Veh1 modulates Ca²⁺ dynamics through VdMid1 and VdMcu1 to control hyphopodium-mediated infection in Verticillium dahliae

Abstract

Verticillium wilt, primarily caused by the fungus Verticillium dahliae, results in substantial crop losses and ecological damage worldwide. During infection, V. dahliae forms a specialized infection structure called the hyphopodium, which penetrates host root cells. Although calcium (Ca²⁺) influx into hyphopodia is essential for hyphopodium-mediated root penetration and subsequent colonization, the mechanism that maintains Ca²⁺ homeostasis in hyphopodia during pathogenesis remains unknown. Here, we show that the transcription factor Veh1 negatively regulates hyphopodium-mediated root penetration by modulating Ca²⁺ homeostasis during plant infection. Deletion of Veh1 increased hyphopodium formation, elevated Ca²⁺ accumulation in hyphopodia, activated VdCrz1-mediated Ca²⁺ signaling, and enhanced virulence. We further demonstrated that Veh1 directly regulates VdMid1, which encodes a putative Ca²⁺ influx component, and VdMcu1, which encodes a mitochondrial Ca²⁺ uniporter that promotes mitochondrial Ca²⁺ sequestration. Through this opposing regulation, Veh1 coordinates cytosolic Ca²⁺ dynamics during hyphopodium-mediated infection. Indeed, deletion of VdMid1 reduced Ca²⁺ accumulation in hyphopodia and attenuated virulence, whereas deletion of VdMcu1 had the opposite effects. Together, these results uncover a Veh1–VdMid1/VdMcu1 regulatory circuit that maintains Ca²⁺ homeostasis during hyphopodium-mediated plant penetration and provide mechanistic insights into fungal pathogenesis.

1. Introduction

Verticillium dahliae is a notorious phytopathogenic fungus that causes wilt disease in over 200 hosts, including economically important agricultural crops and trees [[1], [2], [3]]. V. dahliae is a typical soil-borne pathogen that initiates host infection at the roots and subsequently colonizes the xylem, enabling systemic spread throughout the host plant [4]. To initiate infection, the distal end of the V. dahliae germ tube produces a specialized infection structure called the hyphopodium [5]. Hyphopodium development is mediated by reactive oxygen species (ROS) and calcium (Ca²⁺) signaling, involving the NADPH oxidase subunit VdNoxB and the tetraspanin VdPls1; the VdSte11–VdSte7–VdKss1 mitogen-activated protein kinase (MAPK) cascade; and cellophane surface-induced 1 (VdCSIN1)-mediated cyclic AMP (cAMP) signaling [[6], [7], [8], [9]]. However, the regulatory mechanism underlying hyphopodium-mediated plant penetration is unclear.

Ca²⁺ is a ubiquitous secondary messenger that regulates multiple aspects of biology in eukaryotes [10]. Within fungal pathogens, the polar growth of specialized infection structures is underpinned by a tightly localized Ca²⁺ gradient at the hyphal apex. This Ca²⁺ gradient is required to sustain the polarized secretion and cytoskeletal dynamics that drive invasive growth [11]. In Magnaporthe oryzae, Ca²⁺ does not affect spore germination but specifically influences appressorium formation, indicating a stage-specific requirement for Ca²⁺ at this developmental stage [12]. Intracellular Ca²⁺ imaging showed that Ca²⁺ levels increased at 6 h during appressorium formation and declined at 8 h, suggesting that the formation of the infection structure is tightly associated with the spatiotemporal regulation of Ca²⁺ dynamics [12]. Ca²⁺ treatment in Phyllosticta ampelicida promotes appressorium formation at low concentrations but inhibits it at high concentrations, further supporting the requirement for precisely controlled Ca²⁺ levels for infection-structure formation [13]. In V. dahliae, a high Ca²⁺ gradient at the tip is observed during the formation of penetration pegs from hyphopodia. Indeed, the VdNoxB and VdPls1 mutants, which are unable to accumulate Ca²⁺, fail to develop penetration pegs [6,8]. However, how V. dahliae fine-tunes cytosolic Ca²⁺ levels during pathogenesis is unknown.

Precise regulation of intracellular free Ca²⁺ requires rapid Ca²⁺ influx to meet nutritional and signaling demands, coupled with immediate clearance or chelation to prevent the accumulation of toxic Ca²⁺ levels [10]. Cellular Ca²⁺ is distributed among organelles, including the endoplasmic reticulum (ER), mitochondria, Golgi apparatus, and vacuoles, forming an integrated system that buffers against cytosolic Ca²⁺ stress. A variety of channels and exchangers located in the membranes of these subcellular structures mediate Ca²⁺ transport. Mitochondria function as key intracellular Ca²⁺ reservoirs, with matrix Ca²⁺ concentrations tightly regulated to maintain cellular homeostasis [14]. A primary mechanism for mitochondrial Ca²⁺ influx is the mitochondrial calcium uniporter complex (MCUc), which is essential for Ca²⁺ homeostasis [15]. This complex consists of pore-forming MCU subunits, Essential MCU Regulator (EMRE), and the regulatory factors Mitochondrial Calcium Uptake 1 (MICU1) and MICU2, which together regulate Ca²⁺ uptake. When cytosolic Ca²⁺ increases, clearance mechanisms, including MCUc-mediated uptake, are activated [16]. Another major Ca²⁺ influx pathway is mediated by voltage-gated Ca²⁺ channels (VGCCs). Budding yeast (Saccharomyces cerevisiae) possesses a functionally similar channel that consists of Calcium channel homolog 1 (Cch1p) and Mating pheromone-induced death 1 (Mid1p), where Cch1p activity depends entirely on Mid1p [17,18]. Mid1p localizes to both the plasma membrane and ER membranes [19]. This functional conservation extends to fungal pathogens: in Metarhizium acridum, Mid1 regulates Ca²⁺ transmission, as the deletion of MaMid1 lowered the intracellular Ca²⁺ concentration, impairing appressorium formation and host-cuticle penetration [20]. Although the MCUc and Cch1/Mid1 Ca²⁺ transport components are essential for cellular Ca²⁺ homeostasis, the precise mechanism by which V. dahliae orchestrates and balances these Ca²⁺ dynamics during hyphopodia formation remains unknown.

In this study, we identified the Verticillium enhanced hyphopodium 1 (Veh1) gene, which when deleted led to increased virulence and hyphopodium formation. Veh1 deletion also caused aberrant cytosolic Ca²⁺ accumulation in hyphopodia and activated the VdCrz1 signaling pathway. Veh1 negatively regulates Ca²⁺ accumulation in hyphopodia by directly targeting genes encoding two key Ca²⁺ transport components, VdMid1 and VdMcu1. Specifically, Veh1 represses VdMid1 expression while promoting VdMcu1 expression, thereby maintaining Ca²⁺ homeostasis. Together, these observations define the Veh1–VdMid1/VdMcu1 regulatory circuit as a key modulator of Ca²⁺-mediated plant penetration by V. dahliae hyphopodia.

