A functional atlas of secondary metabolite biosynthetic gene clusters governing growth, stress adaptation, and pathogenicity in Fusarium graminearum
Hao Qi, Liwen Zhao, Luona Xu, Chao Liu, Haolan Cheng, Xingmin Han, Yiyi Ren, Chenghui Xu, Jiayue Yan, Chao Jiang, Bin Ma, Zhonghua Ma, Yun Chen
TL;DR
This study explores how secondary metabolite gene clusters in Fusarium graminearum affect growth, stress adaptation, and pathogenicity, revealing their essential roles beyond just virulence.
Contribution
The study provides a genome-scale functional analysis of 53 SM-BGCs in Fusarium graminearum, revealing their roles in growth, stress adaptation, and pathogenicity.
Findings
Secondary metabolite biosynthetic gene clusters are essential for growth, development, and stress adaptation in Fusarium graminearum.
Two previously uncharacterized gene clusters, BGC36 and BGC47, are critical for virulence and DON production.
SM-BGCs show spatiotemporal regulation during infection, indicating an ecological role in pathogenesis.
Abstract
Filamentous fungi harbor a vast potential for secondary metabolite (SM) biosynthesis, yet the biological functions of numerous biosynthetic gene clusters (BGCs) remain obscure. In Fusarium graminearum, a devastating cereal pathogen, SMs are best known as virulence factors, but their broader contributions to fungal physiology are poorly defined. Here, we present a genome-scale functional dissection of 53 predicted SM-BGCs by constructing a knockout library targeting cluster backbone genes and systematically quantifying 24 phenotypic traits, generating 1,272 phenotypic measurements. This dataset reveals that secondary metabolism is not a dispensable metabolic burden; instead, SM-BGCs are broadly integrated into vegetative growth, asexual development, and abiotic stress adaptation. Transcriptome analyses further uncover pronounced spatiotemporal regulation and tissue-dependent requirements…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —National Key R&D Program of China
- —Pioneer and Leading Goose R&D Program of Zhejiang
- —China Agriculture Research System
- —http://dx.doi.org/10.13039/501100012476Fundamental Research Funds for Central Universities of the Central South University
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMycotoxins in Agriculture and Food · Fungal and yeast genetics research · Plant Pathogens and Fungal Diseases
Introduction
Filamentous fungi are prolific producers of structurally diverse, bioactive secondary metabolites (SMs) [1, 2]. These low-molecular-weight compounds, although generally not required for primary metabolism, often confer major fitness advantages by mediating ecological competition and adaptation to specific niches. The genes for SM biosynthesis are typically organized into biosynthetic gene clusters (BGCs) [3]. A typical cluster contains at least one core synthase, such as polyketide synthase (PKS), non-ribosomal peptide synthase (NRPS), or terpene synthase (TPS), along with a series of auxiliary genes responsible for structural diversification, transport, and regulation. With advances in genome sequencing, it has become clear that most genomes harbor a far greater number of BGCs than previously recognized and that a substantial fraction are weakly expressed or remain “silent” under standard laboratory conditions [4]. BGC activity is regulated at multiple levels, including pathway-specific transcription factors, global regulatory mechanisms (e.g., the LaeA/VeA/VelB complex) [5, 6], and epigenetic modifications (e.g., histone acetylation and methylation) [7–11]. Environmental cues, such as light, nutrient status, and inter-species interactions, are relayed through conserved signaling pathways (e.g., MAPK and cAMP-PKA cascades) [12] to reshape transcriptional and epigenetic states, thereby activating or repressing cluster expression. Functionally, fungal SMs exhibit diverse biological activities, serving as sources for invaluable pharmaceuticals such as antibiotics (e.g., penicillin) [13], immunosuppressants (e.g., cyclosporine A) [14], anti-cancer agents (e.g., taxol) [15], and cholesterol-lowering drugs (e.g., statins) [16]. In plant-associated fungi, SMs also frequently act as phytotoxins or virulence factors in host–pathogen interactions. Consequently, defining the repertoire of fungal BGCs, elucidating their regulatory logic, and assigning biological functions remain central goals with broad implications for medicine and agriculture.
The genus Fusarium, widely distributed in soils and plant-associated habitats, represents a particularly important and diverse lineage for fungal secondary metabolism. Numerous distinct SMs have been identified from Fusarium species. These metabolites serve diverse functions, acting as virulence determinants during plant infection, mediating microbial competition, and enhancing survival under environmental stress [17, 18]. For example, in F. oxysporum, the phytotoxin fusaric acid contributes to wilt symptoms across diverse hosts and is implicated as a virulence factor [19]. Some Fusarium species produce bikaverin, which inhibits competing fungi and can confer a fitness advantage in complex microbial communities [20, 21]. In addition, pigments such as those produced by F. fujikuroi protect the fungus from UV-induced damage [22]. Notably, many Fusarium SMs are mycotoxins that contaminate agricultural commodities worldwide, threatening food and feed safety and, consequently, animal and human health. Accordingly, Fusarium-derived mycotoxins are detected at high frequencies in global surveys. Deoxynivalenol (DON), T-2 toxin, fumonisins (FUM), and zearalenone (ZEA) are among the most prevalent examples [23, 24]. Together, these roles, especially the dual capacity of Fusarium SMs to contribute to virulence while also posing mycotoxin risks, highlight their importance for both crop production and public health.
F. graminearum is the causal agent of Fusarium head blight (FHB) in cereals, leading to yield losses and grain contamination by mycotoxins, most prominently DON [25]. Genomic analyses predict ~ 50-60 secondary metabolite biosynthetic gene clusters (SM-BGCs) in F. graminearum, highlighting its capacity to produce a wide range of SMs. To date, only a minority of these SMs have been chemically characterized. Characterized metabolites include major mycotoxins such as trichothecenes (e.g., DON, nivalenol) and zearalenone, as well as metabolites implicated in diverse biological processes, including aurofusarin (an insect antifeedant) [26], neurosporaxanthin and rubrofusarin (overwintering and ascospore production) [27], gramillins A and B (corn leaf infection) [28], and fusaoctaxin A (intercellular hyphal growth during wheat coleoptile infection). However, the biological functions of many predicted BGCs remain unknown. Even when cluster products have been identified (e.g., PKS15 for FDDP, PKS14 for orcinol, and NPS6 for triacetylfusarin) [29, 30], their physiological roles are frequently unclear. Moreover, systematic functional evidence linking F. graminearum SM-BGCs to core biological processes, such as hyphal development, sporulation, conidial germination, pathogenicity, and abiotic stress tolerance, remains limited. This knowledge gap limits a comprehensive understanding of F. graminearum biology and constrains the development of improved strategies for FHB management and mycotoxin mitigation.
Here, we address these gaps by (i) constructing a genome-scale knockout resource targeting backbone genes from 53 predicted SM-BGCs; (ii) systematically quantifying their contributions to vegetative growth, asexual development, abiotic stress tolerance, and pathogenicity; (iii) prioritizing clusters with reproducible effects on virulence; and (iv) characterizing key virulence-associated clusters.
