Fungal strain and culture conditions

The F. graminearum wild-type strain Z-3639 [45] was used in all experiments. The culture media were prepared following the Fusarium laboratory manual [46]. All of the wild-type strain and transformants used in this study were grown in CM agar plates at 25 °C and stored as mycelial suspensions in 20% glycerol at −80 °C. Conidia of all strains were induced either in carboxymethyl cellulose (CMC) medium or on yeast malt agar (YMA). To obtain mycelia, conidial suspensions of the wild-type strains were inoculated in liquid CM (5 × 10⁶ conidia/mL) and mycelia were harvested after 24 h of incubation on a rotary shaker (200 rpm).

Nucleic acid manipulations, Southern blotting, and PCR

For genomic DNA isolation, each strain was cultured in 5 ml of CM for 3 days in a rotary shaker at 200 rpm, and genomic DNA was extracted according to the Fusarium laboratory manual [46]. Restriction endonuclease digestion and agarose gel electrophoresis were performed following standard protocols. [47] Southern blot hybridization was performed with the North2South Biotin Random Prime Labeling Kit and the North2South Chemiluminescent Hybridization and Detection Kit (Thermo Fisher Scientific, Waltham, MA, USA). The PCR primers used in this study (S1 Table) were synthesized by an oligonucleotide synthesis facility (Bioneer, Daejeon, Korea).

Genetic manipulations and fungal transformations

The double-joint (DJ) PCR method [48] was used to generate the fusion PCR products that were required for the gene deletion, complementation, and overexpression. The fungal transformation was performed as previously described [49].

To construct the deletion mutants, the 5′ and 3′ flanking regions of the target genes were amplified from the genomic DNA, and the Geneticin resistance gene cassette (GEN) was amplified from pII99 [50]. Three fragments were fused via the DJ PCR method, and the final constructs were amplified using the nested primers. The resulting amplicons were transformed into the fungal wild-type protoplasts, using PEG-mediated transformation [51]. The mutant was confirmed via Southern hybridization with a flanking region probe.

Vegetative growth, conidiation, and sexual development

Radial growth and colony morphology were assayed on CM and MM 4–5 days after inoculation. For the conidiation assays, a fresh mycelial plug from CM was inoculated in 5 mL of CMC for 5 days on a rotary shaker (200 rpm). The number of conidia was measured with a hemacytometer.

For sexual development, fungal strains were grown on carrot agar for 5 days, and aerial mycelia were removed with 0.4 mL of a sterile 2.5% Tween 60 solution. The plates were then incubated under near-UV light (wavelength: 365 nm; Sankyo Denki, Tokyo, Japan). The number of perithecia, maturation, ascospore morphology, and ascospore discharge were observed after 7–9 days.

For outcrosses, the female strain was fertilized with 1 mL of a conidial suspension from the male strain 5 days after inoculation on carrot agar.

Cellophane membrane penetration assay

To examine the penetration ability, each strain was inoculated on cellophane-overlaid CM. Cellophane membranes were removed after 36 and 48 h post-inoculation, respectively, and the resulting media were observed after removal. All penetration experiments were conducted three times.

Stress response

The stress responses were evaluated by assaying vegetative growth on CM supplemented with various stress agents: oxidative stress (7 mM hydrogen peroxide and 38 μM menadione), osmotic stress (1 M NaCl, 1 M KCl, 1.5 M sorbitol), ionic stress (5 mM FeSO₄), pH stress (pH 4 and pH 11), cell wall stress (400 mg/L Congo red and 100 mg/L sodium dodecyl sulfate), DNA synthesis inhibition (8.6 mg/L iprodione), the inhibition of mitogen activated protein kinase (0.023 mg/L fludioxonil), the inhibition of meiosis and intracellular transportation (0.1 mg/L benomyl), the inhibition of glycan synthesis (0.2 μg/ml caspofungin, 0.02 μg/ml micafungin, 0.03 μg/ml anidulafungin, and 2.5 μg/ml dimethomorph), the ER stress agents (10 mM DTT and 2.5 μg/ml tunicamycin) and azole fungicide (0.1 mg/L tebuconazole) [36].

Virulence test and mycotoxin analysis

The point inoculation method was performed to assay fungal virulence as previously described [49]. Conidia were harvested from CMC cultures, and 10 μL of each suspension (10⁵ conidia/mL) was injected into the center spikelet of a wheat head (cultivar: Eunpamil). The inoculated wheat plants were incubated in a humid chamber for 3 days and grown in a greenhouse for an additional 18 days. The number of diseased spikelets was measured 21 days after inoculation.

The DON and ZEA extraction was performed as previously described [52]. A fresh mycelial plug was inoculated on 1.5 g of rice substrate for 3 weeks. The rice culture was harvested, ground, and mixed vigorously with 6 mL of 84% acetonitrile for 30 min. After phase separation, the upper phase was filtered through a 0.45 μm syringe filter. Reverse-phase HPLC on a Prominence HPLC system (Shimadzu, Kyoto, Japan) with a C18 column was used for the analysis, with a simple modification of the previous methods [53,54]. For ZEA detection, the mobile phase was 70% aqueous methanol, and the flow rate was 1 mL min⁻¹. For DON detection, the mobile phase was 10% aqueous acetonitrile (ACN), and the flow rate was 1 mL min⁻¹. A gradient elution program was applied as follows: after 10% ACN was maintained for 11 min, it was linearly increased to 30% ACN at 12 min. It was then linearly decreased from 30% ACN at 12 min to 10% ACN at 18 min. Subsequently, 10% ACN was held for 17 min for the reequilibration of the column before the injection of the next sample, giving a total run time of 35 min. Diode-array detection was applied (ZEA and DON were detected at a wavelength of 235 nm).

HPLC analysis of N&O-linked glycans from cell wall mannoproteins

In our study, we performed HPLC-based glycan profiling to obtain information on structures of glycans, which are separated based on their physical and chemical properties. The retention times of glycan peaks were compared with those of standard glycans.

cwMPs were isolated from F. graminearum mycelia and conidia as described previously [9,55,56]. N-linked glycans were released from the purified cwMPs using PNGase F (New England Biolabs, Ipswich, MA, USA) and were purified over a Carbograph Extract-Clean (Grace Davison Discovery Science, Bannockburn, IL, USA) column. The N-glycans were labeled with 2-aminobenzoic acid (2-AA; MilliporeSigma, Burlington, MA, USA) and purified using a Cyano Base cartridge (Bond Elut-CN-E; Agilent, Santa Clara, CA, USA) (100 mg) to remove excess 2-AA. 2-AA-labeled oligosaccharides were analyzed with a Waters 2690 HPLC system and a 2475 fluorescence detector with excitation and emission wavelengths of 360 nm and 425 nm, respectively. Data were collected using Empower 2 software (Waters, Milford, MA, USA).