2. Results

2.1. Veh1 negatively regulates hyphopodium-mediated penetration

To identify key regulatory elements involved in the early stages of V. dahliae infection, we focused on transcription factors, which act as master regulators of gene expression during fungal development and differentiation [21]. By screening a collection of transcription factor deletion mutants using a cellophane membrane assay, we identified one mutant carrying a targeted deletion of VDAG_10210 that exhibited rapid penetration (Fig. S1A). Specifically, the wild-type strain XS11 was able to penetrate the cellophane membrane at 36 h post-inoculation (hpi) and continued to grow following cellophane removal (Fig. 1A). By contrast, the ΔVDAG_10210 mutant displayed accelerated penetration as early as 30 hpi (Fig. 1B). Furthermore, ΔVDAG_10210 showed more extensive colonization beneath the membrane than XS11 (Fig. 1C). To investigate the basis for this enhanced penetration, we examined the formation of infection structures in both the wild-type and mutant strains. Compared to XS11, ΔVDAG_10210 differentiated denser hyphopodia at the hyphal tips (Fig. 1D). When examined under a microscope in the same field of view, ΔVDAG_10210 produced significantly more hyphopodium-derived penetration pegs than XS11 (Fig. 1D and E). These results suggest that the deletion of VDAG_10210 promotes the formation of hyphopodium-mediated penetration pegs in V. dahliae. We therefore named this gene Verticillium enhanced hyphopodium 1 (Veh1). Veh1 encodes a 328-amino acid protein containing a basic leucine zipper (bZIP) domain and is conserved in Verticillium and Colletotrichum species (Fig. S1C and D).

Fig. 1.

Deletion of Veh1 promotes hyphopodium-mediated penetration. A-B Penetration capacity of the wild-type strain XS11 (A) and ΔVeh1 (B) examined using a cellophane assay. A 2-mm mycelium block was placed on minimal medium (MM) covered with cellophane, incubated at 25 °C, and photographed at different time points (above, scale bar, 2 mm). The cellophane was removed, and the cultures were incubated for 5 more days and photographed (below, scale bar, 5 mm). C Penetration diameter of XS11 and ΔVeh1 colonies after penetrating the cellophane. Data are presented as means ± standard deviation (SD) from three biological replicates. Significant differences were determined by two-way ANOVA followed by Sidak's multiple comparisons test (∗P < 0.05; ∗∗∗P < 0.001; ns, not significant). D Microscopy images showing hyphopodia and penetration pegs produced by XS11, ΔVeh1, and the complemented strain ΔVeh1-C on cellophane at 72 h post inoculation (hpi). The asterisks represent the hyphopodia, and the arrows represent narrow penetration pegs formed by the expanding ends of hyphae at 100× magnification. Scale bars, 30 μm in 40× and 10 μm in 100×. E Number of penetration pegs on cellophane from the indicated genotypes at 72 hpi. The numbers were determined by ImageJ using default settings at 40× magnification. Boxplots show the median (line), 25th–75th percentiles (box), and minimum–maximum range (whiskers) from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗P < 0.01; ns, not significant). F Relative Veh1 expression levels during cellophane infection in MM by V. dahliae at the indicated time points. Data are presented as means ± standard deviation (SD) from three biological replicates.

We generated the complementation strain ΔVeh1/Veh1 (ΔVeh1-C) and determined that its penetration level was restored to that of XS11, supporting the genetic complementation of the mutant by intact Veh1 (Fig. 1D, E, S1B and S2B). These results suggest that Veh1 negatively regulates penetration and hyphopodium formation. We then investigated Veh1 expression during V. dahliae infection. Veh1 was significantly induced from 24 to 72 h post-inoculation (hpi), peaking at 48 hpi (Fig. 1F), suggesting that Veh1 functions mainly during later stages of infection.

2.2. Veh1 negatively regulates virulence in smoke tree and Nicotiana benthamiana

We then investigated whether Veh1 affects hyphal growth and virulence. Compared to XS11 and the complemented strain ΔVeh1-C, the ΔVeh1 strains (ΔVeh1-31 and ΔVeh1-33) showed greater radial growth on both potato dextrose agar (PDA) and complete medium (CM) (Fig. 2A and B). Smoke tree (Cotinus coggygria) seedlings inoculated with ΔVeh1 developed more severe disease symptoms than those inoculated with the other strains, including accelerated leaf necrosis and pronounced vascular browning in stems (Fig. 2C and E). Disease index analysis confirmed that the ΔVeh1-inoculated plants (52.2–55.6%) had higher infection rates than the XS11-inoculated plants (41.1–43.3%, P < 0.01) (Fig. 2D). The successful re-isolation and cultivation of V. dahliae mycelia from necrotic roots or stems further confirmed that each strain successfully colonized smoke trees and caused Verticillium wilt (Fig. 2F). Fungal biomass showed a corresponding increase, rising more than 7.4-fold and 11.2-fold at 3 dpi and 4 dpi, respectively, in plants inoculated with the ΔVeh1 strain (Fig. 2G). This hypervirulent phenotype persisted through late phases of infection (Fig. S3A).

Fig. 2.

Deletion of Veh1 enhances virulence. A Growth of the XS11, ΔVeh1, and ΔVeh1-C strains on potato dextrose agar (PDA) or complete medium (CM) plates, photographed after 10 days of culture. Scale bar, 1 cm. B Colony diameter of each strain at the indicated time points. Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Dunnett's multiple comparisons test (∗∗∗P < 0.001; ns, not significant). C Disease symptoms on smoke tree seedlings inoculated with the indicated strains and uninoculated seedlings as control (CK). Images were taken at 35 days post-inoculation (dpi). Scale bars, 10 cm. D Disease indices of infected smoke tree seedlings described in (C). Violin plots show the median (dashed lines) and quartiles (dotted lines) from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗P < 0.05; ns, not significant). E Cross-sections and longitudinal sections of rootstock junctions from smoke tree seedlings inoculated with each strain and control uninoculated seedlings. Scale bars, 0.5 cm. F Re-isolation of each strain from the stems of infected smoke tree seedlings after incubation for 3 days and control uninoculated seedlings. Scale bar, 0.5 cm. G Fungal biomass from the XS11 and ΔVeh1 strains on smoke tree seedlings as determined by quantitative PCR (qPCR). Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Sidak's multiple comparisons test (∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001). H Disease symptoms of N. benthamiana plants inoculated with the different strains. Images were taken at 15 dpi. Scale bars, 5 cm. I Disease symptom categories of plants described in (H). The bars represent the frequency of plants within each disease index class. J Relative fungal biomass of each strain in N. benthamiana plants as determined by qPCR. Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Tukey's multiple comparisons test (∗∗P < 0.01; ∗∗∗P < 0.001; ns, not significant). K Survival curve of N. benthamiana plants upon V. dahliae infection with the indicated strains and control uninfected plants (CK). Data are presented as means ± SD from three biological replicates. Significant differences were determined by the log-rank (Mantel–Cox) test (∗∗∗P < 0.001; ns = not significant).

