UDP-Galactopyranose Mutase Mediates Cell Wall Integrity, Polarity Growth, and Virulence in Fusarium graminearum

ABSTRACT

CHY1 is a zinc finger protein unique to microorganisms that was found to regulate polarized tip growth in Fusarium graminearum, an important pathogen of wheat and barley. To further characterize its functions, in this study we identified CHY1-interacting proteins by affinity purification and selected UDP-galactofuranose (Galf) mutase (UGMA) for detailed characterization, because UGMA and UDP-Galf are unique to fungi and bacteria and absent in plants and animals. The interaction between CHY1 and UGMA was confirmed by yeast two-hybrid assays. Deletion of UGMA in F. graminearum resulted in significant defects in vegetative growth, reproduction, cell wall integrity, and pathogenicity. Infection with the ΔugmA mutant was restricted to the inoculated floret, and no vomitoxin was detected in kernels inoculated with the ΔugmA strain. Compared to the wild type, the ΔugmA mutant produced wide, highly branched hyphae with thick walls, as visualized by transmission electron microscopy. UGMA tagged with green fluorescent protein (GFP) mainly localized to the cytoplasm, consistent with the synthesis of Galf in the cytoplasm. The Δchy1 mutant was more sensitive, while the ΔugmA mutant was more tolerant, to cell wall-degrading enzymes. The growth of the ΔugmA mutant nearly ceased upon caspofungin treatment. More interestingly, nocodazole treatment of the ΔugmA strain attenuated its highly branched morphology, while caspofungin inhibited the degree of the twisted Δchy1 mycelia, indicating that CHY1 and UGMA probably have opposite effects on cell wall architecture. In conclusion, UGMA is an important pathogenic factor that is specific to fungi and bacteria and required for cell wall architecture, radial growth, and caspofungin tolerance, and it appears to be a promising target for antifungal agent development.

IMPORTANCE The long-term use of chemical pesticides has had increasingly negative impacts on the ecological environment and human health. Low-toxicity, high-efficiency and environmentally friendly alternative pesticides are of great significance for maintaining the sustainable development of agriculture and human and environmental health. Using fungus- or microbe-specific genes as candidate targets provides a good foundation for the development of low-toxicity, environmentally friendly pesticides. In this study, we characterized a fungus- and bacterium-specific UDP-galactopyranose mutase gene, ugmA, that contributes to the synthesis of the cell wall component Galf and is required for vegetative growth, cell wall integrity, deoxynivalenol (DON) production, and pathogenicity in F. graminearum. The ugmA deletion mutant exhibited increased sensitivity to caspofungin. These results demonstrate the functional importance of UGMA in F. graminearum, and its absence from mammals and higher plants constitutes a considerable advantage as a low-toxicity target for the development of new anti-Fusarium agents.

INTRODUCTION

Fusarium head blight (FHB) is one of the most devastating wheat diseases worldwide and is mainly caused by Fusarium graminearum (1, 2). FHB not only causes serious yield losses but also contaminates grains with mycotoxins such as deoxynivalenol (DON) and zearalenone, which are difficult to degrade, are teratogenic and carcinogenic, and pose a great threat to human and animal health (3–5). Due to the lack of highly resistant wheat cultivars, chemical control remains the major strategy for controlling FHB, and the available fungicides are largely limited to tebuconazole, prothioconazole, phenamacril, and carbendazim (6). However, the long-term intensive application of fungicides frequently causes F. graminearum resistance (7–9). Moreover, the problems of lingering pesticide residues, ecological water pollution, and food safety caused by the continuous and large-scale use of fungicides are becoming increasingly serious (10, 11). Therefore, it is of great significance to strengthen the research on fungus-specific pathogenic genes as targets to develop low-toxicity, environmentally friendly pesticides.

The cell wall is an important extracellular structure that protects fungal organisms, serves as the interface between the internal and external environments of the cell, and is essential for fungal growth, survival, stress tolerance, and virulence (12, 13). Many cell wall components are absent in mammals and are therefore attractive drug targets for antifungal-drug development. Cell wall inhibitors destroy the cell wall, causing the cell to disintegrate or collapse, mainly by inhibiting the synthesis of or directly binding to cell wall components. At present, fungicides or biological agents with practical relevance in this context consist mainly of compounds that affect the synthesis of cell wall components; for example, kitazine and polyoxins inhibit chitin synthesis, echinocandin agents (e.g., caspofungin) inhibit β-1,3-glucan synthesis, and dimethomorph inhibits cellulose synthesis (14). Galactofuranose (Galf) is an important constituent of the microbial cell surface. In Aspergillus fumigatus, Galf residues account for 5% of the wall and are important for wall structure and pathogenicity (15). Different Galf-containing molecules, including the polysaccharides galactomannan, glycosylphosphatidylinositol (GPI)-anchored lipophosphogalactomannan, N- and O-linked glycans of glycoproteins, and several sphingolipids, are found in both the inner and outer layers of the cell wall or bound to the plasma membrane (16, 17). Galf-containing molecules contribute to wild-type growth and virulence in pathogenic bacteria, fungi, and protozoa and are ubiquitous in nonmammalian species (18), making enzymes in the corresponding biosynthetic pathway attractive drug targets.

