Herbicides as fungicides: Targeting heme biosynthesis in the maize pathogen Ustilago maydis

Abstract

Pathogens must efficiently acquire nutrients from host tissue to proliferate, and strategies to block pathogen access therefore hold promise for disease control. In this study, we investigated whether heme biosynthesis is an effective target for ablating the virulence of the phytopathogenic fungus Ustilago maydis on maize plants. We first constructed conditional heme auxotrophs of the fungus by placing the heme biosynthesis gene hem12 encoding uroporphyrinogen decarboxylase (Urod) under the control of nitrogen or carbon source-regulated promoters. These strains were heme auxotrophs under non-permissive conditions and unable to cause disease in maize seedlings, thus demonstrating the inability of the fungus to acquire sufficient heme from host tissue to support proliferation. Subsequent experiments characterized the role of endocytosis in heme uptake, the susceptibility of the fungus to heme toxicity as well as the transcriptional response to exogenous heme. The latter RNA-seq experiments identified a candidate ABC transporter with a role in the response to heme and xenobiotics. Given the importance of heme biosynthesis for U. maydis pathogenesis, we tested the ability of the well-characterized herbicide BroadStar to influence disease. This herbicide contains the active ingredient flumioxazin, an inhibitor of Hem14 in the heme biosynthesis pathway, and we found that it was an effective antifungal agent for blocking disease in maize. Thus, repurposing herbicides for which resistant plants are available may be an effective strategy to control pathogens and achieve crop protection.

Keywords: biotrophy; corn smut disease; fungal pathogen; phytopathogenesis.

FIGURE 1

Heme biosynthesis is essential in Ustilago maydis and exogenous hemin rescues growth of heme auxotrophs. (a) Model of heme biosynthesis in U. maydis based on the pathway in Saccharomyces cerevisiae. Heme biosynthesis begins and ends in the mitochondria, with succinyl‐CoA and glycine as starting substrates. The gene for Hem12 (red) was placed under the control of the regulatable promoters Pcrg1 or Pnar1. SucCoA – succinyl CoA; ALA – δ‐aminolevulinic acid; UROgenIII – uroporphyrinogen III; CPgenIII – coproporphyrinogen III; PPgenIX – proporphyrinogen IX; PPIX – protoporphyrin IX. (b) Diagram of the constructs used to generate heme auxotrophs. Pcrg1 promotes transcription of hem12 in the presence of arabinose and represses transcription in the presence of glucose while Pnar1 promotes transcription of hem12 in the presence of nitrate and represses it in the presence of ammonium. (Hyg – hygromycin resistance marker, Nat – nourseothricin resistance marker). (c) Growth of heme auxotrophs regulated by the nar1 (left) and crg1 promoters (right) in permissive (ON) and restrictive (OFF) conditions with increasing concentrations of hemin. Plates were imaged after 72 h. (d) Growth of heme auxotrophs regulated by the nar1 promoter in permissive (ON) and restrictive (OFF) liquid media containing increasing concentrations of hemin. Cells were counted using a haemocytometer after 72 h of growth. Statistics were performed using a two‐tailed unpaired t test with Welch's correction (*p ≤ 0.05).

FIGURE 2

Regulatable heme auxotrophs are avirulent in maize seedlings. (a) Virulence of the strains with regulatable Pnar1 (left) and Pcrg1 (right) expression of hem12 in Zea mays seedlings infected 7 days after germination. The disease index (DI) ± standard deviation and p‐values are included above each infection and are based on the symptom rating scheme indicated by the colour chart. Mating was also assayed on charcoal plates (top) and the × over the centre colony indicates the mixture of the 001 and 002 mating partners. Statistics were performed using a two‐tailed, unpaired t test. (*p ≤ 0.05, **p ≤ 0.01). (b) Co‐staining of the plant cell wall with propidium iodide (red) and chitin in the Ustilago maydis cell wall with WGA‐AF488 (green) (i and iii) at 4 days post‐inoculation (dpi). Chlorosis caused by both strains at 4 dpi (ii and iv). i, ii = SG200; iii, iv = Pcrg1::Hem12. Scale bar represents 25 μm.