For release of the O-glycans, the completely dried cwMPs (50 μg) were resuspended in 100 μL of hydrazine monohydrate (Tokyo Chemical Industry, Tyoko, Japan), and the mixture was incubated at 60 °C for 4–6 h. The reactants were dried to remove the hydrazine monohydrate, and the pellets were dissolved in 100 μL of saturated NaHCO₃ (Sigma, Burlington, MA, USA), mixed with 10 μL of (CH₃CO)₂O, and incubated on ice for 30 min without shaking. The O-glycans were purified by using Dowex 50WX8–400 resins (H⁺ form; Sigma), and the isolated O-glycans were labeled with 2-aminobenzoic acid (2-AA; Sigma) and purified using a cyano base cartridge (Bond Elut-CN-E; Agilent) (100 mg). The HPLC analysis of the 2-AA-labeled O-glycan was conducted on a TSKgel Amide-80 column (0.46 by 25 cm, 5 μm; Tosoh, Tokyo, Japan) at a flow rate of 1.0 mL/min. 2-AA-Oligosaccharides were detected with a 2475 fluorescence detector (Waters) at excitation and emission wavelengths of 360 and 425 nm, respectively. Data were collected using Empower 2 chromatography data software (Waters).

Microscopic observation

Microscopic observation was performed with a DM6 B microscope (Leica Microsystems, Wetzlar, Germany) equipped with a Leica DMC6200 camera using an L5 fluorescence filter (excitation 480/40 nm; emission 527/30 nm) for GFP, an N2.1 filter (excitation 515–60 nm; emission 590 nm) for RFP, and a DAP filter (excitation 355–425 nm; emission 470 nm) for UV and violet fluorescent.

Western blotting

Conidial suspensions were inoculated in CM at 5 × 10⁶ conidia per milliliter. Mycelia were harvested 24 h after incubation on a rotary shaker (200 rpm) and washed twice with distilled water. For total protein extraction, fresh mycelia were ground in liquid nitrogen and re-suspended in 1 mL extraction buffer containing protease inhibitors. Lysates were sonicated and centrifuged at 13,000 rpm for 20 min. The resulting supernatants were quantified with Pierce 660 nm Protein Assay Reagent (Thermo Scientific,  San Jose, CA) and were used for western blot analysis. To detect the overall protein N-glycosylation level, total proteins were mixed with SDS-PAGE loading buffer, denatured at 95 °C for 10 min, and then subjected to 8% SDS polyacrylamide gels and transferred to nitrocellulose membranes using the Trans-Blot Turbo Transfer System (Bio-Rad, Hercules, CA, USA). Glycoproteins were detected with a Con A-HRP (1:10,000) as an antibody and photographed using a Chemi-Doc imaging system (Bio-Rad). Relative protein abundance was calculated using Image Lab 4.1 software (Bio-Rad) by comparing the adjusted total lane volume of Con A-HRP-detected glycoproteins with that of the corresponding stain-free gel-detected total protein bands.

To analyze N-glycosylation level of each target glycoproteins from wild type and deletion mutant strains, each target glycoproteins were detected using an anti-GFP antibody (1:5,000; Abcam, Cambridge, UK) as the primary antibody. Protein samples were digested with PNGaseF (P0704; New England Biolabs) for 1 h to cleave between the innermost GlcNAc and asparagine residues of high-mannose, hybrid, and complex oligosaccharides. The digested samples were then mixed with the loading buffer. A theoretically smaller protein band of digested target was detected by immunoblot analysis as previously described [20].

Cell wall components analysis

The cell wall isolation and purification were prepared as described previously [57]. Conidia and mycelium, harvested by vacuum filtration and immediately frozen in liquid nitrogen, were ground into a fine powder under liquid nitrogen. This powder was then transferred to a 2.0 mL screw cap tube (Thermo Fisher Scientific) and disrupted with silica beads (BioSpec Product, Bartlesville, OK) using a Fastprep-24 5G machine (MP Biomedicals, Santa Ana, California, USA). This procedure was carried out until complete cell breakage was achieved (room temperature for eight periods of 1 min on ice), as confirmed by examination under a phase-contrast microscope and by the failure of extracts to produce viable colonies when they were plated on CM agar medium. Cell lysates were separated into a cell wall fraction (pellet) and a soluble cytoplasmic fraction (supernatant) by centrifugation at 1000xg and 4 °C for 10 min. To reduce cytoplasmic contaminations, washing procedures were as follows: First, the cell wall fraction was washed three times in 1 × PBS. Next, the cell wall pellet was washed three times each in 5, 2.5, and 1.25% sodium chloride (NaCl) solution. Finally, native cell walls were obtained by further incubating overnight in 0.1 M sodium carbonate (Na₂CO₃) solution and rinsed in 1 × PBS.

For monosaccharide analysis, 100 mg of cell wall material was hydrolyzed under different conditions: in 6 N HCl at 100 °C for 4 h for amino sugar (glucosamine and galactosamine) analysis, in 2 M trifluoroacetic acid at 100 °C for 4 h for neutral sugar (fucose, galactose, glucose, and mannose) analysis, and in 0.1 N HCl at 80 °C for 1 h for sialic acid analysis.

The sialic acid, neutral and amino sugars were separated and quantitated on a CarboPac PA10 column (4.0 x 250 mm, Dionex Co., Sunnyvale, CA, USA) and CarboPac PA1 column (4.5 x 250 mm), with an ICS-5000 with electrochemical detection using a CarboPac PA1 (4.5 x 50 mm) and CarboPac PA10 (4.0 x 50 mm) cartridge. 16 mM NaOH was used as an eluant at a flow rate of 1.0 ml/min. Duplicated data was analyzed using PeakNet on-line software. The analysis was carried out at the Korea Basic Science Institute (KBSI, located in Ochang, South Korea).

Preparation of glycoprotein samples for proteomic analysis

The glycoproteins were obtained as described previously with a slight modification [23,24]. The total proteins were dissolved with 10 mL of Con A binding buffer (20 mM Tris-HCl, pH 7.4, 0.5 M NaCl, 1 mM CaCl₂, 1 mM MnCl₂) and incubated with 1 mL of Con A-Sepharose beads (GE Healthcare, Chicago, IL) in a column for 2 h with slow rotation. The beads were then washed with 10 mL of Con A binding buffer, and glycoproteins were eluted by the addition of 5 mL of 0.5 M methyl-α-d-mannopyranoside and 0.5 M methyl α-glucopyranoside. The whole cell extracts (total proteins) and glycoproteins were quantified with Pierce 660 nm Protein Assay Reagent (Thermo Scientific). After boiling sample, the denatured proteins were separated on SDS polyacrylamide gels and stained with Coomassie Brilliant Blue. Bands of interest were excised, and trypsin digested.

Tandem mass spectrometry analysis

Tryptic peptides were analyzed using a Q Exactive system equipped with EASY-nLC-1000 nano flow liquid chromatography system (Thermo Scientific). Samples were trapped in an Acclaim PepMap 100 C18 column (75 µm × 2 cm, Thermo Scientific) and separated on an EASY-Spray column (75 µm × 15 cm, particle size ≤3 m, pore size 100 Å, PepMap RSLC C18, Thermo Scientific) using a 90 min gradient of solvent A and B (0.1% FA in ACN). The gradient included 5–40% solvent B at a flow rate of 300 nL/min and a column temperature of 35 °C. The eluted peptides were ionized at 2 kV. A full MS scan was performed from 400 to 2000 m/z with a full width a (FWHM) resolution of 70,000, followed by a data-dependent HCD MS/MS scan of the top 10 ions. Parameters used for MS/MS scan were resolution = 17,500, loop count = 10 (Top 10), isolation window = m/z 2.0, normalized collision energy = 27.

