The FRP1 F‐box gene has different functions in sexuality, pathogenicity and metabolism in three fungal pathogens

SUMMARY

Plant‐pathogenic fungi employ a variety of infection strategies; as a result, fungi probably rely on different sets of proteins for successful infection. The F‐box protein Frp1, only present in filamentous fungi belonging to the Sordariomycetes, Leotiomycetes and Dothideomycetes, is required for nonsugar carbon catabolism and pathogenicity in the root‐infecting fungus Fusarium oxysporum. To assess the role of Frp1 in other plant‐pathogenic fungi, FRP1 deletion mutants were generated in Fusarium graminearum and Botrytis cinerea, and their phenotypes were analysed. Deletion of FgFRP1 in F. graminearum led to impaired infection of barley roots, but not of aerial plant parts. Deletion of BcFRP1 in B. cinerea did not show any effect on pathogenicity. Sexual reproduction, however, was impaired in both F. graminearum and B. cinerea FRP1 deletion mutants. The mutants of all three fungi displayed different phenotypes when grown on an array of carbon sources. The F. oxysporum and B. cinerea deletion mutants showed opposite growth phenotypes on sugar and nonsugar carbon sources. Replacement of FoFRP1 in F. oxysporum with the B. cinerea BcFRP1 resulted in the restoration of pathogenicity, but also in a switch from impaired growth on nonsugar carbon sources to impaired growth on sugar carbon sources. This effect could be ascribed in part to the B. cinerea BcFRP1 promoter sequence. In conclusion, the function of the F‐box protein Frp1, despite its high sequence conservation, is not conserved between different fungi, leading to differential requirements for pathogenicity and carbon source utilization.

INTRODUCTION

Fungi associated with plants employ different strategies to colonize or infect host plants. In addition, they exhibit lifestyles ranging from saprotrophic to biotrophic or necrotrophic growth. As a result of this broad variety of strategies and lifestyles, the requirement for certain cellular processes and proteins involved in the infection mechanism probably varies between fungi.

The plant‐pathogenic fungi Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea are examples of fungi with different lifestyles and infection strategies. The soil‐borne hemibiotrophic fungus F. oxysporum infects its specific host through the roots and frequently colonizes the xylem vessels, causing wilt disease (Di Pietro et al., 2003; Gordon and Martyn, 1997; Michielse and Rep, 2009; Roncero, 2003). Fusarium graminearum, a putative hemibiotrophic fungus, is dispersed mainly as sexual ascospores and predominantly infects aerial plant parts, causing head blight or crown rot on cereal plants (Goswami and Kistler, 2004; Stephens et al., 2008). Botrytis cinerea is a necrotrophic fungus with over 200 hosts, infecting field crops, ornamental flowers and harvested fruits and vegetables. Botrytis cinerea can enter plant tissue immediately after conidial germination via cuticle penetration through the formation of an appressorium‐like structure (Choquer et al., 2007; van Kan, 2006; Williamson et al., 2007). Despite different infection strategies and distant taxonomy, plant‐pathogenic fungi still share conserved genes that are important during infection, such as genes encoding proteins of conserved signalling pathways. For instance, deletion of one or more mitogen‐activated protein (MAP) kinase or G‐coupled protein genes shows an effect on pathogenicity and/or cell wall‐degrading enzyme (CWDE) gene expression in many fungi, including F. oxysporum, F. graminearum and B. cinerea (Di Pietro et al., 2001; Doehlemann et al., 2006; 2002, 2003, 2005; Jenczmionka et al., 2003; Rui and Hahn, 2007; Schulze Gronover et al., 2001; Segmuller et al., 2007; Xu, 2000; Yu et al., 2008; Zheng et al., 2000).

In addition to conserved genes fulfilling a role in pathogenicity, fungi also require genes for their specific pathogenic lifestyle and/or infection strategy. For instance, the melanin‐deficient mutant buf1 and the cyclic AMP‐dependent protein kinase A mutant cpkA in Magnaporthe grisea (Sesma and Osbourn, 2004) and isocitrate lyase (ICL1) mutants in different plant‐pathogenic fungi (Idnurm and Howlett, 2002; Jonkers et al., 2009; Wang et al., 2003) are deficient in leaf infection, but still invade roots. Conversely, FOW1, a gene required for root infection in F. oxysporum, is not required for appressorium formation and leaf penetration in M. grisea (Inoue et al., 2002; Sesma and Osbourn, 2004).

Frp1 is a fungal protein required for pathogenicity in F. oxysporum (Duyvesteijn et al., 2005; Jonkers et al., 2009). Frp1 belongs to the family of F‐box proteins that commonly facilitate the degradation of target proteins via ubiquitination through a Skp, Cullin, F‐box protein (SCF)‐containing complex (Jonkers and Rep, 2009a; Willems et al., 2004). Fungal F‐box proteins have been shown to be involved in the control of the cell division cycle, glucose sensing and mitochondrial connectivity (reviewed by Jonkers and Rep, 2009a). In Neurospora crassa, the F‐box protein Fwd1 has been shown to be required for degradation of the circadian clock protein FREQUENCY (FRQ) via ubiquitination through the SCF complex (He and Liu, 2005).

The FoFRP1 deletion mutant shows impaired root colonization and penetration defects, which are related to growth impairment on a broad spectrum of nonsugar carbon sources, such as organic acids, amino acids and cell wall polymers. The FoFRP1 deletion mutant also shows impaired expression of CWDE genes and other carbon catabolite repression (CCR) genes (Jonkers and Rep, 2009b; Jonkers et al., 2009). Mutation of CRE1, a gene controlling CCR gene expression, in the FoFRP1 deletion mutant restored the above‐described impairments (Jonkers and Rep, 2009b).

At present, it is unknown whether Frp1 homologues in other fungi also regulate CCR and carbon source assimilation, and whether these processes are required for pathogenicity in fungi with a different infection strategy and lifestyle. Therefore, the F. graminearum and B. cinerea orthologues of the F‐box protein Frp1 were studied to reveal whether they have similar functions as in F. oxysporum. In addition, functional conservation of FRP1 was assessed by cross‐species' complementation.

RESULTS

Conservation of FRP1 in filamentous fungi

To examine the conservation of FRP1 in the fungal kingdom, we searched for Frp1 homologues in available fungal sequences at the Broad Institute (http://www.broadinstitute.org), the DOE Joint Genome Institute (http://genome.jgi‐psf.org) and the Podospora anserina Genome Project (http://podospora.igmors.u‐psud.fr) using the blastp algorithm (Altschul et al., 1997). Using a cut‐off score of 1 × 10E⁻¹⁰, single Frp1 homologues were found in fungi from classes Sordariomycetes, Leotiomycetes and Dothideomycetes, all belonging to the Ascomycetes. The only putative Frp1 homologue belonging to the Eurotiomycetes retrieved using the blast search, CIMG_02198 from Coccidioides immitis, had a score (1 × 10E⁻⁶) below the cut‐off and was therefore not included.