FIGURE 3

Hemin uptake is partially clathrin‐dependent and high hemin concentrations induce iron‐independent reactive oxygen species accumulation and lipid peroxidation. (a) Percentage of the wild‐type (WT) single cell population stained with DCFDA after treatment with or without hemin for 6 h, as assayed by flow cytometry. Statistics were performed using one‐way analysis of variance (ANOVA) (***p ≤ 0.001). (b) Percentage of the population shifted upon BODIPY 581/591 C11 staining of WT cells treated with or without hemin for 6 h, assayed using flow cytometry. Statistics were performed using one‐way ANOVA (***p ≤ 0.001). (c) Diagram depicting the mode of action of chlorpromazine (CPZ) on clathrin‐mediated endocytosis (CME) (top). Quantification of the internal hemin concentration in WT cells treated with hemin supplied with or without CPZ for 1 h (bottom). The fluorescence (Ex. 410 nm, Em. 610 nm) of cells treated with hemin alone compared to those treated with hemin and CPZ was measured. Statistics were performed using a paired t test. (*p ≤ 0.05). (d) Reverse transcription‐quantitative PCR (RT‐qPCR) (log scale) of cells treated with hemin or iron for 1 h to examine transcript levels for vps23 and chc1. Statistics were performed using a one‐way ANOVA test with a Tukey procedure as a post hoc test. (***p ≤ 0.001). (e) RT‐qPCR (log scale) of cells treated with hemin ± CPZ for 1 h to examine transcript levels for vps23 and chc1. Statistics were performed using a one‐way ANOVA test with a Tukey procedure as a post hoc test. (***p ≤ 0.001). (f) Growth of the regulatable heme auxotroph on permissive (ON) and restrictive (OFF) media supplemented with increasing concentrations of hemin ± CPZ at 72 h of incubation.

FIGURE 4

Transcriptome response to treatment with hemin versus FeCl₃ after hemin starvation. (a) Flowchart of the sample preparation for the RNA‐seq experiment. The Pnar1 heme auxotroph was grown in permissive conditions (ON) for 24 h, then starved for 24 h in restrictive conditions (OFF). Cells were subsequently transferred to ON, OFF, OFF +15He (15 μM hemin) and OFF +15Fe (15 μM FeCl₃). (b) Regulation of the heme biosynthesis pathway comparing 1 h iron treatment versus 1 h hemin treatment. The hem1 and hem12 genes were the only two heme biosynthesis genes that were significantly differentially downregulated upon hemin treatment. Statistics were performed using a Wald test (***p ≤ 0.001). (c) Heat map showing an overview of the RNA‐seq data and the upregulation of the ergosterol biosynthesis pathway upon 1 h of hemin treatment. (d) Gene Ontology (GO) enrichment analysis of 637 differentially upregulated genes upon hemin treatment versus FeCl₃ where colour represents log₁₀(false discovery rate) and circle size represents the number of genes in the GO term. The figure was made using ShinyGO v. 0.80 (Ge et al., 2019). (e) GO term enrichment clustering analysis using REVIGO for the 637 genes upregulated upon hemin versus FeCl₃ treatment (red represents the most significant and yellow the least).