Data analysis

The raw MS spectra files were searched against the F. graminearum protein database (47,879 entries downloaded from UniProtKB on Apr. 4^th, 2024) using Proteome Discoverer (v2.4.1.15; Thermo Fisher Scientific). The search settings were as follows: Sequest HT search engine, trypsin (full), maximum two missed cleavage sites, MS tolerance (10 ppm), MS/MS tolerance 0.02 Da, carbamidomethylation (C) for fixed modification, and oxidation (M) for variable modification. Peptide/protein identification and quantitation were conducted on Scaffold (v5.0.1, Proteome Software Inc.) with a SEQUEST XCorr threshold of at least 1.8 for charge 1, 2.5 for charge 2, and 3.5 for charges 3 and 4. Total spectrum count was used for semiquantitative analysis. For normalization, the endogenous housekeeping protein GAPDH was used as an internal control. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository under the accession number PXD063761.

Bioinformatic analysis

Gene Ontology (GO) analysis was performed using the REVIGO in FungiDB [58] database and ClueGO [59] software tool. The Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://geneontology.org/) was used to annotate the pathways enriched in each protein group. These results were visualized using the R program.

Glycoproteins that were differentially regulated by Alg3 (Fg26583) and Alg12 (Fg03053) were selected to establish the PPI network. PPI network was visualized using Cytoscape software (version 3.10.0) performed using the psych R-package with standard settings. The principal component analysis (PCA) was performed using the psych R-package with standard settings. Spearman’s rank correlation analysis was performed using the psych R package, with significance determined at P < 0.05 (two-tailed).

Identification of protein glycosylation-related genes in F. graminearum

To identify genes potentially involved in protein glycosylation in F. graminearum, we employed a multi-step approach. Initially, we retrieved genes from three Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway databases (map00510, N-Glycan biosynthesis; map00513, Various types of N-glycan biosynthesis; map00515, Mannose type O-glycan biosynthesis; map00514, Other types of O-glycan biosynthesis). Additionally, leveraging the well-established protein glycosylation pathways in S. cerevisiae and S. pombe, we supplemented our search with protein glycosylation-related genes from the Saccharomyces Genome Database (SGD) [60] and PomBase [61]. Utilizing BLASTp, we compared the sequences of S. cerevisiae and S. pombe genes to identify orthologs in the genomes of various fungal species, including F. graminearum. Through these analyses, we identified 65 putative protein glycosylation-involved genes in F. graminearum (Fig 1a, b and S2 Table). The genes were further classified based on both their function and the type of oligosaccharides they transfer.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Comparative analysis of protein glycosylation-related genes among fungal species.

(a) Pie chart indicating the classification and distribution of putative protein glycosylation-related genes in the fungal species, F. graminearum, S. cerevisiae, S. pombe, C. neoformans, N. crassa, M. oryzae, and A. nidulans. The genes were classified based on both their function and the type of oligosaccharides they transfer. (b) Dot plot shows the homologues of genes related to protein glycosylation across the different fungal species. The number of genes in each species is represented by the areas of the circles. The identity of protein sequence values in each species were represented based on S. cerevisiae and S. pombe. Genes are categorized and represented with distinct colors based on their respective functions, employing the same strategies as illustrated in Fig 1a.

https://doi.org/10.1371/journal.ppat.1013325.g001

Our investigation yielded several noteworthy observations. First, the proportion of classified genes varied among species (Fig 1a). Mannosidases were notably more abundant in filamentous fungi than in yeast-form fungi, whereas glucosidases were less represented in S. cerevisiae and S. pombe compared to other fungal species. Additionally, mannosyltransferases were significantly more prevalent in S. cerevisiae compared to other fungal species. Interestingly, the overall ratio of classified genes was similar in filamentous fungi, whereas it differed considerably in yeast-form fungi. In the case of C. neoformans, its overall ratio was more similar to that of filamentous fungi, except for glucosidases. Second, our analysis revealed the presence of certain glycosyltransferases that are species-specific (Fig 1b). For example, galactosyltransferase only existed in M. oryzae, N. crassa, A. nidulans, and S. pombe but not in other fungi; β-1,4-mannosyl-glycoprotein 4-β-N acetylglucosaminyltransferase (MGAT3) was found only in filamentous fungi; the number of genes encoding M-pol I (VAN1 and MNN9) and M-pol II (MNN9, ANP1, MNN10, MNN11, and OCH1) subunits were varied across species. Third, the overall protein sequence similarity (E-value) was comparable across species. These findings highlight the diversity and specificity of protein glycosylation-related genes across fungal species, suggesting potential targets for future comparative functional studies.

Construction of the F. graminearum mutant library involved in protein glycosylation

To delve into the biological functions of the 65 genes involved in protein glycosylation in F. graminearum, we systematically constructed deletion mutants for each of the N- and O-glycosylation-related genes, followed by comprehensive functional characterization. With the exception of SEC59 (Fg00930), which was generated in a previous study [44], we generated gene replacement constructs for all other glycosylation-related genes using the split-marker approach. These constructs were then transformed into protoplasts of the F. graminearum wild-type strain Z-3639. Through homologous recombination, we successfully deleted 52 out of 65 putative glycosylation-related genes (S3 Table, Fig A in S1 Text).

For the remaining 13 putative glycosylation-related genes (S3 Table), despite screening over dozens of transformants from at least three independent transformation experiments, we were unable to identify knockout mutants, indicating that the deletion of these genes might be lethal. Notably, ten of these genes have orthologs that are essential in both S. cerevisiae and S. pombe, indicating their critical functions across fungal species. The remaining five genes showed species-specific lethality. Intriguingly, ten genes exhibited lethality exclusively in S. cerevisiae and S. pombe, while they remained non-lethal in F. graminearum. Among the ten genes, only three were found to have paralogs in filamentous fungi (S2 Table). For example, in the case of CWH41/MOGS (Fg17395), two paralogous genes (MGG_04045 and MGG_09721) were identified in M. oryzae. Notably, MGG_04045 has been previously characterized as GLS1, which was shown to be non-essential [24]. For PSA1 (Fg17687), two homologous copies are present even within F. graminearum itself (including Fg10789). Similarly, M. oryzae possesses two PSA1 homologs (MGG_01288 and MGG_05936), and A. nidulans has two (AN1911 and AN5586). N. crassa is unique among the species examined in containing only a single PSA ortholog (NCU06003). Lastly, for STT3 (Fg01117), two copies were identified in M. oryzae (MGG_02821 and MGG_04773). These observations suggest that the presence of paralogs in filamentous fungi may contribute to functional compensation, potentially explaining why some of these genes appear non-essential in F. graminearum. Among the remaining seven genes, ALG2 (Fg04001), RFT1 (Fg06329), and ALG11 (Fg20645) exhibited severe growth defects upon deletion. These genes may thus be considered as quasi-essential or fitness-critical, as their essentiality differs across species, but they likely play key roles in cellular viability and growth. These findings imply the presence of functional divergence in certain glycosylation genes among different fungal species or the presence of functionally redundant glycosylation genes within F. graminearum.

Profiling the collective phenotypic traits of the F. graminearum mutant library

Given the variation in the presence and proportion of protein glycosylation-related genes among fungal species, we compared our mutant phenotypic data, particularly focusing on growth and virulence, with relevant studies on glycosylation in other species [15,17,62–65] (S3 Table and Fig B in S1 Text). While studies on protein glycosylation-related genes in fungal species like M. oryzae, G. graminicola, and C. neoformans are less abundant compared to well-studied model organisms such as S. cerevisiae, S. pombe, and C. albicans, our comparative analysis revealed both unique and conserved functions in regulating growth and virulence across fungal species.

