expressed in dikaryotic fruiting body Schuren and Wessels (1990); Wessels et al. (1991) SC3a Basidiomycetes Schizophyllum commune + Involved in aerial hyphae formation and able to attach to hydrophobic surfaces Wösten et al. (1994); Van Wetter et al. (1996) SC4a Basidiomycetes Schizophyllum commune − Lines gas channels in dikaryotic fruiting body Schuren and Wessels (1990); Wessels et al. (1991); Wessels (1997) ssgAb Ascomycetes Metarhizum anisopliae − Unknown. Expressed during appressorium development St. Leger et al. (1992) rodAc Ascomycetes Aspergillus nidulans + Conidial spore wall protein Stringer et al. (1991) dewAc Ascomycetes Aspergillus nidulans + Conidial spore wall protein Stringer and Timberlake (1995) EASd Ascomycetes Neurospora crassa + Conidial spore wall protein Bell-Pederson et al. (1992); Lauter et al. (1992) Hydrophobin-encoding genes were obtained from: a Dr Frank Schuren (University of Groningen); b Dr Ray St. Leger (Cornell University); c Dr W.E.Timberlake (University of Georgia); d Drs Deborah Bell-Pederson and Jay Dunlap (Dartmouth Medical School). All the hydrophobins belonged to class I. Genomic clones of SC1, SC3, SC4, rodA, dewA and EAS were selected containing all necessary components for expression in the corresponding native fungus (see Table II). Each genomic fragment was subcloned to an M.grisea transformation vector containing a selective marker conferring resistance to either bleomycin or bialophos (Table II). The resulting plasmids were then transformed into the mpg1::Hph gene replacement mutant 53-R-39 (Talbot et al., 1996). Transformants were analysed by Southern blot to identify transformants carrying single-copy integrations of each hydrophobin gene (data not shown), and two recombinant transformants expressing each heterologous hydrophobin were selected for phenotypic analysis (Table II). Table 2. Hydrophobin expression vectors generated in this study and corresponding M.grisea transformants Hydrophobin gene Promoter Selectable marker Recombinant construct Recombinant strains Genotype SC1 SC1 bleomycina pNJT7 MJK7.8 mpg1::Hph;SC1(p)::SC1::ble MJK7.16 SC3 SC3 bleomycin pS330 SSC3.1 mpg1::Hph;SC3(p)::SC3 ::ble SSC3.3 SC4 SC4 bleomycin pMJK58 SSC4.5 mpg1::Hph;SC4(p)::SC4 ::ble SSC4.8 rodA rodA bialophosb pMJK20 ARODA.24 mpg1::Hph;rodA(p)::rodA::bar ARODA.29 dewA dewA bleomycin pNJT26 ADEWA. mpg1::Hph;dewA(p)::dewA::ble ADEWA. EAS EAS bleomycin pTON5 NEAS. 5 mpg1::Hph;EAS(p)::EAS::ble NEAS.25 SC1 MPG1c bialophos pNJT7 MSC1.2 mpg1::Hph;MPG1(p)::SC1::bar MSC1.7 SC3 MPG1c bialophos pMJK85 MSC3.10 mpg1::Hph;MPG1(p)::SC3::bar MSC3.13 SC4 MPG1c bialophos pMJK79 MSC4.3 mpg1::Hph;MPG1(p)::SC4::bar MSC4.10 ssgA MPG1c bialophos pMJK64 MSSGA.5 mpg1::Hph;MPG1(p)::ssgA::bar MSSGA.14 rodA MPG1c bialophos pMJK10 MRODA.4 mpg1::Hph;MPG1(p)::rodA::bar MRODA.5 dewA MPG1c bialophos pMJK146 MDEWA. 12 mpg1::Hph;MPG1(p)::dewA::bar MDEWA.15 EAS MPG1c bialophos pMJK55 MEAS.9 mpg1::Hph;MPG1(p)::EAS::bar MEAS.13 MPG1 MPG1 bleomycin pMJK26 MJK.13d mpg1:Hph; MPG1(p)::MPG1::ble a Bleomycin resistance gene, ble, from pAN8-1 (Punt et al., 1987). b Bialophos resistance gene, bar, from pCB1265 (Sweigard et al., 1996). c All expressed as translational fusions under the control of a 1.2 kb PstI–NcoI MPG1 promoter fragment from pNJT190 (see Materials and methods). d Transformant described previously by Talbot et al. (1996). At the same time, genomic clones of SC1, SC3, SC4, rodA, dewA and EAS, and a cDNA clone of ssgA were selected for construction of translational fusions with the MPG1 promoter fragment. Each hydrophobin gene was subcloned in-frame to the MPG1 promoter fragment (see Materials and methods) and introduced into the M.grisea transformation vector pCB1265 conferring bialophos resistance (Sweigard et al., 1997). Bialophos-resistant transformants were selected and Southern blotted to identify those with single-copy