Transcription factor Amr1 induces melanin biosynthesis and suppresses virulence in Alternaria brassicicola

Abstract

Alternaria brassicicola is a successful saprophyte and necrotrophic plant pathogen. Several A. brassicicola genes have been characterized as affecting pathogenesis of Brassica species. To study regulatory mechanisms of pathogenesis, we mined 421 genes in silico encoding putative transcription factors in a machine-annotated, draft genome sequence of A. brassicicola. In this study, targeted gene disruption mutants for 117 of the transcription factor genes were produced and screened. Three of these genes were associated with pathogenesis. Disruption mutants of one gene (AbPacC) were nonpathogenic and another gene (AbVf8) caused lesions less than half the diameter of wild-type lesions. Unexpectedly, mutants of the third gene, Amr1, caused lesions with a two-fold larger diameter than the wild type and complementation mutants. Amr1 is a homolog of Cmr1, a transcription factor that regulates melanin biosynthesis in several fungi. We created gene deletion mutants of Δamr1 and characterized their phenotypes. The Δamr1 mutants used pectin as a carbon source more efficiently than the wild type, were melanin-deficient, and more sensitive to UV light and glucanase digestion. The AMR1 protein was localized in the nuclei of hyphae and in highly melanized conidia during the late stage of plant pathogenesis. RNA-seq analysis revealed that three genes in the melanin biosynthesis pathway, along with the deleted Amr1 gene, were expressed at low levels in the mutants. In contrast, many hydrolytic enzyme-coding genes were expressed at higher levels in the mutants than in the wild type during pathogenesis. The results of this study suggested that a gene important for survival in nature negatively affected virulence, probably by a less efficient use of plant cell-wall materials. We speculate that the functions of the Amr1 gene are important to the success of A. brassicicola as a competitive saprophyte and plant parasite.

Figure 1. Verification of Δamr1 deletion mutants.

A. Replacement of the Amr1 coding region with the selectable marker, Hygromycin B transferase (HygB) resistance cassette. On the HygB-probed blot, the expected 4.5 Kb band indicates a single copy insertion of the Hyg B resistance cassette in lane d4. Additional bands in the other lanes suggest multiple insertions of the cassette. B. Replacement of the Amr1 coding region with a green fluorescent protein (GFP) coding region and a HygB cassette. The upper panel of both A and B is a schematic diagram of wild-type (wt) and mutant loci and the lower panel consists of five Southern blots showing loss of the Amr1 gene in selected mutants. Dots (•) indicate lanes of mutants used in this study. Probe regions are marked by P_h and P_a. Abbreviations: E = EcoRI enzyme digestion site. d1 = Δamr1-1, d4 = Δamr1-4, d5 =  Δamr1-5, and D3 = Δamr1:Amr1p-GFP. e1-e5, E1, E3, and E4 are ectopic insertion mutants.

Figure 2. Increased virulence of Δamr1 mutants on individual host plants (Brassica oleracea).

A. Lesions on B. oleracea 5 days after inoculation with ∼2,000 conidia of the wild type, a Δamr1 mutant, and three complemented mutants with two types of constructs. B. Lesions produced by ∼440 and ∼400 conidia respectively of the wild type (left) and a Δamr1 mutant (right). C. Lesions produced by ∼330 and ∼300 conidia respectively of the wild-type (left) and a Δamr1 mutant (right). The three images in panels A–C are replicate experiments with the same pattern of inoculation. Note that the size of lesions caused by the mutant is similar on all leaves, while lesions caused by the wild type are smaller on young leaves than on old leaves. Δamr1-1 and Δamr1-4 produced similar results. D. Pathogenicity assay results for three mutants (Δamr1-1, Δamr1-4, and Δamr1-5) compared to the wild type. E. Pathogenicity assay results with average lesion diameter (mm) and standard deviation. Abbreviations: wt = wild-type A. brassicicola; Δamr1 = Amr1 deletion mutant; compl = mutant complemented with a native allele of the Amr1 gene (Δamr1:Amr1); cons = mutant complemented with chimeric constructs of the TrpC promoter and a native allele of the Amr1 coding region (Δamr1:TrpCp-Amr1).