To comprehensively elucidate the biological functions of protein N- and O-glycosylation processes in F. graminearum, we functionally characterized all resulting gene deletion mutants for defects in virulence, vegetative growth, mycotoxin production, as well as asexual and sexual development. The phenotypic dataset of the mutant collection was organized based on the sequential order of glycan synthesis and visually represented using a color scale (Figs 2, 3, and S3 Table). This systematic phenotypic analysis revealed that approximately 87.6% of the deletion mutants (57 out of 65) exhibited at least one discernible phenotypic change compared to the wild-type strain.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Profiling overall protein N-glycosylation pathway genes in F. graminearum.

(a) Schematic diagram of the protein N-glycosylation pathway in F. graminearum. The sequences of all genes participating in N-glycan biosynthesis from the budding yeast S. cerevisiae, S. pombe and Candida albicans were used for BLASTp searches to identify their orthologs in the genome of F. graminearum. (b) Phenotypic traits of genes involved in protein N-glycosylation at the ER (endoplasmic reticulum) and Golgi (golgi apparatus) were examined. virulence, vegetative growth, mycotoxin, and asexual development were categorized as continuous phenotypic traits and scored on a 11-point scale (e: essential, -4: reduced rate between 0 and 0.29, -3: reduced rate between 0.3 and 0.49, -2: reduced rate between 0.5 and 0.69, -1: reduced rate between 0.7 and 0.89, 0: no difference compared with wild-type, + 1: enhanced rate between 1.10 and 0.29, + 2: enhanced rate between 1.30 and 1.49, + 3: enhanced rate between 1.50 and 1.69, +4: enhanced rate between 1.70 and 1.89, +5: enhanced rate over 1.90). Sexual developments were categorized as discrete phenotypic traits and scored on a 4-point scale (a: no difference compared with wild-type, b: decreased number of perithecia and ascospore or delayed perithecia maturation or decreased number of discharged ascospores, c: no perithecia and ascospore formation or no ascospore discharge, e: essential genes). More than three biologically independent experiments were performed for each phenotypic trait. Abbreviations: WH wheat head, CM complete media, MM minimal media, DON deoxynivalenol, ZEA zearalenone, CO conidia, PE perithecia, DI discharge, ACO ascospore.

https://doi.org/10.1371/journal.ppat.1013325.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Profiling overall protein O-glycosylation pathway genes in F. graminearum.

(a) Schematic diagram of the protein O-glycosylation pathway in F. graminearum. The sequences of all genes participating in O-glycan biosynthesis from the budding yeast S. cerevisiae, S. pombe and Candida albicans were used for BLASTp searches to identify their orthologs in the genome of F. graminearum. (b) Phenotypic traits of genes involved in protein O-glycosylation at the ER (endoplasmic reticulum) and Golgi (golgi apparatus) were examined. virulence, vegetative growth, mycotoxin, and asexual development were categorized as continuous phenotypic traits and scored on a 11-point scale (e: essential, -4: reduced rate between 0 and 0.29, -3: reduced rate between 0.3 and 0.49, -2: reduced rate between 0.5 and 0.69, -1: reduced rate between 0.7 and 0.89, 0: no difference compared with wild-type, + 1: enhanced rate between 1.10 and 0.29, + 2: enhanced rate between 1.30 and 1.49, + 3: enhanced rate between 1.50 and 1.69, +4: enhanced rate between 1.70 and 1.89, +5: enhanced rate over 1.90). Sexual developments were categorized as discrete phenotypic traits and scored on a 4-point scale (a: no difference compared with wild-type, b: decreased number of perithecia and ascospore or delayed perithecia maturation or decreased number of discharged ascospores, c: no perithecia and ascospore formation or no ascospore discharge, e: essential genes). More than three biologically independent experiments were performed for each phenotypic trait. Abbreviations: WH wheat head, CM complete media, MM minimal media, DON deoxynivalenol, ZEA zearalenone, CO conidia, PE perithecia, DI discharge, ACO ascospore.

https://doi.org/10.1371/journal.ppat.1013325.g003

The severity of phenotypic alterations increased notably with the deletion of mutants involved in early glycan biosynthesis. Furthermore, the indispensability of key genes was confirmed, including GDP-mannose pyrophosphorylase (PSA1) (Fg10789), responsible for synthesizing GDP-mannose from GTP and mannose-1-phosphate; dolichol phosphate mannose (Dol-P-Man) synthase (DPM) complex subunits (Fg24245, Fg09545, Fg14327), catalyzing the formation of Dol-P-Man from Dol-P and GDP-Man; the oligosaccharyltransferase (OST, STT) complex subunits (Fg20575, Fg01277, Fg23937), responsible for catalyzing asparagine-linked glycosylation of newly synthesized proteins; and the protein O-mannosyltransferase (PMT) complex subunits (Fg03745, Fg01543), which transfers mannose from dolichyl phosphate-D-mannose to serine and threonine residues of target proteins. These findings underscore the profound impact of glycan structure on various biological functions in F. graminearum.

To ascertain correlations among various phenotypes, in silico N-/O-glycosylation pathway of F. graminearum, including the structure of glycan, were constructed, based on the putative glycosylation-related genes (Figs 2a and 3a). Spearman’s rank correlation was computed for glycan mutants exhibiting multiple mutant phenotypes (Fig C in S1 Text). The analysis revealed robust correlations among virulence, vegetative growth, and both asexual and sexual reproduction. Similarly, glycan maturation in the O-glycosylation pathway also showed a strong correlation with virulence, vegetative growth, and sexual reproduction, whereas a few genes associated with the N-glycosylation pathway exhibited only a moderate correlation. Noteworthy findings include the identification of OST (Fg01117), MNS1 (Fg01555), MNL2 (Fg01661), and HOC1 (Fg06423) genes, which play distinct roles in virulence independent of vegetative growth. Additionally, DON and ZEA exhibited no significant correlation with both N- and O-glycosylation pathways. We found that specific truncated glycan structures are closely associated with sexual development, including those: (1) N-glycans present before mannosidase-mediated trimming, (2) N-glycans lacking sequential extension by Mnn2/5, Mnn14, Mnn4, Mnn1, and Mgat1, and (3) N-/O-glycans not yet modified by Mnn1 and Mnt2–4. These results indicate that the early phase of glycan mgaturation could affect more diverse phenotypes, including virulence, vegetative growth, and sexual reproduction, and could also have a greater impact on phenotypic severity compared to the late phase of glycan maturation.

As glycan maturation exhibited a discernible correlation with various phenotypic traits except mycotoxin production and asexual development, we analyzed the transcript levels of genes involved in protein glycosylation across these stages (S4 Table and Fig D in S1 Text). The results showed that the genes encoding Ost subunits, which are responsible for transferring Glc₃Man₉GlcNAc₂ to an Asn residue in newly synthesized proteins, were upregulated during vegetative growth and the overall virulence stage. However, these same genes were either downregulated or exhibited no significant difference during sexual and asexual developmental stages.