We examined the virulence of the ΔVeh1 strains on Nicotiana benthamiana seedlings. Consistent with the above observations, the two ΔVeh1 strains (#31 and #33) induced more severe disease in 4-week-old plants than either XS11 or ΔVeh1-C at 15 dpi (Fig. 2H). Plants inoculated with each mutant strain (#31 and #33) developed more severe chlorosis and wilting symptoms, which were quantitatively confirmed by disease index scoring (Fig. 2I). These differences between the mutant strains and the wild type and complemented strains persisted at 35 dpi, as plants inoculated with the ΔVeh1 strains exhibited the most severe symptoms (Fig. S3B) and highest disease indices (Fig. S3C). Quantitative PCR analysis using the V. dahliae internal transcribed spacer (5.8S-ITS) region and N. benthamiana GAPDH as the reference gene [22] revealed significantly higher fungal biomass in the stems of plants inoculated with the ΔVeh1 strains at 15 dpi and 35 dpi (Fig. 2J). Mortality rates were higher in plants infected with the ΔVeh1 strains than in plants infected with XS11 or ΔVeh1-C (Fig. 2K). Collectively, the hyperpenetration and hypervirulence phenotypes of the ΔVeh1 strains suggest that Veh1 inhibits a signaling pathway required for infection structure development.

2.3. Veh1 regulates cytosolic Ca²⁺ homeostasis in the hyphopodium

To explore how deletion of Veh1 enhances hyphopodium formation and virulence, we performed RNA-seq on ΔVeh1 and the wild-type strain XS11 grown on cellophane-covered minimal medium (MM) for 72 h. We identified 544 differentially expressed genes (DEGs; Fig. 3A and S4A, Table S1), which were strongly enriched for gene ontology (GO) terms related to transporter activity, binding, and catalytic activity (Fig. S4B). Notably, GO enrichment analysis revealed that these DEGs were significantly enriched for terms related to ion transport, particularly Ca²⁺ binding (P < 0.01) (Fig. 3B, Table S1). Among these, expression of VDAG_07555, which encodes a vacuolar Ca²⁺ ATPase that depletes cytosolic Ca²⁺, was significantly downregulated in ΔVeh1 (Fig. 3C). GO enrichment analysis of terms in the Biological Process category, along with a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, further highlighted metal ion homeostasis and transmembrane transport pathways (Fig. S4C–E). These results indicate that Veh1 modulates ion transport, with a specific influence on Ca²⁺-associated genes during hyphopodium formation.

Fig. 3.

Deletion of Veh1 raises cytosolic Ca²⁺ levels in the hyphopodium. A Hierarchical clustering analysis of upregulated and downregulated differentially expressed genes (DEGs) in ΔVeh1 relative to XS11. The color scale from blue to orange represents row-scaled Z-scores of gene expression levels (TPM), with orange indicating high expression and blue indicating low expression. B Gene Ontology (GO) term enrichment analysis of DEGs (Molecular Function category). C Heatmap showing the expression levels of the indicated DEGs encoding calcium ion–binding proteins according to TPM based on the RNA-seq data (|log₂(fold-change)| > 1, P-adjust <0.05). D Detection of Ca²⁺ levels during cellophane infection in MM by V. dahliae at different times post inoculation. This ratio is proportional to the cytosolic free Ca²⁺ concentration. E ΔVeh1 shows greater sensitivity to Ca²⁺ treatment than XS11. Scale bar, 1 cm. F Percentage of mycelial growth inhibition at 10 dpi of the colonies described in (E). Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Tukey's multiple comparisons test (∗∗∗P < 0.001; ns, not significant). G Ca²⁺ imaging using Fluo-4 AM. Fluorescence was detected using confocal microscopy. Scale bar, 25 μm. H Quantification of relative fluorescence intensity from the Ca²⁺ probe in XS11, ΔVeh1, and ΔVeh1-C described in (G). Data are presented as means ± SD from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (ns, not significant). I Relative Ca²⁺ contents in ΔVeh1 compared with XS11 and ΔVeh1-C during hyphopodia formation at 72 hpi. Data are presented as means ± SD from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗∗P < 0.001; ns, not significant). J Subcellular localization of Veh1-GFP in the V. dahliae ΔVeh1 strain in response to Ca²⁺ treatment. Scale bars, 10 μm. K Penetration ability of XS11 and ΔVeh1 into MM covered with cellophane under 1 mM EGTA treatment at 48 or 72 hpi. Scale bars, 2 mm (above), 5 mm (below). L Relative VdCrz1 expression levels in XS11 and ΔVeh1 examined by RT-qPCR. Data are presented as means ± SD from three biological replicates. Significant differences were determined by Student's t-test (∗∗∗P < 0.001, ns, not significant). M Heatmap showing the expression levels (reported in TPM) of VdCrz1 target genes based on the RNA-seq data.

Ca²⁺ plays an essential role in appressorium formation and undergoes a transient increase, followed by a decline, during this stage [12]. Ca²⁺ accumulation has also been observed in V. dahliae hyphopodia [6,8]; however, whether Ca²⁺ levels change dynamically during hyphopodium formation, and how such Ca²⁺ dynamics are regulated, remain unknown. Given that Veh1 controls the expression of genes involved in Ca²⁺ transport and binding, we examined Ca²⁺ dynamics during hyphopodium formation using the Ca²⁺ probe Fura-2AM. Ca²⁺ levels exhibited a characteristic pattern: an initial increase, peaking at 48 hpi, followed by a gradual decline (Fig. 3D), closely paralleling the expression profile of Veh1. To determine whether Veh1 regulates Ca²⁺ dynamics during hyphopodium formation, we next tested whether Veh1 responds specifically to Ca²⁺. Because ROS production from the base of the hyphopodium in V. dahliae promotes Ca²⁺ influx and affects the Ca²⁺ concentration in the hyphopodium [6], we also examined the response of Veh1 to ROS. ΔVeh1 was hypersensitive to Ca²⁺ but not H₂O₂ treatment (Fig. 3E, F, S5A and B), indicating that Veh1 may not be involved in ROS-mediated hyphopodium penetration. However, analysis using a Ca²⁺-sensitive fluorescent probe did not reveal a significant difference in Ca²⁺-dependent fluorescence intensity between XS11 and ΔVeh1 hyphae (Fig. 3G and H). Collectively, these results suggest that Veh1 responds specifically to Ca²⁺ and regulates Ca²⁺ dynamics in the hyphopodium.

Given the decrease in Ca²⁺ levels at later stages of infection (Fig. 3D) and the concurrent high Veh1 expression (Fig. 1F), we reasoned that Veh1 might limit Ca²⁺ accumulation in the hyphopodium. Analysis of hyphopodia collected at 72 hpi from each strain confirmed that ΔVeh1 accumulates significantly more Ca²⁺ than XS11 and ΔVeh1-C (Fig. 3I). Veh1 expression was upregulated further upon Ca²⁺ stimulation (Fig. S5C), consistent with its infection-induced expression. In addition, in a subcellular localization assay, Veh1-GFP exhibited a diffuse distribution throughout the cytoplasm and nucleus under normal conditions. However, upon Ca²⁺ treatment, Veh1-GFP translocated into the nucleus (Fig. 3J). Thus, Veh1 is a Ca²⁺-responsive transcription factor that negatively regulates Ca²⁺ accumulation during later stages of infection by regulating the expression of genes encoding proteins involved in Ca²⁺ transport.

To investigate whether Veh1-dependent regulation of Ca²⁺ levels is required for plant penetration, we treated hyphopodia with the Ca²⁺ chelator EGTA. Under 1 mM EGTA treatment, fungal growth was unaffected (Fig. S6A and B), but the penetration of XS11 was delayed from 48 hpi to 72 hpi, whereas ΔVeh1 had penetrated the cellophane layer at both time points (Fig. 3K). However, higher EGTA concentrations completely blocked hyphopodium formation (Fig. S6C). These results indicate that Ca²⁺ availability is required for penetration and that loss of Veh1 partially bypasses the inhibitory effect of limited extracellular Ca²⁺.

