Comparative transcriptomic analysis reveals infection strategies of Botrytis deweyae on Polygonatum cyrtonema Hua

Summary

Botrytis deweyae causes gray mold on Polygonatum cyrtonema Hua, but its infection mechanism remains unknown. Microscopy (optical, SEM, and TEM) revealed that B. deweyae invades P. cyrtonema leaves either by direct penetration, characterized by the secretion of gelatinous substances and formation of infection cushions, or through stomatal invasion. Transcriptome analysis revealed that glyceride metabolism, oxidoreductase activity, and MAPK signal transduction facilitated the infection of P. cyrtonema leaves by B. deweyae. Glyceride metabolism was associated with lipid transport, lipid localization, sterol transport, and other material transport within the cell membrane. Host-derived high osmotic pressure triggered the MAPK signaling pathway in B. deweyae, leading to the expression of downstream antioxidant enzymes for osmotic regulation. The significant enrichment of Hog1 (K04441) within the MAPK pathway indicates its potential role in modulating the osmotic stress response signaling pathway in B. deweyae. This study provides a theoretical framework for the pathogenesis of B. deweyae and aids in the development of targeted treatments for gray mold.

Graphical abstract

Highlights

• •Glyceride metabolism and MAPK signaling facilitate fungal infection
• •Hog1 (K04441) may serve as a key regulator of osmotic stress adaptation

Natural sciences; Microbiology; Plant bioinformatics; Plant biology

Introduction

Botrytis deweyae is a member of the genus Botrytis and exhibits facultative pathogenicity. Initially described by Grant-Downton as the causal agent of “spring sickness” (Foliar Disease) in cultivated Hemerocallis L.,¹ this fungus has recently been reported to infect Polygonatum cyrtonema Hua.² and Fragaria × ananassa (Duchesne ex Weston),³ causing gray mold and fruit rot, respectively. These findings suggest that, under global climate change, certain potentially pathogenic fungi may be evolving toward full pathogenicity. Therefore, understanding the pathogenic mechanisms of such fungi is crucial for effective disease prevention and control.^4,5

Research on the pathogenic mechanisms of Botrytis spp. has predominantly focused on the model pathogen B. cinerea, while the virulence factors of B. deweyae remain largely unexplored. Despite the close relationship between B. cinerea and B. deweyae, this difference in pathogenicity is fundamentally genetic rather than morphological in origin. For instance, while morphological variations can distinguish varieties with different host ranges across plant families, formae speciales are often morphologically identical and are differentiated primarily by their specific pathogenicity to different genera within the same plant family.⁶ This genetic basis for virulence is exemplified by the fusaria⁷: Fusarium graminearum produces potent B-type trichothecene mycotoxins (e.g., deoxynivalenol, DON), which serve as major virulence factors in Fusarium head blight of wheat.⁸ In contrast, F. oxysporum employs distinct pathogenic strategies, such as causing vascular wilt, and does not produce these trichothecenes. Notably, the capacity for toxin synthesis can vary substantially not only between species but even among different formae speciales within the same species.⁹ This principle underscores the importance of gene-level analysis in understanding the pathogenic potential of less-studied species such as B. deweyae.

The invasion of host plants by B. cinerea can be delineated into four distinct stages: host penetration, tissue necrosis, tissue decomposition, and spore formation. First, conidia attach to the host surface by recognizing its hydrophobic and carbohydrate properties and then germinate to produce germ tubes that directly penetrate the host epidermis¹⁰; protruding hyphae, which form appressoria through the accumulation of melanin, glycerol, and other substances; and invading structures, such as infection pads.¹¹ Furthermore, toxins, pathogenic factors, and enzymes that promote programmed cell death (PCD) are produced. Following successful invasion by B. cinerea, plant tissues disintegrate,¹² allowing the fungus to produce conidia in the diseased tissues that spread via rain and airflow, becoming a source of subsequent infections.¹³ Ultrastructural analysis of rapeseed leaves infected by B. cinerea revealed that the infection pads formed by hyphae secrete gelatinous substances to degrade the cell wall of epidermal cells, allowing hyphae to penetrate the leaves through damaged epidermal cells and grow within mesophyll cells.¹⁴ Gray mold directly penetrates blueberry fruit via hyphae; therefore, investigating the modes of invasion of pathogenic fungi and ultrastructural changes in host plants is necessary for effective disease control.¹⁵