Additionally, distinct glycosyltransferase genes exhibited stage-specific expression patterns. ALG5 (Fg19779), MNT2 (Fg09595), and GLS1 (Fg17395) were elevated in their transcript levels during vegetative growth, while ALG13 (Fg06927), ALG14 (Fg04029), ALG12 (Fg03053), MNN10 (Fg23425), and MNN14 (Fg27773) were upregulated during sexual developmental stages. In the asexual developmental stage, the transcript level of MNL2 (Fg04573) showed specific elevation at 12 h. Notably, in the virulence stage, PSA1 (Fg10789), ANP1 (Fg14905), and MNN11 (Fg15225) were specifically upregulated. Regarding O-glycosylation, PMT complex genes, transferring mannose residues from dolichyl phosphate-D-mannose to protein serine/threonine residues at the ER, displayed downregulation during sexual development but upregulation during the virulence stage. These findings imply that distinct or varied glycan structures may be produced at each developmental stage.

Glycan maturation does correlate with the various stress responses

We examined the responses of deletion mutants under 20 distinct in vitro stress conditions, including oxidative stress, osmotic stress, ionic stress, pH stress, cell wall stress, MAPK inhibition condition, DNA synthesis inhibition condition, meiosis and transport inhibition condition, cell wall glycan synthesis inhibition condition, ER stress, and several fungicide treatment conditions (Figs 4 and E in S1 Text, and S5 Table). The stress response dataset of mutants involved in protein glycosylation was quantitatively represented using a color scale. Apart from essential genes, our analysis revealed that approximately 76.9% of the deletion mutants (40/52) exhibited at least one discernible phenotypic defect. Interestingly, the majority of stress responses demonstrated resistance (72.8%) rather than susceptibility (27.2%). Nevertheless, we did not observe any significant correlation between glycan maturation and stress responses (Fig F in S1 Text).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Characterizing in vitro phenotypic traits of genes involved in protein glycosylation.

In vitro phenotypic traits were assessed across 20 different growth conditions. The growth score was determined by normalizing the relative growth rates (compared to complete media) to the wild-type and correcting them based on the maximum absolute value observed under each stress condition. The color code is as follows: the light grey indicates essential genes, red indicates reduced or susceptible genes, and blue indicates enhanced or tolerant genes. Each phenotypic trait was evaluated in more than three biologically independent experiments. Abbreviations: H2O2 CM + 7 mM H2O2, Menadione CM + 38 μM Menadione, NaCl CM + 1 M NaCl, KCl CM + 1 M KCl, Sorbitol KCl CM + 1.5 M Sorbitol, FeSO4 CM + 5 mM FeSO4, SDS CM + 100 mg/L sodium dodecyl sulfate, Congo Red CM + 400 mg/L Congo Red, Fludioxonil CM + 0.023 mg/L Fludioxonil, Iprodione CM + 8.6 mg/L Iprodione, Benomyl CM + 0.1 mg/L Benomyl, Dimethomorph CM + 2.5 μg/ml Dimethomorph, Caspofungin CM + 0.2 μg/ml Caspofungin, Anidulafungin CM + 0.03 μg/ml Anidulafungin, Micafungin CM + 0.02 μg/ml Micafungin, DTT CM + 10 mM dithiothreitol, Tunicamycin CM + 2.5 μg/ml Tunicamycin, Tebuconazole CM + 0.1 mg/L Tebuconazole.

https://doi.org/10.1371/journal.ppat.1013325.g004

The composition of cell wall components and glycan structures showed notable differences between conidia and mycelia

A significant enrichment of glycoproteins was observed in the cell membrane and extracellular regions, including the cell wall [66,67]. In a recent study, we highlighted the distinct regulation of cell wall proteins across developmental stages [57]. Given the variability in gene expression patterns of the protein glycosylation pathway across developmental stages (S4 Table and Fig D in S1 Text), we explored whether the glycan structures of cell wall glycoproteins also exhibited variations across these stages. Due to the intricate internal structures of perithecia, representing the sexual reproduction stage, sampling from this stage was not feasible. Furthermore, samples were not collected from F. graminearum-infected wheat tissue, representing the virulent stage, due to challenges in separating invasive hyphae from host tissue. Consequently, our investigation focused on mycelia and conidia, representing vegetative growth and asexual development, respectively.

To elucidate the monosaccharide composition of glycan structures, we conducted an analysis of cell wall components (Fig 5a and S6 Table). This analysis unveiled higher proportions of glucosamine, the chitin precursor, in mycelia compared to conidia. Conversely, the proportion of glucose, the glucan precursor, was higher in conidia than in mycelia. No significant differences were observed in the proportions of mannose and galactose between mycelia and conidia, and fucose and galactosamine were not detected. We additionally analyzed the monosaccharide composition of cell wall glycoproteins to determine whether galactose is present in the cell wall itself or in the glycan structures attached to cell wall glycoproteins. This analysis revealed that galactose was not detected in the cell wall glycoproteins, implying that it was only present in the cell wall (Fig 5c and S6 Table).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. The proportion of cell wall components and glycan structures differed between conidia and mycelia.

(a) Monosaccharides composition of cell walls between conidia and mycelia. Fucose and galactosamine were not detected. (b) Different glycoprotein levels in mycelia and conidia were assessed by western blot using the Con A-HRP antibody. MY, mycelia; CO, conidia. The intensity of each lane, as observed after lectin blotting, was normalized by the intensity of the respective sample from the stain-free gel. (c) Schematic diagram illustrating the differences in cell wall components, glycoproteins, and glycan structures between mycelia and conidia. (d-e) HPLC profiles of N-glycans of cell wall mannoproteins (cwMPs) from the mycelia and conidia of the wild-type strain Z-3639 of F. graminearum. Normal-phase HPLC was conducted using a Shodex Asahipak NH2P-50 4E column (Showa Denko) (0.46 by 25 cm) with solvent A (2% glacial acetic acid–1% tetrahydrofuran–acetonitrile) and solvent B (5% glacial acetic acid–3% triethylamine–1% tetrahydrofuran–HPLC water) at a flow rate of 1.0 ml/min. After sample injection, the proportion of solvent B was maintained at 30% for 5 min and then increased linearly to 70% over 80 min. + α-1-2,3 MNS, N-glycans treated with α-1-2,3 MNS. (f-g) O-glycans of cwMPs from the mycelia and conidia of the wild-type strain Z-3639 of F. graminearum. Normal-phase HPLC was conducted using a TSKgel Amide-80 column (Tosoh Bioscience) (0.46 by 25 cm) with solvent A (2% glacial acetic acid–1% tetrahydrofuran–acetonitrile) and solvent B (5% glacial acetic acid–3% triethylamine–1% tetrahydrofuran–HPLC water) at a flow rate of 1.0 ml/min. After sample injection, the proportion of solvent B was maintained at 10% for 5 min and then increased linearly to 90% over 60 min. + α-1-2,3 MNS, O-glycans of treated with α-1-2,3 MNS.

https://doi.org/10.1371/journal.ppat.1013325.g005

We further conducted a comparative analysis of glycosylation levels in the two different stages. Lectin blotting using concanavalin A (Con A) revealed no difference in the ratio of cell wall glycoprotein (Fig 5b). However, more glycoproteins with high molecular weight were detected in conidia than in mycelia. These results highlight not only the differences in the proportion of chitin and glucan matrix between mycelia and conidia but also the higher abundance of glycoproteins with high molecular weight in conidia compared to mycelia (Fig 5c).