Figure 3. Confocal microscopic images showing Amr1 gene expression and its protein localization in fungal nuclei during conidiation and pathogenesis.

A–C. Green fluorescence indirectly shows localization of Amr1 protein in fungal nuclei: GFP is expressed as the Amr1-GFP fusion protein. D–F. Green fluorescence in fungal tissues indicates an abundance of GFP, whose gene is under the regulation of the Amr1 promoter in the Δamr1:Amr1p-GFP mutant. G. Fluorescence shows co-localization of Amr1-GFP proteins, and 4′,6-diamidino-2-phenylindole (DAPI) stains that bind to nucleotides. Pink color represents auto-fluorescence of plant tissues. Arrowheads indicate conidia. Arrows point to fungal hyphae with localized green fluorescence in the nuclei during late plant infection. hpi =  hours post-inoculation.

Figure 4. Melanin deficiency and its effects on fungal morphology and responses to stressors.

A–D. Conidial chains produced by each strain. Insets: fungal growth of each strain showing colony color and pink pigment secreted by Δamr1 mutants during saprophytic growth on PDA. E. Pink exudate from the tips of conidial chains of the Δamr1 mutant. F. Germination and hyphal growth after UV irradiation. G. Growth comparisons between mutants and wild type either under high temperature (33°C), or in the presence of the indicated chemical. Abbreviations: Δamr1 = Amr1 deletion mutant; Δamr1:Amr1 = Δamr1 mutant complemented with the Amr1 allele; Δamr1:TrpCp-Amr1 = mutant that constitutively expresses the Amr1 gene under control of the TrpC promoter.

Figure 5. Effect of pectin on the vegetative growth of four strains of Δamr1 mutants and wild-type Alternaria brassicicola.

A–D. Data show mean dry weight in milligrams. Each graph is the result of an independent experiment. The poor correlation between the absolute amount of dry biomass and the length of the incubation period was partially due to differences in the number of conidia used in each experiment. Values above the bars indicate the percent increase in dry biomass of the mutant compared to the wild type. Hours under each chart show the length of the incubation period. ** indicates p<0.01. Error bars represent standard deviation. wt = wild type.

Figure 6. Comparison of differentially expressed genes in the Δamr1 and Δabvf19 mutants compared to wild-type Alternaria brassicicola.

A. Comparison between up-regulated genes in the Δamr1 mutant and down-regulated genes in the Δabvf19 mutant during the late stage of infection. B. Number of differentially expressed glycoside hydrolase genes. C. Comparison of expression levels for the 24 genes differentially expressed in the Δamr1 and Δabvf19 mutants compared to the wild type. Y-axis shows Log₂ (mutant)_expression-Log₂ (wild type)_expression).

Figure 7. Expression of melanin biosynthesis-associated genes and four hydrolytic enzyme-coding genes.

A–C. Relative transcription abundance of each gene was determined in comparison to actin gene transcripts in the same tissue. Y-axes show relative abundance of the transcripts compared to the actin gene. D. Expression ratio between the Δamr1 and wild type during late stage of infection. A total of three biological replicates (N = 3) were used for this study. Bars represent standard error. wt = wild type, Δa = Δamr1:Amr1p-GFP, GY = glucose yeast extract broth. SCD1 = Scytalone dehydratase, Brn1 = T3HN reductase, Brn2 = T4HN reductase, Cbh7 = cellobiohydrolase, Amr1 = Alternaria melanin regulation, chymo = chymotrypsin.

Figure 8. Schematic diagram of the PCR strategy used to make each construct.

A. Construct for replacement of the Amr1 gene with a Hygromycin B resistance cassette. B. Construct for replacement of the Amr1 gene with a GFP coding region and Hygromycin B resistance cassette. C. Amplification of the wild-type allele of the Amr1 gene. D. Construct for the constitutive Amr1 expression cassette. E. Construct for the Amr1-GFP fusion protein expression.