UDP-Galf exists in equilibrium with UDP-galactopyranose (UDP-Galp) in solution and is the precursor of Galf-containing molecules (19). Briefly, UDP-Galf is synthesized in the cytoplasm, in which UDP-glucose is converted to UDP-Galp by UDP-glucose 4-epimerase (UGE), UDP-Galp is converted to UDP-Galf by UDP-galactopyranose mutase (UGM), and UDP-Galf is eventually transferred to the Golgi apparatus by Galf transferases (UGT) to participate in Galf-containing sugar chain biosynthesis in fungi (20). In the Galf synthesis pathway, UGM is a unique enzyme that catalyzes the conversion of UDP-Galp to UDP-Galf. In Aspergillus niger and A. nidulans, UDP-Galf mutase (UGMA) contributes to cell wall integrity and wall structure (19, 21). In addition, two other sequentially acting genes (UGEA and UGTA) in the Galf biosynthesis pathway have been identified and functionally analyzed in A. niger and A. fumigatus (16, 18, 22). At present, there are no available studies on the Galf synthesis pathway and related genes in F. graminearum.

CHY1 is a novel zinc finger protein that was first identified in F. graminearum and regulates rapid linear growth (23). To identify new CHY1-related proteins that regulate polarity growth, pulldown assays were performed using FLAG-Trap_A beads. Among the candidate interacting proteins, UGMA as the key enzyme for Galf synthesis has been shown to be involved in polarized mycelial growth, and deletion of this gene results in highly branched hyphae (19, 21). However, UGMA has not been reported in plant-pathogenic fungi; therefore, in this study, we focused on UGMA for further functional evaluation in F. graminearum. CHY1 is a nonmammalian protein that is conserved in filamentous fungi and is responsible for specific hyphal polarity growth. UGMA is a noteworthy CHY1′ protein candidate that is also ubiquitous in nonmammalian/plant species and is therefore probably involved in fungus/microbe-specific biological processes. In this study, we verified the interaction between CHY1 and UGMA in a yeast two-hybrid assay. UGMA was not essential in F. graminearum, but the deletion of UGMA caused cell wall defects, with highly branched mycelia and significantly reduced virulence. This study also showed that CHY1 and UGMA probably have opposite effects on cell wall architecture and fungicide tolerance. The ΔugmA mutant was more sensitive to the inhibitor of β-1,3-glucan synthesis caspofungin, suggesting that it may be a potential fungicide target.

RESULTS

UGMA interacts directly with CHY1.

CHY1 encodes a novel zinc finger in F. graminearum, as reported in our previous study (23). To identify CHY1-related proteins involved in polarity growth, we carried out a pulldown analysis. A total of 129 putative CHY1-interacting proteins were finally obtained after filtering out nonspecific protein targets that were also pulled down in the green fluorescent protein (GFP)-only control. To screen reliable interacting proteins, 32 proteins with an average of at least 3 unique peptides and a score of 13 or higher were retained (Table 1). As shown in Table 1, the CHY1-interacting proteins obtained included microtubulins β2Tub, β1Tub, and α1Tub and the Hsp70 heat shock proteins Ssa and Ssb, which have been reported to modulate β2Tub folding in F. graminearum (24). These results further support our previous finding that CHY1 is a microtubule-regulatory protein.

TABLE 1.

Putative CHY1-interacting genes identified by affinity purification

Gene ID	Unique peptides (two repeats)		Annotations	Names in references
	1	2
FGSG_00838	18	28	Hsp70 heat shock protein	FgSsa (24)
FGSG_00921	13	19	Hsp70 heat shock protein	FgSsb (24)
FGSG_06611	12	15	Tubulin	FgTub2/β2Tub (39, 40)
FGSG_07281	7	6	CHY-type zinc finger	FgChy1 (23)
FGSG_07075	10	20	Biotin carboxylase
FGSG_09491	7	13	K homology domain
FGSG_02233	6	12	Armadillo-type fold
FGSG_01248	14	14	Aminoacyl-tRNA synthetase
FGSG_05530	9	13	CDC48 family
FGSG_10940	6	13	Helix hairpin bin domain superfamily
FGSG_00728	8	8
FGSG_04181	8	12	ABC transporter-like
FGSG_01591	7	10	Phosphate carrier protein
FGSG_09530	9	8	Beta tubulin	FgTub1/β1Tub (39, 40)
FGSG_00639	9	8	Alpha tubulin	FgTub4/α1tub (39)
FGSG_09638	13	9	Aspartate carbamoyltransferase
FGSG_16145	4	9	ATP-dependent RNA helicase C
FGSG_05422	5	10	Eukaryotic translation initiation factor
FGSG_00646	5	6	Mitochondrial carrier
FGSG_09266	6	8	Hydroxymethylglutaryl coenzyme A synthase
FGSG_06270	3	8	Uncharacterized protein
FGSG_00278	8	6	Biotin synthase
FGSG_07939	3	6	Coatomer alpha subunit
FGSG_16171	4	6	HMGL-like domain
FGSG_00637	2	6	Vacuolar ATP synthase subunit B
FGSG_05972	2	5	Nucleoside diphosphate kinase
FGSG_09966	3	4	Uncharacterized protein
FGSG_05985	2	7	Uncharacterized protein
FGSG_05997	1	6	UDP-galactopyranose mutase
FGSG_05610	3	6	Uncharacterized protein
FGSG_03238	4	4	Polyketide synthase
FGSG_16213	6	4	AMP-dependent synthetase/ligase