Results
Genome-wide identification and comparative analysis of SM-BGCs in F. graminearum
To delineate SM-BGCs in F. graminearum, we first mined the PH-1 reference genome using antiSMASH, which predicted 49 candidate clusters. To improve annotation accuracy and completeness, we subsequently re-annotated PH-1 with a focus on identifying key backbone enzymes of secondary metabolism, including PKSs, NRPSs, TPSs, and cyclodipeptide synthases (CDPSs). By integrating these refined annotations with previously reported SM-BGCs in F. graminearum, we compiled a final catalog of 53 canonical SM-BGCs in PH-1. NRPS-type clusters were the most prevalent (n = 23), followed by TPS (n = 13) and PKS (n = 11) clusters. We also identified five hybrid clusters harboring more than one class of backbone enzymes and one CDPS cluster (Fig. 1a-b). Notably, products from 15 of the 53 clusters have been previously characterized, with chemical structures reported (Table S1).Fig. 1. Comparative genomic analysis of secondary metabolite biosynthetic gene clusters (SM-BGCs) across Fusarium species. a Phylogenetic distribution and abundance of SM-BGCs. A maximum likelihood phylogenetic tree was constructed based on 4,492 single-copy orthologous genes from 19 plant-pathogenic Fusarium species. Bubbles indicate the number of BGCs classified by core biosynthetic enzyme class: PKS, NRPS, TPS, Hybrid, and others (e.g., CDPS). Bubble sizes are scaled within categories. b Classification of SM-BGCs in F. graminearum PH-1. The 53 canonical SM-BGCs identified in strain PH-1 are classified according to their backbone enzyme type: NRPS (n = 23), TPS (n = 13), PKS (n = 11), Hybrid (n = 5), and CDPS (n = 1). c Conservation patterns of F. graminearum SM-BGCs across Fusarium species. Visualization of the presence and conservation of the 53 F. graminearum BGCs across the 19 Fusarium species based on the sequence similarity of key backbone enzymes. Abbreviations: BGCs, biosynthetic gene clusters; CDPS, cyclodipeptide synthases; NRPS, non-ribosomal peptide synthetase; PKS, polyketide synthase; SM-BGCs, secondary metabolite biosynthetic gene clusters; TPS, terpene synthase
We next compared the distributions of SM-BGCs across F. graminearum and 18 additional plant-pathogenic Fusarium species. On average, the analyzed species encoded 47 SM-BGCs per genome, and we observed no significant correlation between genome size and SM-BGC number (Pearson r = 0.16, * P* = 0.506). For example, F. solani, despite having the largest genome among the tested species, harbored only 44 SM-BGCs, below the cross-species average. This decoupling is consistent with the evolutionary dynamics of SM-BGCs within the Fusarium genus (i.e., lineage-specific patterns of retention, gain and loss of clusters, and in some cases, duplication and divergence) (Fig. 1a). To examine how the conservation of SM-BGCs varies with phylogenetic distance, we assessed the presence and sequence similarity of backbone enzymes corresponding to each of the 53 PH-1 clusters across the 19 Fusarium species panel. Conservation was strongly associated with phylogenetic proximity: closely related species retained highly similar clusters, whereas more distant lineages exhibited pronounced presence-absence variation for specific SM-BGCs. For instance, the terpene-associated cluster BGC50 was detected only in a subset of plant pathogens closely related to F. graminearum. In contrast, twelve clusters (BGC06, BGC09, BGC11, BGC24, BGC32, BGC39, BGC40, BGC46, BGC47, BGC48, BGC52, and BGC53) were comparatively conserved across all 19 species, with backbone enzyme sequences showing > 70% identity (Fig. 1c). This conservation pattern is consistent with vertical inheritance and suggests that these SM-BGCs have been retained due to functional constraints. One illustrative example is BGC53, which encodes the biosynthesis of pigments; these pigments contributes to ascosporic pigmentation and are important for perithecium formation during sexual development in Fusarium.
A genome-wide SM-BGC knockout library reveals condition-dependent impacts on vegetative growth
Following the identification of 53 SM-BGCs in PH-1, we asked to what extent these BGCs contribute to core physiological traits. We constructed a genome-wide SM-BGC knockout library by individually deleting the key backbone enzyme-encoding genes for each cluster (PKS, NRPS, TPS, CDPS, or hybrid) via homologous recombination. For each targeted gene, at least three independent deletion transformants were obtained, PCR-verified (Figure S1), and used for subsequent phenotyping, providing a robust resource for functional analysis.
Given that secondary metabolism can interface with primary physiology in a context-dependent manner, we quantified radial hyphal expansion of all mutants across six media, representing rich, complex, and minimal conditions: potato dextrose agar (PDA), complete medium (CM), yeast malt agar medium (YMA), carrot agar medium (CA), wheat head tissue medium (WA), and minimal medium (MM). Overall, 24 of the 53 mutants showed significantly altered colony growth relative to the wild-type in at least one medium (Fig. 2a and Table S2), indicating that disruption of a substantial fraction of SM-BGCs impacts vegetative growth. Notably, growth effects were strongly medium-dependent, consistent with pronounced genotype-by-environment interactions. For instance, ΔFg03747, ΔFg04694, ΔFg17487, ΔFg10933, and ΔFg06631 consistently formed smaller colonies on PDA, CA, and YMA (Fig. 2b-c). In contrast, these same mutants grew comparably or indistinguishably from the wild-type on WA and MM (Fig. 2a and Table S2). Together, these results show that a substantial subset of SM-BGCs influence vegetative growth in F. graminearum, with effects that depend on nutritional context.Fig. 2. Secondary metabolite biosynthetic gene clusters (SM-BGCs) influence vegetative growth of F. graminearum in a medium-dependent manner. a Medium-dependent vegetative growth phenotypes of SM-BGC knockout mutants. Radial colony diameters of all 53 SM-BGC backbone gene deletion mutants were measured and normalized relative to the wild-type strain (WT, PH-1) after 3 days at 25 °C on six distinct media: PDA, CA, YMA, CM, WA, and MM. The heatmap indicates the difference in colony diameter (mutant mean/WT mean). Red signifies enhanced growth; blue signifies reduced growth. b, c Representative examples of altered hyphal growth on selected media. Panel (b) shows colony morphology, and panel (c) presents quantified colony diameters (means ± SD) for WT and five selected SM-BGC mutants (ΔFg03747, ΔFg04694, ΔFg17487, ΔFg10933, and ΔFg06631) on PDA, CA, YMA and MM. Asterisks within panels (a) and (c) denote statistical significance relative to the wild-type (*P < 0.05, **P < 0.01, ***P < 0.001; one-way ANOVA with Dunnett's multiple comparisons test, n = 3 biological replicates). Abbreviations: CA, carrot agar medium; CM, complete medium; MM, minimal medium; PDA, potato dextrose agar; SM-BGCs, secondary metabolite biosynthetic gene clusters; WA, wheat head tissue medium; YMA, yeast malt agar medium
SM-BGCs modulate conidiogenesis, conidial morphogenesis and germination