Based on the analysis of cell wall components, we anticipate that the glycan structure of cell wall glycoproteins would mainly consist of mannose, galactose, glucose, and glucosamine. To gain insight into the general structure of N- and O-glycans in F. graminearum, we conducted high-performance liquid chromatography (HPLC)-based glycan analysis on both mycelia and conidia. The glycans assembled on cell wall mannoproteins (cwMPs) from the wild-type (WT) strain Z-3639 of F. graminearum were obtained, labeled with 2-aminobenzoic acid, and analyzed by Normal-phase HPLC. The results revealed distinctive N-glycan profiles between mycelia and conidia, ranging from approximately 8–20 monosaccharides in size. In mycelial profile, the peaks corresponding to the N-glycans containing eight (M8, GlcNAc₂Man₈) or more mannose residues were distributed from 40 min to 55 min range, whereas in conidia N-glycan profile, the peaks larger than M8 were predominantly detected within the 40–50 min range. The treatment with α-1,2- and α-1,3-mannosidase (MNS), specifically cleaving α-1,2- and α-1,3-mannose linkages, led to an overall reduction in glycans observed between 40 and 48 min, converting most glycan peaks to a smaller peak corresponding to four mannose residues (M4) in both mycelia and conidia. In contrast, the peaks observed after 51 min in the mycelia were not significantly affected by the enzyme treatment (Fig 5d-e). The results indicated that the N-glycans of F. graminearum observed between 40 and 48 min are mostly composed of mannose residues in α-1–2 and 1–3 linkages. However, the remaining peaks suggest the presence of another mannose linkage, such as α-1, 6, or the addition of other monosaccharide besides mannose.

The O-glycan profiles of F. graminearum revealed that O-glycans are relatively small, typically consisting of one to six monosaccharide residues (Fig 5f-g). Two distinct peaks were detected as the glycan peak composed of two mannose residues (M2 and M2*), suggesting the presence of two isomer forms in different linkages. As observed in the N-glycan profiles, the overall O-glycan peak patterns of mycelia and conidia displayed a few noticeable differences. Besides the differential proportion of M2 isomers, mycelia appear to contain less extended O-glycans (M3 ~ M6) compared to conidia. In contrast to N-glycans, the O-glycan peaks exhibited only partial alterations upon α-1–2, -3 MNS treatment, indicating that the terminal mannose linkages of F. graminearum O-glycan might be α-1,6 rather than α-1,2 or α-1,3, or suggesting that other monosaccharide besides mannose might be present as terminal sugar.

Next, we investigated whether the changes in glycan abundance resulted from alterations in the transcript levels of corresponding biosynthetic enzymes. Notably, analysis of previously published transcriptome data revealed higher expression levels of most N-mannosyltransferase-encoding genes at the mycelial stage compared to conidia (Fig D in S1 Text), consistent with the more abundant presence of hypermannosylated -type of N-glycans in mycelia (Fig 5d-e). Also, the expression levels of most O-mannosyltransferase-encoding genes were downregulated in mycelia compared to conidia, consistent with the O-glycna profiles exhibiting more extended O-glycans in conidia. This correlation between the transcriptome data with glycan profile data strongly indicated that the abundance of glycans with different structures is mainly determined by the expression levels of glycan processing genes at transcription.

ALG3 (Fg26583) and ALG12 (Fg03053) are implicated in various developmental processes and N-glycan biosynthesis

While most glycosyltransferases involved in the initial N-glycosylation at the ER are encoded by a single gene for each glycan synthesis reaction, the late N-glycosylation at the Golgi and O-glycosylation involve multiple glycosyltransferases. This implies that these complex subunit genes are genetically redundant. Based on a previous study in C. neoformans, where deletion of ALG3, ALG9, or ALG12—encoding dolichyl-phosphate-mannose (Dol-P-Man)-dependent mannosyltransferases involved in ER luminal lipid-linked N-glycan biosynthesis—led to the accumulation of truncated core-N-glycans [15], we aimed to investigate the roles of the corresponding genes in F. graminearum. As shown in Figs 2a and 6c, Alg3 (Fg26583), Alg9 (Fg17447), and Alg12 (Fg03053) function sequentially in the N-glycan biosynthesis pathway: Alg3 initiates the addition of an α-1,3-mannose to Dol-PP-GlcNAc₂Man₅, followed by the addition of α-1,2- and α-1,6-mannoses by Alg9 and Alg12, respectively. Notably, Alg9 has a dual role—adding both the initial α-1,2-mannose and capping the α-1,6-mannose branch with another α-1,2-mannose—placing it both upstream and downstream in the pathway. Deletion of ALG3 and ALG12, but not ALG9, led to significant phenotypic defects. This is likely due to the critical positioning of their enzymatic steps in core N-glycan assembly: Alg3 acts early to initiate α-1,3 mannose branching, while Alg12 performs the penultimate α-1,6-mannose addition required for completing the core structure. In contrast, the terminal α-1,2-mannose addition catalyzed by Alg9 may be partially dispensable or compensated by redundant mechanisms via additional α-1,2-mannosyltransferases in the Golgi such as MNN2/5, resulting in milder phenotypes upon deletion. Therefore, our functional analysis was focused on the genes involved in the biosynthesis of core N-glycans in the ER, particularly by examining the roles of ALG3 (Fg26583) and ALG12 (Fg03053), as their deletion resulted in various phenotypic defects including virulence (Fig 2 and S3 Table).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. F. graminearum Fg03053 (Alg12) and Fg26583 (Alg3) are involved in N-glycan biosynthesis within the ER lumen.

(a) HPLC profiles of N-glycans of cell wall mannoproteins (cwMPs) from the mycelia and conidia of the wild-type, Δfg03053, and Δfg26583. Normal-phase HPLC was conducted using a Shodex Asahipak NH2P-50 4E column (Showa Denko). + JBM, N-glycans treated with α- jack bean α-1,2,3,6 MNS. (b) Partial schematic diagram illustrating the difference in final glycan structures between the wild-type and each mutant strain Δfg03053, and Δfg26583. (c) Partial schematic diagram illustrating the difference in final N-glycan structures between the wild-type and each mutant strain Δfg03053, and Δfg26583. (d) Subcellular localization of Fg03053 (Alg12) and Fg26583 (Alg3). Fg03053 and Fg26583 were fused with green fluorescent protein (GFP) and Hmr1, an endoplasmic reticulum (ER) marker, was tagged with red fluorescent protein (RFP). The yellow color in the merged images indicates co-localization. Scale bar = 10 μm.

https://doi.org/10.1371/journal.ppat.1013325.g006

Alg12 (Fg03053) encodes 552 amino acids and contains the Dol-P-Man:Man₇GlcNAc₂-PP-Dol α-1,6-mannosyltransferase domain (IPR039485), while Alg3 (Fg26583) encodes 432 amino acids and contains the glycosyltransferase, Alg3 domain (IPR007873). Both proteins contain 11 and 5 transmembrane domains, respectively, and are predicted to be localized in the ER. We further examined the degree of sequence and domain conservation between F. graminearum and the other fungal species (S3 Table). Although the protein sequences of yeast-form fungi exhibited lower conservation compared to filamentous fungi, the sequence identities in all species and their functional domains were retained throughout evolution. Given this level of sequence conservation, we assume that Alg12 (Fg03053) and Alg3 (Fg26583) function as α-1,6-mannosyltransferase and Dolichol-P-Man-dependent α-1,3-mannosyltransferase, respectively.