Colonization of B. cinerea on plants is an important process.¹⁶ The most representative cell death process in plant disease resistance is the reactive oxygen species (ROS)-induced allergic reaction. This process limits the growth of pathogenic fungi. B. cinerea can also produce several cell death-inducing proteins, which are recognized by plants through pathogenesis-related protein receptors.¹⁷ Pattern-triggered immunity is initiated to manipulate the PCD mechanism and induce local host cell death so that the pathogen can obtain nutrition.¹⁸ However, when the ROS concentration in the host exceeds the tolerance limit of B. cinerea, the fungus can regulate cytochrome c-peroxidase (POD) and superoxide dismutase (SOD) to scavenge host-derived ROS.^19,20 Catalase (CAT) (hydrogen peroxide) accumulates in hyphae tips and infection cushions,²¹ facilitating permeation into host cells by promoting the oxidation of stratum corneum polymers. In addition, the expression of B. cinerea virulence factors requires functional signal transduction cascades, especially the heterotrimer G protein cascade, mitogen-activated protein kinase (MAPK) cascade, and calcineurin pathway.²² Yang et al. found that mitogen protein kinase (BcOs4) is necessary for the nutritional differentiation and pathogenicity of gray mold.²³ Pathogenic fungi can use many different virulence genes and strategies to cause diseases simultaneously. In particular, the infection mechanism of B. cinerea utilizes many pathways and active virulence components and is thus very complex. Examining the structural and infectious strategies adopted by pathogenic fungi to infect host plants is beneficial to deeply understand the pathogenic mechanisms employed by pathogens and evaluate the severity of diseases and resistance of host plants.²⁴ Transcriptome sequencing (RNA sequencing) can reveal these pathogenic mechanisms at the molecular level, identify influencing factors, and provide a reference for the development of drugs that target gray mold. The structure and mode of infection of B. deweyae, as well as transcriptomic changes in pathogenic genes during the infection process, remain unknown. In this study, we analyzed the structure and mode of infection of B. deweyae through optical, SEM, and TEM, and the key pathogenic pathways were identified using transcriptome sequencing and analysis.

Results

Hyphal penetration and development progress of B. deweyae

Following inoculation with B. deweyae, the disease progression was monitored. No hyphae were detectable at 0 hpi (Figure 1A), with only stomata visible under optical microscopy (Figure 1E). The signs of infection emerged at 24 hpi (Figure 1B), characterized by the development of water-soaked lesions and the appearance of hyphae extending toward stomata (Figure 1F). Lesion expansion and the formation of white aerial hyphae coincided with hyphal penetration into the tissue via stomata by 48 hpi (Figures 1C–1G). By 96 hpi, severe wilting, plant decline, and mortality were observed, accompanied by extensive occlusion of stomata by dense hyphae (Figures 1D–1H).

Figure 1.

Symptoms of gray mold in Polygonatum cyrtonema and the infection process of Botrytis deweyae at different time points

(A–D) Development of gray mold disease in P. cyrtonema at different time points post-inoculation.

(E–H) Observation of B. deweyae hyphal infection under optical microscopy at four time points. Scale bars indicate 40 μm.

SEM observation revealed that no hyphal growth occurred from sterile agar plugs on P. cyrtonema leaves at the 0 hpi, with stomata clearly visible on the leaf surface. (Figure 2A). The infection process progressed distinctly over time. At 12 hpi, hyphal attachment and the development of infection cushions were observed, accompanied by lytic secretions (Figure 2B). Following the initial stomatal penetration by hyphae at 18 hpi (Figure 2C), which marked the initial breach of host defense, the development of mature infection cushions was evident by 24 hpi (Figure 2D). Further hyphal proliferation within the epidermis and intercellular spaces culminated in widespread hyphal aggregation and host cell death (Figure 2E).

Figure 2.

Scanning electron micrographs of Botrytis deweyae infection in Polygonatum cyrtonema at different stages

(A) Stomatal structure on the leaf surface in the absence of hyphal infection at 0 hpi.

(B) Development of infection cushions and infectious hyphae at 12 hpi.

(C) Hyphae penetrate the leaf epidermis via stomatal pores at 18 hpi.

(D) Hyphal attachment to the leaf surface and expanding infection cushions at 24 hpi.

(E) The hyphae penetrate the leaf epidermis and extend along intercellular spaces. Note: S, stomata; IC, infection cushion; IH, infectious hyphae. Scale bars are 50 μm in panels A–C and 40 μm in panels D and E.

TEM revealed the cellular dynamics of infection within P. cyrtonema leaves. At 0 hpi, no fungal structures were detected in the epidermis, mesophyll, or vascular bundles; organelles remained intact, and cell walls were uniformly electron-dense (Figures 3A and 3B). By 24 hpi, primary hyphae had traversed the cuticle and penetrated epidermal cells (Figure 3C). Intracellular hyphae then expanded, causing plasmolysis and dilation of the host cell wall (Figure 3D). Hyphae were absent from vascular bundles (Figures 3E and 3F). Secondary hyphae subsequently crossed into neighboring cells (Figure 3G). Continued proliferation led to widespread hyphal colonization of the mesophyll, accompanied by mitochondrial swelling and ribosome aggregation (Figure 3H). Deeper invasion followed: Hyphae reached the vascular bundles, induced dissolution of the xylem middle lamella (Figures 3I and 3J). Additionally, stomatal entry was also documented, with hyphae traversing the substomatal cavity and degrading guard-cell walls (Figures 3K and 3L).

Figure 3.

Transmission electron micrographs of Botrytis deweyae infection in Polygonatum cyrtonema at different stages

(A and B) No fungal hyphae were observed in the epidermis, mesophyll cells, or vascular bundles.

(C) Primary hyphae penetrated the epidermal cells.

(D) Hyphae expanded intracellularly.

(E and F) No hyphae were observed in the vascular bundles at 24 hpi.

(G) Secondary hyphae extended into adjacent cells.

(H) Hyphae continued to expand and were distributed throughout the mesophyll cells.

(I) Hyphae progressively invaded deeper tissues and penetrated the vascular bundles.

(J) Secondary hyphae invaded the vascular bundles, causing cell lysis and death.