Analysis of publicly available transcriptome datasets revealed pronounced stage-specific expression dynamics of SM-BGC backbone genes during conidiation and conidial germination (Fig. 3a). During conidiation, 8 genes were significantly upregulated, and 6 were significantly downregulated (adjusted P < 0.05 and |log_2_ fold change|≥ 1). During conidial germination (24 h), 6 genes were upregulated, and 4 were downregulated (≥ twofold) (Fig. 3a and Table S3). These distinct expression patterns suggest that a subset of SM-BGCs may be engaged during these developmental transitions. To investigate this, we quantified the conidial production of all 53 backbone-gene deletion mutants in mung bean medium. Five mutants exhibited significant alterations in conidial production compared to the wild-type (Fig. 3b). Conidial yields for ΔFg07798 and ΔFg03066 decreased by > 85%, whereas ΔFg03747 and ΔFg09381 produced no detectable conidia under our assay conditions. Importantly, although several mutants exhibited mild vegetative growth phenotypes (Fig. 2a and Table S2), the severity of the conidiation defects far exceeded the extent of growth impairment, suggesting that reduced conidiation is unlikely to be a secondary consequence of impaired vegetative growth. We next examined conidial morphogenesis by quantifying conidial length and septation. Relative to the wild-type strain PH-1 (average 63–80 μm), 12 mutants produced significantly shorter conidia; for example, ΔFg06631 conidia averaged 33.50 μm. In contrast, ΔFg17745 produced significantly longer conidia (Fig. 3c). Septation was also altered: ΔFg06631 conidia contained ~ 4 septa on average, fewer than PH-1 (~ 5 septa), whereas ΔFg16873 conidia averaged nearly 6 septa (Fig. 3d). Conidial germination assays further revealed significant delays in ΔFg06631 and ΔFg10464. Under conditions where the wild-type reached ~ 90% germination, both mutants remained below 10% germination (Fig. 3e and Figure S2). Collectively, these data indicate that specific SM-BGCs in F. graminearum are involved not only in regulating conidial abundance and morphogenesis but also in influencing conidial germination.Fig. 3SM-BGCs influence conidiation, conidial morphology, and conidial germination. a Transcriptional dynamics of 53 SM-BGC backbone genes during conidiation and germination. The heatmap shows the log_2_ fold change relative to 24 h mycelia cultured in YEPD; time points are expressed as hours post-inoculation. “Con” denotes the conidiation stage, and “Germination” denotes the conidial germination stage, non-significant genes are shown in gray. b Conidiation of SM-BGC deletion mutants in mung bean medium, quantified at 5 days post-inoculation. n = 3 independent biological replicates. c Conidial length distributions and mean length for the wild-type strain (PH-1) and 53 SM-BGC mutants (n = 100 conidia per strain). d Proportions of conidia with the indicated numbers of septa for the wild-type and mutant strains. e Percentage of germinated conidia for each strain at 6 h after inoculation in 1/2 YEPD medium. Data in (b, c, e) are presented as the means ± SD. Asterisks indicate significant differences compared to the wild-type (*P < 0.05, **P < 0.01, ***P < 0.001; one-way ANOVA with Dunnett's multiple comparisons test). Abbreviations: SM-BGC, secondary metabolite biosynthetic gene cluster; YEPD, yeast extract peptone dextrose
SM-BGCs diversely mediate abiotic stress tolerance and fungicide responses
Fungal SM-BGCs are often minimally expressed during vegetative growth but can be elicited by environmental cues. As such induction is typically interpreted as a downstream consequence of stress signaling, we reasoned that stress-responsive SM programs may also contribute to cellular homeostasis during environmental fluctuations. To systematically evaluate this possibility, we quantified the radial growth of all 53 backbone-gene knockout mutants together with the wild-type strain across 15 distinct stress conditions, generating 810 strain-by-condition phenotypic measurements. The panel encompassed ionic stress (Ca^2+^, Mg^2+^), osmotic stress (NaCl, KCl), oxidative stress (menadione, H_2_O_2_), cell wall-perturbing agents (Congo red, SDS), pH stress (4.0, 11.0), temperature stress (20 °C, 30 °C), and fungicide challenges (carbendazim, tebuconazole, phenamacril) (Fig. 4a; Table S4).Fig. 4. Abiotic stress and fungicide responses are modulated by SM-BGCs in F. graminearum. a Phenomic screen of 53 SM-BGC knockout mutants and the growth of the wild-type strain PH-1 under 15 abiotic stress conditions resulted in 810 phenotypic measurements (54 strains × 15 conditions). The heatmap shows the mean mycelial growth inhibition from ≥ 3 biological replicates. Statistical significance was assessed by one-way ANOVA with Dunnett's post-hoc test relative to the wild-type, with *P < 0.05, **P < 0.01, ***P < 0.001. Stress Conditions: Ionic: 0.5 M MgCl_2_, 0.5 M CaCl_2_ (on PDA). Osmotic: 1 M NaCl, 1 M KCl (on PDA). Oxidative: 0.5 mg/mL menadione (VK3), 0.05% H_2_O_2_ (on PDA). Cell Wall: 0.02% Congo red, 0.1% SDS (on MM). pH: pH 4.0, pH 11.0 (on PDA). Temperature: 20 °C, 30 °C (on PDA). Fungicides: 0.5 μg/mL carbendazim, 0.3 μg/mL tebuconazole, 0.3 μg/mL phenamacril (on MM). b Comparative mycelial growth inhibition for WT and five selected mutants (ΔFg17745, ΔFg10097, ΔFg17487, ΔFg10933, and ΔFg16873) under CaCl_2_ and MgCl_2_ stress. c Comparative mycelial growth inhibition for WT and five selected mutants (ΔFg01790, ΔFg02324, ΔFg03340, ΔFg03066, and ΔFg16873) under KCl and NaCl stress. d Comparative mycelial growth inhibition for WT and ΔFg03747 under oxidative stress induced by vitamin K3 or H_2_O_2_. e Comparative mycelial growth for WT, ΔFg07798, and ΔFg17168 at 20 °C and 30 °C. f Comparative mycelial growth inhibition for WT, ΔFg15872, ΔFg08795, ΔFg08378, ΔFg10933, and ΔFg06507 in the presence of carbendazim, tebuconazole, and phenamacril. Abbreviations: CR, Congo red; Car, carbendazim; MM, minimal medium; PDA, potato dextrose agar; Phe, phenamacril; SM-BGC, secondary metabolite biosynthetic gene cluster; Teb, tebuconazole; VK3, vitamin K3; WT, wild-type
Overall, 17 of the 53 mutants (32.1%) displayed significantly altered sensitivity to at least one stressor compared to the wild-type, indicating a broad functional coupling between secondary metabolism and abiotic stress adaptation. Notably, ΔFg10933 and ΔFg16873 showed pleiotropic sensitivity to three distinct stress factors. Under Ca^2+^ stress (0.5 M CaCl_2_ in PDA), the wild-type strain PH-1 growth was inhibited by ~ 60%, whereas five mutants (ΔFg17745, ΔFg10097, ΔFg17487, ΔFg10933, and ΔFg16873) exhibited significantly enhanced tolerance, each showing less than 30% inhibition (Fig. 4b), suggesting a functional role for these SM-BGCs in calcium-associated homeostasis. In osmotic assays, ΔFg01790, ΔFg02324, ΔFg03066, and ΔFg16873 displayed hypersensitivity to NaCl and/or KCl, whereas ΔFg03340 displayed enhanced resistance specifically to NaCl (Fig. 4c). Oxidative stress responses were largely comparable to the wild-type across the mutant library, with the notable exception of ΔFg03747, the core-gene mutant of a siderophore-associated cluster, which was specifically hypersensitive to vitamin K3 (Fig. 4d), consistent with the established role of siderophore systems in mitigating oxidative damage by modulating redox-active metal availability. In addition, several SM-BGCs influenced temperature-associated growth, with ΔFg07798 (fusarin C biosynthesis) and ΔFg17168 (perithecial pigment) growing faster than the wild-type at 20 °C (Fig. 4e).