Elevated Ca²⁺ levels in hyphopodia activate Crz1-mediated signaling to promote penetration peg formation [6]. We therefore examined whether deleting Veh1 would alter VdCrz1 signaling. VdCrz1 expression was significantly upregulated in ΔVeh1 during hyphopodium induction (Fig. 3L). Among downstream Ca²⁺ transporter genes regulated by VdCrz1 [23], the expression of the vacuolar Ca²⁺-transporting ATPase VdPMC1 and Golgi Ca²⁺/Mn²⁺-transporting ATPase VdPMR1 was downregulated, whereas that of vacuolar transient receptor potential (TRP) Ca²⁺ channel VdYVC1 and the ER P-type Ca²⁺-transporting ATPase VdCOD1 was upregulated in ΔVeh1 relative to XS11 (Fig. 3M and S6D). Together, these results indicate that Veh1 restricts Ca²⁺ accumulation in the hyphopodium, thereby suppressing Crz1 signaling.

2.4. Veh1 fine-tunes Ca²⁺ homeostasis in the hyphopodium via VdMid1 and VdMcu1

To determine how Veh1 transcriptionally regulates Ca²⁺ homeostasis, we used chromatin immunoprecipitation sequencing (ChIP-seq) to identify the direct target genes of Veh1. Using two independent biological replicates of the ΔVeh1/Veh1-GFP strain, we identified 5462 Veh1-IP peaks across the genome, associated with 1244 genes (Fig. 4A and B, Table S2). The peaks were predominantly located near transcription start sites (TSSs, 31.6%), within exons (30.3%), promoter regions (from −1 kb to the TSS, 23.1%), intergenic regions (10.4%), and introns (Fig. 4C). Based on the ChIP-seq data, we identified a highly enriched cis-element containing the 5′-ACGT-3′ core motif targeted by bZIP transcription factors, as well as several over-represented motifs targeted by Veh1 (Fig. 4D and S7). Integrating these results with the RNA-seq datasets identified 79 genes that are both predicted to be bound by Veh1 and differentially expressed (Fig. 4E, Table S3). GO term enrichment analysis of these candidate genes highlighted VdMcu1 (VDAG_03789), which is associated with mitochondrial Ca²⁺ homeostasis and Ca²⁺ channel activity (Fig. S8A, Table S4). Further examination of Veh1-bound genes also identified VdMid1 (VDAG_01399), encoding a Ca²⁺ channel (Fig. S8B, Table S4). BLASTP searches confirmed that VdMid1 is a homolog of the yeast Ca²⁺ influx-promoting protein Mid1p. InterPro analysis revealed a conserved MID domain (IPR024338) in VdMid1, which is also present in Mid1p, a plasma-membrane protein that mediates α-factor-induced Ca²⁺ influx during mating (Fig. S8C). Likewise, VdMcu1 contains a conserved C-terminal mitochondrial calcium uniporter domain (IPR006769), which is characteristic of inner-mitochondrial membrane transporters that mediate Ca²⁺ uptake (Fig. S8D).

Fig. 4.

Veh1 controls Ca²⁺ homeostasis by directly regulating VdMid1 and VdMcu1 expression. A-B Metaplots (top) and heatmaps (bottom) of ChIP-seq signals for the Veh1 immunoprecipitation (Veh1-IP; A) and Input (B) samples across 2-kb windows showing enrichment of Veh1 near the transcription start site (TSS). C Distribution of Veh1-binding sites to various genic and intergenic features of the V. dahliae genome. D Sequence logo showing the enriched Veh1-binding motif identified by ChIP-seq. E Venn diagram showing the overlap between the RNA-seq and ChIP-seq data, defining putative direct Veh1 target genes. F-G Integrative Genome Viewer (IGV) windows showing the peaks from the Veh1-IP sample along the promoters of VdMid1 (F) and VdMcu1 (G) identified by ChIP-seq. P1 indicates the region examined by ChIP-qPCR. H ChIP-qPCR showing the enrichment of Veh1 at the promoters of VdMid1 and VdMcu1. A rabbit IgG was used as a negative control. The ChIP signals are shown as percentages of the input. Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Sidak's multiple comparisons test (∗∗∗P < 0.001; ns, not significant). I Dual luciferase assay showing the transcriptional activity of Veh1 toward the LUC reporter gene driven by the VdMid1 or VdMcu1 promoter. The pGreenII 62-SK empty vector was used as the negative control. Data are presented as means ± SD from three biological replicates. Significant differences were determined by Student's t-test (∗P < 0.05; ns, not significant). J Relative expression levels of VdMid1 and VdMcu1 in XS11 and ΔVeh1 examined by RT-qPCR. Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Sidak's multiple comparisons test (∗∗P < 0.01).

ChIP-seq revealed distinct Veh1-binding peaks in the promoter regions of both genes: a 348-bp-wide peak at the VdMid1 promoter and a 378-bp-wide peak at the VdMcu1 promoter (Fig. 4F and G). Within the peak regions of VdMid1 and VdMcu1, we identified a core 5′-ACGT-3′ motif for VdMid1 and the G-box motif 5′-CACGTG-3′ for VdMcu1 (Fig. S9). To validate these interactions, we performed ChIP-qPCR assays on the ΔVeh1/Veh1-GFP strain using an anti-GFP antibody for immunoprecipitation, with IgG as a negative control. Compared to the IgG control, Veh1-GFP was significantly enriched at both promoters (Fig. 4H). In an electrophoretic mobility shift assay (EMSA), Veh1 directly bound to the 5′-ACGT-3′ elements in the promoter regions of VdMid1 and VdMcu1 (Fig. S9). To investigate the effects of Veh1 on the transcriptional activity of the target promoters, we performed a dual luciferase reporter assay in N. benthamiana leaves. In contrast to the empty vector control, co-expression with Veh1 suppressed the firefly luciferase (LUC) activity derived from the VdMid1 promoter (VdMid1pro:LUC) but elevated the LUC activity driven by the VdMcu1 promoter (VdMcu1pro:LUC). These effects were abolished when the 5′-ACGT-3′ sequence was mutated to 5′-AAAA-3' (Fig. 4I). In agreement with these results, VdMid1 expression was elevated and VdMcu1 expression was lower in the ΔVeh1 mutant compared to the wild-type strain XS11 (Fig. 4J). Together, these results demonstrate that VdMid1 and VdMcu1 are direct targets of Veh1 in V. dahliae and that Veh1 may fine-tune Ca²⁺ homeostasis in vivo by coordinating the expression of genes involved in Ca²⁺ influx and mitochondrial sequestration pathways.