A ClustalW alignment of predicted Frp1 proteins that were retrieved from the various databases was used to create a phylogenetic tree (Fig. 1A). This tree is based on amino acids 175–278 of the alignment, which includes and extends beyond the F‐box domain (Fig. S1, see Supporting Information). The putative Frp1 homologues from the Dothideomycetes are substantially shorter and less similar than those in the Sordariomycetes and Leotiomycetes; for example, the S. nodorum Frp1 homologue is 338 amino acids long, in contrast with the 527 amino acids of F. oxysporum.

Figure 1.

Occurrence of Frp1 homologues in various fungi. (A) A phylogenetic tree was created with MegAlign software using amino acids 175–278 from the middle domain of the Frp1 homologues, which includes the F‐box domain. The fungal classes to which the respective fungi belong are given next to the tree. Frp1 homologues of the following organisms were used in this alignment: Ab, Alternaria brassicicola (AB01633.1); Bc, Botrytis cinerea; Cg, Cheatomium globosum (CHGG_00736.1); Fg, Fusarium graminearum (FGSG_01326.3); Fo, Fusarium oxysporum (FOXG_00058.2); Fs, Fusarium solani (Nectria haematococca) protein ID 103049; Fv, Fusarium verticillioides (FVEG_01458.3); Hj, Hypocrea jecorina (Trichoderma reeesei) protein ID 120583; Mg, Magnaporthe grisea (MGG_06351); Nc, Neurospora crassa (NCU09899); Pa, Podospora anserine (Pa_1‐12760); Ptr, Pyrenophora tritici‐repentis (PTRG_0644.1); Sn, Stagonospora nodorum (SNOG_14560.1); Ss, Sclerotinia sclerotiorum (SS1G_14401). For B. cinerea Frp1, an alternative gene model was used. (B) An alignment of the F. oxysporum, F. graminearum and B. cinerea Frp1 proteins was created using MacVector software. The F‐box domain is underlined.

The sequence conservation, including the F. oxysporum, F. graminearum and B. cinerea Frp1 homologues, is unequally distributed over the protein. In particular, the regions near the N‐terminus (amino acids 38–175) and after the central domain (amino acids 425–495) are weakly conserved, whereas the regions in the central domain (amino acids 175–425) and the C‐terminus (amino acids 495–595) show greater conservation (Fig. 1B).

Generation of F. graminearum and B. cinerea FRP1 deletion mutants

To investigate the function of Frp1 in fungi other than F. oxysporum, deletions of FRP1 were made in F. graminearum and B. cinerea, a phylogenetically and phenotypically more distant fungus. The entire FgFRP1 open reading frame (ORF) (FGSG_01326) was replaced in F. graminearum strain Z03643 by a geneticin resistance cassette (Han et al., 2007) (Fig. S2A, see Supporting Information). Correct homologous recombination in two independent mutants was confirmed by Southern analysis (Fig. S2B). Vegetative growth of the mutants was similar to that of the wild‐type, but conidiation was reduced in the mutants when grown on CMC medium (see Experimental details). The addition of 1% glucose to the medium restored conidiation to the wild‐type level (data not shown).

In B. cinerea, the BcFRP1 ORF (BC1G_09967) was replaced in strain B05.10 by a nourseothricin resistance cassette (Fig. S3A, see Supporting Information). Correct homologous recombination in two independent mutants was confirmed by polymerase chain reaction (PCR) (Fig. S3C) and Southern analysis (Fig. S3E). In addition, one deletion mutant was complemented with the BcFRP1 locus (Fig. S3B), resulting in replacement of the nourseothricin resistance cassette by the BcFRP1 locus and the hygromycin resistance cassette. Again, homologous recombination was verified by PCR (Fig. S3D) and Southern analysis (Fig. S3E). The vegetative growth and conidiation of these mutants were similar to those of the wild‐type (data not shown).

FRP1 is not required for natural infection in F. graminearum or B. cinerea

As the F. oxysporumΔfrp1 mutant displays a loss of pathogenicity, the F. graminearum and B. cinereaΔfrp1 mutants were also assessed for their ability to cause disease on their natural hosts. Wild‐type F. graminearum, two ΔFgfrp1 mutants and an ectopic transformant were point inoculated in wheat ears. Two weeks after inoculation, the total number of infected spikes on each ear was counted. The two independent ΔFgfrp1 mutants caused disease similar to the wild‐type and the ectopic transformant (Fig. 2A), indicating that FgFRP1 is dispensable for ear infection.

Figure 2.

The Δfrp1 mutant strains of Fusarium graminearum and Botrytis cinerea are not affected in the infection of commonly infected plant organs. (A) Fusarium graminearum wild‐type (WT), ectopic mutant and two ΔFgfrp1 mutant strains were assessed in a wheat infection assay. Bars indicate the total amount of infected spikes after 2 weeks. (B) Botrytis cinerea WT and two ΔBcfrp1 mutants were assessed on primary leaves of bean plants. Photographs were taken 72 h post‐inoculation (hpi). (C) Botrytis cinerea WT and mutant ΔBcfrp1#4 were assessed in leaf and fruit infection assays. Bars indicate the lesion diameter.

The wild‐type and two ΔBcfrp1 mutants of B. cinerea were inoculated on the primary leaves of bean plants (Fig. 2B). In addition, one deletion mutant was extensively tested on various plants and fruits, such as detached leaves of tobacco and tomato plants, apple, pepper and tomato fruits. In none of these assays a significant difference in lesion size or appearance was observed between the wild‐type and ΔBcfrp1 mutant strains (Fig. 2C), indicating that, also in B. cinerea, BcFRP1 is not required for pathogenicity.

FgFRP1 is required for root infection by F. graminearum

Although the common infection route of F. graminearum on cereals is believed to be by the dispersion of spores on cereal ears, infections of stems, seeds and roots have also been reported (Lysøe et al., 2006; Stephens et al., 2008). Therefore, the ability of the F. graminearumΔFgfrp1 mutant to cause disease on barley roots was assessed. Barley seeds were incubated with a spore suspension of the wild‐type, the ectopic mutant or the F. graminearumΔFgfrp1 mutants and, after 2 weeks of germination, lesions on the roots and fungal growth within the roots were determined. The roots developed from seeds infected with the F. graminearumΔFgfrp1 mutants showed less brown spots and lesions compared with roots developed from seeds infected with the wild‐type and ectopic mutant (Fig. 3). When seeds and attached roots were surface sterilized, cut into pieces of approximately 5 mm and placed on agar plates, mycelial outgrowth from all root pieces (∼5.5 cm in total) could be observed from wild‐type infected roots. The outgrowth of the ΔFgfrp1 mutant from infected roots was restricted to the seed and the first ∼1 cm of the root (Table 1). This suggests that the F. graminearumΔFgfrp1 mutant, like the F. oxysporumΔFofrp1 mutant, has a reduced ability to colonize and grow inside roots, resulting in a reduced pathogenicity phenotype.