(K and L) Hyphae entered the leaf tissue through stomata on the lower epidermis. Note: PH, primary hyphae; SH, secondary hyphae; VE, vessel; GC, guard cell; EC, epidermal cell; MC, mesophyll cell; CU, cuticle; ST, sieve tube. Scale bars are 2 μm in panel B and 4 μm in panels A and C–L.

Transcriptome analysis of B. deweyae

In total, 76.28 GB of clean data were obtained through RNA-seq analysis of 12 samples; each sample yielded 6.10 Gb of clean data. The percentage of Q30 bases was approximately 95% (Table S3). While detecting differentially expressed genes (DEGs), the screening criteria were set as a differential expression ratio (fold change) ≥ 2 and a false discovery rate (FDR) < 0.01. To quantify the gene expression levels in each sample, a box diagram was drawn to determine the gene expression distribution. The box diagram of the overall gene expression level (FPKM) distribution in different samples is shown in Figure S1A.

Sample correlation evaluation and differential gene number statistics

Spearman’s correlation coefficient (R²) was used to evaluate the index of biological repeat correlation.²⁵ Within each treatment group, a strong correlation was observed among the samples (Figure S1B), indicating that the repeatability of the test was reliable.

From the three groups (G1, G2, and G3), 4925 genes were differentially expressed across the gene sets, with 2605 genes upregulated and 2320 genes downregulated. G2 group had the largest number of DEG sets (1958) (Figure 4A). To determine the relationships among DEGs in the three differential gene sets, a Wayne diagram analysis was performed, and an Euler diagram was constructed via the Bioinfo Intelligent Cloud network platform (Figures 4B and 4C). In total, 281 DEGs were upregulated, and 121 genes were downregulated.

Figure 4.

Statistics for the number of differentially expressed genes (DEGs)

(A) Histogram shows the number of DEGs.

(B) Euler plot shows the differentially upregulated genes.

(C) Euler plot shows the differentially downregulated genes. See also Figure S2.

Enrichment analysis of DEGs in the early stage of infection

To understand the changes in DEGs and the pathways involved in the early stage of infection, the changes in the expression of DEGs in G1 group gene concentration were visualized using a volcanic map and expression cluster ThermoMap (Figures S2A and S2B). KEGG and GO enrichment analyses of 629 upregulated genes were conducted.

The KEGG classification map showed that nine genes were annotated in the peroxisome pathway in cellular processes. The MAPK signaling pathway was the most annotated item in environmental information processing, and protein processing in the endoplasmic reticulum was the most involved pathway in genetic information processing (Figure 5A). The KEGG enrichment map revealed carbon metabolism, glycolysis/gluconeogenesis, and glycerolipid metabolism as the most annotated pathways involved in metabolism. KEGG analysis revealed the enrichment of genes involved in metabolic pathways. The KEGG enrichment map also confirmed that glyceride metabolism plays a crucial role in the entire infection process (Figure 5B).

Figure 5.

Kyoto Encyclopedia of Genes and Genomes (KEGG) classification and enrichment

(A) Line chart of the KEGG pathway classification.

(B) Bubble chart of the top 20 KEGG-enriched differentially expressed genes. The ordinate represents the KEGG pathway. The abscissa represents the Rich factor. The larger the Rich factor, the greater the enrichment. The larger the point, the greater the number of differential genes enriched in the pathway. The redder the color of the dots, the lower the q-value (Benjamini–Hochberg-adjusted, p value < 0.05), indicating higher statistical significance after multiple testing correction.

The GO enrichment map showed that oxidoreductase activity was the most significantly enriched in biological process, alongside genes related to lipid, sterol, and steroid binding. The expression of genes involved in lipid transport, lipid localization, and sterol transport in molecular function (MF) increased (Figure 6A). The topGO directed acyclic map of DEGs revealed a close relationship between lipid transport and sterol transport, with a high degree of enrichment (Figure 6B). Lipid transport and sterol transport were involved in lipid binding and oxidoreductase activity, respectively. Furthermore, EAE98_008811 was involved in sterol and glucosterol binding (Figure 6C).

Figure 6.

Gene ontology (GO) enrichment analysis of the up-regulated differentially expressed genes (DEGs)

(A) GO enrichment bubble chart of biological process and molecular function. The ordinate represents the GO pathway. The abscissa represents the rich factor. The larger the rich factor, the greater the enrichment. The larger the point, the greater the number of differential genes enriched in the pathway. The redder the color of the dots, the lower the q-value (Benjamini–Hochberg-adjusted, p value < 0.05), indicating higher statistical significance after multiple testing correction.

(B) Top 10 GO-directed acyclic graph (topGO) for the DEGs in biological processes about lipid transport and sterol transport, which also includes their hierarchical relationships across different levels. Each box (or ellipse) contains the description of the GO term and its enrichment significance value. Different colors represent different levels of enrichment significance; darker colors indicate higher significance. For topGO significance, p values are calculated using the Kolmogorov-Smirnov statistical test.

(C) GO enrichment chord diagram of the upregulation DEGs. The left half represents different genes and their relative expression levels, the darker the color, the more obvious the up-regulation of the modified genes, and the right half represents the annotation to different molecular function pathways, which are represented by different color patches and connecting lines.