To search for new polarity growth regulatory proteins associated with CHY1, one UDP-galactopyranose mutase, UGMA, drew our attention as a candidate CHY1-related protein (Table 1), even though the number of unique peptides was not very high, because UGMA has been reported to be involved in hyphal polarity branching and to regulate cell wall integrity in A. nidulans (19). To further verify the interaction between CHY1 and UGMA, we generated UGMA bait and CHY1 prey constructs and transformed them into yeast strain AH109. Transformants expressing the prey and bait constructs were able to grow on SD-Trp-Leu-His plates and showed beta-galactosidase (LacZ) activity (Fig. 1), indicating that UGMA may directly interact with CHY1 in F. graminearum.

FIG 1.

CHY1 interacts with UGMA in Fusarium graminearum. The interaction of CHY1 and UGMA was detected using yeast two-hybrid assays. Different concentrations (cells per milliliter) of the yeast transformants expressing CHY1 prey and UGMA bait constructs were assayed for growth on selective plates (SD-Leu-Trp-His) and for β-galactosidase (LacZ) activities.

UGMA is conserved in filamentous fungi.

The ugmA gene was identified in the EnsemblFungi F. graminearum genome database; it contains 1,832 bp and encodes a 516-amino-acid protein with one amino oxidase domain. Sequence alignment and phylogenetic analyses revealed that UGMA homologs are well conserved in filamentous fungi and some yeasts (e.g., Cryptococcus neoformans) but are absent from mammals and plants (Fig. 2A; also, see Fig. S1A in the supplemental material), suggesting that UGMA may be a promising low-toxicity target for fungicide development.

FIG 2.

Phylogenetic analysis of UGMA from filamentous fungi and functional identification of UGMA in growth and hyphal branches. (A) Phylogenetic tree was viewed using Mega6. A neighbor-joining tree with 10,000 bootstrap replicates in different organisms was created. (B) Colony morphology of the wild-type strain PH-1, the ΔugmA mutant, and the complemented strain (ΔugmA::ugmA) grown on PDA and CM at 25°C for 3 days. (C) Hyphal morphology at the edge of PH-1, ΔugmA mutant, and ΔugmA::ugmA strain colonies (top) and 12-h germlings of each strain (bottom).

Deletion of ugmA results in major growth defects.

To investigate the biological function of UGMA, we generated ugmA deletion mutants using the split-marker approach (25). The hygromycin-resistant transformants were confirmed by PCR with four pairs of primers (Fig. S1B). We obtained five transformants with the same phenotype and chose one for further experiments in this study. Compared with the wild-type strain, the growth rate of the ΔugmA mutant was significantly decreased, and the cells formed compact colonies without aerial hyphae on potato dextrose agar (PDA) and complete medium (CM) plates after 3 days of incubation (Fig. 2B; Table 2). The observation of colony edges revealed that the ΔugmA mutant showed hyperbranching, and more severe hyperbranching was observed in the 12-h germlings in liquid yeast extract-peptone-glucose (YEPD) medium (Fig. 2C), indicating that the ΔugmA mutant exhibited extremely serious polarity defects, causing severe vegetative growth inhibition.

TABLE 2.

Vegetative growth, conidiation, and disease indices of Fusarium graminearum strains^a

Strain	Growth rate (mm/day)b		Conidiationc (106 conidia/mL)	Disease indexd
	PDA	CM		Corn silk	Wheat spike
PH-1	10.99 ± 0.31 a	9.81 ± 0.31 a	1.91 ± 0.12 a	6.25 ± 0.26 a	16.0 ± 2.9 a
ΔugmA mutant	1.02 ± 0.15 b	1.05 ± 0.27 b	0.08 ± 0.06 b	0.0 b	0.2 ± 0.2 b
ΔugmA::ugmA mutant	10.86 ± 0.08 a	9.76 ± 0.30 a	1.87 ± 0.16 a	6.18 ± 0.19 a	12.5 ± 3.0 a

^aValues are means and standard deviations calculated with results from three independent replicates. The same letter indicates that there was no significant difference. Different letters mark statistically significant differences (P ≤ 0.05).