Figure 3.

Fusarium graminearumΔFgfrp1 mutants show impaired infection of barley roots. Fusarium graminearum wild‐type (WT), an ectopic transformant and two ΔFgfrp1 strains were assessed in a barley root infection assay. Bars indicate the percentage of roots containing lesions. The two ΔFgfrp1 mutant strains show a significantly smaller number of diseased roots when compared with the wild‐type.

Table 1.

Fungal outgrowth of roots infected with wild‐type or ΔFgfrp1 mutants.

Strain	Root length displaying fungal outgrowth (cm)
Wild type PH‐1	5.5 ± 0.5
ΔFgfrp1 #10	1 ± 0.5
ΔFgfrp1 #14	1 ± 0.5
Control (water)	0

FRP1 is required for sexual reproduction in F. graminearum and B. cinerea

As sexual reproduction of F. oxysporum remains to be established under laboratory conditions and has not been observed in nature, the role of Frp1 during this process has not been evaluated previously. However, both F. graminearum and B. cinerea are capable of a sexual life cycle, and the role of Frp1 during the sexual life cycle was assessed in these two fungi. As a homothallic fungus, F. graminearum can develop perithecia and ascospores without the need for a mating partner. When placed on carrot agar plates, which induces the sexual reproduction process in the wild‐type, the F. graminearumΔFgfrp1 mutant proved to be sterile (Fig. 4A). This suggests that Frp1 is required for sexual reproduction in F. graminearum.

Figure 4.

The Δfrp1 mutants from Fusarium graminearum and Botrytis cinerea are sterile. (A) The ΔFgfrp1 mutant from F. graminearum does not produce perithecia when placed on carrot agar, whereas the wild‐type (WT) strain does. (B) Botrytis cinerea WT and ΔBcfrp1 mutant strains are able to produce sclerotia that are morphologically slightly different (left panels). Apothecia are produced when wild‐type sclerotia are fertilized with microconidia of B. cinerea wild‐type or ΔBcfrp1 mutant strain (male, top right panel). Sclerotia of the wild‐type B. cinerea strain can produce apothecia after fertilization with microconidia from a wild‐type strain, whereas sclerotia of ΔBcfrp1 mutant strain are female sterile after fertilization with microconidia from a wild‐type strain (female, bottom right panel). Note the slimy appearance of the mutant sclerotia, which produce copious amounts of extracellular polysaccharide.

Botrytis cinerea is a heterothallic fungus with a biallelic mating system. This requires for sexual reproduction: the fertilization of sclerotia (melanized mycelial mass structures) of one isolate (defined as ‘female’) by microconidia (defined as ‘male’) of the opposite mating type (Faretra et al., 1988). The B. cinereaΔBcfrp1 mutant is able to produce sclerotia, although with a delay of approximately 5 days when compared with the wild‐type (Fig. 4B, top left). Moreover, sclerotia of the ΔBcfrp1 mutant show a slightly different appearance relative to the wild‐type recipient strain B05.10; they are smaller and have an intensely black colour and a rough‐edged surface. By contrast, sclerotia of the wild‐type are typically larger, have a smoother surface and a shiny, silvery colour (Fig. 4B, bottom left). When the ΔBcfrp1 mutant (generated in strain B05.10 of MAT1‐1 identity) was used as a male in a cross with wild‐type strain SAS405 (MAT1‐2 identity), acting as female, it resulted in the production of apothecia of normal appearance, similar to those obtained in crosses of the wild‐type strain B05.10 with SAS405 (Fig. 4B, top right, and Table 2). However, when sclerotia of the ΔBcfrp1 mutant were fertilized with microconidia of SAS405, no apothecia were produced, whereas the fertilization of sclerotia of B05.10 with microconidia of SAS405 resulted in the formation of apothecia of normal appearance. The sclerotia of the ΔBcfrp1 mutant formed a thick gelatinous polysaccharide layer, and mycelium grew into the air (Fig. 4B, bottom right, and Table 2) and occasionally formed conidiophores with macroconidia (data not shown). This peculiar phenotype has not been observed in other transformants or mutants analysed to date (Rui and Hahn, 2007; Segmuller et al., 2008; J. van Kan, unpublished data). We conclude that the B. cinereaΔBcfrp1 mutant is female sterile.

Table 2.

Botrytis cinerea crosses and apothecia formation with wild‐type strains SAS405 and B05.10 and with SAS405 and the ΔBcfrp1 mutants as male and female.

Cross no.	No. of biological replicates	Sclerotia (♀)	Microconidia (♂)	Formation of apothecia
1	5	SAS405	B05.10	+
2	1	SAS405	—	−
3	3	B05.10	SAS405	+
4	5	Δfrp1#4	SAS405	−
6	1	Δfrp1#4	—	−
7	6	SAS405	Δfrp1‐4‐1	+

Deletion of FRP1 results in a unique carbon source utilization phenotype for each of the examined fungi

The F. oxysporumΔFofrp1 mutant shows significant growth reduction on a broad variety of carbon sources, including root exudate and cell wall components, which we correlated with a loss of pathogenicity (Jonkers et al., 2009). As the deletion of FRP1 in the two other fungi does not result in a loss of pathogenicity on aerial plant parts, it raises the question whether Δfrp1 mutants of the other two fungi display the same growth phenotype as the F. oxysporumΔFofrp1 mutant. To investigate this, the three fungal species (wild‐type strains and the corresponding Δfrp1 mutants) were analysed in a BiologFF Microtiter plate™ assay. This method allows simultaneous growth assessment on 95 different carbon sources (Khalil and Alsanius, 2009).

The growth behaviour of the F. oxysporumΔFofrp1 mutant largely corresponds to the growth phenotype observed earlier on agar plates (Jonkers et al., 2009). Strong growth reduction is seen on carboxylic acids and amino acids and on a few polysugars and sugar acids. Enhanced growth is observed on two cycle polysaccharides, two methylglycosides and two sugar alcohols (Fig. 5 and Figs S4–S9, see Supporting Information). The F. graminearumΔFgfrp1 mutant is reduced in growth on several mono‐ and polysugars and on a few amino acids and carboxylic acids. Enhanced growth is observed on two sugar alcohols and two polysugars (Fig. 5 and Figs S4–S9). In general, deletion of FgFRP1 in F. graminearum leads to a less severe growth phenotype on various carbon sources than the deletion of FoFRP1 in F. oxysporum. In contrast, the B. cinereaΔBcfrp1 mutants show no growth reduction on any carbon source, but enhanced growth on many, mainly mono‐, di‐ and trisaccharides, two sugar alcohols and two glucosides (Fig. 5 and Figs S4–S9). Growth of the complemented ΔBcfrp1 mutant on the above‐mentioned carbon sources was indistinguishable from that of the wild‐type (data not shown). Both F. graminearum and B. cinereaΔfrp1 strains grow normally on agar plates containing ethanol or cell wall components (data not shown). This contrasts with the growth impairment on these carbon sources seen with the F. oxysporumΔFofrp1 mutant (Jonkers et al., 2009). We conclude that the function of Frp1 in growth on nonsugar carbon sources is not conserved in F. graminearum and B. cinerea.