HPLC profiles were conducted on N-glycans obtained from cwMPs of mycelia and conidia samples for each mutant strain and compared to that of the WT glycans (Fig 6a, b). Compared to the WT glycans, the N-linked oligosaccharides, specifically the M8 and hypermannosylated peaks (larger than M8), were markedly reduced in both mutants. Conversely, the M3 to M6 peaks (GlcNAc₂Man_3–5) and M6-M7 peaks (GlcNAc₂Man_6–7) were increased in the ALG3 (Fg26583) and ALG12 (Fg03053) deletion mutant strains, respectively, compared to that of the WT strain. The overall N-glycan profiles of mycelia and conidia became quite similar upon deletion of ALG3 (Fg26583) and ALG12 (Fg03053), due to the almost absence of outer chain biosynthesis causing differential degree of hypermannosylation. Treatment with jack bean MNS, cleaving α-1,2,3,6-linked mannose residues, led to the shift of the N-glycan peaks to M1 (GlcNAc₂Man₁) in AGL3 deletion or to M1 ~ M3 in the AGL11 deletion mutant strain. These results suggest that ALG3 (fg26583) deletion strain is defective in the conversion of Dol-PP-GlcNAc₂Man₅ to Dol-PP-GlcNAc₂Man₆, which constitutes the initial step in attaching a mannose residue to the lipid-linked N-oligosaccharide in the ER lumen, and ALG12 (fg03053) deletion strain is defective in the conversion of Dol-PP-GlcNAc₂Man₇ to Dol-PP-GlcNAc₂Man₈ (Fig 6c).

Because glycans core N-glycans are synthesized in the ER, we generated transgenic strains expressing Alg12 or Alg3 fused with green fluorescent protein (GFP) under the control of a constitutive promoter RP27 [68], alongside the ER marker FgHmr1 fused with red fluorescent protein (RFP) [69]. Both Alg12-GFP and Alg3-GFP were co-localized with the Hmr1-RFP, indicating that the subcellular localization of Alg12 and Alg3 is in the ER, consistent with their orthologs in S. cerevisiae (Fig 6c).

Deletion of ALG3 (Fg26583) and ALG12 (Fg03053) affect the glycosylation pattern of Cnb1, Gic1, and Glx, which are essential for the virulence of F. graminearum

N-glycosylation plays pivotal roles in glycoprotein modification and the ER-mediated protein quality control (ERQC) system [24,26]. Lectin blotting using Con A revealed that the total intensities of bands detecting glycoproteins of alg3 (fg26583) and alg12 (fg03053) were reduced by more than 2-fold compared to the wild type in both mycelia and conidia (Fig 7a and b). This finding suggests that the loss of ALG3 (Fg26583) and ALG12 (Fg03053) are not only involved in glycan synthesis but also in modulating the total amount of glycoprotein. This might be probably due to decreased stability of glycoproteins with truncated glycans or inefficient transfer of truncated core-N-glycans to proteins via Ost, resulting overall reduction of glycosylation.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Fg03053 and Fg26583 are involved in the N-glycosylation of glycoproteins Cnb1, Gic1, and Glx, which are required for the virulence of F. graminearum.

(a,b) Lectin blotting analysis of mycelia and conidia stages. Gycoprotein levels of the wild-type, Δfg03053, and Δfg26583 were assessed by western blot using the Con A-HRP antibody. The intensity of each lane in the lectin blot was normalized by the intensity of the respective sample from the stain-free gel. Mean and standard deviation of the intensity of each strain were estimated with data from three (n = 3) independent biological replicates (marked with white dots on the bars). Different letters indicate significant differences based on analysis of variance (ANOVA) followed by Bonferroni test (P = 0.01). (c-e) Western blot analysis using an anti-GFP antibody for detection of under-glycosylated cell wall proteins Cnb1, Gic1, and Glx in Δfg03053, and Δfg26583. SP, signal peptide domain; WSC, cell wall integrity and stress response component domain; branch symbol, putative N-glycosylation site; circle, putative O-glycosylation site. N- or O-glycosylation sites were predicted using NetNGlyc 1.0 and NetOGlyc 4.0. (f-h) Subcellular localization of Cnb1, Gic1, and Glx. Mycelia were grown in liquid complete media for 24 h. Calcofluor white (CFW) staining visualizes cell walls and septa. GFP and CFW fluorescence images were superimposed in the ‘Merge’ panel. Scale bar = 10 μm.

https://doi.org/10.1371/journal.ppat.1013325.g007

To investigate whether ALG3 (Fg26583) and ALG12 (Fg03053) are involved in the glycosylation of their target glycoproteins, we focused on the glycosylation patterns of three glycoproteins identified from our previous cell wall glycoproteome data [57], known for their involvement in the virulence of F. graminearum: glyoxal oxidase (Glx), which catalyzes hydrogen peroxide production [70], and the two phosphatases (Cnb1 and Gic1) [38,70]. Cnb1, Gic1, and Glx were predicted to harbor 4, 6, and 7 glycosylation sites, respectively, and signal peptides at the N terminus. Immunoblot analysis revealed that the GFP-tagged Cnb1, Gic1, and Glx in alg3 (fg26583) and alg12 (fg03053) mutant strains displayed lower molecular weights compared to those in the wild-type strain (Fig 7c-e). Upon PNGase F treatment, which cleaves N-linked oligosaccharides from glycoproteins, the protein sizes converged to the same molecular weight in both mutant strains and the wild-type strain. This indicates that truncated N-glycans with reduced size were attached to Cnb1, Gic1, and Glx proteins in the alg3 (fg26583) and alg12 (fg03053) mutant strains.

Additionally, we confirmed the subcellular localization of Cnb1, Gic1, and Glx under a fluorescence microscope with Calcofluor White staining, showing that all three proteins were primarily localized in the cell wall (Fig 7f-h). This finding suggests that Cnb1, Gic1, and Glx are cell wall glycoproteins modified with the attachment of N-glycans, which are synthesized by Alg3 (Fg26583) and Alg12 (Fg03053)-mediated early N-glycosylation pathway in F. graminearum.

Quantitative glycoproteomics profiling and functional classification of identified glycoproteins in F. graminearum

To elucidate downstream target glycoproteins, whose stability might be affected by truncation of N-glycans due to the loss of Alg3 (Fg26583) and Alg12 (Fg03053), we employed a label-free quantitative glycoproteomic approach [71]. To determine whether the difference in peptide counts between the WT strain and the mutants originated from the glycoprotein degradation in the deletion mutants or arose from altered glycoprotein binding affinity to Con A due to abnormal glycan formation in the deletion mutants, we utilized the global proteome as a control and loaded an equal amount of protein for each sample analysis (Fig G in S1 Text). If a protein was enriched more than two-fold in the Con A-eluted sample compared to the global proteome, we defined it as a glycoprotein. We also included proteins orthologous to or overlapping with the N-glycoproteome of G. graminicola, M. oryzae, or F. graminearum in previous studies [23,24,26]. Following detection and identification via liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis, we identified 406 glycoproteins among 1159 total proteins, representing 8.2% of the total proteins in F. graminearum, with high confidence (S7 Table).