^bAverages and standard deviations calculated from at least three independent measurements.

^cConidiation in 5-day-old CMC cultures.

^dDisease was rated by the number of symptomatic spikelets 14 days after inoculation. At least 10 wheat heads were examined in each replicate.

We further generated a complemented ΔugmA::ugmA strain by reintroducing the UGMA-GFP fusion construct into the ΔugmA mutant. Normal vegetative growth, reproduction, and pathogenicity phenotypes were observed in this strain, indicating that the phenotypic defects of the ΔugmA strain were due to the deletion of ugmA (Table 2 and Fig. 2; also see Fig. 4 and 5). GFP labeling was used to observe UGMA localization.

FIG 4.

UGMA is required for asexual and sexual reproduction in F. graminearum. (A) Conidial morphology and septum stained with calcofluor white (CFW). Bar = 25 μm. (B) Conidiation of each stain. Different letters above the bars indicate significant differences. (C) Number of septa of each strain. (D) Carrot agar mating plates of PH-1 and the ΔugmA and ΔugmA::ugmA strains. Arrows and arrowheads point to perithecia and ascospores, respectively. Bar = 20 μm.

FIG 5.

UGMA is essential for infection and virulence in F. graminearum. (A) Subcellular distribution of UGMA in conidia and hyphae. Nuclei and vacuoles are colored with DAPI and CMAC. Bar = 25 μm. (B) Flowering wheat heads were inoculated with conidia of each strain for 14 days postinoculation (dpi). (C) Corn silks were inoculated with blocks of cultures of the same set of strains. Photographs were taken 5 dpi. (D) DON production of each strain assayed with diseased wheat kernels from symptomatic spikelets. Different letters above the bars indicate significant differences.

UGMA is involved in cell wall architecture.

Strains were stained with calcofluor white (CFW) by using the same dye concentrations and imaging conditions for all strains (19). As shown in Fig. 3A, the fluorescence intensity of the ΔugmA mutant was obviously higher and the cell walls were thicker than those of the wild type. Hyphal wall thickness was further examined using transmission electron microscopy (TEM) (Fig. 3B). Cross-sections of growing hyphae of both PH-1 and the ΔugmA mutant showed abundant, well-preserved cytoplasmic organelles, suggesting that the internal structure of the ΔugmA strain cells developed normally. However, the hyphal walls of the ΔugmA mutant were almost 3-fold thicker than those of the wild-type strains, which was consistent with the CFW staining results. Furthermore, the outer layer of the cell wall was almost invisible in the ΔugmA mutant (Fig. 3B).

FIG 3.

Hyphae of PH-1 and ΔugmA mutant strains as shown by light microscopy and TEM. (A) Hyphae of PH-1 and the ΔugmA mutant were stained with calcofluor white and examined by light microscopy under UV irradiation. Bar = 20 μm. (B) TEM of cross-sectioned F. graminearum hyphae. Bars = 200 nm (top) and 500 nm (bottom). Black arrows indicate the outer layer of the cell wall. White arrows mark the thickness of the cell wall.

UGMA regulates asexual and sexual reproduction.

Conidia and ascospores play key roles in disease occurrence and rapid dissemination as primary and secondary infection sources, respectively (26). Therefore, conidiation abilities and sexual reproduction were assessed in this study. The quantification of ΔugmA mutant conidia showed a dramatic reduction (98.6%) compared with conidial growth in the wild-type strain PH-1 (Fig. 4A and B). As shown in Fig. 4C, the conidia of the ΔugmA mutant were shorter, with 0 to 3 septa (100%), than those of the wild type, with 4 to 6 septa (67.4%). In the sexual reproduction assay, PH-1 and the complemented ΔugmA::ugmA strain formed abundant black perithecia on carrot agar plates after fertilization. However, no perithecia were produced in the ΔugmA mutant (Fig. 4D). These results suggest that ugmA is essential for asexual and sexual reproduction.

Localization of UGMA in different developmental stages.

The complementary ΔugmA::ugmA strain was used to study the localization of UGMA in F. graminearum. In conidia, GFP signals were observed in the cytoplasm but not in nuclei; however, unlike the situation in conidia, UGMA-GFP mainly accumulated in the cytoplasm, but not in the vacuole, in the mycelium (Fig. 5A). The results indicated that UGMA is probably transferred among the cytoplasm, nuclei, and vacuoles in different developmental stages. Considering the interaction of CHY1 and UGMA, we further observed the location of UGMA in the chy1 gene deletion mutant by transforming UGMA-GFP into the Δchy1 mutant. The results showed that the localization of UGMA-GFP in Δchy1 was the same as that in the wild type (Fig. S2), indicating that CHY1 is dispensable for the localization of UGMA.

UGMA plays an important role in pathogenicity.