Given our prior work demonstrating that sublethal fungicide exposure can activate the DON biosynthetic gene cluster, we further investigated whether fungicide stress triggers a broader remodeling of SM pathways. Transcript profiling of 53 SM-BGC core genes following treatment with three mechanistically distinct fungicides (carbendazim, tebuconazole, and phenamacril) revealed the significant induction of 10, 2, and 1 core genes, respectively (Figure S3 and Table S5), indicating widespread transcriptional reprogramming of the secondary metabolome during fungicide stress. To determine whether this transcriptional response confers functional fungicide tolerance, we assessed mutant sensitivities and found that five mutants (9.4%) showed altered responses to at least one fungicide. Specifically, ΔFg15872 and ΔFg10933 were less sensitive to tebuconazole, while ΔFg08795 and ΔFg06507 were hypersensitive. Additionally, ΔFg08378 showed increased sensitivity to phenamacril (Fig. 4f). Collectively, these results indicate that a subset of SM-BGCs plays a crucial role in abiotic stress tolerance and fungicide response, with individual clusters exhibiting specialized roles under distinct environmental challenges in F. graminearum.
SM-BGCs are spatiotemporally induced during wheat infection and play tissue-dependent roles in F. graminearum virulence
RNA-seq profiling of F. graminearum during infection of wheat heads and coleoptiles revealed pronounced spatiotemporal regulation of all 53 SM-BGC backbone genes. In wheat heads, SM genes were broadly activated at 3 days post-inoculation (dpi), representing the strongest global induction. This included eight genes with greater than twofold upregulation (P < 0.05) (Fig. 5a and Table S6), most notably Fg03537, a core gene within the DON biosynthetic pathway. By 10 dpi, the number of upregulated genes declined. In contrast, coleoptile infection followed an inverse trajectory, with early repression followed by strong late induction. Specifically, nine genes were significantly upregulated at stage 4 of coleoptile infection, including Fg03537 (~ 136-fold), Fg03245 (~ 557-fold), and Fg16873 (~ 1100-fold) (P < 0.05) (Table S6). In addition, clear tissue-specific differences were observed among SM genes. For instance, Fg04588 was significantly induced during early wheat head infection but showed minimal change across coleoptile infection stages (Fig. 5a). Together, these expression profiles support tissue- and stage-specific control of SM pathways during infection. Therefore, we systematically evaluated the virulence of all 53 SM-BGC core-gene knockout mutants on both wheat heads and coleoptiles.Fig. 5. Spatiotemporal expression of SM-BGCs and their tissue-specific contributions to virulence. a Spatiotemporal transcriptomic profiling of 53 SM-BGC backbone genes during infection of wheat heads and coleoptiles. The heatmap displays transcript levels (log_2_ fold change) relative to mycelia (for wheat heads) or fresh conidia (for coleoptiles). Developmental stages 2-4 correspond to stage 2, polar growth; stage 3, doubling of the long axis; and stage 4, first hyphal branching; non-significant genes are shown in grey. b, c, d, e Pathogenicity analysis of SM-BGC mutants on wheat heads and coleoptiles. b Quantification of infected spikelets per wheat head at 14 dpi. c Representative disease symptoms on wheat heads at 14 dpi, scale bar = 2 cm. d Representative lesions on wheat coleoptiles at 7 dpi, scale bar = 1 cm. e Quantification of lesion length on coleoptiles at 7 dpi. Data in (b) and (e) are presented as means ± SD. Asterisks indicate significant differences relative to the wild-type strain PH-1 (*P < 0.05, **P < 0.01, ***P < 0.001; one-way ANOVA with Dunnett's multiple comparisons test). Abbreviations: dpi, days post-inoculation; SM-BGC, secondary metabolite biosynthetic gene cluster; WT, Wild-type
Wild-type PH-1 inoculation resulted in typical FHB symptoms on wheat heads at 14 dpi, with an average of ~ 16–17 infected spikelets per head. Under the same conditions, six mutants (11.3%, 6/53) showed significantly attenuated virulence, characterized by slower fungal spread and infection of only 1-5 spikelets (Fig. 5b). These attenuated mutants were ΔFg10397, ΔFg03747, ΔFg03537, ΔFg05794, ΔFg17487, and ΔFg06631. Consistent with their reduced head blight symptoms, ΔFg03537, ΔFg17487, ΔFg06631, and ΔFg05794 also caused significantly shorter lesions on wheat coleoptiles than the wild-type. In contrast, ΔFg10397 and ΔFg03747 were indistinguishable from PH-1 in the coleoptile assay, indicating tissue-dependent requirements for these clusters (Fig. 5c-e). In summary, beyond the well-characterized DON biosynthetic cluster, several additional SM-BGCs contribute to the full virulence of F. graminearum during wheat head infection. Moreover, the requirement for specific SM-BGCs can vary markedly between wheat heads and coleoptiles.
The PKS-type cluster BGC36 is required for full virulence and impacts DON biosynthesis
From the initial screen of 53 SM-BGC backbone gene knockouts, six showed significantly reduced virulence on wheat heads. Four of these (ΔFg10397, ΔFg03747, ΔFg03537, ΔFg17487) correspond to BGCs that have been previously linked to pathogenicity. We therefore focused subsequent analyses on two previously uncharacterized SM-BGCs: BGC36 (harboring Fg05794, encoding a polyketide synthase) and BGC47 (harboring Fg06631, encoding an acyl-CoA ligase).
AntiSMASH prediction indicated that BGC36 comprises 26 contiguous genes (Fg05785–Fg05809) (Fig. 6a). The cluster encodes putative transcription factors (Fg05789), a polyketide synthase (Fg05794), a carboxypeptidase (Fg05797), and a P450 monooxygenase (Fg05806). To further evaluate the contribution of BGC36 to virulence, we generated mutants for four additional genes within the cluster (ΔFg05789, ΔFg05791, ΔFg05793, and ΔFg05797) as well as a cluster-deletion mutant (ΔCluster). Although these mutants grew comparably to the wild-type on PDA (Fig. 6b), each showed significantly reduced virulence on both wheat heads and coleoptiles, resembling the mutant ΔFg05794 (Fig. 6c-f). Virulence was restored to near-wild-type levels in the complementation strain ΔFg05794-C (Fig. 6c-f). Consistent with infection-associated expression, Fg05794 transcripts were strongly induced in planta relative to YEPD-grown mycelia, with ~ 50-fold in heads and ~ 8-fold in coleoptiles at 72 h post-inoculation (Fig. 6g-h). Collectively, these data indicate that SM-BGC36 is induced during infection and is required for full virulence in F. graminearum.Fig. 6BGC36 is critical for full virulence and positively regulates DON biosynthesis. a Schematic representation of BGC36 gene cluster. Arrows indicate gene orientation and predicted functions. b Vegetative growth phenotypes. Colony morphology of the wild-type strain and BGC36 mutants grown on PDA for 3 days at 25 °C. c, d, e, f Virulence assessment of SM-BGC36 mutants. Representative disease symptoms (c) and quantification of infected spikelets (d) on wheat heads at 14 dpi, scale bar = 2 cm. Representative lesions (e) and quantification of lesion length (f) on coleoptiles at 7 dpi, scale bar = 1 cm. g, h Transcriptional induction of BGC36 genes in planta. Relative expression levels of the indicated genes during wheat head (g) and coleoptile (h) infection relative to vegetative mycelia. i, j, k Impact of ΔFg05794 on DON biosynthesis. i Relative expression of key TRI genes (TRI1, TRI5, TRI101) in WT and ΔFg05794 grown in trichothecene biosynthesis induction medium for 48 h. j DON-toxisome formation indicated by Tri1-GFP in the WT and ΔFg05794 after 48 h in TBI, scale bar = 10 μm. k Quantification of DON production in 7-day TBI culture filtrates. Data in (d), (f), (i), and (k) are presented as means ± SD from three independent biological replicates. Asterisks indicate significant differences relative to WT (P < 0.05, *** P < 0.01, **** P* < 0.001; one-way ANOVA with Dunnett's multiple comparisons test). Abbreviations: BGC, Biosynthesis gene cluster; DON, deoxynivalenol; dpi, days post-inoculation; PDA, potato dextrose agar; PKS, polyketide synthase; TBI, trichothecene biosynthesis induction; WT, wild-type
All tested mutants in BGC36 initiated infection at the inoculated spikelet but showed limited spread across the rachis, a phenotype often associated with reduced DON production. Given that ΔFg05794 showed no significant differences in sensitivity to osmotic, oxidative, or cell wall stressors compared to the wild-type (Fig. 4a), we examined whether the reduced virulence of BGC36 mutants is associated with altered DON biosynthesis. We therefore assessed the expression of representative DON biosynthesis genes (TRI1, TRI5, TRI101) in the ΔFg05794 mutant under trichothecene-inducing medium (TBI). As shown in Fig. 6i, the expression of all three tested TRI genes was significantly reduced in ΔFg05794 compared with the wild-type. Concurrently, using Tri1-GFP as a marker for the DON-toxisome, we observed that the wild-type strain formed characteristic spherical DON-toxisomes after 48 h of shaking culture in TBI, whereas ΔFg05794 showed no detectable toxisomes and lacked GFP signal (Fig. 6j). Consistently, total DON production in ΔFg05794 was significantly lower than that in the wild-type strain (Fig. 6k). Together, these findings support a functional connection between BGC36 and the trichothecene biosynthetic program. Disruption of BGC36 is associated with reduced DON biosynthesis and limited spread within host tissues, implying that the impaired production of DON contributes to the observed reduction in virulence.