2.5. VdMid1 and VdMcu1 play opposite roles in Ca²⁺ homeostasis, plant penetration, and virulence

To elucidate the roles of VdMid1 and VdMcu1 in cytosolic Ca²⁺ homeostasis, plant penetration, and virulence, we generated single knockout and complementation strains for the corresponding genes (Fig. S8). The ΔVdMid1 mutant exhibited slower vegetative growth compared with the wild type (Fig. 5A). To investigate the roles of VdMid1 and VdMcu1 in regulating Ca²⁺ levels, we examined the responses of each mutant and their complemented strains to elevated Ca²⁺ concentrations. ΔVdMid1 was less sensitive to Ca²⁺ treatment, whereas ΔVdMcu1 showed greater sensitivity to the same treatment, as determined by the extent of growth inhibition in response to Ca²⁺ present in the plates (Fig. 5B and S9A). We assessed intracellular Ca²⁺ levels using fluorescence staining. ΔVdMid1 hyphae showed a lower Ca²⁺-dependent fluorescence signal than those of XS11 and the complemented strain ΔVdMid1-C, whereas ΔVdMcu1 hyphae showed elevated Ca²⁺ accumulation compared to the corresponding controls (Fig. 5C). In agreement with this observation, fluorescence intensity from the Ca²⁺ probe was lower in ΔVdMid1 but higher in ΔVdMcu1 compared to the controls (Fig. 5D). These results demonstrate that VdMid1 and VdMcu1 antagonistically regulate cytosolic Ca²⁺ levels.

Fig. 5.

VdMid1 and VdMcu1 play opposing roles in cytosolic Ca²⁺ accumulation and penetration. A Photographs of the wild-type strain XS11, the mutants ΔVdMid1 and ΔVdMcu1, and their complemented strains, after growth on PDA plates for 10 days. Scale bar, 1 cm. B ΔVdMid1 is less sensitive to Ca²⁺ treatment, whereas ΔVdMcu1 is more sensitive to Ca²⁺. The percentage of mycelial inhibition was calculated at 10 dpi. Data are presented as means ± SD from three biological replicates. Significant differences were determined by two-way ANOVA followed by Sidak's multiple comparisons test (∗∗P < 0.01; ∗∗∗P < 0.001; ns, not significant). C Ca²⁺ imaging analysis of each strain using Fluo-4 AM. Fluorescence was detected using confocal microscopy. Scale bar, 25 μm. D Relative fluorescence intensity of the Ca²⁺ probe of the strains described in (C). Data are presented as means ± SD from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗∗P < 0.001; ns, not significant). E-F Penetration ability of ΔVdMid1 (E) and ΔVdMcu1 (F) assayed using a cellophane assay. Scale bars, 2 mm (above), 5 mm (below). G Microscopy images showing the hyphopodia and penetration pegs produced by each strain on cellophane. The asterisks indicate the hyphopodia, and the arrows indicate narrow penetration pegs formed at the expanding ends of hyphae at 100× magnification. Scale bar, 20 μm. H Number of penetration pegs on cellophane. The numbers were determined by ImageJ under default settings at 40× magnification. Boxplots show the median (line), 25th–75th percentiles (box), and minimum–maximum range (whiskers) from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗P < 0.01; ∗∗∗P < 0.001; ns, not significant). I Relative Ca²⁺ contents in each strain during hyphopodium formation. Data are presented as means ± SD from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗P < 0.01; ∗∗∗P < 0.001; ns, not significant). J-K Expression analysis of VdMid1 (J) and VdMcu1 (K) during cellophane infection by V. dahliae at the indicated time points.

To explore their roles in plant penetration, we conducted cellophane assays. ΔVdMid1 failed to penetrate the cellophane until 48 hpi, indicating delayed infection (Fig. 5E). Conversely, ΔVdMcu1 exhibited accelerated penetration, beginning at 30 hpi (Fig. 5F). ΔVdMid1 produced significantly fewer penetration pegs, whereas ΔVdMcu1 formed more pegs than XS11 and the complemented strains (Fig. 5G and H). To directly link these phenotypes to the regulation of Ca²⁺ levels, we measured relative Ca²⁺-dependent fluorescence in hyphopodia. ΔVdMid1 hyphopodia contained lower Ca²⁺ levels, whereas ΔVdMcu1 hyphopodia contained higher Ca²⁺ levels compared to the controls (Fig. 5I). Furthermore, both genes were specifically induced during V. dahliae infection (Fig. 5J and K). Collectively, these observations demonstrate that VdMid1 and VdMcu1 play opposite roles in regulating Ca²⁺ accumulation in hyphopodia and modulating fungal penetration.

To determine the contributions of VdMid1 and VdMcu1 to virulence, we performed pathogenicity assays on smoke tree (C. coggygria) seedlings. Plants infected with ΔVdMid1 retained more green leaves and developed milder Verticillium wilt symptoms than those infected with XS11 or ΔVdMid1-C (Fig. 6A). Conversely, plants infected with ΔVdMcu1 showed more severe disease symptoms, including pronounced leaf abscission and wilting, than those infected with XS11 or ΔVdMcu1-C (Fig. 6A). Disease index analysis confirmed the significantly reduced disease severity in ΔVdMid1-infected plants and increased disease severity in ΔVdMcu1-infected plants relative to the respective controls (Fig. 6B). Histological analysis revealed weaker vascular browning in ΔVdMid1-infected plants and darker browning in ΔVdMcu1-infected plants compared with controls (Fig. 6C). Re-isolation tests confirmed successful colonization by all strains (Fig. 6D). Quantification of fungal biomass revealed lower fungal loads in ΔVdMid1-infected plants and higher loads in ΔVdMcu1-infected plants (Fig. 6E). These virulence phenotypes were reproduced in N. benthamiana infection assays. ΔVdMid1 exhibited attenuated virulence, whereas ΔVdMcu1 showed enhanced virulence, accompanied by corresponding differences in fungal biomass, in N. benthamiana (Fig. 6F–H, S11B–C). Collectively, these results demonstrate that deleting VdMid1 compromises V. dahliae pathogenicity, whereas deleting VdMcu1 enhances virulence in the host plants.

Fig. 6.

VdMid1 and VdMcu1 have opposing effects on virulence. A Verticillium wilt symptoms on smoke tree seedlings infected with XS11, ΔVdMid1, ΔVdMcu1, ΔVdMid1-C, or ΔVdMcu1-C. Images were taken at 35 dpi. Scale bars, 10 cm. B Verticillium wilt index in smoke tree seedlings inoculated with each strain described in (A). Violin plots show the median (dashed lines) and quartiles (dotted lines) from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗∗P < 0.001; ns, not significant). C Cross-sections and longitudinal sections of rootstock junctions from smoke tree seedlings infected with the different strains described in (A). Scale bars, 0.5 cm. D Re-isolation of V. dahliae strains from the stems of smoke tree seedlings inoculated with the strains described in (A) at 3 dpi. Scale bars, 0.5 cm. E Relative fungal biomass of each strain in inoculated smoke tree seedlings as determined by qPCR. Data are presented as means ± SD from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗P < 0.05; ∗∗∗P < 0.001; ns, not significant). F Verticillium wilt symptoms of N. benthamiana plants infected with the different strains described in (A) and control uninfected plants (CK). Images were taken at 35 dpi. Scale bars, 5 cm. G Number of plants described in (F) with Verticillium wilt symptoms. The bars represent the frequency of plants within each disease index class. H Fungal biomass of each strain in N. benthamiana as determined by qPCR. Data are presented as means ± SD from three biological replicates. Significant differences were determined by one-way ANOVA followed by Dunnett's multiple comparisons test (∗∗P < 0.01; ∗∗∗P < 0.001; ns, not significant).