integrations of each MPG1 promoter gene fusion (not shown). Two transformants carrying single-copy integrations for each of the seven MPG1 promoter–hydrophobin fusions were then selected for phenotypic analysis. Details of the 26 recombinant strains generated in this study are given in Table II. RNA gel blot analysis was carried out to ensure MPG1-like patterns of expression in each gene fusion transformant. Starvation stress is known to induce MPG1 expression (Talbot et al., 1993; Beckerman and Ebbole, 1996: Lau and Hamer, 1996) and, therefore, transformants were exposed to starvation stress by transfer of fungal mycelium to medium lacking either a nitrogen or carbon source for 12 h of growth. At this time, RNA was extracted, fractionated by gel electrophoresis, blotted and probed with the corresponding hydrophobin gene for each transformant. An example, showing expression of the 846 bp rodA transcript of transformant ARODA.24 during nitrogen and carbon source starvation, is shown in Figure 2. All transformants containing MPG1 promoter fusions were tested in this way with identical results. Figure 2.MPG1 promoter-regulated expression of A.nidulans rodA in an M.grisea mpg1− mutant. Total RNA was extracted from cultures of M.grisea which had been subjected to nitrogen starvation stress for 18 h and was then fractionated by gel electrophoresis, blotted and probed with a 4.3 kb XhoI–PstI fragment of the rodA genomic clone, pMS10 (Stringer et al., 1991). A loading control hybridization was carried out with an M.grisea rDNA probe, pMG1. Lanes contain total RNA from wild-type strain Guy-11; mpg1::Hph mutant, 53-R-39; MPG1(p)::rodA:: bar transformant, MRODA.4; MPG1(p)::rodA:: bar transformant MRODA.5. Download figure Download PowerPoint Expression of heterologous hydrophobin genes regulated by their own promoters was not detected by RNA gel blot analysis (using total RNA) in transformants either during infection-related development or during starvation stress (data not shown). Complementation of the conidiation-deficient phenotype of an mpg1::Hph deletion mutant by heterologous hydrophobin-encoding genes Conidiation is severely reduced in mpg1::Hph deletion mutants, and conidiating cultures have an 'easily wettable' phenotype (Talbot et al., 1993). The easily wettable phenotype is due to loss of the MPG1-encoded rodlet protein layer found on the surface of M.grisea conidia (Talbot et al., 1996). This also reduces the efficiency of conidiogenesis, allowing only a single conidium to differentiate from each conidiophore, in contrast to conidiophores of isogenic wild-type strains of M.grisea which produce 4–5 conidia in a sympodial array (Ou, 1985; Talbot et al., 1996). Deletion of MPG1 also causes a reduction in pathogenicity, resulting in ∼80% reduction in the usual number of disease lesions on infected rice seedlings (Talbot et al., 1993, 1996). This is associated with reduction in the ability of mpg1::Hph mutants to elaborate appressoria (Talbot et al., 1993, 1996). Recombinant transformants expressing each hydrophobin gene were therefore examined to determine their ability to substitute for MPG1 in each of its attributed functions. As a control in these experiments, two transformants were also analysed routinely where MPG1 had been re-introduced into an mpg1− mutant (Table II). We previously reported full complementation of all mpg1− mutant phenotypes in these transformants (Talbot et al., 1996), which are phenotypically indistinguishable. Conidiogenesis was assessed by taking plate cultures of M.grisea grown for a uniform time, flooding them with water and gently removing conidia with a glass rod. When transformants expressing hydrophobins under control of their own promoters were analysed, the only transformants to show partial complementation of conidial numbers were those expressing the S.commune hydrophobin, SC1 (Figure 3A). Two