Subsequently, an enrichment analysis was performed on the down-regulated DEGs during the early infection stage (group G1). The KEGG pathway analysis revealed a significant down-regulation (q-value <0.05) of ten genes associated with ABC transporters (Figures 7A and 7B). In the GO enrichment analysis, the twenty-five genes associated with the “host cell nucleus” term was significantly enriched within the cellular component category, while “DNA-binding transcription factor activity” and “ATPase activity” were enriched under the MF category (Figure 7C). These results suggest that the infection strategy of B. deweyae may involve the suppression of host resistance factors and transport-related signaling molecules.

Figure 7.

Gene ontology (GO) enrichment analysis of the downregulated differentially expressed genes (DEGs) at the early stage of infection

(A) The top 20 significantly enriched KEGG pathways.

(B) Chord diagram displays the genes associated with the enriched KEGG pathways.

(C) GO enrichment of cellular component and molecular function chord diagram of the downregulation DEGs. The redder the color of the dots, the lower the q-value (Benjamini–Hochberg-adjusted, p value < 0.05), indicating higher statistical significance after multiple testing correction.

Enrichment analysis of DEGs during infection

To explore the functions associated with the continuous upregulation of DEGs at 0–96 h, GO enrichment analysis was performed on 281 co-expressed genes shared by the three groups (G1, G2, and G3) upregulated genes (Figure 4B). The MF enrichment map of the GO pathway indicated significant enrichment in oxidoreductase activity (Figure 8A). The topGO directed acyclic map revealed that the GO node (term) was enriched by the top five DEGs and their hierarchical relationships. Forty-three genes were involved in oxidoreductase activity, as shown in the KEGG network diagram (Figure 8B). The topGO directed acyclic map also showed that the MFs included antioxidant processes, with 2 CAT- and 3 POD-related genes participating in the antioxidant reduction process (Figure 8C).

Figure 8.

Enrichment map for the differentially expressed genes during the infection of Botrytis deweyae

(A) Gene ontology (GO) enrichment bubble chart. The redder the color of the dots, the lower the q-value (Benjamini–Hochberg-adjusted, p value < 0.05), indicating higher statistical significance after multiple testing correction.

(B) Network enrichment map for GO molecular function. The line color represents the different pathways through which the annotation is reached. The size of the dots represents the number of genes. The larger the size, the greater the enrichment; GO classification of the oxidoreductase process.

(C) Different colors represent different levels of enrichment significance; darker colors indicate higher significance. For topGO significance, p values are calculated using the Kolmogorov-Smirnov statistical test.

MAPK signaling pathway

Fourteen genes were involved in the MAPK cascade signal transduction pathway in response to osmotic pressure, of which five were downregulated, and nine were upregulated (Figure 9). In this study, a high osmotic pressure environment was shown to activate the MAPK cascade reaction. B. deweyae perceives changes in osmotic pressure through membrane surface receptors. This perception inhibits Sln1, which, under standard conditions, phosphorylates Ypd1 to promote gene expression. With Sln1 inhibited, a MAPK cascade signaling pathway is activated. This activation, despite a short-term disruption in signal transduction, ultimately leads to the mediation of osmolyte synthesis. Besides Sln1, Stel1 also participates in the MAPK cascade process by downregulating gene expression. Hog1 (EAE98_007802, K04441) plays an important role as an intermediate regulatory gene between the cascade reaction and synthesis of downstream osmolytic substances.

Figure 9.

Diagram of the MAPK signaling cascade pathway

Each gene is represented in the wireframe. Red shading represents upregulation of the gene, green represents downregulation, and the heat map displays the enrichment level of the gene at four time points. The redder the color, the greater the enrichment of the gene at that time point in the pathway.

Glyceride metabolic pathway

To explore the genes involved in glyceride metabolism, the KEGG database containing 17 upregulated genes involved in glyceride metabolism was used to construct a glyceride metabolic pathway map. To visually observe the expression of 17 genes in each sample, the heatmap of 17 genes was mapped using the Baimaike cloud platform and displayed in combination with the KEGG pathway map (Figure 10), in which EAE98_001653 (K01046) promoted the synthesis of many metabolites, including 1, 2-diacyl-sn-3-phosphate (1, 2-diacyl-sn-glycerol-3-phosphate), triacylglycerol (TAG), and monoacylglycerol (1-Acyl-sn-glycerol-3-phosphate).

Figure 10.

Diagram of the glycerolipid metabolism pathway

Kyoto Encyclopedia of Genes and Genomes (KEGG) classification map of the differentially expressed genes. Each gene is represented in the wireframe. Red shading represents upregulation of the gene, green represents downregulation, and the heat map displays the enrichment level of the gene at four time points. The redder the color, the greater the enrichment of the gene at that time point in the pathway.

Network core gene of Hog1 in PPI

The PPI network diagram (Figure 11) shows that EAE98_007802 interacts with many pathways, including the stress response, lipid transport, and MAPK cascade reaction pathways. This finding was consistent with that of the MAPK signaling pathway analysis. B. deweyae responds to host-generated toxic metabolites and the high-osmosis glycerol MAPK (HOG-MAPK) pathway, indicating that EAE98_007802 significantly impacts toxin production, cell membrane composition, and pathogenicity in B. deweyae. EAE98_001658 (K00799) interacts with the POD gene EAE98_005198 (K13338). B. deweyae is involved in POD synthesis and cellular detoxification. In addition, EAE98_009286 (K00467), which is involved in lipid transport, interacts with four genes involved in MAPK signaling, POD activity, and stress response, indicating that B. deweyae utilizes multiple pathways with synergistic effects to cope with the high osmotic pressure produced by host cells. Furthermore, sterol transport and lipid activity on cell membranes are frequent.