3. Discussion

In this study, we identified Veh1 as a Ca²⁺-responsive transcriptional regulator that suppresses hyphopodium-mediated plant penetration by negatively regulating Ca²⁺ accumulation in V. dahliae hyphopodia. As illustrated in Fig. 7, perception of a root by the pathogen triggers hyphopodium development and subsequent penetration peg formation in V. dahliae. During this transition, cytoplasmic Ca²⁺ levels fluctuate, initially increasing then gradually decreasing, which coincides with the induction of Veh1 expression. Once activated, Veh1 directly represses the expression of VdMid1, which encodes a plasma membrane–associated Ca²⁺ influx component, while promoting the expression of VdMcu1, which encodes a mitochondrial Ca²⁺ uniporter, thereby collectively reducing cytosolic Ca²⁺ levels. Reduced Ca²⁺ accumulation in the hyphopodium suppresses penetration peg formation, likely by attenuating Crz1-mediated signaling and impairing Ca²⁺-dependent developmental programs. Together, these results define a Veh1–VdMid1/VdMcu1 module that fine-tunes Ca²⁺ homeostasis in V. dahliae infection structures and restrains invasive growth. By uncovering this Ca²⁺-responsive transcriptional circuit, our work provides a mechanistic framework for understanding how fungal pathogens coordinate Ca²⁺ dynamics with the differentiation of infection structures.

Fig. 7.

Model depicting the role of Veh1-mediated Ca²⁺ homeostasis in hyphopodia in the regulation of plant infection by V. dahliae. Upon sensing the host root, the germ tube emerging from a conidium adheres tightly to the root epidermis and differentiates into a hyphopodium, which subsequently gives rise to a penetration peg to initiate host invasion. During early infection, cytosolic Ca²⁺ progressively accumulates in the hyphopodium, promoting formation of a penetration peg through VdCrz1 signaling. At later stages of infection, elevated Ca²⁺ levels induce Veh1 expression. In turn, Veh1 activates expression of the mitochondrial Ca²⁺ uniporter gene VdMcu1 while repressing that of the plasma membrane Ca²⁺ influx channel gene VdMid1. The coordinated regulation of VdMcu1 and VdMid1 lowers cytoplasmic Ca²⁺ accumulation, thereby restoring Ca²⁺ homeostasis in the hyphopodium. Therefore, the Veh1–VdMid1/VdMcu1 regulatory pathway fine-tunes Ca²⁺ dynamics to balance penetration peg formation with successful plant infection.

Recent work in model phytopathogens, particularly M. oryzae, has uncovered multilayered regulatory mechanisms governing infection, ranging from phosphorylation cascades to transcriptional networks [21]. Transcription factors, including C₂H₂ and bZIP family proteins, have emerged as key regulators of infection structure development [24]. Instead of appressoria, V. dahliae employs specialized hyphopodia to breach tough root layers. Analyses of gene function have identified regulators with distinct roles in this developmental stage. For example, the ΔVdAtf1 mutant, which lacks Activating transcription factor 1, penetrates cellophane but fails to expand into neighboring onion (Allium cepa) cells, whereas loss of Heme activator protein X (VdHapX) function delays penetration without affecting hyphopodium formation [25,26]. In this study, a genetic screen of V. dahliae transcription factor mutants identified Veh1 as a previously unrecognized negative regulator of hyphopodium formation, as ΔVeh1 mutants exhibited excessive penetration peg formation. Phylogenetic analysis placed Veh1 in the bZIP family. Consistent with observations in M. oryzae, our results demonstrate that the V. dahliae bZIP transcription factor Veh1 is important for plant penetration, suggesting that bZIP-mediated regulation of infection structures may be evolutionarily conserved across phytopathogenic fungi.

Specialized infection structures are essential for breaching host barriers and establishing disease. Consistent with this paradigm, mutants defective in regulators of infection structure formation typically display impaired penetration and reduced virulence [27,28]. In contrast to such loss-of-function phenotypes, deleting Veh1 enhanced hyphopodium-mediated penetration and invasive growth, ultimately accelerating host colonization and exacerbating disease severity (Fig. 2, S3). These observations establish Veh1 as a pivotal negative regulator of fungal penetration and pathogenicity.

Cytosolic Ca²⁺ is a universal second messenger that integrates environmental cues with developmental outputs [10,29]. In the fungal pathogen Cytospora chrysosperma, ROS-driven Ca²⁺ influx modulates redox homeostasis [30]. In V. dahliae, VdPls1 and VdNoxB stimulate ROS production to mediate Ca²⁺ accumulation and promote penetration peg formation [6]. Here, we demonstrate that Veh1 negatively regulates cytosolic Ca²⁺ accumulation by modulating the transcription of Ca²⁺-related genes. Because elevated Ca²⁺ levels activate downstream effectors, including the conserved calmodulin–Crz1 signaling pathway [31], and Crz1 promotes penetration peg formation in V. dahliae [6,32], the upregulation of VdCrz1 and its targets in ΔVeh1 provides a mechanistic explanation for the hyper-penetration phenotype of this mutant. However, Veh1 does not directly bind to the promoters of VdCrz1 or its target genes (Fig. S6E; Table S2), suggesting that Veh1 indirectly suppresses Ca²⁺–Crz1 signaling to inhibit penetration peg formation.

Analysis of Ca²⁺ dynamics revealed a biphasic Ca²⁺ profile in hyphopodia: an initial increase in accumulation followed by a decrease, indicating stringent temporal control during infection. Consistent with a previous report in M. oryzae showing that both insufficient and excessive Ca²⁺ levels impair appressorium formation [13], our results indicate that Ca²⁺ homeostasis is similarly critical for V. dahliae. Chelation experiments confirmed that although limited Ca²⁺ restricted penetration in wild-type strains, ΔVeh1 retained the ability to penetrate cellophane, underscoring the notion that Veh1-dependent regulation of Ca²⁺ levels is a decisive factor in penetration competence.

Interestingly, Veh1 is transcriptionally induced by Ca²⁺, despite its inhibitory role in penetration, forming a canonical negative-feedback module [33]. We propose that increased Ca²⁺ levels initially promote penetration through positive regulators such as VdCrz1; however, prolonged Ca²⁺ accumulation becomes detrimental to this process. At later infection stages, Ca²⁺-mediated induction of Veh1 represses Ca²⁺ influx while promoting Ca²⁺ sequestration, thereby restoring Ca²⁺ homeostasis in hyphopodia and preventing excessive penetration.

An analysis combining RNA-seq and ChIP-seq data identified VdMid1 and VdMcu1 as direct targets of Veh1. Veh1 activates VdMcu1 expression, promoting mitochondrial Ca²⁺ uptake, while it represses the expression of VdMid1, which encodes a channel required for Ca²⁺ influx (Fig. 4, S8). By repressing VdMid1-mediated Ca²⁺ influx and activating VdMcu1-mediated mitochondrial Ca²⁺ uptake, Veh1 lowers cytosolic Ca²⁺ levels through complementary mechanisms, thereby fine-tuning Ca²⁺ homeostasis. This parallel regulatory strategy resembles observations in the insect pathogenic fungus Beauveria bassiana, where Hemocoel colonization-associated regulator 1 (BbHcr1) either activates or represses hypothetical protein 1 (HP1) and hypothetical protein (HP2) to modulate fungal pathogenicity [34].