The NRPS-type cluster BGC47 contributes to virulence and is associated with cell wall integrity and DON biosynthesis
Based on antiSMASH prediction and protein annotation, Fg06631 (encoding an acyl-CoA ligase) is located within SM-BGC47 together with nine additional genes, including Fg06627 (CAP-Gly domain-containing protein) and Fg06634 (monooxygenase). Most genes in this cluster were significantly induced during F. graminearum infection of wheat heads (Fig. 7a). To further evaluate the contribution of SM-BGC47 to virulence, we generated a deletion mutant of Fg06627, another gene within the cluster, and assessed pathogenicity on wheat heads and coleoptiles. Compared to ΔFg06631, ΔFg06627 showed significantly reduced virulence on both wheat heads and coleoptiles (Fig. 7b-e), supporting a role for SM-BGC47 in full virulence. Stress phenotype screening revealed that both ΔFg06627 and ΔFg06631 mutants were more sensitive to the cell wall stressor Congo red and the cell membrane stressor SDS than wild-type PH-1 (Fig. 7f-g). Consistent with compromised envelope integrity, mycelia of both mutants were more readily degraded by a cell wall-lytic enzyme cocktail, resulting in increased protoplast release. After 0.5 h of treatment at 30 °C, mutant mycelia released abundant protoplasts, whereas most wild-type PH-1 mycelia remained intact and released only a few protoplasts (Fig. 7h). Western blotting revealed a marked reduction in Mgv1 in ΔFg06631 (Fig. 7i-j), consistent with altered activity of the CWI pathway. Given the established requirement of DON for interspikelet spread, we next assessed DON biosynthesis in ΔFg06631 in TBI. ΔFg06631 showed reduced expression of representative TRI genes (TRI1, TRI5, TRI101), failed to form DON-toxisomes, and produced substantially less DON than wild-type PH-1 (Fig. 7k-l). Collectively, these data identify SM-BGC47 as a virulence-associated cluster that is associated with both CWI pathway activity and trichothecene biosynthesis; disruption of SM-BGC47 is accompanied by cell wall/membrane stress sensitivity, impaired DON production, and reduced virulence.Fig. 7BGC47 modulates virulence by maintaining cell wall integrity and facilitating DON biosynthesis. a Heatmap showing log_2_ fold changes of 12 genes within the SM-BGC47 cluster during wheat head infection relative to YEPD-grown mycelia. b, c, d,e Virulence assessment of SM-BGC47 mutants. Representative symptoms (b) and quantification of infected spikelets (c) on wheat heads at 14 dpi, scale bar = 2 cm. Representative lesions (d) and quantification of lesion length (e) on coleoptiles at 7 dpi, scale bar = 1 cm. (f-g) Hypersensitivity to cell wall and membrane stress. f Colony morphology of the wild-type and mutants grown on MM supplemented with 0.02% SDS or 0.02% Congo red for 3 days. g Quantification of mycelial growth inhibition relative to the corresponding control in (f). h Sensitivity of WT and ΔFg06631 to cell wall-degrading enzymes. Microscopic observation of protoplast release after treating mycelia with a cell wall-degrading enzyme cocktail for 30 min at 30 °C, scale bar = 5 μm. i Phosphorylation of FgMgv1 (P-FgMgv1, ~ 42 kDa) in WT and ΔFg06631 determined by western blotting with an anti-phospho-p44/42 antibody. Total FgMgv1 (~ 42 kDa) protein levels were detected using an anti-p44/42 antibody. P-FgMgv1, phosphorylated FgMgv1; FgMgv1, total FgMgv1. Relative band intensity was normalized to WT. j Relative expression of key TRI genes in TBI cultures at 48 h. (k) Visualization of DON toxisomes (Tri1-GFP) in hyphae grown in TBI for 48 h, scale bar = 10 μm. (l) Quantification of DON production in 7-day TBI culture filtrates. Data in (c), (e), (g), (j), and (l) are presented as means ± SD from three independent biological replicates. Asterisks indicate significant differences relative to WT (*P < 0.05, **P < 0.01, ***P < 0.001; one-way ANOVA with Dunnett's multiple comparisons test). Abbreviations: BGC,biosynthesis-related gene cluster; CR, Congo red; DON, deoxynivalenol; dpi, days post-inoculation; MM, minimal medium; SM-BGC, secondary metabolite biosynthetic gene cluster; YEPD, yeast extract peptone dextrose; WT, wild type
Discussion
In this study, we constructed a deletion library targeting the core genes of 53 predicted SM-BGCs in F. graminearum, enabling genome-scale functional assessment. The pathogenic lifecycle of fungi depends on coordinated regulation of primary and secondary metabolism [31]. In diverse pathosystems, SMs contribute to virulence and fitness through distinct modes of action, including immune suppression by gliotoxin in Aspergillus fumigatus [32], cuticle penetration by melanin and host manipulation by pyriculol in Magnaporthe oryzae [33, 34], and infection facilitation by p-coumaroyl-agmatine ethyl ester (p-CAEE) in Valsa mali [35]. Beyond these established virulence roles, our systematic profiling reveals that multiple SM-BGCs contribute to core physiological traits, including vegetative growth, asexual development, and stress responses (Table S7). These results argue against a simple view of SM biosynthesis as a dispensable "luxury" and instead support broader functional integration of SM-BGCs with fungal physiology. By identifying additional virulence-associated clusters and connecting SM-BGC functions to growth, development, and cell-envelope-related phenotypes, this work expands the functional landscape of the predicted SM repertoire in F. graminearum.