To determine the role of UGMA in virulence, infection assays were carried out by inoculating a conidial suspension of each strain on flowering wheat heads. After 2 weeks of inoculation, PH-1 (disease index, 16.0) and the complemented strain (disease index, 12.5) caused typical blight symptoms and spread to the other spikelets. However, only 25% of ΔugmA mutant-inoculated kernels (disease index, 0.2) showed limited scab symptoms, which occurred only in the palea and lemma and failed to spread to the glume or rachis (Fig. 5B; Table 2). Accordingly, DON was not detected in the ΔugmA mutant-inoculated kernels, whereas the DON concentrations of the wild-type strain and complementation strain were 300 μg g⁻¹ and 273 μg g⁻¹, respectively (Fig. 5D). In infection assays with corn silks, the ΔugmA mutant showed no visible extensions after 7 days of inoculation. Under the same conditions, the discoloration extended more than half the length of the corn silks of PH-1 and ΔugmA::ugmA strains (Fig. 5C). These results indicate that UGMA is essential for F. graminearum infection and extension.

ugmA is responsible for tolerance to caspofungin.

The ΔugmA mutant had thick hyphal walls that lacked the Galf component, suggesting that they might be sensitive to wall-targeting agents. Therefore, we examined the sensitivity of the ΔugmA mutant to different wall- and membrane-inhibiting agents, including caspofungin, polyoxins, and tebuconazole. Caspofungin had a more significant inhibitory effect on the ΔugmA mutant than the wild type after germination on YEPD for 20 h. There were no obvious differences in germination inhibition between PH-1 and the ΔugmA mutant upon polyoxin B and tebuconazole treatment; instead, tebuconazole caused increased branching and expansion in the ΔugmA mutant. More obviously, the ΔugmA mutant almost stopped growing even after prolonged incubation for 30 h or more (data not shown). These results indicated that the ΔugmA mutant was more sensitive to the glucan synthesis inhibitor caspofungin but not to chitin synthesis-targeting polyoxin B or cytomembrane-targeting tebuconazole (Fig. 6).

FIG 6.

The ugmA gene deletion mutant exhibits increased sensitivity to caspofungin. Conidia of each strain were inoculated into liquid YEPD medium for 20 h with or without the wall (membrane)-inhibiting agents. CK, the conidia inoculated in liquid YEPD without the inhibiting agents.

UGMA and CHY1 regulate the cell wall composition in opposite ways.

Considering the physical interaction of CHY1 and UGMA, we further explored their connection to cell wall function. PH-1 and the Δchy1 and ΔugmA strains were treated with cell wall lysis buffer containing the lysing enzymes Driselase and chitinase. After incubation for 1 h at 30°C, the hyphae of the Δchy1 mutant were almost completely digested and released abundant protoplasts, whereas half of the hyphae of PH-1 were released as protoplasts, and almost no protoplasts of the ΔugmA mutant were visible (Fig. 7A). After incubation for 2 h, the hyphae of PH-1 were almost completely digested and released abundant protoplasts; however, only a small fraction of the ΔugmA mutant was digested, and few protoplasts were observed. These results indicate that Δchy1 strain is more sensitive while ΔugmA strain is more tolerant to cell wall-degrading enzymes.

FIG 7.

UGMA and CHY1 reverse regulate cell wall integrity. (A) Mycelia of PH-1, the ΔugmA mutant, and the Δchy1 mutant treated with Driselase, lysozyme, and chitinase at 30°C. Bar = 20 μm. White arrows mark protoplasts of the ΔugmA mutant. (B) PH-1, the ΔugmA mutant, and the Δchy1 mutant were treated with the microtubule inhibitor nocodazole and the cell wall inhibitor caspofungin. Bar = 20 μm.

To further investigate the relationship between CHY1 and UGMA, the Δchy1 mutant was treated with the glucan synthesis inhibitor caspofungin, and the ΔugmA mutant was treated with the microtubule inhibitor nocodazole. Interestingly, the hyphae of the Δchy1 strain were less curved after treatment with caspofungin in YEPD. Analogously, the number of high branches in the ΔugmA strain was reduced when the mutant was treated with nocodazole (Fig. 7B). These results indicated that CHY1 and UGMA are negatively regulated.

DISCUSSION

Polarity growth, the dominant growth form of filamentous fungi, is regulated by a large number of genes that are common in fungi but absent from plants and animals. With the development of green agriculture, research on low-toxicity, environmentally friendly pesticides has attracted increasing attention. Genes that are unique to fungi and absent from mammals and plants offer considerable advantages as low-toxicity, eco-friendly targets for antifungal drug development. CHY1 was previously reported to be conserved in filamentous fungi but absent from plants and animals and to regulate specific polarity tip extension in fungi. In this study, we aimed to explore other CHY1-related genes that regulate fungus-specific polarized tip growth and do not exist in plants and animals and to elucidate their biological functions.