Channels of the VGCC family contribute to Ca²⁺ homeostasis and biological functions across fungi [23]. In yeast, Mid1p localizes to the plasma membrane and is essential for Ca²⁺ influx during mating [17]. Mid1 influences conidiation, hyphal polarity, and cell wall composition in Aspergillus nidulans [35]. In entomopathogenic fungi, Mid1 affects intracellular ion homeostasis and contributes to virulence by affecting the initial penetration stages [20]. These observations suggest that Mid1 is a conserved regulator of Ca²⁺ homeostasis and development. In V. dahliae, VdMid1 functions as a major route for Ca²⁺ entry, and its loss of function compromises penetration and virulence, highlighting its role as a key effector downstream of Veh1. Conversely, the MCU complex in mitochondria acts as a Ca²⁺ sink [15]. In mammals, MICU1 regulates MCU activity and is essential for mitochondrial Ca²⁺ uptake [36]. Filamentous fungi such as Aspergillus fumigatus possess McuA, which encodes an MCU homolog crucial for mitochondrial Ca²⁺ uptake and stress responses; its deletion significantly increases cytoplasmic Ca²⁺ concentrations [37]. We identified VdMcu1 as the fungal MCU homolog responsible for limiting cytosolic Ca²⁺ levels during infection. Deleting VdMcu1 increased cytosolic Ca²⁺ accumulation and enhanced penetration and virulence, phenocopying ΔVeh1 and demonstrating that mitochondrial sequestration is an integral component of Ca²⁺ homeostasis in the hyphopodium. Together, these results place Veh1 at the core of a Ca²⁺-responsive transcriptional circuit that links environmental Ca²⁺ dynamics to infection-structure competence. By coordinately tuning Ca²⁺ influx and sequestration through VdMid1 and VdMcu1, Veh1 prevents excessive penetration and limits pathogenicity, revealing a previously unappreciated layer of negative regulation in root-invading fungi.

Our study identified a mechanism by which Veh1-mediated transcriptional regulation of VdMid1 and VdMcu1, which encode two crucial Ca²⁺ transport proteins, maintains Ca²⁺ homeostasis in hyphopodia to control plant infection. However, several key questions remain unanswered. For example, although the ΔVeh1 mutant produced more infection structures than WT under the same conditions, we cannot exclude the possibility that this enhanced penetration ability is partially attributable to its accelerated growth and increased biomass accumulation. Therefore, future studies are needed to exclude the influence of vegetative growth on penetration efficiency. In addition, the precise mechanisms by which Veh1-interacting proteins participate at this stage, and how their interactions differentially modulate VdMid1 and VdMcu1 transcript accumulation, are unclear. Moreover, although this study showed that Veh1 suppresses VdCrz1 signaling, the molecular basis by which Veh1 indirectly modulates VdCrz1 signaling is unclear. Finally, given that Veh1 is a key regulator of ion transport, whether other potential Veh1 target genes identified by ChIP contribute to Ca²⁺ transport remains to be investigated.

4. Materials and methods

4.1. Fungal strains, mutant generation, complementation, and vector construction

V. dahliae strain XS11 was used as the wild type in this study. Genes were disrupted in V. dahliae using our previously described protocol [38]. Briefly, fresh mycelia were treated with driselase (D9515, Sigma, St. Louis, MO), lysozyme (RM1027, RYON, Shanghai, China), and cellulase (RM1030, RYON, Shanghai, China) to obtain V. dahliae protoplasts. The flanking sequences of the Veh1, VdMid1, and VdMcu1 genes were amplified from XS11 genomic DNA and fused to either side of the hygromycin resistance cassette via fusion PCR (Table S5). Subsequently, the replacement constructs were directly introduced into XS11 protoplasts via transformation. For complementation, the full-length coding and promoter regions of Veh1, VdMid1, and VdMcu1 were amplified and used to transform protoplasts of the respective deletion mutants. Mutants and their successful complementation strains were verified by PCR using gene-specific primers.

To express Veh1-GFP, the full-length Veh1 coding sequence was amplified and cloned into the pCAMBIA1300-GFP vector by homologous recombination using the ClonExpress II One Step Cloning Kit (Vazyme, Beijing, China). The resulting Veh1-GFP construct, together with a geneticin resistance cassette, was used to transform ΔVeh1 protoplasts. Transformants were selected on geneticin-containing medium and confirmed by fluorescence microscopy and immunoblotting.

4.2. Virulence assays and observation of infection structures

Fresh conidial suspensions of each V. dahliae strain were prepared by culturing the strains in liquid complete medium (CM) for 5 days. The resulting cultures were filtered through four layers of cheesecloth to remove mycelia, and the conidial suspensions were diluted to 1 × 10⁶ conidia mL⁻¹ with sterile deionized water. The roots of one-month-old N. benthamiana plants and one-year-old smoke tree (Cotinus coggygria) seedlings were dipped in conidial suspensions of the indicated strains for 10 and 30 min, respectively. The inoculated seedlings were transplanted into sterile nutrient soil and grown in a greenhouse at 25 °C. All plants were observed at 15 or 35 days post-inoculation and evaluated for wilt symptoms on a scale of 0 to 4 as follows [39]: 0 = no symptoms; 1 = leaves on one branch show wilt symptoms; 2 = leaves on 2–3 branches show wilt symptoms; 3 = more than half of the leaves are yellowing or wilting, with some defoliation; and 4 = the entire plant is wilted or dead. To detect the pathogen in the infected plants, root and stem segments were rinsed with distilled water, soaked in 3% (v/v) sodium hypochlorite for 3 min, and immersed in 75% (v/v) ethanol for 1 min. The samples were then washed three times with sterile water and incubated on solid potato dextrose agar (PDA) plates containing 100 μg/mL ampicillin to isolate V. dahliae from the respective tissues. These pathogenicity assays were performed three times.

To observe infection structures, cellophane membranes inoculated with each strain were removed from the minimal medium (MM) and rinsed with sterile water. The colony-covered side of the cellophane was mounted on a glass slide and covered with a coverslip for microscopy (DM1000 LED, Leica) observation. Infection structures were examined at the four edges of the colony, arranged in a square pattern. Images of hyphal pegs at each edge were captured at 40× magnification, with three images per edge (12 total). The experiments were performed three times independently.

4.3. Transcriptome sequencing and data analysis

Total RNA was extracted from XS11 and ΔVeh1 cultures grown on cellophane for 72 h using TRIzol® Reagent (Invitrogen), followed by DNase I (TaKaRa) treatment. RNA quality was assessed by agarose gel electrophoresis and Bioanalyzer analysis, using OD_260/280 ratios of 1.8–2.2 and RIN values ≥ 8.0 as quality criteria. RNA was quantified using a NanoDrop spectrophotometer. High-quality total RNA (>1 μg) was used to construct Illumina TruSeq libraries via polyA selection, cDNA synthesis (SuperScript kit), end-repair, and size selection (300 bp). PCR-amplified libraries (TruSeq™ RNA Sample Preparation Kit, Illumina, USA) were sequenced on the Illumina NovaSeq 6000 platform (2×150 bp). Raw reads were processed with fastp (https://github.com/OpenGene/fastp), aligned to the reference genome [40] using HISAT2 (http://ccb.jhu.edu/software/hisat2/index.shtml), and assembled with StringTie (https://ccb.jhu.edu/software/stringtie/). Sequencing and library preparation were performed by Majorbio Bio-pharm (Shanghai). To identify differentially expressed genes (DEGs), gene expression levels were calculated using transcripts per million (TPM). Genes with |log₂(fold-change)| ≥ 1 and p-adjust ≤0.05, as determined by the R package DESeq2 [41], were considered to be DEGs. In addition, GO functional-enrichment analysis (Gene Ontology, http://www.geneontology.org) was performed to identify the enriched GO terms and metabolic pathways among DEGs at p-adjust ≤0.05 compared to the whole-genome background. GO enrichment was carried out using Goatools (https://github.com/tanghaibao/Goatools).