Figure 5.

Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinereaΔfrp1 mutants on various carbon sources. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants.

Introduction of the BcFRP1 gene in the F. oxysporum FRP1 deletion mutant results in partial growth restoration

A different molecular function of Frp1 between the two Fusarium species and Botrytis might explain the observed differences in carbon source utilization and pathogenicity. To test this, we introduced FgFRP1 or BcFRP1 in the F. oxysporumΔFofrp1 mutant. The F. oxysporumΔFofrp1 mutant was transformed not only with the BcFRP1 gene driven by its native promoter sequence, but also with the BcFRP1 gene driven by the FoFRP1 promoter sequence. Fusarium oxysporum wild‐type, ΔFofrp1 and the different transformants were tested in the microtitre assay mentioned earlier to assess growth on different carbon sources. Full growth restoration was observed for the F. oxysporumΔFofrp1 mutant transformed with FoFRP1 or FgFRP1. In addition, the transformants with the BcFRP1 gene controlled by the FoFRP1 promoter showed normal growth on all carbon sources tested (Fig. 6A). However, the strain with the BcFRP1 gene behind its native promoter showed a different phenotype. Growth restoration was observed on organic acids (2‐keto‐d‐gluconic‐acid, α‐keto‐glutaric acid, l‐malic acid, quinic acid, d‐saccharic acid and succinic acid) and amino acids (l‐alanine, l‐aspartic acid, l‐glutamic acid, l‐serine, l‐threonine), but lack of full restoration or even growth reduction was observed on several sugar carbon sources, including the monosaccharides l‐arabinose, d‐fructose, d‐galactose, α‐d‐glucose, l‐rhamnose and d‐xylose and on α‐d‐lactose and glycogen (Fig. 6A). This growth reduction, however, was not seen with all sugar carbon sources, as growth enhancement was observed on sucrose and raffinose. These experiments suggest that Frp1 of B. cinerea and F. graminearum are functional homologues of F. oxysporum Frp1, but that the promoter sequence of the BcFRP1 gene affects the ability of this protein to restore a wild‐type growth phenotype in F. oxysporumΔFofrp1. When the relative transcript levels of the two different BcFRP1 versions were compared with each other by quantitative reverse transcription‐polymerase chain reaction (RT‐PCR) using the constitutively expressed FEM1 gene (Schoffelmeer et al., 2001) as reference, the level of BcFRP1 driven by the FoFRP1 promoter was about five to ten times higher than the level of BcFRP1 driven by its native promoter (Fig. 6B). This suggests that a low BcFRP1 transcript level limits growth on simple sugars, in contrast with either no transcript or higher transcript levels.

Figure 6.

Growth of the Fusarium oxysporumΔFofrp1 mutant is restored by transformation with FgFRP1, but not with BcFRP1, controlled by its native promoter. (A) Growth indicated by the absorbance at 600 nm (OD₆₀₀) of: F. oxysporum wild‐type (WT), ΔFofrp1 and transformants bearing the FgFRP1 gene, the BcFRP1 gene or BcFRP1 driven by the FoFRP1 promoter (pFoFRP1‐BcFRP1) on various carbon sources. (B) Relative transcript levels of BcFRP1 in transformants bearing the BcFRP1 gene or BcFRP1 driven by the FoFRP1 promoter (pFoFRP1‐BcFRP1). FEM1 was used as reference to calculate the relative expression levels using the ΔC _t method.

Both FgFRP1 and BcFRP1 can complement the pathogenicity defect of the F. oxysporumΔFofrp1 mutant

To determine whether the FRP1 genes from the different fungal species can restore pathogenicity to the F. oxysporumΔFofrp1 strain, the F. oxysporum wild‐type, ΔFofrp1 strain and the transformants containing FgFRP1 and BcFRP1 genes controlled by their native promoters were tested in a bioassay. All strains tested were pathogenic to the same level as the wild‐type (Fig. 7), confirming that the FgFRP1 and BcFRP1 genes are functional homologues of FoFRP1. Interestingly, the F. oxysporum transformant harbouring BcFRP1 controlled by its native promoter is fully pathogenic, in spite of its reduced growth on several sugars.

Figure 7.

The FgFRP1 or BcFRP1 genes restore virulence of the Fusarium oxysporumΔFofrp1 mutant. Bars indicate the disease index scores of tomato plants infected by F. oxysporum wild‐type (WT), ΔFofrp1, four transformants bearing the FgFRP1 gene and two transformants bearing the BcFRP1 gene.

DISCUSSION

Most F‐box proteins play a crucial role in the ubiquitination and degradation of proteins that have become redundant. They act by recruiting target proteins to an SCF complex where they are marked with ubiquitin for proteasomal degradation (Willems et al., 2004). However, some F‐box proteins, including Frp1 from F. oxysporum, seem not to comply with this hypothesis (Jonkers and Rep, 2009a). Although the molecular function of Frp1 remains to be discovered, this work sheds new light on the biological function of Frp1 in different fungi. Frp1 conservation is restricted to filamentous fungi from the classes Sordariomycetes, Leotiomycetes and Dothideomycetes, with homologues from the latter class showing much less conservation. That this specific F‐box protein is not conserved in Eurotiomycetes could be caused by gene loss. Fungi have many F‐box proteins and it cannot be ruled out that a functional counterpart of Frp1 is present in Eurotiomycetes. The fact that over 50 F‐box proteins, many of unknown function, can be found in a single fungal genome underlines the complexity of processes requiring F‐box proteins.

In this study, we compared the role of Frp1 among three plant‐pathogenic fungi in pathogenicity, carbon source assimilation and sexual reproduction. For this purpose, in addition to the ΔFofrp1 mutant of the root pathogen F. oxysporum, deletions of FRP1 in the putative hemibiotrophic wheat ear pathogen F. graminearum (Sordariomycete) and the necrotrophic broad host range pathogen B. cinerea (Leotiomycete) were created, followed by extensive comparisons of the deletion phenotypes. As observed earlier in F. oxysporum, the deletion of FgFRP1 in F. graminearum resulted in impaired infection of roots. This reduction in pathogenicity was not observed on wheat ears. Deletion of BcFRP1 in B. cinerea had no effect on its ability to cause disease on the various plant leaves and fruits tested. These differences could be related to different requirements of metabolic pathways in roots and aerial plant parts, or, for B. cinerea, a different role of Frp1 in the regulation of metabolism.