FIGURE 5

The ∆abc1 mutants show increased susceptibility to hemin. (a) Spot assays of ∆abc1 mutants in both wild‐type (WT) and heme‐auxotrophic backgrounds on permissive and restrictive conditions supplemented with increasing concentrations of hemin after 72 h of incubation. (b) Spot assays of ∆abc1 mutants in both WT and heme‐auxotrophic backgrounds on permissive and restrictive conditions supplemented with increasing concentrations of hemin and CPZ after 72 h of incubation. (c) Growth of the ∆abc1 deletion mutant on minimal medium (MM) after 72 h (unless otherwise indicated) supplemented with the indicated xenobiotics (4NQO – 4‐nitroquinoline 1‐oxide at 0.025 μg/mL; FLC – fluconazole at 2.5 μg/mL; TEB – tebuconazole at 0.05 μM; MIC – miconazole at 0.1 μg/mL; KET – ketoconazole at 0.05 μg/mL; BEN – benomyl at 1 μg/mL; AmpB – amphotericin B at 1 μg/mL; RAP – rapamycin at 100 nM). (d) Relative transcript levels of abc1 in the abc1 overexpression strains as measured by reverse transcription‐quantitative PCR. Statistics were performed using an unpaired t test with Welch's correction (*p ≤ 0.05, **p ≤ 0.01). (e) Growth of the abc1 overexpression strains on complete medium (CM) supplemented with high concentrations of hemin after 48 h of incubation.

FIGURE 6

Abc1 is localized to the cell surface and ∆abc1 mutants show altered susceptibility to cell surface stresses. (a) Localization of Abc1‐GFP in the Pnar1 regulatable heme auxotroph at the cell surface in punctate structures as determined by confocal microscopy. Cells were grown overnight in ON medium, starved in OFF medium for 24 h and treated with 15 μM hemin for 2 h before visualization. i. Differential interference contrast (DIC) for wild type (WT). ii. Fluorescence for WT. iii. DIC for Abc1‐GFP. iv. Fluorescence for Abc1‐GFP. The panel on the right shows an enlarged 3D‐rendered image of fluorescence for Abc1‐GFP over 42 z‐stacks (150 nm separation). Scale bar represents 5 μm. (b) Growth of the ∆abc1 and ∆erg3 mutants on complete medium (CM) supplemented with high concentrations of hemin and grown for 48 h. (c) Growth of the Δabc1 mutant on CM supplemented with different cell surface stressors: TEB – tebuconazole, AmpB – amphotericin B, CAS – caspofungin, CFW – calcofluor white. Growth was assessed after 48 h unless indicated otherwise. (d) Melanization defects of the Δabc1 mutant in two different melanizing conditions. i. Colour of cultures after 72 h growth in minimal medium supplemented with glucose and malate (top) and culture viscosity after 3 days measured using a viscometer and reported as time to travel (bottom). Statistics were performed using one‐way analysis of variance (***p ≤ 0.001). ii. Melanization of the indicated strains on solid l‐DOPA medium.

FIGURE 7

The ∆abc1 mutant has a slight virulence defect in planta. (a) Virulence assay of mutants evaluated 2 weeks post‐inoculation. The disease index ± SD and p‐values are included above each infection and symptoms were evaluated with the indicated colour scheme. Statistics were performed using an unpaired t test (*p ≤ 0.05). (b) Mating of the indicated strains on DCM charcoal plates. The × indicates the mixture of the mating partners in the centre colonies. (c) Melanin accumulation in teliospores was assessed 2 weeks post‐inoculation.

FIGURE 8

BroadStar inhibits heme biosynthesis and reduces Ustilago maydis virulence in planta. (a) Growth in liquid potato dextrose broth supplemented with increasing concentrations of the herbicide BroadStar with or without hemin after 24 h (starting inoculum: 2 × 10⁵ cells/mL). Statistics were performed using a paired t test (*p ≤ 0.05, **p ≤ 0.01). (b) Growth on potato dextrose agar (PDA) supplemented with increasing concentrations of the herbicide BroadStar with or without 30 μM hemin. (c) Growth on PDA supplemented with increasing concentrations of flumioxazin, the active ingredient in the herbicide BroadStar, with or without 30 μM hemin. (d) Virulence of wild‐type strains on Zea mays seedlings treated with BroadStar at the indicated dilutions of the recommended herbicidal dose (recommended dose, RD = 168 kg/ha) (left). Plants were infected 7 days after germination. The disease index ± SD as well as p‐values are included above each infection (*p ≤ 0.05, **p ≤ 0.01). Mating was also assayed on charcoal plates supplemented with two concentrations of the herbicide (right). The × indicates the mixture of the mating partners in the colony on the right.