4.4. RT-qPCR

Total RNA was extracted from samples using a Plant Tissue RNA Rapid Extraction Kit (Tiangen, Beijing, China) according to the manufacturer's protocol. First-strand cDNA was synthesized from the RNA using a PrimeScript RT Kit (ABclonal, Wuhan, China). SYBR Green-based qPCR assays (Yeason, Shanghai, China) were performed on an ABI 7500 Real-Time PCR System (Applied Biosystems, USA) following standard protocols. The thermal cycling conditions consisted of an initial denaturation at 95 °C for 15 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 32 s. Gene expression levels were quantified via the 2^−ΔΔCT method [42], with the V. dahliae β-tubulin gene serving as the endogenous control according to Duressa et al. [43]. All experimental procedures strictly adhered to manufacturers' guidelines. Primers used in this assay and brief descriptions are listed in Table S5.

4.5. Determination of fungal biomass

At the designated time points after inoculation indicated in the figures, DNA was extracted from the roots or stems of three individual seedlings per treatment group using a Plant Genomic DNA Kit (Tiangen). Prior to extraction, the samples were thoroughly rinsed with sterile water. Fungal biomass was quantified using a qPCR assay targeting the V. dahliae ITS, with plant (tobacco or smoke tree) GAPDH used as the control for plant biomass [22]. All primers used are listed in Table S5. The experiment was performed with three independent biological replicates.

4.6. Measurement of relative Ca²⁺-dependent fluorescence signals

To measure intracellular Ca²⁺ content, wild-type, mutant, and complemented strains were cultured in CM for 24 h. The mycelia were filtered, collected, and washed three times with phosphate-buffered saline (PBS). Fluo-4 AM stock solution (Beyotime Biotechnology, S1061S) was diluted in PBS to prepare a 5 μM working solution, which was then added to the mycelia, followed by incubation at room temperature for 30 min in the dark. The samples were washed three times with PBS and incubated for 15 min to complete the intracellular conversion of Fluo-4 AM to Fluo-4. Fluorescence imaging was performed using a confocal laser scanning microscope (TCS SP8, Leica; excitation wavelength, 490 nm; emission wavelength, 525 nm). The experiment was performed with three independent biological replicates. For each biological replicate, five images were taken of each strain to measure and analyze its fluorescence intensity.

To measure Ca²⁺ levels during later infection stages, hyphopodia grown on cellophane for 72 h were collected and loaded with Fluo-4 AM following the protocol described above. After incubation, relative fluorescence units (RFUs) were measured using a fluorescence microplate reader (Spark, Tecan; excitation, 490 nm; emission, 525 nm). The experiment was performed with three independent biological replicates. In each independent experiment, hyphopodial samples collected from three separate cellophane cultures were measured in parallel for each strain.

To measure Ca²⁺ dynamics during hyphopodium formation, hyphopodia incubated on cellophane were collected and loaded with 5 μM Fura-2AM (Yeasen Biotech Co., Ltd., Shanghai, China) for 45 min in the dark. After washing, intracellular free Ca²⁺ levels were measured using excitation wavelengths of 340 nm and 380 nm, with emission at 510 nm. A control without the probe was performed to account for autofluorescence. Data from both excitation wavelengths (F₃₄₀/F₃₈₀) were compared to correct for loading efficiency and sample heterogeneity, enabling quantitative comparison of relative Fura-2AM Ca²⁺ signals. The experiment was performed with three independent biological replicates.

4.7. ChIP-seq and ChIP-qPCR

ChIP was performed as described by Li et al. [44] with some modifications. Briefly, mycelia were subjected to crosslinking with 1% (v/v) formaldehyde and quenched with 125 mM glycine. The samples were ground in liquid nitrogen, resuspended in nuclear extraction buffer, and the chromatin sheared into ∼300-bp fragments using an ultrasonic disrupter (Xiaomei Ultrasound, China). Following centrifugation, the supernatant was diluted with ChIP dilution buffer for immunoprecipitation with anti-GFP or anti-IgG antibodies (ABclonal, Wuhan, China). Protein A/G magnetic beads were sequentially washed with low-salt, high-salt, LiCl, and TE buffers. The DNA was eluted from the beads, crosslinking was reversed, and the DNA was purified. The enriched DNA was subjected to sequencing (Annoroad Genetic Technology, Beijing, China) or quantitative PCR. Relative enrichment values were calculated as a percentage of input. ChIP-seq was performed in two independent biological replicates, and ChIP-qPCR was performed using three independent biological replicates.

4.8. Dual luciferase assay

Dual-luciferase reporter assays were performed using the pGreenII 62-SK and pGreenII 0800-LUC vectors. The full-length coding sequence of Veh1 and the promoter regions of VdMid1 and VdMcu1 were cloned into the respective vectors. Agrobacterium (Agrobacterium tumefaciens) strain GV3101 was transformed with the resulting plasmids. Positive colonies were cultured in liquid LB medium containing 50 μg/mL kanamycin and 50 μg/mL rifampicin, resuspended in MgCl₂ buffer (10 mM MgCl₂, 10 mM MES, and 200 μM acetosyringone), and incubated for 4 h at 28 °C. The bacterial suspensions harboring the reporter and effector constructs were adjusted to OD₆₀₀ = 0.4, mixed in equal volumes, and co-infiltrated into the leaves of N. benthamiana plants. A Dual-Luciferase Reporter Assay System kit (Beyotime, RG027) was used to measure relative luciferase activity as described in the instruction manual.

4.9. EMSA

The Veh1 coding sequences were inserted into the pGEX-4T-1 vector, and the protein product was purified using GST-tag Purification Resin (Beyotime, P2251). The promoter fragments of VdMid1 and VdMcu1 containing the native ACGT site were synthesized and labeled with biotin. The direct binding affinity of proteins to probes was detected using a Chemiluminescent EMSA Kit (Beyotime, GS009) following the manufacturer's protocol.

4.10. Statistical analysis

Statistical analyses were performed using GraphPad Prism 10. Statistically significant differences between control and experimental groups were assessed using ordinary one-way analysis of variance (ANOVA) assuming a Gaussian distribution, or by two-way ANOVA when two factors were introduced. Paired comparisons were analyzed by Student's t-test (∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001; ns, not significant).

CRediT authorship contribution statement

Xiao He: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review & editing. Chen Tang: Methodology. Jiahui Wang: Investigation. Krishna V. Subbarao: Writing – original draft, Writing – review & editing. Yonglin Wang: Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article.

Data availability

RNA-seq and ChIP-seq data generated in this study are available in the NCBI Sequence Read Archive (SRA) under accession numbers PRJNA1376500 and PRJNA1376495.