Figure 11.

Interaction network diagram for the differentially expressed genes of Botrytis deweyae

PPI diagram of the DEGs. The size of each node represents the Edge count, and the width of the line represents the degree of interaction score, which was automatically scored by the database. The shape of the line represents the interaction relationship, and the node shape reflects the function of the genes in the KEGG and GO pathways.

Verification of the transcriptome of DEGs through qPCR

To verify the accuracy and reliability of the gene expression levels obtained from transcriptome sequencing, DEGs were randomly selected for qPCR verification and correlation analysis. Most of the 12 genes were well correlated (Figure 12). In addition, EAE98_000017 (K17103), involved in lipid transport, and EAE98_007802, involved in osmotic pressure regulation, were basically consistent with transcriptome sequencing results, confirming the reliability of transcriptome data.

Figure 12.

Validation of 12 differentially expressed genes identified from the Botrytis deweyae transcriptome by quantitative reverse-transcription PCR

Data are represented as mean ± SEM (n = 3 biological replicates), R² indicates the goodness of linear fit between gene expression levels (FPKM) and qPCR data.

Discussion

Several pathogens produce infection structures, such as infection pads, appressorium,²⁶ and other specialized formations, to invade their hosts²⁷ by penetrating the plant leaf epidermis or directly through wounds and stomata. B. cinerea invades plants by forming an infection pad, which accumulates cell wall-degrading enzymes, and softens plant cell tissue to facilitate invasion.¹¹ In this present, we integrated optical, SEM, and TEM to follow the entire penetration and developmental sequence of B. deweyae on P. cyrtonema leaves. Light microscopy clearly revealed stomatal entry, and SEM further confirmed this route while also showing that the pathogen preferentially invades through the epidermis or stomata; hyphal tips swell into infection cushions and secrete a mucilage-like matrix comparable to that described in oilseed rape,¹³ facilitating host penetration.²⁸ TEM showed that primary hyphae quickly penetrate epidermal cells, causing plasmolysis and wall swelling, then spread between cells. Subsequently, secondary hyphae move into neighboring mesophyll and vascular tissues, triggering host-cell death (Figure 3). This phenomenon of stomatal entry has also been observed in B. elliptica during the infection of tetraploid and diploid lily (Lilium rosthornii).²⁹ Thus, the infection strategy of B. deweyae mirrors the enzymatic mode reported for B. cinerea, but its cell impact on P. cyrtonema is now resolved at the ultrastructural level.

Lipase production is a critical component of the fungal cuticle-degrading arsenal, with pathways involved in triglyceride metabolism and phospholipid homeostasis contributing significantly to host invasion.³⁰ The infection process involves dynamic lipid exchanges between the pathogen and host, leading to alterations in fatty acid composition in both organisms upon fungal attachment,³¹ a process that can be initiated by re-modulating fatty acid uptake.³² For instance, the necrotrophic pathogen B. cinerea secretes the lipase Lip1, which is crucial for penetrating intact plant structures.³³ In this study, glyceride metabolism and TAG synthesis genes were upregulated. Of these, EAE98_001653 (K14674) was involved in the synthesis of various intermediates in the glyceride metabolism pathway, indicating that complex regulatory mechanisms were involved in the regulation of cell membrane homeostasis and substance transport. The GO term map showed that this process was closely related to lipid transport and sterol binding. KEGG enrichment analysis further revealed that genes related to lipid transport and sterol binding were enriched in the gene network map. Sterol, an important component of the cell membrane, maintains the integrity of the cell membrane, the activity of membrane-bound enzymes, membrane fluidity, and the transport of cell materials. It can also affect hyphal growth, cell differentiation, and conidia formation. The deletion of the sterol reductase gene MoErg4 in M. grisea leads to the loss of cell membrane integrity and destroys the formation of the lipid raft in the plasma membrane.³⁴ EAE98_008811 (K20456) participated in these metabolic pathways, indicating that sterol transport and lipid transport are necessary for maintaining cell membrane integrity and life activities in B. deweyae. However, the specific functions of EAE98_001653 and EAE98_008811 need to be further studied.

Furthermore, glycerol synthesis, a key metabolic process under stress, is linked to these pathways. Glycerol-3-phosphate, a direct precursor to glycerol, is produced from dihydroxyacetone phosphate (a glycolytic product) by glycerol-3-phosphate dehydrogenase (GPDH), encoded by the GPD1 gene. The expression of GPD1 is, in turn, regulated by the hypertonic glycerol (HOG) pathway under osmotic stress.³⁵ The central role of the HOG pathway, particularly the MAPK Hog1, in coordinating stress responses (e.g., to osmotic, oxidative, and heavy metal stress), morphogenesis, and virulence has been extensively elucidated in B. cinerea³⁶ and a number of other plant pathogens.³⁷ However, the genes involved in these critical pathways remain uncharacterized in the recently described pathogen B. deweyae.