The UDP-galactofuranose mutase (UGMA) was identified as a candidate CHY1 interaction protein by pulldown assays. Yeast two-hybrid assays confirmed the interaction between CHY1 and UGMA. BLASTP searches of the F. graminearum genome with UGMA of A. nidulans showed that the F. graminearum genome contained a single sequence (516 amino acids) with high amino acid sequence identity (79.11%) with UGMA. Multiple-sequence alignment and phylogenetic analyses suggested that UGMA homologs are specific to fungi and bacteria and are not found in plants and animals. Galf, a five-member ring form of galactose in whose synthesis UGMA participates, is also restricted to fungi, bacteria, and protozoa. Accordingly, UGMA or the Galf synthesis pathway is an ideal target for developing antifungal agents that can specifically inhibit fungi with almost no side effects.

To investigate the contribution to virulence, the ugmA gene was deleted from F. graminearum strain PH-1. The ΔugmA mutant was viable, but its growth rate was severely inhibited. Similar to the ΔugmA strain of A. nidulans, F. graminearum ΔugmA produced wide, thick-walled, highly branched hyphae, which was probably the main reason for the limited vegetative growth of ΔugmA. The deletion of the ugmA gene in F. graminearum also resulted in an aberrant conidial morphology, reduced sporulation, sexual reproduction defects and reduced pathogenicity. In addition, UDP-Galp mutase has been identified in other fungal species, such as Cryptococcus neoformans, A. fumigatus, and A. niger, and plays important roles in growth or pathogenicity (18, 19), indicating that UGMA is functionally conserved in different fungi. The Galf synthesis pathway relies on UDP-glucose 4-epimerase (UgeA), UDP-galactofuranose mutase (UGMA/B), and galactofuranosyl (Galf) transferases (UgtA/B) sequentially; the deletion of any of these components in A. nidulans causes increased wall thickness, attenuated virulence, and reduced growth (18, 19, 27), indicating that a lack of Galf-containing sugars affects vegetative growth, cell wall structure, conidiation, virulence, and sexual reproduction in filamentous fungi.

Caspofungin (which blocks wall β-glucan synthesis) is a fungal cell wall inhibitor, and Galf forms the side chains of many glycoconjugates of the cell wall (14, 17). We tested the caspofungin sensitivity of the ΔugmA mutant and the wild-type strain. Consistent with the ΔugmA mutant of A. nidulans, the ΔugmA mutant showed increased sensitivity to caspofungin. In addition, we compared caspofungin with the antifungal drugs polyoxin B (which targets chitin synthesis), calcofluor white (which binds to cellulose and chitin in the cell), and tebuconazole (which targets ergosterol biosynthesis). The results indicated that the ΔugmA mutant did not show any sensitivity or tolerance to these drugs, unlike A. nidulans ΔugmA and ΔugtA mutants, which were significantly more sensitive to calcofluor white than the wild type. Therefore, ugmA in F. graminearum is probably specifically involved in the sensitivity to glucan synthesis inhibitor agents. Galf in fungi probably contributes to β-glucan synthesis, or Galf-containing sugar regulates cell wall architecture in coordination with glucan.

In the present study, the thickened wall of the ΔugmA mutant was more tolerant to stress imposed by a mixture of enzymes including cellulose, chitinase and pectinase. This result also explains why the ΔugmA mutant does not show polyoxin or calcofluor white sensitivity. Considering these results together, we suspect that the content of chitin in the ΔumgA mutant was higher than that in the wild type. The lack of Galf and the increase in chitin are probably the main reasons for the observed abnormality and thickening and the enhanced mechanical strength of the cell wall.

Solid-state nuclear magnetic resonance (NMR) has revealed that galactomannan forms a highly dynamic region with α-1,3-glucan, galactosaminoglycan, glycoprotein, etc., in the outer-layer cell wall (12, 28). In this study, TEM showed that the outermost layer of the ΔugmA mutant disappeared, indicating that the synthesis of Galf is important for the construction of the outermost layer of the cell wall in F. graminearum. These results are consistent with reports that Galf is involved in the composition of galactomannan in A. fumigatus (15, 17).