A key finding of this work is the identification of two previously uncharacterized clusters, SM-BGC36 and SM-BGC47, that are required for full virulence on both wheat heads and coleoptiles. The core synthase of SM-BGC36 (Fg05794) shows high homology to the gibepyrone A biosynthetic gene in F. fujikuroi, a reported virulence factor [36], raising the possibility that F. graminearum produces a related metabolite during infection. Notably, mutants in BGC36 and BGC47 exhibited concordant defects in TRI gene induction and DON biosynthesis, suggesting that these clusters influence regulatory programs that govern trichothecene activation. In F. graminearum, TRI output is controlled by integrated transcriptional and chromatin-based programs that respond to environmental cues, including global secondary metabolism regulators and epigenetic control of SM loci [37]. PRC2-associated H3K27me3 repression and its reader proteins (e.g., BP1) can restrict the expression of secondary metabolism genes [38], whereas induction of the trichothecene cluster under DON-promoting conditions is associated with histone acetylation programs, including Gcn5-dependent activities and permissive chromatin organization at the TRI locus [39, 40]. In addition, coordination between H3K36 methylation and H3K27me3 has been implicated in shaping secondary-metabolism expression states [41]. These established layers provide plausible points of convergence for BGC36 and BGC47. For BGC36, the requirement of multiple cluster genes, including a putative transcription factor within the cluster, is consistent with a model in which cluster-encoded regulation and/or BGC36-derived metabolites facilitate full TRI induction, possibly by promoting a permissive transcriptional or chromatin state at the trichothecene locus. For BGC47, the observed defects in cell envelope integrity and reduced Mgv1 phosphorylation suggest that altered CWI pathway activity could indirectly dampen TRI activation through signaling-to-chromatin coupling under inducing conditions rather than through a simple linear pathway. Defining the specific transcription factors and epigenetic regulators that mediate these connections will be an important direction for future work.
Our study further emphasizes that fungal pathogenicity is shaped by spatiotemporal deployment of specific SMs across distinct host tissues. Transcriptome analysis revealed distinct SM-BGC expression patterns during infection of wheat heads versus coleoptiles, and these differences were supported by tissue-specific virulence phenotypes in our mutant library. For instance, ΔFg03747 (SM-BGC27) showed attenuated virulence on wheat heads but was indistinguishable from wild-type PH-1 in the coleoptile assay. This tissue specificity may reflect adaptation to contrasting infection microenvironments. The wheat head represents a complex microbial and chemical context in which iron acquisition and redox buffering (e.g., via siderophores) may be particularly important, whereas coleoptile infection may impose different physical and nutritional constraints that favor alternative metabolic programs [53]. Together, these observations support an ecological dimension in which distinct SM-BGCs are deployed in a tissue- and stage-dependent manner to support fitness within specific host microenvironments [42].
Our genome-scale genetic survey argues against a simple “energy drain” model and instead indicates context-dependent contributions of SM-BGCs to vegetative fitness and development. Although PKS deletions have been reported to alter morphology or even accelerate growth in Setosphaeria turcica [43] and Botrytis cinerea [44], we observed multiple mutants with clear defects in vegetative growth and asexual development, inconsistent with a cost-only interpretation. Notably, growth retardation in mutants such as Δpks2 (SM-BGC34) was medium-dependent, evident on rich media but not on minimal media. This pattern suggests that these SM pathways may contribute to nutrient utilization and/or signaling under specific conditions rather than acting solely as an energetic burden. We also found cluster-specific roles in asexual reproduction. For example, SM-BGC27 (siderophore) [45] and SM-BGC52 (terpene) were required for conidiation, while SM-BGC47 profoundly influenced conidial morphology and germination. Similar pleiotropic phenotypes have been reported in B. cinerea, supporting the view that secondary metabolism can be integrated with core developmental programs rather than functioning solely as an accessory biosynthetic layer.
Consistent with this broader physiological integration, our finding that 32.1% of SM-BGC mutants exhibit altered sensitivity to diverse abiotic stressors highlights SMs as an important determinant of environmental fitness in F. graminearum. The regulatory logic of fungal secondary metabolism has traditionally been framed in a largely “top-down” manner, wherein conserved signaling cascades, particularly MAPK pathways, sense environmental inputs and orchestrate transcriptional activation of specific BGCs [46]. Increasing evidence, however, indicates that secondary metabolism is not merely a downstream output but can also provide feedback to shape stress signaling, with such bidirectional coupling described in fungal chemical biology [31]. Our data are consistent with three non-mutually exclusive ways in which SMs may support abiotic stress adaptation. First, some SMs may provide direct protection. The oxidative-stress hypersensitivity of the siderophore-deficient mutant ΔFg03747 (SM-BGC27) is consistent with the loss of iron sequestration and redox buffering, which can exacerbate iron-catalyzed hydroxyl radical formation via Fenton chemistry, a process increasingly appreciated as critical for Fusarium environmental fitness [47]. Second, our data suggest regulatory crosstalk between secondary metabolism and canonical stress pathways. The reduced Mgv1 phosphorylation observed in the SM-BGC47 mutant background is consistent with altered CWI pathway activity and may contribute to the associated envelope-stress phenotypes [48]. One possibility is that BGC-derived metabolites modulate envelope-derived cues or the stability and activation of signaling complexes under stress. Third, the broad sensitivity of several PKS/NRPS mutants to ionic and membrane-perturbing conditions is consistent with roles for SMs in maintaining envelope and membrane homeostasis under stress [49]. Together, these observations argue that secondary metabolism is not simply a dispensable “luxury” but can function as a physiological buffer supporting envelope homeostasis and survival under fluctuating conditions.
In summary, our study indicates that many predicted SM-BGCs in F. graminearum make measurable contributions to the fungal life cycle. Across the mutant collection, SM-BGCs contributed to vegetative growth, asexual development, stress tolerance, and virulence-associated traits, including DON biosynthesis and CWI. The mutant library provides a resource for future dissection of SM functions and chemical outputs in Fusarium. Future work will aim to identify the metabolites produced by key clusters such as SM-BGC36 and SM-BGC47 and to define the regulatory pathways through which they influence TRI activation, stress responses, and virulence. These insights may inform new strategies to control FHB and reduce mycotoxin contamination.
Materials and methods
Fungal strains and culture conditions
The wild-type F. graminearum strain PH-1 was used as the genetic background strain for all genetic manipulations and as the reference strain in phenotypic assays. PH-1 and all transformants were cultured on potato dextrose agar (PDA, 200 g/L potato, 20 g/L glucose, 10 g/L agar). For mycelial growth assays, strains were cultured on complete medium (CM) (10 g/L glucose, 2 g/L peptone, 1 g/L yeast extract, 1 g/L casamino acids, nitrate salts, 50 mL/L 20 × nitrate salts, 0.1% trace elements, 0.1% vitamin elements, 10 g/L agar, pH 6.5) and minimal medium (MM) (0.5 g/L KCl, 2 g/L NaNO_3_, 1 g/L K_2_HPO_4_, 0.5 g/L MgSO_4_.7H_2_O, 0.01 g/L FeSO_4_.7H_2_O, 30 g/L sucrose, 20 μL/L trace elements, 10 g/L agar, pH 6.9) [50]. Carrot agar medium (CA) was prepared by boiling 200 g of carrots for 20 min. The mixture was then homogenized in a blender, adjusted to a final volume of 1 L with deionized water, and solidified with 15 g/L agar before autoclaving. Wheat head tissue medium (WA) was prepared by boiling 50 g of wheat tissue in water for 20 min. The plant debris was removed by filtration through cheesecloth, and the filtrate was adjusted to a final volume of 1 L and solidified with 15 g/L agar [51]. Mung bean medium was used for conidiation assays [52]. For the induction of trichothecene biosynthesis, fungal cultures were grown in liquid TBI medium (30 g/L sucrose, 1 g/L KH_2_PO_4_, 0.5 g/L MgSO_4_.7H_2_O, 0.5 g/L KCl, 0.01 g/L FeSO_4_.7H_2_O, 0.8 g/L putrescine hydrochloride, 20 μL/L trace elements). The pH was adjusted to 4.5 using 1 M HCl [53].