The HOG signaling pathway falls under the MAPK cascade. The transmembrane proteins Sho1 and Sln1 act as biosensors of the eukaryotic HOG pathway and regulate the response of organisms to various extracellular stimuli. BcHpt is a core element of the HOG transduction pathway of B. cinerea. The BcHpt acetylation at the BcHpt lysine acetylation site (Lys161) significantly inhibits B. cinerea by reducing its growth rate and weakening its pathogenicity.³⁸ Sln1 in Magnaporthe grisea can sense swelling pressure, negatively regulate melanin and swelling pressure formation, and trigger appressorium repolarization, thereby affecting the pathogenicity of rice blast fungus.³⁹ In the present study, the KEGG pathway was significantly enriched in response to high osmotic pressure, and the sensing protein on the Sln1 biofilm was weakly expressed in response to the osmotic pressure produced by the host and regulated osmotic pressure through the cascade reaction. BcSkn7, a gene involved in oxidative stress in B. cinerea, also plays an important role in regulating trophic differentiation and various stress responses.⁴⁰ The redox process analysis showed that B. deweyae can tolerate the high osmotic pressure environment induced in P. cyrtonema leaves through the MAPK signaling cascade pathway, which may be initiated by antifungal active substances secreted by host cells or a high ROS concentration. Hog1 (EAE98_007802), which is expressed in B. deweyae, plays a crucial role in the pathogenesis.

In plants, a complex feedback loop exists between ROS and MAPK signaling: ROS can activate MAPK cascades, which in turn regulate ROS production and signaling. However, high levels of ROS become detrimental, necessitating a balanced antioxidant system involving enzymes such as SOD, CAT, and POD to maintain redox homeostasis.⁴¹ Pathogens, particularly biotrophic fungi, exploit this system by manipulating the host’s antioxidant enzymes to modify the redox environment, thereby suppressing defenses and facilitating colonization.^42,43 In this study, the significant enrichment of CAT and POD in B. deweyae suggests a similar strategy. These enzymes are crucial for scavenging ROS and preventing oxidative damage. This role is conserved in pathogens such as F. graminearum and M. grisea, where PODs eliminate host-derived ROS to promote infection.⁴⁴ Similarly, in B. cinerea, the peroxisome pathway (involving BcPex genes) is vital for stress tolerance and pathogenicity.⁴⁵ The enrichment of CAT and POD in B. deweyae likely facilitates the invasion of P. cyrtonema by scavenging excessive leaf ROS and resisting osmotic pressure, mirroring the function of secreted catalases in Puccinia striiformis that confer resistance to host oxidative stress. This indicates that B. deweyae employs a well-established pathogenic tactic of deploying antioxidant enzymes to neutralize plant defenses and establish infection.

ATP-binding cassette (ABC) transporters are ubiquitous proteins that utilize ATP to transport a wide array of substrates across membranes, including lipids and antifungal compounds.^46,47 While often studied in the context of multidrug resistance (PDR/MDR), Data from this study revealed a significant downregulation of ABC transporter-related genes in B. deweyae during early infection (24 hpi) of P. cyrtonema leaves. This suggests a potential role beyond efflux, possibly interfering with host immune signaling to facilitate infection. Concurrently, the significant enrichment of terms related to the “host cell nucleus” (cellular component) and “DNA-binding transcription factor activity” and “ATPase activity” (MF) points to a critical strategy of nuclear manipulation. Plant immunity is heavily reliant on precise transcriptional regulation, and successful pathogens such as M. oryzae secrete nuclear effectors to reprogram host defense genes by targeting transcription factors.^48,49 The enrichment of nucleus-associated genes in B. deweyae, which appear to originate from fungal effectors,⁵⁰ strongly suggests a similar tactic of subverting host transcription. Although this study identifies key pathways, future functional validation through proteomics and genetic approaches is essential to confirm their precise roles in virulence.

Collectively, this study elucidated the mechanisms of B. deweyae infection in P. cyrtonema, including the formation of infection cushions and invasion through stomata. Evidence suggests that B. deweyae responds to host-induced osmotic pressure through the high osmotic glycerol (HOG) signaling pathway and the MAPK cascade, with Hog1 playing an important role in pathogenesis. In addition, the antioxidant enzymes CAT and POD were implicated in scavenging ROS and resisting high osmotic pressure, whereas the upregulation of glyceride metabolism and TAG synthesis pathways maintained cell membrane integrity and facilitated substance transport.

Limitations of the study

This study provides insights into the genes actively expressed during the infection of P. cyrtonema by B. deweyae, yet it underscores the inherent limitations of the transcriptomic approach. A comprehensive understanding of the pathogen’s whole genome, particularly regarding complex developmental processes such as conidiation, necessitates a complete genomic blueprint. The inability of B. deweyae to sporulate readily in vitro, as observed in this and previous studies, points to potentially unique regulatory mechanisms. Therefore, securing a complete genome sequence for B. deweyae is a critical objective for future work.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to, and will be fulfilled by, the lead contact, Guoli Li, (20130198@cqtgmc.edu.cn).

Materials availability

All data associated with this study are present in the paper or the supplemental information and can also be obtained from the corresponding authors upon request.

Data and code availability

• •The datasets generated or analyzed in the current study are available in the NCBI database under the BioProject ID PRJNA1115085 (https://www.ncbi.nlm.nih.gov/sra/PRJNA1115085).
• •This study did not generate original code.
• •Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request.