Both CHY1 (23) and UGMA regulate polarity growth. In pulldown and yeast hybrid assays, CHY1 physically interacted with UGMA, and the localization of CHY1 and UGMA in mycelia was identical, indicating that CHY1 and UGMA probably have a close regulatory relationship in mycelial growth. However, the Δchy1 mutant is characterized by mycelial bending (23), while the ΔugmA mutant is characterized by high branching of mycelia, corresponding to two different types of polarity growth. These results do not clarify how CHY1 and UGMA functionally regulate each other. The following observations shed light on the puzzle from one point of view. Under the same conditions in terms of the weight of mycelia and the concentration of cell wall lysis buffer, the protoplast release rate showed the order Δchy1 mutant > PH-1 > ΔugmA mutant, indicating that cell wall strength was increased in turn. Interestingly, microtubule inhibitor treatment of the ΔugmA mutant attenuated its highly branched morphology, and caspofungin treatment of the Δchy1 mutant reduced the degree of mycelial twisting. Considering the direct interaction between CHY1 and UGMA, it was expected that CHY1 and UGMA would show a negative regulatory relationship in controlling the composition of the cell wall. The presence of CHY1 and UGMA ensures that cell wall thickness and strength are maintained at an intermediate and normal level. Considering that the deletion of CHY1 does not affect the localization of UGMA, the interaction of CHY1 and UGMA probably regulates the dynamic localization or expression of other proteins, thus affecting the cell wall architecture. The cell wall structure and mycelial polarity growth are closely related in filamentous fungi, and CHY1 and UGMA in F. graminearum are mainly responsible for the branched and linear extension of hyphae, respectively. We further speculate that a weakened cell wall may result in the twisted extension of mycelia, and the thickened cell wall is probably the reason for the highly branched hyphae.

In conclusion, we know that neither UGMA nor Galf exists in mammals and higher plants and that the deletion of UGMA inhibits the synthesis of Galf and Galf-containing molecules in the cell wall, resulting in cell wall thickening. The thickened cell wall affects the establishment of initial branching sites of mycelia. The highly branched hyphae also severely limit the radial extension of mycelia and further affect the infection ability and pathogenicity of the fungi (Fig. 8). Changes in the cell wall composition make the ΔugmA mutant more sensitive to the wall-targeting drug carpofungin. Therefore, UGMA or genes involved in Galf synthesis are promising targets for new low-toxicity antifungal agents. Furthermore, the inhibition of fungus-specific biological pathways such as cell wall synthesis and mycelial polarity growth will be an effective way to develop low-toxicity drug targets.

FIG 8.

Proposed functions of UGMA in Fusarium graminearum. UGMA is a unique enzyme that catalyzes the conversion of UDP-Galp to UDP-Galf, which is an important constituent of cell wall polysaccharide and plays an essential role in wall architecture, radial growth, asexual/sexual development, and pathogenicity.

MATERIALS AND METHODS

Strains and growth conditions.

The wild-type strain PH-1 (29) and mutants of F. graminearum generated in this study were cultured on PDA at 25°C. Colony morphology and sexual reproduction were analyzed on CM and carrot agar plates as previously described (30). Conidiation and germination were analyzed in liquid carboxymethyl cellulose (CMC) and YEPD (1% yeast extract, 2% peptone, 2% glucose) cultures, respectively (31). Protoplast preparation and polyethylene glycol (PEG)-mediated transformation were performed as described previously (32). For transformant selection, hygromycin B (MDBio; CAS, 31282-04-9) and Geneticin (Sigma, St. Louis, MO, USA) were added to final concentrations of 300 μg/mL and 400 μg/mL, respectively.

Pulldown assays.

For affinity purification, fresh mycelia harvested from Δchy1::chy1-FLAG transformants were ground and suspended in extraction buffer for protein extraction as previously described (33). Total proteins were mixed with anti-FLAG M2 beads (Sigma), followed by rotation at 25°C for 2 h. After a series of washing steps, the bound proteins were eluted from the beads in accordance with the manufacturer’s instructions (Sigma). Eluted protein samples were fractionated by 10% SDS-PAGE and stained with Coomassie brilliant blue R-250. Bands of interest were excised and sent to LuMing Tech (Shanghai, China) for mass spectrometry with an LTQ Orbitrap Velos Pro spectrometer (Thermo Finnigan). Liquid chromatography-tandem mass spectrometry (LC-MS/MS) data were additionally searched against the annotations of the F. graminearum genome from the EnsemblFungi database (http://fungi.ensembl.org/index.html) by using Mascot 2.3 software. The experiment was conducted with two biological replicates. Nonspecific proteins bound to the anti-FLAG antibody were excluded by the wild-type control.

Generation of the ΔugmA mutant.

ugmA gene replacement constructs were generated by a split-PCR approach as described previously (25). The PCR products of flanking sequences of ugmA fused with hygromycin phosphotransferase fragments were transformed into protoplasts of the wild-type PH-1 strain. The hygromycin-resistant transformants were screened by PCR with four primer pairs (Fig. S1B and C). All the primers used in this study are listed in Table 3. Four deletion mutants of ugmA with consistent phenotypes were obtained. One positive transformant was selected for phenotypic analyses.

TABLE 3.