In F. oxysporum, Frp1 is required for growth on a broad array of nonsugar carbon sources, such as polysaccharides, and for the expression of CWDE genes (Jonkers et al., 2009), qualitative traits which are likely to be required for root infection. Cross‐species' complementation of the Fofrp1 mutant with the BcFRP1 gene restores pathogenicity and growth on nonsugar carbon sources, such as amino acids and carboxylic acids. However, growth on monosaccharides is not restored and, on some sugar sources, growth is even repressed. This indicates that a loss of pathogenicity in the Fofrp1 mutant is not a result of inefficient growth on the monosaccharides l‐arabinose, l‐rhamnose or d‐xylose. In addition, growth on the monosaccharides d‐fructose, d‐galactose and α‐D‐glucose is apparently not a prerequisite of plant pathogenicity. In contrast with the F. oxysporumΔFofrp1 mutant, the F. graminearumΔFgfrp1 mutant grows normally on agar plates containing various nonsugar carbon sources and polysaccharides. This suggests that the F. graminearumΔFgfrp1 mutant secretes sufficient amounts of CWDEs and uses the degradation products to generate energy. However, the expression of the CWDE gene xylanase 4 (XYL4) in the F. graminearumΔFgfrp1 mutant grown on oat spelt xylan is undetectable, whereas in the wild‐type it is observed (Fig. S10, see Supporting Information). This suggests that the F. graminearumΔFgfrp1 mutant also shows altered CWDE gene expression, possibly leading to compromised root infection, but not to reduced growth on agar plates and cereal ear infection.

The major differences between the Δfrp1 mutants of the two Fusarium spp. and B. cinerea relative to the wild‐type strains concern carbon catabolism. The B. cinereaΔBcfrp1 mutant shows enhanced growth on specific simple sugars, whereas the F. oxysporumΔFofrp1 mutant shows reduced growth on nonsugar carbon sources. This suggests that, in B. cinerea, Frp1 suppresses growth on simple sugar sources, whereas, in F. oxysporum, Frp1 is required to activate growth on nonsugar sources. During the infection of leaves and fruits, however, no faster spread or disease development was observed. Apparently, the suppression of growth on simple sugars by Frp1 has no role during the infection process in B. cinerea, and the relevance of this suppression remains to be elucidated.

Based on the cross‐species' complementation experiments, where the B. cinerea and Fusarium FRP1 genes were introduced into the F. oxysporumΔFofrp1 mutant strain, it seems that B. cinerea Frp1 functions differently in carbon source utilization in F. oxysporum than Frp1 from the Fusarium species. This difference could be attributed in part to the promoter sequence. Usage of the native B. cinerea promoter led to lower transcript levels and reduced growth on simple sugars when compared with the usage of the F. oxysporum promoter. How these lower transcript levels influence the activity and levels of the protein is unknown. As reported earlier, the promoter sequence of FoFRP1 contains an upstream ORF that starts 163 bp upstream of ATG and is 33 bp long (Duyvesteijn et al., 2005). We found evidence that, in F. oxysporum, the upstream ORF suppresses the expression levels of FoFRP1 (W. Jonkers and M. Rep, unpublished data). In the upstream region of the BcFRP1 gene, a possible upstream ORF is also present (W. Jonkers, personal observations), which starts 175 bp upstream of ATG and is 96 bp long. Whether this upstream ORF also affects the expression of BcFRP1 in these different complemented strains is unknown.

Considering the above, it seems that the molecular function of Frp1 is conserved in the different fungi, but is used by each of the fungi in a different manner. Whether an interplay between Frp1 and Cre1, as seen in F. oxysporum (Jonkers and Rep, 2009b), is also present in F. graminearum and B. cinerea remains to be elucidated, and, if there is an interplay, whether it contributes to the phenotypes observed. Another conserved protein that plays a role in CCR and pathogenicity, but is used differently in different fungi, is the protein kinase Snf1. In yeast, this kinase is known to phosphorylate the Cre1 counterpart Mig1 (Treitel et al., 1998). A role of Snf1 in regulating Cre1 in filamentous fungi has still not been established (Cziferszky et al., 2003). Snf1 is required for pathogenicity in different plant‐pathogenic fungi but deletion of the SNF1 gene results in different phenotypes in each fungus (Lee et al., 2009a; Ospina‐Giraldo et al., 2003; Tonukari et al., 2000; Yi et al., 2008). Both the F. oxysporum and F. graminearumΔsnf1 mutants show reduced expression of CWDE genes, but they differ in their growth phenotypes on sugars. Both mutants show reduced growth on galactose, trehalose and xylan, but only the F. oxysporumΔsnf1 mutant shows reduced growth on arabinose, fructose and xylose. This suggests that, in these two fungi, and similar to Frp1, Snf1 is required in common as well as different metabolic pathways. The B. cinerea SNF1 deletion mutant shows a more severe phenotype; it grows poorly on all carbon sources, including glucose. In fact, the growth phenotype of this mutant on any carbon source is comparable with growth on water agar: it only produces very thin aerial mycelium. This indicates that SNF1 in B. cinerea is required for basal growth and/or carbon source sensing, and not specifically for growth on alternative carbon sources or the expression of CWDE genes (J. Schumacher and B. Tudzynski, University of Münster, Münster, Germany, personal communication). It appears, therefore, that CCR is regulated differently in B. cinerea compared with other fungi, although the proteins involved are similar and, in the case of Frp1, have been proven to be functional homologues of each other.

This work also attributes a novel function to Frp1; it is required for sexual reproduction in F. graminearum and B. cinerea. How Frp1 regulates the sexual cycle in F. graminearum is not known. Using gene expression profiling, FgFRP1 was found to be down‐regulated in a mat1‐2 mutant in F. graminearum, suggesting that FgFRP1 itself might be controlled by mating‐type locus genes (Lee et al., 2006). Alternatively, the function of FgFrp1 in carbon assimilation may explain its requirement in the sexual cycle. Such a relationship has been shown for the Δsnf1 and glyoxylate and methylcitrate cycle mutants Δicl1 and Δmcl1 in F. graminearum (2009a, 2009b). The genes deleted in the last two mutants are required for growth on C2 carbon sources (ICL1) or propionate (MCL1). The double‐deletion mutant (Δicl1/ Δ mcl1) shows impaired sexual reproduction.

The B. cinerea BcFRP1 deletion mutant displayed a specific impairment in female fertility, but not in male fertility. To our knowledge, this is the first description of a B. cinerea mutant that is capable of producing sclerotia, but fails to develop apothecia. Some previously described mutations in B. cinerea also result in female sterility, but, in these cases, the mutants are entirely defective in the production of sclerotia (Rui and Hahn, 2007; Segmuller et al., 2008) and are therefore, by definition, female sterile. It is unclear whether the mating defect in the B. cinerea BcFRP1 deletion mutant occurs during fertilization of the sclerotia by microconidia of the mating partner, or during plasmogamy. The mechanism underlying the defect in sexual reproduction in the B. cinerea BcFRP1 deletion mutant also remains to be resolved. It is possible that Frp1 is required to target specific proteins to an SCF complex during sexual reproduction. The identification of Frp1 target proteins in F. graminearum and B. cinerea might clarify its role during sexual reproduction.