Acknowledgments

This research was financially supported by the Science and Technology Research Program of Chongqing Municipal Education Commission (KJZD-K202201207 and KJZD-M202202701), the Project of Chongqing Science and Technology Bureau (CSTB2023TIAD-ZXX0014), the Innovative Team Project for Tumor Immunology Research (TIG202303), and the Chongqing Key Laboratory Project for the Development and Utilization of Genuine Medicinal Materials in the Three Gorges Reservoir Area (Sys202100074).

Author contributions

M.L. and W.L.M. designed the experiments and wrote this paper. W.L.M. analyzed the data. Z.N.X. and Y.X.L. performed qRT-PCR and measured hormone contents. T.T.G. planted and managed the seedlings of Polygonatum cyrtonema. Y.Y. and S.T.Z. contributed to the editing of the manuscript. G.L.L. and F.Q.Y. reviewed the manuscript critically.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Biological samples
Raw data of transcriptomic sequencing	NCBI BioSample	Accession: SRX27466050- SRX27466061
Critical commercial assays
Transcriptomic sequencing	Biomarker Technologies Co., LTD.	N/A
Deposited data
Raw data of transcriptomic sequencing	NCBI BioProject	SRA: PRJNA1115085
Experimental models: Organisms/strains
Botrytis deweyae strain1 4-2	College of Biological and Food Engineering	N/A
Software and algorithms
Transcriptome data analysis	BMKCloud (www.biocloud.net)	N/A
Graphical abstract	BioRender (https://www.biorender.com)	N/A
Network analysis (PPI)	Cytoscape (https://cytoscape.org/)	N/A
KEGG enrichment analysis	clusterProfiler 4.16.0 (R package)	N/A
Primers designed for qPCR	Primer Premier 6	N/A

Experimental model and study participant details

Plant and pathogen model

B. deweyae was originally isolated from the leaves of P. cyrtonema in an herb plantation located in Longju Town, Wanzhou District, Chongqing, China. In adddition, the tubers of P. cyrtonema were collected from a disease-free plantation of 3-year-old plants in Longju Town (30°38′1″N, 108°42′27″E).

P. cyrtonema was formally identified by Fu-Qiang Yin. A voucher specimen was deposited in the Herbarium of Chongqing Three Gorges University under the accession number CTGU20240956.

All necessary permissions and licenses to collect P. cyrtonema were obtained from the Forestry Bureau of Wanzhou District, Chongqing. The collection was performed in accordance with local regulations and guidelines.

Method details

Optical, scanning electron microscopy and transmission electron micrographs of pathogen penetration and development progress

The fungal inoculum consisted of mycelial plugs due to the consistent inability of B. deweyae to produce macroconidia in sufficient quantities under a range of tested culture conditions, a trait noted in previous studies.¹ B. deweyae was incubated at 25°C for 3 days, and a 5-mm diameter plugs was selected as inoculated. The leaves eaves of healthy P. cyrtonema were punctured along the symmetrical parts of the veins using the acupuncture method. The colony slice was turned upside down at the acupuncture sites, and the plants were then transferred to a greenhouse under the conditions of 95% ± 1% relative humidity, 25°C ± 1°C, and under alternating light and dark periods (12 h light and 12 h darkness). A sterile agar plug was used as the control. The leaves cut into 0.5 × 0.8-cm pieces at the junction of diseased and healthy tissues, following fixation for 24 h in a 1:1 (v/v) mixture of 95% ethanol and glacial acetic acid, leaf tissues were cleared in saturated chloral hydrate solution. The transparent tissues were then rinsed with water, stained with dilute aqueous aniline blue for 5–10 min, mounted in glycerol, and observed under a light microscope. For SEM, the prepared samples (0, 12, 18, 24, and 48 h post inoculation) placed on a stub, gold-coated in a sputter coater, and observed through Cryo-SEM (Hitachi SU3500, Tokyo, Japan). For TEM observation, the samples (0, 24, 48, and 96 h post inoculation) were embedded in pure EMBed 812 resin and polymerized in a 70°C oven for 12–48 h. Ultrathin sections of 70–90 nm thickness were prepared from the resin blocks using an ultramicrotome (Leica UC7). The sections were flattened with a fine brush, transferred onto copper grids, and examined under Hitachi HT-7800 (Hitachi, Tokyo, Japan).

Transcriptome samples preparation

Varieties of tubers with similar sizes were selected as materials and treated with 50% carbendazim to obtain disease-free seedlings. Danish peat soil (Pindstrup sphagnum) was used as the seedling substrate. The leaves were punctured evenly along the central vein and middle sections three times via acupuncture. The inoculation method is the same as above.

Samples were collected at 0, 24, 48, and 96 hpi. The fungi plugs were gently collected in a 1.5-mL centrifuge tube using sterile tweezers and immediately preserved in liquid nitrogen. Three samples were collected at each time point. The sample numbers at 0, 24, 48, and 96 hpi were Tbotry1, Tbotry2, Tbotry3, and Tbotry4, respectively.

RNA extraction, quantitative detection and transcriptome sequencing

Total RNA from B. deweyae was extracted using TRlzol Reagent (Life Technologies, California, USA). RNA concentration and purity were measured using NanoDrop 2000 (Thermo Fisher Scientific, Wilmington, Delaware, USA). RNA integrity was evaluated using an RNA Nano 6000 detection kit on an Agilent Bioanalyzer 2100 system (Agilent Technologies, California, USA). Once RNA quality was qualified, one portion was used for sequencing, whereas the remainder was used for quantitative polymerase chain reaction (qPCR).