PCR primers used in this study

Primer	Sequence (5′–3′)	Description or reference
ugmA-1/F	ATGGAGTAAGCGGTGACGA	ugmA deletion
ugmA-1/R	TTGACCTCCACTAGCTCCAGCCAAGCCAGTGGAGCGACTGGAAAGA	ugmA deletion
ugmA-2/F	GAATAGAGTAGATGCCGACCGCGGGTTATTGAGTTGGGCACGTTGT	ugmA deletion
ugmA-2/R	CCTTCCTGTCCATCCCTTT	ugmA deletion
HYG/F	GGCTTGGCTGGAGCTAGTGGAGGTCAA	41
HY/R	GTATTGACCGATTCCTTGCGGTCCGAA	41
YG/F	GATGTAGGAGGGCGTGGATATGTCCT	41
HYG/R	AACCCGCGGTCGGCATCTACTCTATTC	41
ugmA-3/F	CCCGAGGGTTTCGTAAGTT	ΔugmA mutant screen
ugmA-3/R	AGATGGTGGCACGGTAGAA	ΔugmA mutant screen
ugmA-4/F	GGAAATTACAAACAACGCTC	ΔugmA mutant screen
ugmA-4/R	GGATCATCTTGGCCCTGTT	ΔugmA mutant screen
H855R	GCTGATCTGACCAGTTGC	41
H856F	GTCGATGCGACGCAATCGT	41
H852	AACTCACCGCGACGTCTGTC	41
H850	TTGTCCGTCAGGACATTGTT	41
ugmA-CM/F	AGGGAACAAAAGCTGGGTACCGGAAATTACAAACAACGCTC	Δchy1 complementation and localization
ugmA-CM/R	GAACAGCTCCTCGCCCTTGCTCACCTGGGCCTTGGAGTGGCTGG	Δchy1 complementation and localization

ugmA gene complementation and subcellular localization.

For complementation, the ugmA open reading frame (ORF) plus its native promoter were amplified and cloned into pKNTG (34) using a ClonExpress II one-step cloning kit (Vazyme), and the construct was transformed into protoplasts of the ΔugmA mutant (23). G418-resistant transformants expressing ugmA-GFP were analyzed by PCR and GFP signal examination.

Staining and microscopy.

The cell walls of PH-1 and the ΔugmA mutant were stained with 10 μg of CFW per ml for 30 s (23). Nuclei and vacuoles of the complemented ΔugmA::ugmA strain with GFP signals were stained with 20 ng of 49,6-diamidino-2-phenylindole (DAPI) per ml for 10 min (23) and 7-amino-4-chloromethylcoumarin (CMAC, Keygen Biotech) at a final concentration of 10 μM for 15 min as described previously (23). Microscopy examination of the conidia, hyphae, and fluorescence signals was performed with a Nikon E400 microscope.

Yeast two-hybrid assays.

The ORFs of ugmA and chy1 were amplified from F. graminearum cDNA derived from mycelial growth on liquid YEPD medium and then cloned into the pGADT7 prey vector and the pGBKT7 bait vector, respectively. The resulting bait and prey vectors were cotransformed into the AH109 yeast strain as previously described (31). The Leu⁺ and Trp⁺ transformants were isolated and assayed for growth on SD-Trp-Leu-His medium and galactosidase activity as described previously (31). The positive and negative controls were provided in the Matchmaker library construction and screening kit (Clontech).

Transmission electron microscopy.

The mycelium grown in YEPD medium for 12 h was examined by transmission electron microscopy (HT7800; Hitachi, Japan). Samples were prepared as described previously (35). Mycelia were fixed for at least 8 h in 2.5% glutaraldehyde and then cleaned three times with ultrapure water. The cells were postfixed for 2 h in 1% osmium tetroxide, washed with ultrapure water three times, and dehydrated in a graded ethanol series. After transfer to 100% anhydrous acetone and dehydration three times, Epon 812 was replaced in the samples, and they were embedded in pure resin, which was polymerized at 60°C for 2 days. Ultrathin sections were prepared by a Leica EM UC7 ultrathin sectioning method and observed with a Hitachi H-7650 transmission electron microscope with an AMT camera system operating at 80 kV after double staining with phosphotungstate and uranium acetate.

For morphometric analysis, hyphal width was measured by the decussation method, and wall thickness measurements were performed where the cell membrane was crisply in focus.

Infection and DON production assays.

For the plant infection assays, flowering wheat heads of cultivar Annong 8455 were inoculated with suspended spores as previously described (36). The inoculated wheat heads were examined at 14 days postinoculation (dpi), and the disease index was estimated. Corn silks at the early anthesis stage were inoculated with hyphal blocks of each strain for 5 days at 25°C as previously described (23). For mycotoxin production assays, inoculated wheat kernels were collected at 14 days postinoculation. DON was extracted and purified as previously described (37). The amount of DON (per milligram of sample) in each sample was determined using an Agilent 1260 system.

Cell wall lysis and antifungal drug treatment.

The 12-h germlings of the wild-type strain PH-1 and the ΔugmA mutant were treated with a cell wall enzyme mixture (Driselase, lysing enzymes, and chitinase) at equal concentrations at 30°C for 1 h, 2 h, and 3 h (38). Conidia of PH-1 and the ΔugmA mutant were inoculated into liquid YEPD medium for 20 h with (and without) caspofungin, polyoxins, and tebuconazole to examine the effect of the wall or membrane inhibitors on conidial germination.