The equal spectral counts of the housekeeping protein glyceraldehyde 3-phosphate dehydrogenase (GAPDH) indicate that all protein samples were analyzed in equal amounts. Principal component analysis (PCA) revealed that the global proteome and glycoproteome from each strain differed in spectral counts patterns, implying that glycoproteome was well-extracted in this experiment (Fig H in S1 Text). Of the 406 glycoproteins identified, 274 had been previously identified in prior N-glycoproteomic studies (S7 Table). These results not only support the efficient extraction of N-glycoproteins but also suggest that the remaining 132 proteins could be considered to have only O-glycans.

To predict the role of glycoproteins in F. graminearum, we conducted Gene Ontology (GO) functional classification and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses. We found that most glycoproteins were localized to extracellular regions, cell wall, plasma membrane and ER. (Fig I in S1 Text). These results align with the understanding that glycoproteins primarily function in the ER, membrane-bounded organelles, and secretory system. The largest proportion of glycoproteins were involved in cell wall organization or biogenesis, carbohydrate metabolism, protein folding and proteolysis. For the molecular functions, the majority of glycoproteins were involved in protein folding and peptidases, supporting the key roles of protein glycosylation associated with protein quality control and amino acid cycling. Proteins with predicted structural molecule activity and pyrophosphatase activity were also glycosylated

KEGG pathway enrichment analysis revealed that glycoproteins are most enriched in the biosynthesis of secondary metabolites, protein processing in the ER, and carbon metabolisms including glycolysis/gluconeogenesis and starch and sucrose metabolisms. Protein processing in the ER, which is closely associated with N-glycosylation as shown in previous studies, plays a crucial role in protein quality control and transport. We identified 17 glycoproteins involved in protein processing in the ER, including Nef and Bip for protein recognition by luminal chaperones; Ost, Gls1, Gls2, Cnx, Uggt, Ergic53, Ero1, Os9, and Pdis for protein quality control in the ER; and Hsp70, Hsp40, Skp1, and P97 for ER-associated protein degradation. Thus, protein glycosylation might play essential roles in protein processing and quality control in the ER by affecting the function, stability, or localization of proteins involved in the ERQC pathway.

Taken together, these results suggest that the glycoproteins of F. graminearum are linked to several important biological processes including protein quality control, amino acid cycling, and carbon metabolism, which are considered key components of pathobiological processes and are primarily localized at the cell wall and extracellular regions.

Glycosyltransferases involved in the early core N-glycosylation pathway influence global protein glycosylation across various biological processes

Pearson correlation analysis revealed no significant correlation between the global proteome and glycoproteome profiles, indicating that differential spectral counts observed in the ∆alg3 and ∆alg12 mutants were not due to general changes in protein expression. Instead, they likely reflect impaired glycoprotein enrichment efficiency, resulting from defective glycan structures and altered binding affinity to Con A (Fig J in S1 Text).

Comparing the spectral counts of Con A-eluted samples between wild-type and mutant strains, we identified 295 (72.6%) and 249 (61.3%) differentially glycosylated proteins in alg3 (fg26583) and alg12 (fg03053) mutants, respectively. Absence of Alg3 (Fg26583) was found to affect more glycoproteins than Alg12 (Fg03053), consistent with the Con A-blotting results showing fewer glycoproteins in the alg3 (fg26583) sample.

Gene Ontology annotation of all identified glycoproteins from total global proteomics revealed their potential roles in biological processes, molecular functions, and cellular components (Fig 8a). Most glycoproteins were predicted to localize to the ER, membrane, and extracellular regions. Notably, glycoproteins localized to the vacuole were exclusively identified in alg3 (fg26583). These glycoproteins were linked to polysaccharide metabolic processes, cell wall organization, and protein quality control (e.g., response to unfolded protein and proteolysis) in both alg3 (fg26583) and alg12 (fg03053). Additionally, proteins unique to alg3 (fg26583) were involved in glycosylation (protein glycosylation and glycoprotein metabolic processes) and the cellular response to chemical stimuli. In molecular function classification, most glycosylated proteins were predicted to have hydrolase activity in both alg3 (fg26583) and alg12 (fg03053). Additionally, glycoproteins with predicted mannosyltransferase activity, unfolded protein binding, and pyrophosphatase activity were exclusively identified in alg3 (fg26583).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Glycoproteomic analysis of wild type.

Δfg03053 (alg12), and Δfg26583 (alg3). The dot plot illustrates enriched Gene Ontology (GO) terms related to biological processes for (a) identified glycoproteins and (b) downregulated proteins in the total proteome of the wild type, Δfg03053, and Δfg26583 strains. The size of the dot represents the percentage of the gene count, and the color indicates the FDR, false discovery rates. (c) Protein-protein interactions (PPIs) of glycoproteins associated with phenotypic defects affected by Fg03053 and Fg26583. 43 glycoproteins were visualized by the Cystoscope program. The functions of the proteins were categorized by shape, and the phenotypes were represented using different colors. Additionally, proteins that might be affected solely by Fg03053 or Fg26583 were marked with blue or red borders, respectively.

https://doi.org/10.1371/journal.ppat.1013325.g008

We also identified 193 (16.7%) and 223 (19.2%) proteins that were downregulated in the alg3 (fg26583) and alg12 (fg03053) mutants, respectively. In contrast to glycoproteins, GO functional classification analysis revealed that most global proteins were localized to intracellular organelles and the cytoplasm (Fig 8b). Regarding biological processes and molecular functions, proteins associated with oxidative stress were predominantly downregulated in alg12 (fg03053), while those involved in protein quality control were primarily downregulated in alg3 (fg26583). Collectively, our proteomic analysis results highlight that the structure of core N-glycans assembled in the ER is crucial for modulating the steady-state levels of glycoproteins involved in various biological processes, primarily localized to extracellular regions, as well as proteins associated with diverse intracellular processes.

Identification of virulence-related glycoproteins affected by early core N-glycosylation in F. graminearum

Among the glycoproteins affected by the loss of Alg3 (Fg26583) or Alg12 (Fg03053), we identified 60 glycoproteins previously characterized in functional studies [36–38,40,44,70, 72–99]. Of these, 24 were associated with virulence, 10 were essential, and 9 were linked to phenotypes not directly related to virulence, including vegetative growth, sexual development, mycotoxin production, and stress sensitivity (Fig 8c and S7 Table). These glycoproteins participate in diverse cellular processes and include transcription factors (Mads001 and Hex1), phosphatases (Cna1, Cnb1, Pp2a, Gic1, Ptc5, Pho8, Psr2, Fg09475, Fg03986, and Fg07304), kinases (Ste7 and Cmk1), peroxidases (Fpx1, Fca6, and Fca7), 70-kDa heat shock proteins (Kar, Ssc, and Sse), lipases (Lip28 and Tag), chitin synthases (Chs5 and Chs7), proteins involved in autophagy (Rab11, Vps74, Atg27), or microtubule function (Tua2, Tub2 and Cap1) (S7 Table).

A protein-protein interaction network analysis revealed that 23 glycoproteins were co-regulated by Alg3 (Fg26583) and Alg12 (Fg03053), while 14 were specifically regulated by Alg3 (Fg26583), and 7 were exclusively regulated by Alg12 (Fg03053). Altogether, these findings suggest that the functional defects of Alg3 (Fg26583) and Alg12 (Fg03053), key players in the early core N-glycosylation pathway, affect a wide range of glycoproteins by generating altered core N-glycan structures, modulating the amount and function of glycoprotein and ultimately impacting virulence in F. graminearum.