PE150 mode sequencing was performed using the Illumina NovaSeq6000 sequencing platform (NovaSeq Xp 4-Lane Kit v1.5, Shanghai, China). High-quality clean reads were generated by aligning RNA-seq reads with a reference genome sequence (GeneBank ID SAMN41509756-SAMN41509767) using the Hisat2 tool.

Analysis of transcriptome data

The National Center for Biotechnology Information (NCBI) Non-redundant Protein Sequence, Pfam, Eukaryotic Orthologous Groups/Clusters of Orthologous Groups of Protein, Swiss-Prot (a manually annotated and reviewed protein sequence database), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Gene Ontology (GO) databases were used to annotate the sequence. Genes with significant differences in expression level across samples were classified as differentially expressed genes (DEGs). The gene sets obtained from differential expression analysis were called differentially expressed gene sets, and were labeled “A_vs_B,” where A represents the control group and B represents the experimental treatment group. The differentially expressed gene sets were grouped as shown in Table S1.

Based on the count value of the genes in each sample, differential analysis software was used to screen for DEGs. DESeq2 uses a model based on a negative binomial distribution to determine differential gene expression.⁵¹ The Benjamini–Hochberg procedure was used to obtain the corrected P-value by controlling the error detection rate.⁵² To identify DEGs, fold-change ≥2 (upregulated) or fold-change ≤0.5 (downregulated) and false discovery rate (FDR) < 0.01 were used as screening criteria.

In-depth data mining analysis was performed using BMKCloud (www.biocloud.net). GO enrichment analysis of the DEGs was performed using the clusterProfiler package based on the Wallenius non-central hypergeometric distribution,⁵³ with p < 0.05 as the threshold. The KEGG database (Kanehisa et al. 2011; http://www.kegg.jp), KOBAS database,⁵⁴ and clusterProfiler software were used to analyze the enrichment of DEGs in the KEGG pathway. The results of the differential expression analysis and the interaction pairs included in the database were combined. The sequences of the DEGs were aligned to the B. deweyae genome using the STRING database (http://string-db.org/) to obtain predicted protein–protein interactions (PPIs) and visualized using Cytoscape 3.10. 1.⁵⁵

Verification of pathogen transcriptome DEGs using qPCR

The DEGs selected from key pathway genes were verified by qPCR, and the qualified RNA extracted from the original materials was using a kit purchased from MightyScript Bioengineering (Shanghai) Co., Ltd. (MightyScript first strand cDNA synthesis Master Mix, B639251). The reverse transcription system was 20 μL, and the dosage of each RNA sample was 2 μL (total RNA, 2 μg). Subsequently, 2 × II M-MLV RT mix was added at 10 μL and topped up to 20 μL with RNase-free ddH₂O. The mixture was incubated at 25°C for 5 min and reverse-transcribed at 48°C for 30 min. Reverse transcriptase was inactivated at 85°C for 5 min. Immediately after the reaction, the mixture was stored at −20°C in a refrigerator for further analysis.

Primers were designed using Primer Premier 6 software, and the sequence is shown in Table S2. The actin and beta-tubulin sequences of the reference genome were used as the internal reference genes. The qPCR system was 20 μL. To this volume, 10 μL Master Premix green fluorescent dye (SGExcel FastSYBR, B532955, Sangon Biotech, Shanghai, China) was added. Subsequently, 0.3 μL of each upstream and downstream primers and 2 μL of cDNA were added, and the final volume was topped up to 20 μL with RNase-free ddH₂O. All samples were prepared in triplicate. Real-time fluorescent qPCR was performed on a Gentier 96E 96-well real-time qPCR instrument (Xi’an Tianlong Technology Co., Ltd. Shaanxi, Xi’an).

The one-step real-time fluorescence qPCR program was as follows: pre-denaturation at 95°C for 15 s; reaction at 95°C for 5 s; and reaction at 58°C for 30 s for 38 cycles. Dissolution was performed at 95°C for 15 s, and then at 60°C for 1 min. When the Threshold cycle (Ct) value was between 15 and 25 cycle, the dissolution curve showed a single peak. When the melting temperature (TM) value was 80°C–90°C, the target gene is considered to be successfully amplified. After obtaining the data through Gentier 96E, the experimental data were preliminarily collated using the built-in software in Gentier 96E, and the relative expression value was obtained using the 2^{−Δ Δ Ct} method. The formula is presented below, and GraphPad Prism 8.0 was used to create the graphs.

Δ Ct = (target gene Ct in the experimental group − internal reference gene Ct in the experimental group) × (target gene Ct in the control group − internal reference gene Ct in the control group).

Quantification and statistical analysis

All statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software), R version 4.2.1, and BMKCloud (www.biocloud.net) for auxiliary bioinformatics analysis. Data are presented as mean ± SEM unless otherwise stated. All experiments were performed with n = 3 biological replicates. For GO and KEGG enrichment analysis, Benjamini–Hochberg-adjusted p-values (q-values) ≤ 0.05 were considered statistically significant. Detailed statistical parameters for each experiment, including the statistical test used, exact n values, and dispersion measures, are provided in the corresponding figure legends or STAR Methods section.