In conclusion, Frp1, a conserved F‐box protein present in some classes of filamentous fungi, plays a role in carbon assimilation and sexual reproduction in different fungal species. The protein itself can be exchanged between the different fungi investigated; however, it is used in each in a specific manner, resulting in different carbon source assimilation phenotypes. This divergence of function may be correlated with differences in fungal lifestyle and infection strategies.

EXPERIMENTAL PROCEDURES

Biomaterials and plate assays

Fusarium oxysporum f.sp. lycopersici race 2 isolate 007 (Mes et al., 1999), the F. oxysporumΔFofrp1 strain (Duyvesteijn et al., 2005), Fusarium graminearum isolate Z03643 and F. graminearumΔFgfrp1 strains (Han et al., 2007) and Botrytis cinerea strain B05.10 (Quidde et al., 1999) were used in this study. Spores from F. oxysporum strains were harvested after 5 days of growth in NO₃ medium [5 mm KNO₃, 0.17% yeast nitrogen base (YNB) without amino acids or ammonia (Difco, Le Pont de Claix, France) and 3% sucrose] after 5 days of shaking on a rotary shaker at 150 rpm and 25 °C. Spores from F. graminearum strains were harvested after 7 days of growth in CMC medium (15 g/L carboxymethylcellulose, 1 g/L yeast extract, 0.5 g/L MgSO₄.7H₂O, 1 g/L NH₄NO₃, 1 g/L KH₂PO₄ and, for the F. graminearumΔFgfrp1 strains, 1% w/v glucose) on a rotary shaker at 150 rpm and 25 °C. Botrytis cinerea spores were harvested from 10‐ to 14‐day‐old PDAB (potato dextrose agar supplemented with 10% w/v hackled bean leaves) plates and resuspended in H₂O for the BiologFF assay or Gamborg's B5 medium (Duchefa, Haarlem, The Netherlands) supplemented with 10 mm glucose and 10 mm KH₂PO₄/K₂HPO₄, pH 6.4, for the plant infection assays.

Fusarium oxysporum plate assays were performed on agar plates containing NO₃ medium, 1.5% bacto agar and 1% w/v variable carbon source instead of sucrose. Fusarium graminearum plate assays were performed on DFM (Yoder and Christianson, 1998) agar plates containing 1% w/v variable carbon source instead of glucose. Plates were inoculated with 2 × 10⁴ spores on the centre of the plate, followed by 1 week of incubation at 25 °C.

Self‐fertilization tests in F. graminearum were performed by growing mycelia of strains on carrot agar for 5–7 days. Strains were mock fertilized as described previously (Leslie and Summerell, 2006). For the B. cinerea outcrossing tests, the heterothallic strain carrying the mat1‐2 deletion (Lee et al., 2003) was used as female and ΔBcfrp1 was used as male, so that all ascospore progeny resulted from heterozygous crosses. Female mycelia on plates were fertilized with a suspension of conidia from the male strains grown on CMC medium. Ascospores and asci in perithecia were observed after incubation for 10–14 days following fertilization.

The B. cinerea plate assays were performed on Gamborg B5 agar plates containing 1% w/v variable carbon source. Plates were inoculated with mycelial plugs. Sclerotia formation was analysed by inoculating CM agar plates (Pontecorvo et al., 1953) with a mycelial plug, followed by incubation in the dark for 2 weeks at room temperature. Sexual crosses of B. cinerea were performed as described previously (Faretra et al., 1988; Van der Vlugt‐Bergmans et al., 1993).

For the BiologFF assay (Biolog FF plates™), each well was filled with 10⁴ spores in 150 µL. Time point zero was taken after 2 h on a spectrophotometer at 600 nm, when the lyophilized carbon sources were dissolved. Further measurements were made in duplicate at time points 24, 48, 72 and 96 h after inoculation.

Plant infections

The tomato cultivar C32, which is generally susceptible to all F. oxysporum f.sp. lycopersici races, was used for inoculations. Ten‐day‐old seedlings were uprooted, dipped into a 10⁷ spore/mL suspension and potted in soil. The biomass and disease index were quantified as described by Mes et al. (1999) 3 weeks after inoculation.

Point inoculation of wheat cultivar Bobwhite was performed as described previously (Seong et al., 2005). Barley seeds of cultivar ‘Tipple’ (Boerengoed, Wilsum, the Netherlands) were inoculated by incubating the sterile seeds for 1 h in a shaking macroconidia spore suspension, and subsequently potted in vermiculite (Lysøe et al., 2006). After 2 weeks of growth, the roots were assessed for brown lesions.

Intact leaves of French bean (Phaseolus vulgaris), tobacco (Nicotiana tabacum), tomato (Solanum lycopersicum) and fruits of apple (Malus domestica‘granny smith’), red bell pepper (Capsicum annuum) and tomato were inoculated with droplets (7–10 µL) of a B. cinerea conidial suspension (2 × 10⁵ spores/mL). For a reproducible infection, the fruits were wounded with a needle and the droplets were placed on the wound. The infected leaves and fruits were incubated in a plastic propagator box at 20 °C under natural illumination.

Deletion and complementation constructs

For the targeted gene deletion of FgFRP1 in F. graminearum, a construct carrying a geneticin resistance gene cassette, flanked by the 5′ and 3′ flanking regions of FgFRP1, was amplified by a double‐joint PCR method (Yu et al., 2004) with two sets of PCR primers, 1 + 2 and 3 + 4 (Table 3), respectively. The primers 2 and 3 (Table 3) carry sequence tails that overlap with the 5′ and 3′ ends of the gene, respectively. The geneticin resistance gene was controlled by the Aspergillus nidulans trpC promoter and terminator, and the gene fragment was amplified from the vector pII99 (Namiki et al., 2001) with the primers 5 and 6 (Table 3). Three amplicons (5′ flanking, 3′ flanking and gene) were mixed and used as a template for the second round of PCR, which was followed by the final round of PCR using the new nested primer pair 7 and 8 (Table 3). The final product was transformed into the wild‐type strain GZ03643 to generate the FgFRP1 deletion mutant ΔFgfrp1.

Table 3.

Primers used in this study.