Identification, characterization, and annotation of SM-BGCs
Reference genome assemblies for 19 plant pathogenic Fusarium species were retrieved from the National Center for Biotechnology Information (NCBI) database. Detailed accession numbers for each genome are provided in Table S8. SM-BGCs were predicted for all 19 Fusarium species using antiSMASH version 7.0 [54]. To generate a high-confidence set of SM-BGCs for F. graminearum (strain PH-1), we used a re-annotated PH-1 genome [55] and integrated the results from the antiSMASH analysis with data from previously published studies. This yielded a final set of 53 high-confidence SM-BGCs containing core enzymes, such as PKS, NRPS, TPS, and CDPS. These 53 clusters were designated SM-BGC01 to SM-BGC53 according to their chromosomal locations. The specific type of core enzyme for each of the 53 SM-BGCs was further analyzed and confirmed using CusProSe [56].
Phylogenetic analysis
A total of 19 Fusarium genomes were downloaded from the NCBI database for phylogenetic reconstruction. Single-copy orthologous genes were identified using BUSCO v5.2.2 [57] based on the hypocreales_odb10 lineage dataset. From this analysis, 4,492 genes conserved across all 19 species were identified. Sequences for each gene were aligned using MAFFT v7.471 [58] with the "--auto" parameter. Ambiguous regions were removed using trimAl v1.4 [59] with the "gappyout" option. The resulting alignments were concatenated into a supermatrix using PhyKIT v1.2.1 [60]. Finally, a maximum likelihood analysis was performed using IQ-TREE (v1.6.12) under the GTR + G4 + F model [61].
Generation of gene knockout and complementation mutants
All gene deletion mutants were generated in the wild-type PH-1 background. Gene replacement constructs were assembled using the double-joint PCR method. Briefly, ~ 1,000 bp upstream and downstream flanking regions of each target gene were amplified from wild-type genomic DNA and fused with the hygromycin resistance gene (hph) (Figure S1). Transformation of F. graminearum was performed via the polyethylene glycol (PEG)-mediated protoplast transformation method [62] using the resulting linear deletion construct. For complementation, the open reading frame of the target gene was fused to GFP under the native promoter and introduced into the corresponding mutant together with a geneticin-resistance marker (Figure S4) [63]. Hygromycin B (Calbiochem, La Jolla, CA, USA) or Geneticin (G418) (Sigma‒Aldrich, St. Louis, MO, USA) was added to the regeneration medium at a final concentration of 100 µg/mL for transformant selection. Correct gene deletion and complementation were confirmed by diagnostic PCR. The primers used in this study are listed in Table S9.
RNA-seq data acquisition and bioinformatic analysis
Publicly available raw transcriptomic sequencing data were retrieved from the NCBI Sequence Read Archive (SRA) database. For reproducibility and traceability, all SRR accession numbers utilized in this study are comprehensively listed in Table S10. The SRA files were converted to FASTQ format using the SRA Toolkit [64]. Raw sequence reads were assessed for quality using FastQC [65]. Trimmomatic [66] was used to remove adapter sequences and trim low-quality reads to obtain high-quality clean data. The processed reads were pseudo-aligned to the F. graminearum reference transcriptome, and transcript abundances were quantified using Kallisto [67]. The subsequent conversion of transcript-level abundances into gene-level counts was performed using TBtools [68].
Differential expression analysis was performed using the DESeq2 R package. To maintain a consistent and comparable framework across datasets, genes were classified as differentially expressed genes (DEGs) if they met both of the following criteria: an adjusted P value < 0.05 and an absolute |log_2_-fold change|≥ 1. Notably, for the conidia germination transcriptome (PRJNA664506) [55], which had only two biological replicates per group, formal statistical significance testing was not performed for this dataset due to the limited number of biological replicates (Table S3). All transcriptomic data were visualized using ChiPlot [69].
Conidiation and conidial germination assay
For conidiation, strains were cultured in mung bean broth as previously described. Briefly, five fresh mycelial plugs (5 mm in diameter) excised from the actively growing colony margin were inoculated into 20 mL of mung bean broth in a 50-mL flask and incubated at 25 °C with shaking at 180 rpm for 5 days. Cultures were filtered through three layers of lens-cleaning paper to remove mycelia. Conidia were collected by centrifugation (6,000 rpm for 4 min), washed twice with sterile distilled water, and resuspended in sterile water. For septation analysis, conidia were stained with calcofluor white (CFW, Sigma-Aldrich) and imaged using an Evident FV4000 confocal microscope (Olympus, Tokyo, Japan). For germination assays, conidial suspensions were adjusted to 1 × 10^5^ conidia/mL and mixed 1:1 with YEPD to a final concentration of 5 × 10^4^ conidia/mL. Samples were incubated at 25 °C, and germinated conidia were scored at 6 h post-inoculation by microscopy. At least 100 conidia were counted per replicate, and each assay included 3 biological replicates.
Determination of sensitivity to abiotic stress agents
Sensitivity to abiotic stress agents was assessed by measuring radial growth on PDA or MM supplemented with the indicated compounds. Mycelial plugs (5 mm in diameter) were taken from the actively growing margins of 2-day-old PDA cultures of the wild-type PH-1 and the indicated mutants and placed onto stress-containing plates. Plates were incubated for 3–4 days, after which colony diameters were measured. Growth inhibition was calculated relative to the corresponding unsupplemented control medium. Stress agents and their final concentrations are provided in the relevant figure legends. For each condition, at least three independent biological replicates were performed, with three technical replicates per biological replicate. Sensitivity to cell wall-lytic enzymes was evaluated by monitoring protoplast release. Conidia were inoculated into liquid YEPD and incubated for 14 h at 25 °C with shaking (180 rpm). Hyphae were collected by filtration through three layers of lens-cleaning paper, washed with 0.7 M NaCl, and incubated in 0.7 M NaCl containing cellulase (30 mg/mL) (D9515, Sigma), lysing enzyme (30 mg/mL) (RM1027, RYON, Shanghai, China) and driselase (20 mg/mL) (RM1030, RYON, Shanghai, China) in 0.7 M NaCl. The mixture was incubated at 30 °C for 30 min with gentle shaking. Protoplast release was assessed by light microscopy (Nikon, Tokyo, Japan). Each experiment was repeated three times independently.
Pathogenicity assay
All virulence assays were performed using the wheat cultivar Jimai 22 (FHB susceptible). Wheat head infection assays were conducted on flowering heads under field conditions as previously described [53]. Briefly, mycelia were harvested from 24-h liquid YEPD cultures, washed with sterile water, and homogenized briefly to generate a fragmented mycelial suspension. A 10 µL aliquot of mycelial suspension was then injected into a central floret of each wheat head (n = 15). Mock controls received sterile distilled water. Inoculated wheat heads were maintained under high humidity for the first 48 h to facilitate infection, and disease severity was quantified at 14 days post-inoculation (dpi) by counting the number of symptomatic (bleached) spikelets per inoculated head. Coleoptile infection assays were performed using a pin-point inoculation method. Wheat seeds were first surface-sterilized with 75% (v/v) ethanol for 1 min, rinsed three times with sterile distilled water, and germinated on double-layered sterile filter paper with sterile water at 25 °C in the dark. Three-day-old etiolated seedlings (2-3 cm) were wounded at the coleoptile tip with a sterile needle, and 10 µL of fragmented mycelial suspension was applied to the wound site. Mock controls were treated with sterile distilled water. Each treatment included 20 seedlings. Seedlings were maintained at 25 °C under a 12 h light/12 h dark photoperiod, and lesion length was measured at 7 dpi to quantify virulence.