transformants, MJK7.8 and MJK7.16 [SC1(p)::SC1::ble], produced mean conidial numbers of 9.5×105/ml compared with the wild-type strain Guy-11, which produced a mean of 8×106 conidia/ml, and the isogenic mpg1::Hph mutant 53-R-39 which produced a mean of 3×105 conidia/ml (Figure 3A). Both of the SC1(p)::SC1::ble transformants were also complemented for the easily wettable phenotype, being able to support a droplet of water on the culture surface (Figure 3C). Figure 3.Bar graph of numbers of conidia produced from cultures of an mpg1− mutant 53-R-39 transformed with heterologous hydrophobin-encoding genes. Conidia were collected by flooding plate cultures and gently scraping aerial mycelium with a glass rod. Suspensions were filtered through cheesecloth to remove mycelial debris and counted. (A) Conidial numbers from mpg1− transformants expressing hydrophobins regulated by their native promoters. (B) Conidial numbers from mpg1− transformants expressing hydrophobins regulated by MPG1(p). Error bars indicate standard error of the mean (n = 5). (C) Wettability of transformant cultures assessed by their ability to support a 10 ml drop of water on the surface of conidiating cultures. (+ indicates wild-type, − indicates an easily wettable water-soaked phenotype). Download figure Download PowerPoint Transformants containing MPG1 promoter gene fusions produced considerably greater numbers of conidia compared with the isogenic mpg1− mutant 53-R-39 (Figure 3B). The highest conidial numbers were found in recombinants expressing MPG1(p)::EAS, MPG1(p)::dewA and MPG1(p)::SC1 gene fusions. Conidial numbers in these transformants were >3×106/ml and therefore not significantly different from wild-type when compared by the Student's t-test (P >0.05). In contrast, MPG1(p)::SC3::bar transformants produced conidial numbers which were significantly different from wild-type (t = 35.1; P 0.05; df = 5). Consistent with this, MPG1(p)::SC3::bar transformants were easily wettable while the remaining transformants were identical to wild-type M.grisea with respect to surface hydrophobicity (Figure 3C). Introduction of heterologous hydrophobin-encoding genes restores pathogenicity of an mpg1::Hph deletion mutant The ability of transformants to cause rice blast disease was determined by infecting rice seedlings of a susceptible rice cultivar CO-39. Conidial suspensions were prepared and adjusted to 1×104 conidia/ml. Fourteen-day-old rice seedlings were then sprayed with the suspension and the disease allowed to progress for 96 h. At this time, rice blast infections with the wild-type M.grisea strain Guy-11 produced large numbers of oval-shaped lesions on rice leaves. In mpg1− mutants, the frequency of lesions was between 10 and 20% of the number generated by an isogenic wild-type strain of M.grisea (Talbot et al., 1993, 1996). To assess pathogenicity of transformants expressing alternative hydrophobins, the lesion number was recorded on 5 cm leaf sections as shown in Figure 4A and B. Pathogenicity was partially restored for SC1(p)::SC1::ble transformants, but none of the other transformants carrying hydrophobin genes expressed under endogenous promoters showed significant increases in lesion number compared with the mpg1::Hph mutant (P >0.1; df = 40) (Figure 4A and C). In contrast, six of the transformants expressing hydrophobin genes controlled by the MPG1 promoter showed partial restoration of pathogenicity. The highest number of disease lesions was produced by MPG1(p)::EAS and MPG1(p)::SC1 transformants which were not significantly different from lesion numbers generated by Guy-11 (Figure 4B and D). MPG1(p)::SC3 transformants were not restored for pathogenicity and showed similar lesion numbers to the mpg1::Hph null mutant (Figure 4B). We conclude that MPG1 promoter-driven expression of dewA, rodA, SC1, SC4, EAS and ssgA can restore the activity of MPG1 in pathogenicity of