Primer no.	Sequence (5′–3′)
1	GAAAGCTGCGGAAGAAACGGGTAA
2	CCTTCAATATCATCTTCTGTCGATTTGGTGTTTGACAAGACAAG
3	GCACAGGTACACTTGTTTAGAGAACGAGAACCTTGCTTTCGCTT
4	GGAACATGCGCCCTTGCTTACTC
5	CGACAGAAGATGATATTGAAGG
6	CTCTAAACAAGTGTACCTGTG
7	GCATCATTCCTTGCATCA
8	TTGCGGTGGTGTTGTATG
9	ATAGTTTAGCGGCCGCGATTGAGTCCGGCTGTATCG
10	GCTCTAGATTATTTACCTGCGGGCTACG
11	CCCAAGCTTAGAGAGGCAGGGAAAGGTTC
12	CCGCTCGAGGAAGAGACGGGAGCGAGAC
13	ATAAGAATGCGGCCGCGAAGAGACGGGAGCGAGAC
14	CCCAAGCTTGATTGAGTCCGGCTGTATCG
15	AAAGGGGCCCCAACGTCAC
16	TTGGATCCTCGCGTTTGATGCTAGAAT
17	AAAACTGCAGCATGGAACCACTTCATCTTACT
18	AAAAAGCTTACCCCCCAATCAATATAAAC
19	AAATCTAGATGTCCCCCCACAGCTGC
20	AAAACTGCAGGGCGTCGAAGAGAAAAC
21	GTCATCGAGAGTTACGGTTC
22	ACTAGTCTGGACATAGATAGAA
23	TCAGTGGCTCATTGGGTGCTCTAA
24	TGTGTCTGGGTGTGTCTCGACAAT
25	AAGCCTTACACCATCCGCTACTCT
26	ACCAGCCTTGTCGGTGATCTTGAA

To generate the B. cinerea BcFRP1 disruption construct, pBcFRP1KO, PCR was performed with primer combinations 9 and 10 (Table 3), in order to amplify a 929‐bp NotI–XbaI fragment corresponding to the upstream region, and primers 11 and 12 (Table 3), in order to amplify a 939‐bp HindIII–XhoI fragment corresponding to the downstream region, using genomic DNA as a template. The fragments were sequentially cloned into pNR1 (Malonek et al., 2004), resulting in pBcFRP1KO. The BcFRP1 complementation construct, pBcFRP1com, was generated by cloning a 3914‐bp HindIII–NotI PCR fragment containing the BcFRP1 ORF, including 929‐bp upstream and 939‐bp downstream sequences obtained with primers 13 and 14 (Table 3), into pOliHP (Rolke et al., 2004), resulting in pBcFRP1com.

The pRD803 plasmid containing HPH and BLE resistance cassettes (Duyvesteijn et al., 2005) was used to create the plasmid containing the F. graminearum FgFRP1 gene pRD803FgFRP1AB. Fusarium graminearum FgFRP1, including a 1347‐bp upstream and 516‐bp downstream region, was amplified using primers 15 and 16 (Table 3) containing ApaI and BamHI sites in their linker, respectively. The amplified F. graminearum FgFRP1 locus was cloned into the pRD803 plasmid previously linearized by ApaI and BamHI.

The pRD803 plasmid was also used to create the plasmid pRD803BcFRP1PH containing the B. cinerea BcFRP1 gene, including 759‐bp upstream and 939‐bp downstream sequences. A BglII/HindIII BcFRP1 fragment isolated from pBcFRP1com was cloned into a BamHI and HindIII linearized pRD803 plasmid. To obtain the final plasmid pRW1pBcFRP1PH, BcFRP1 was digested from pRD803BcFRP1PH using MluI and HindIII restriction sites, and cloned into plasmid pRW1p containing a BLE resistance cassette (Houterman et al., 2009), which had been linearized previously by MluI and HindIII.

To create the vector containing the BcFRP1 ORF preceded by the FoFRP1 promoter sequence, a three‐point ligation strategy was used. The ORF of BcFRP1 was amplified with primers 17 and 18 (Table 3) containing PstI and HindIII linkers, respectively. The promoter sequence was amplified with primers 19 and 20 (Table 3) containing XbaI and PstI linkers, respectively. The two PCR fragments were cloned into vector pRW1p, which was linearized using restriction sites XbaI and HindIII, thereby creating pRW1ppFoBcFRP1XH.

Fungal transformations

Agrobacterium‐mediated transformation was performed as described previously (Takken et al., 2004). Putative transformants were transferred to Czapex Dox agar (CDA, Oxoid, Basingstoke, Hampshire, England) plates equilibrated with 100 mm Tris‐HCl, pH 8.0, containing 100 µg/mL zeocin, 100 µg/mL hygromycin and 200 µm cefotaxim (Duchefa). After 4 days, when the colonies appeared, spores from putative transformants were suspended in 10 µL of sterile water and spread onto potato dextrose agar (PDA, Difco) plates containing 100 µg/mL zeocin, 100 µg/mL hygromycin and 200 µm cefotaxim. Single‐spore colonies were punched out from the plates by ringcaps (Hirschmann, Eberstadt, Germany) and placed on fresh PDA plates. Genomic DNA was obtained from mycelium by extraction with 200 µL of glass beads, 300 µL of phenol–chloroform and 400 µL of TE buffer, pH 8.0, per 1 cm² of mycelium. The presence of the different complemented FRP1 genes in the transformants was verified by PCR, and at least two correct individual transformants were used.

To perform protoplast transformation of F. graminearum, inoculation of mycelial blocks of the wild‐type strain GZ03643 was started in CMC liquid medium (Cappellini and Peterson, 1965) and grown at 25 °C for 3 days. Fungal conidia (1 × 10⁶) produced in CMC were inoculated into 50 mL of YPG (3 g yeast extract, 10 g peptone and 20 g glucose per litre) liquid medium and grown at 25 °C for 12 h. Mycelia harvested by filtration were incubated in 80 mL of 1 m NH₄Cl containing Driselase (10 mg/mL; InterSpex Products, Inc., Foster City, CA, USA) to generate protoplasts. Approximately 5 µg of fusion PCR product obtained by double‐joint PCR was added directly, together with 1.2 mL of 60% (w/v) polyethyleneglycol (PEG) (M _W= 3350), to protoplast suspensions. Fungal transformants carrying the geneticin resistance cassette were selected on regeneration medium (1 g yeast extract, 1 g casein enzymatic hydrolysate, 342 g sucrose and 15 g agar per litre) containing geneticin (75 µg/mL; Sigma‐Aldrich, Zwijndrecht, The Netherlands). Homologous recombination at the FgFRP1 locus was analysed by Southern analysis. For this, genomic DNA was digested with EcoRI and hybridized with an [α‐³²P]dCTP‐labelled probe corresponding to the downstream region of FgFRP1, as described by standard procedures (Sambrook and Russell, 2001).

Protoplast transformation of B. cinerea was performed as described previously (Schulze Gronover et al., 2001). Fungal genomic DNA of the transformants was isolated as described previously (Cenis, 1992). Homologous recombination at the BcFRP1 locus was analysed by PCR and Southern analysis. For the latter, genomic DNA was digested with PstI and hybridized with a [α‐³²P]dCTP‐labelled probe corresponding to the upstream region of BcFRP1, as described previously (Segmuller et al., 2007).

Northern blotting

Wild‐type and a F. graminearumΔFgfrp1 mutant were grown in 100 mL NO₃ medium containing 1% glucose for 5 days with shaking on a rotary shaker at 150 rpm and 25 °C; mycelium was harvested over two layers of miracloth (Calbiochem, Darmstadt, Germany), washed with sterile water and transferred to 100 mL NO₃ medium containing 1% xylan as a carbon source. After 2 days, the mycelium and spores were harvested, followed by RNA extraction and Northern blotting as described previously (Jonkers et al., 2009). Probes for XYL4 were obtained by PCR using primers 21 and 22 (Table 3).