DON production, toxisome observation and TRI gene expression analysis
To induce DON biosynthesis, all strains were initially cultured in YEPD medium at 25 °C with shaking (180 rpm) for 24 h in the dark, then transferred to TBI liquid medium and incubated at 28 °C with shaking (150 rpm) under dark conditions. For DON quantification, culture supernatants were collected after 7 days in TBI medium, and DON was measured using a commercial ELISA Quantification Kit (Wis008, Wise Science, Zhenjiang, China) following the manufacturer's instructions. A Δtri5 mutant was included as a negative control for the assay. All experiments were performed with three independent biological replicates. For toxisome visualization, Tri1-GFP-labeled strains were cultured in TBI liquid medium for 40 h and examined by confocal microscopy. For TRI gene expression analysis, mycelia were harvested after 48 h in TBI medium for RNA extraction and subsequent qRT-PCR analysis of representative DON biosynthesis genes.
RNA Extraction and Quantitative Real-Time PCR (qRT‒PCR) analysis
Total RNA was extracted from fungal mycelia and infected plant tissues using TRIzol Reagent. For analysis of TRI gene expression, mycelia were harvested from TBI cultures as described above. For analysis of fungal gene expression in planta, infected wheat coleoptile samples were collected at 72 h post-inoculation (hpi). Samples were immediately frozen in liquid nitrogen and ground into a fine powder, and total RNA was extracted following the manufacturer's protocol (TaKaRa Biotechnology, Dalian, China). First-strand cDNA was synthesized from 1 μg of total RNA using HiScript II reverse transcriptase (HiScriptII Q RT Kit, Vazyme, R223-01). qRTPCR was performed using SYBR Green Supermix (ChamQ SYBR qPCR Master Mix, Vazyme, Q311-02) on a CFX real-time PCR system (Bio-Rad). Each 20-µL reaction contained 10 µL of 2 × Master Mix, 0.4 µL of each primer (10 µM), and 2 µL of the diluted cDNA template. Actin was used as the internal reference gene. Relative transcript levels were calculated using the 2^-ΔΔCt method as previously described. All assays were performed with two biological replicates and three technical replicates per sample. Primer sequences are listed in Table S9.
Western blotting assay
Total protein was extracted as previously described [70]. Equal amounts of protein from each sample were separated on a 10% SDS-PAGE gel and transferred to a polyvinylidene fluoride membrane using a Bio-Rad electroblotting apparatus. The membrane was blocked for 1 h at room temperature in 5% (w/v) non-fat dry milk dissolved in TBST. To detect phosphorylated FgMgv1 and FgGpmk1, the membrane was incubated overnight at 4 °C with a primary antibody specifically recognizing dually phosphorylated p44/42 MAPK (Erk1/2) (phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) antibody; Cell Signaling Technology, Boston, MA, USA), which has been previously validated for detecting phosphorylated FgMgv1 and FgGpmk1. The upper and lower bands represent phosphorylated FgMgv1 and FgGpmk1, respectively. Total FgMgv1 protein levels were detected using an anti-Mpk1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Secondary antibody incubation and chemiluminescent detection were performed as previously described [70]. Each experiment was conducted with three independent biological replicates.
Statistical analysis
All experiments were performed in at least three independent biological replicates. Data are presented as the means ± SD for each condition. Statistical significance for comparisons between mutants and the wild-type was assessed using one-way ANOVA followed by Dunnett's test. The significance levels were denoted as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.
Supplementary Information
Supplementary Material 1: Figure S1. Generation and PCR verification of gene deletion mutants. (a) Schematic representation of the gene deletion strategy. The target gene was replaced by a hygromycin resistance cassette (HPH) via homologous recombination. Arrows indicate the binding sites of primers used for verification. (b) DNA ladder indicating fragment sizes for PCR verification. (c) PCR verification of representative gene deletion mutants that exhibited at least one measurable phenotype under the tested conditions. (d) PCR verification of representative gene deletion mutants with no detectable phenotypes under the tested conditions. Three independent deletion transformants (T1/2/3) were used for verification. Primers hph-F and ID1-R or ID1-F/R were used to confirm HPH cassette insertion, whereas primers ID2-F/R, internal to the target gene open reading frame, were used to confirm the absence of the native gene. The mock lane represents a negative control using water as the PCR template. Figure S2. Defective conidial germination in specific SM-BGC mutants. Conidial germination assay. Representative images of germinating conidia from the wild-type, ΔFg10464, and ΔFg06631 strains at 6 h post-incubation in 1/2 YEPD liquid medium. scale bar = 50 μm. Figure S3. Transcriptional responses of SM-BGC core genes to fungicide stress. The heatmap displays the log2-fold changes in transcript abundance of 53 SM-BGC core genes in mycelia treated with three distinct fungicides, carbendazim (Car), tebuconazole (Teb), and phenamacril (Phe), relative to untreated controls. Red and blue indicate significantly upregulated and downregulated genes, respectively, whereas gray denotes genes without statistically significant differential expression. Differential expression was defined as adjusted P < 0.05 and |log2 fold change| ≥ 1. Figure S4. Generation and PCR verification of gene deletion mutants and complemented strains in SM-BGC36 and SM-BGC47. (a) Gene replacements were performed by homologous recombination following the strategy shown in Figure S1. For each gene, successful deletion was verified using primer pairs ID1-F/ID1-R and ID2-F/ID2-R. (b) Schematic representation of the complementation construct. The construct contains the full-length target gene driven by its native promoter together with a neomycin resistance gene (neo). (c) PCR verification of the Fg05794 complemented strain (ΔFg05794-C). The primer pair ID1-F/R was used to verify correct integration of the complementation cassette, and the primer pair ID-Gene-F/ID-GFP-R was used to detect the presence of the reintroduced gene-GFP fusion construct. For all gels: T1/2/3, independent transformants; WT, wild-type; Δ, deletion mutant; Mock, negative control using water as the PCR template.Supplementary Material 2: Table S1. Characteristics of the 53 SM-BGCs identified and analyzed in F. graminearum PH-1.Supplementary Material 3: Table S2. Colony diameter differences and statistical analysis of 53 SM-BGC mutants compared with the wild type.Supplementary Material 4: Table S3. Expression profiles of genes significantly regulated during conidiation and conidial germination.Supplementary Material 5: Table S4. Mycelial growth inhibition rate of SM-BGC mutants under abiotic and fungicide stresses.Supplementary Material 6: Table S5. Expression profiles of genes significantly regulated during treatment with carbendazim, tebuconazole, and phenamacril.Supplementary Material 7: Table S6. Expression profiles of genes significantly regulated during infection of wheat heads and coleoptiles.Supplementary Material 8: Table S7. Phenotypic profiling of 53 SM-BGC deletion mutants in F. graminearum.Supplementary Material 9: Table S8. Summary of genome assembly statistics, accession numbers, and host information for the 19 Fusarium species analyzed.Supplementary Material 10: Table S9. Primers used in this study.Supplementary Material 11: Table S10. List of RNA-seq datasets used in this study.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Ding Z, Yang L, Wang G, Guo L, Liu L, Wang J, et al. Fusaric acid is a virulence factor of Fusarium oxysporum f. sp. cubense on banana plantlets. Trop Plant Pathol. 2018;43:297–305. 10.1007/s 40858-018-0230-4