M.grisea. Figure 4.Pathogenicity of mpg1− mutant 53-R-39 transformed with heterologous hydrophobin-encoding genes. Fourteen-day-old rice seedlings of susceptible cultivar CO-39 were inoculated with conidial suspensions of M.grisea and allowed to develop rice blast symptoms for 4 days. (A) Bar graph of mean rice blast lesion density per 5 cm leaf tip from mpg1− transformants expressing hydrophobins regulated by their native promoters. (B) Bar graph of mean rice blast lesion density per 5 cm leaf tip from mpg1− transformants expressing hydrophobins regulated by MPG1(p). Error bar indicates standard error of the mean (n = 40). (C and D) Photographs of leaves from plants inoculated with each transformant indicated. Download figure Download PowerPoint Restoration of appressorium development by expression of heterologous hydrophobins in an mpg1::Hph mutant Reduced pathogenicity of mpg1− mutants is associated with a decrease in the ability of the fungus to elaborate appressoria (Talbot et al., 1993). Transformants were therefore tested for their ability to develop appressoria on hydrophobic surfaces. Consistent with restoration of other Mpg1 mutant phenotypes, SC1(p)::SC1::ble transformants showed partial restoration of this ability (Figure 5A). None of the remaining transformants, however, produced significantly greater numbers of appressoria compared with the isogenic mpg1::Hph mutant. Transformants expressing hydrophobin genes under control of the MPG1 promoter showed restoration of appressorium formation on hydrophobic surfaces, with the exception of MPG1(p)::SC3 transformants. These produced numbers of appressoria which were not significantly different from the mpg1− mutant, 53-R-39 (t = 2.96; P >0.04; df = 5). Appressoria were produced in the greatest numbers by MPG1(p)::EAS, MPG1(p)::SC1 and MPG1(p)::dewA transformants (Figure 5B). Appressorium production was therefore restored to near wild-type levels by expression of EAS, SC1 and dewA under control of the MPG1 promoter, and partially restored by expression of SC4, ssgA and rodA. Figure 5.Infection-related development with heterologous hydrophobin-encoding genes. Appressoria were allowed to form on hydrophobic Teflon membranes (DuPont) for 24 h. (A) Bar graph of mean appressorium formation by mpg1− transformants expressing hydrophobins regulated by their native promoters. (B) Bar graph of mean appressorium formation by mpg1− transformants expressing hydrophobins regulated by MPG1(p). MJK16 is an mpg1− mutant complemented with MPG1 (Talbot et al., 1996). The error bar indicates standard error of the mean (n = 5). (C) Appressorium development by MPG1(p)::SC3 transformant, MSC3.10. (D) Appressorium development by MPG1(p)::SC1 transformant, MSC1.2. Bar = 40 μm. Download figure Download PowerPoint A control experiment was carried out to determine whether the variation in appressorium formation was associated with introduction of the heterologous hydrophobins. Because transformation in M.grisea is integrative, it has the potential to cause mutations and, therefore, a control experiment was necessary to test whether appressorium deficiency was due simply to the insertion of plasmid DNA. Appressorium development in M.grisea can be induced by exogenous application of cAMP (Lee and Dean, 1993) and, previously, we and others showed that MPG1 acts upstream of the cAMP signalling pathway as appressorium deficiency in mpg1− mutants can be restored by exogenous cAMP (Beckerman and Ebbole, 1996; Talbot et al., 1996). A 10 mM aliquot of cAMP was therefore added to germinating conidia of all hydrophobin transformants. Under these conditions, all transformants generated in this study were able to form appressoria (data not shown). Complementation of mpg1− by heterologous hydrophobins is correlated with production of a rodlet protein Rodlet proteins are well known ultrastructural characteristics of aer