Quantitative RT‐PCR analysis

RNA was isolated from two individual complemented strains containing the BcFRP1 gene driven by its native promoter sequence, and from three individual strains containing the BcFRP1 gene driven by the FoFRP1 promoter sequence, growing on complete medium for 2 days. The subsequent cDNA obtained by RT‐PCR was used as template for quantitative PCR (qPCR), which was performed in three replicates with a DyNamo™ SYBR® Green qPCR (Finnzymes, Woburn, MA, USA) using a DNA‐Engine Peltier thermal cycler (Bio‐Rad, Hercules, CA, USA) equipped with a Chromo4™ real‐time PCR detector and MJ Opticon Monitor™ analysis software (Bio‐Rad, Hercules, CA, USA). To quantify mRNA levels of BcFRP1 and the constitutively expressed FEM1 gene, primers 23 and 24, and 25 and 26, respectively, were used (Table 3).

Supporting information

Fig. S1 Alignment of Frp1 protein domains used for phylogenetic tree construction. An alignment of the various Frp1 proteins was created using MacVector software. The alignment is based on the domain of the Frp1 proteins showing the highest homology, which includes and extends beyond the F‐box domain (underlined).

Fig. S2 Analysis of transformants deleted for FgFRP1 in Fusarium graminearum. A knockout construct containing a geneticin resistance cassette was introduced into the wild‐type strain. (A) Schematic representation of the knockout strategy for FgFRP1 drawn to scale. (B) Southern analysis was performed to verify correct homologous recombination at the FgFRP1 locus in the FgFRP1 disruptant. To this end, chromosomal DNA of the various mutants was digested with EcoRI, blotted and hybridized with the indicated probe corresponding to the FgFRP1 terminator. The FgFRP1 locus of the wild‐type strain (W) is visible as a 5.8‐kb fragment corresponding to the FgFRP1 terminator region and to the 3′ part of the FgFRP1 open reading frame (ORF). In the ectopic mutant (1), an extra band corresponding to a fragment of 10.0 kb is visible and, in the FgFRP1 disruption mutant (2), introduction of the gene replacement cassette by homologous recombination led to the expected replacement of the 5.8‐kb fragment by a fragment of 10.0 kb.

Fig. S3 Analysis of transformants deleted for BcFRP1 in Botrytis cinerea. A knockout construct containing a nourseothricin resistance cassette was introduced into the wild‐type strain. (A) Schematic representation of the knockout strategy for BcFRP1 drawn to scale. (B) Schematic representation of the complementation strategy drawn to scale. (C) Polymerase chain reaction (PCR) analysis was performed to verify homologous recombination and the absence of wild‐type nuclei in the Bcfrp1 deletion strains. (D) PCR analysis was performed to verify homologous recombination and the absence of nuclei lacking BcFRP1 in the BcFRP1 complementation strains. (E) Southern analysis was performed to verify correct homologous recombination at the BcFRP1 locus in the Bcfrp1 deletion and complementation mutants. To this end, chromosomal DNA of the various mutants was digested with PstI, blotted and hybridized with a probe corresponding to the BcFRP1 upstream region. The FRP1 locus of the wild‐type strain (WT) is visible as a 7.4‐kb fragment. In the two BcFRP1 deletion mutants, introduction of the gene replacement cassette by homologous recombination led to the expected replacement of the 7.4‐kb fragment by a fragment of 2.5 kb. Introduction of the BcFRP1 complementation cassette by homologous recombination in the Bcfrp1 deletion mutant led to the expected replacement of the 2.5‐kb fragment by a fragment of 5.0 kb.

Fig. S4 Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea Δfrp1 mutants on monosaccharides. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants. (D) Schematic representation of the observed growth phenotypes of the various wild‐type and corresponding mutant strains. Empty cell, similar growth of wild‐type and Δfrp1 mutants; –, Δfrp1 mutant growth is reduced when compared with the wild‐type; +, Δfrp1 mutant growth is enhanced when compared with the wild‐type.

Fig. S5 Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea Δfrp1 mutants on di‐, tri‐ or cyclic (poly)saccharides. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants. (D) Schematic representation of the observed growth phenotypes of the various wild‐type and corresponding mutant strains. Empty cell, similar growth of wild‐type and Δfrp1 mutants; –, Δfrp1 mutant growth is reduced when compared with the wild‐type; +, Δfrp1 mutant growth is enhanced when compared with the wild‐type.

Fig. S6 Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea Δfrp1 mutants on amino acids. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants. (D) Schematic representation of the observed growth phenotypes of the various wild‐type and corresponding mutant strains. Empty cell, similar growth of wild‐type and Δfrp1 mutants; –, Δfrp1 mutant growth is reduced when compared with the wild type; +, Δfrp1 mutant growth is enhanced when compared with the wild type.

Fig. S7 Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea Δfrp1 mutants on carboxylic acids. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants. (D) Schematic representation of the observed growth phenotypes of the various wild‐type and corresponding mutant strains. Empty cell, similar growth of wild‐type and Δfrp1 mutants; –, Δfrp1 mutant growth is reduced when compared with the wild‐type; +, Δfrp1 mutant growth is enhanced when compared with the wild‐type.

Fig. S8 Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea Δfrp1 mutants on sugar acids and sugar alcohols. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants. (D) Schematic representation of the observed growth phenotypes of the various wild‐type and corresponding mutant strains. Empty cell, similar growth of wild‐type and Δfrp1 mutants; –, Δfrp1 mutant growth is reduced when compared with the wild‐type; +, Δfrp1 mutant growth is enhanced when compared with the wild‐type.

Fig. S9 Growth of the Fusarium oxysporum, Fusarium graminearum and Botrytis cinerea Δfrp1 mutants on phosphate sugar, polyamine, methylglycosides, benzyl alcohol, purines, pyrimidine and glucosides. Growth is indicated by the absorbance at 600 nm (OD₆₀₀) of: (A) F. oxysporum wild‐type (WT), ΔFofrp1 and ΔFofrp1 complemented with FRP1; (B) F. graminearum and two independent ΔFgfrp1 mutants; and (C) B. cinerea and two independent ΔBcfrp1 mutants. (D) Schematic representation of the observed growth phenotypes of the various wild‐type and corresponding mutant strains. Empty cell, similar growth of wild‐type and Δfrp1 mutants; –, Δfrp1 mutant growth is reduced when compared with the wild‐type; +, Δfrp1 mutant growth is enhanced when compared with the wild‐type.

Fig. S10 Expression of XYL4 is undetectable in the Fusarium graminearum ΔFgfrp1 mutant. Using Northern analysis, the expression of the cell wall‐degrading enzyme (CWDE) gene XYL4 was assessed in F. graminearum wild‐type (WT) and ΔFgfrp1 mutant during noninducing (glucose) and inducing (xylan) conditions.

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item