The powdery mildew-resistant Arabidopsis mlo2 mlo6 mlo12 triple mutant displays altered infection phenotypes with diverse types of phytopathogens

Abstract

Arabidopsis thaliana mlo2 mlo6 mlo12 triple mutant plants exhibit complete immunity against infection by otherwise virulent obligate biotrophic powdery mildew fungi such as Golovinomyces orontii. While this phenotype is well documented, the interaction profile of the triple mutant with other microbes is underexplored and incomplete. Here, we thoroughly assessed and quantified the infection phenotypes of two independent powdery mildew-resistant triple mutant lines with a range of microbes. These microorganisms belong to three kingdoms of life, engage in diverse trophic lifestyles, and deploy different infection strategies. We found that interactions with microbes that do not directly enter leaf epidermal cells were seemingly unaltered or showed even enhanced microbial growth or symptom formation in the mlo2 mlo6 mlo12 triple mutants, as shown for Pseudomonas syringae and Fusarium oxysporum. By contrast, the mlo2 mlo6 mlo12 triple mutants exhibited reduced host cell entry rates by Colletotrichum higginsianum, a fungal pathogen showing direct penetration of leaf epidermal cells comparable to G. orontii. Together with previous findings, the results of this study strengthen the notion that mutations in genes MLO2, MLO6 and MLO12 not only restrict powdery mildew colonization, but also affect interactions with a number of other phytopathogens.

Figure 1

The G. orontii resistance phenotype of the mlo2-6 mlo6-4 mlo12-8 triple mutant is indistinguishable from the mlo2-5 mlo6-2 mlo12-1 triple mutant. Six-week-old Arabidopsis plants were touch-inoculated with G. orontii conidiospores. (A) Scheme depicting the T-DNA insertion sites in MLO2, MLO6 and MLO12. Rectangles represent exons, black lines introns. Triangles symbolize the T-DNA insertion sites of the various mlo alleles. Lines flanked by inverted arrows (primer binding sites) below the gene models indicate the RT-PCR amplicons used to test for MLO transcript accumulation in the mutant lines. (B) RT-PCR analysis of MLO2, MLO6 and MLO12 transcript accumulation. Primer pairs covering the regions indicated in panel A were used to amplify the respective transcript amplicons from cDNA of lines mlo2-5 mlo6-2 mlo12-1 and mlo2-6 mlo6-4 mlo12-8 (two individuals each) as well as Col-0 wild-type plants (positive control). RT-PCR reactions without reverse transcription (control 1) and amplification without template (control 2) served as negative controls. White arrowheads indicate RT-PCR products of the expected size in case of Col-0 wild-type plants. (C) Representative macroscopic infection phenotypes at 8 dpi. (D) Light micrographs visualizing fungal pathogenesis at 48 hpi. Leaf samples were cleared in destaining solution and fungal infection structures subsequently stained with Coomassie Brillant Blue. Bars = 100 µm. (E) Quantitative assessment of host cell entry. Data show the mean ± standard error of the mean (SEM) from three experiments. In each experiment, at least 100 interaction sites from 1-3 leaves of five independent plants per genotype were assessed (total of > 500 interaction sites per genotype and experiment). *** Indicates a statistically significant difference from Col-0 (P < 0.001) according to a GLM test (binomial distribution) on n = 3 independent experimental replicates.

Figure 2

The mlo2 mlo6 mlo12 triple mutants show an unaltered H. arabidopsidis phenotype. Sixteen-day-old seedlings were spray-inoculated with H. arabidopsidis (isolate Noco2) and spore formation (spores per g FW) was quantitatively assessed at 7 dpi. Sporulation was normalized relative to accession Col-0 set as 100%. Data show the mean ± SEM based on five independent experiments. Accession Ler served as resistant control, the mutant eds1-2 as super-susceptible control. ** Indicates a statistically significant difference from Col-0 (P < 0.01) according to a Wilcoxon-Mann-Whitney rank sum test on n = 5 independent experimental replicates.

Figure 3

The mlo2 mlo6 mlo12 triple mutants show an unaltered A. laibachii infection phenotype. Six-week-old plants were spray-inoculated with A. laibachii (isolate Nc14) and infection phenotypes were assessed at 10 dpi. (A) Representative examples of inoculated rosette leaves of accessions Col-0, Ksk-1 and the two mlo2 mlo6 mlo12 triple mutant lines. Insets show a magnification of an inoculated leaf area. (B) Quantitative evaluation of the infection rate. Data show the mean ± standard deviation (SD) of the proportion of rosette leaves with disease symptoms (macroscopically visible white rust pustule formation) based on 5-6 plants per genotype and 12-30 evaluated leaves per plant. ** Indicates a statistically significant difference from Col-0 (P < 0.01) according to a Wilcoxon-Mann-Whitney rank sum test.

Figure 4

The mlo2 mlo6 mlo12 triple mutants show decreased host cell entry by C. higginsianum. Infection phenotypes of Col-0, mlo2-5 mlo6-2 mlo12-1 and mlo2-6 mlo6-4 mlo12-8 at 3 dpi with C. higginsianum (isolate IMI349063A). Plants were spray-inoculated with spore suspension (5 × 10⁵ spores ml⁻¹). (A) Representative examples of whole leaves cleared and stained with Trypan blue (left column) and light micrographs showing leaf epidermal cells after clearing in choral hydrate (right column). Biotrophic hyphae (arrows) are visible beneath some melanised appressoria. Bars = 30 μm. (B) Quantitative assessment of host cell entry. Data show the mean ± SD from counts of at least 140 appressoria from each leaf (one leaf from each of 3 different plants), i.e. at least 420 appressoria per plant genotype. ***Indicates a statistically significant difference from Col-0 (P < 0.001) according to a GLM test (Poisson distribution) on n = 3 technical replicates (individual plants). The experiment was repeated once with similar results (Fig. S1).

Figure 5

The mlo2 mlo6 mlo12 triple mutants show enhanced disease symptoms upon challenge with F. oxysporum. Two-week-old Arabidopsis seedlings were inoculated with a spore suspension of F. oxysporum (isolate Fo5176). (A) Infection phenotypes were scored at 5, 7 and 10 dpi by assigning a disease index on a 0 (no symptoms) to 5 (severe disease symptoms) scale. Data shown are from a representative experiment and based on 15-20 seedlings per genotype. Each symbol in the categorical scatter plot (circle, square or triangle) represents the infection phenotype of one seedling. The crosses indicate the mean values ± SEM. *, ** and *** indicate statistically significant differences from Col-0 (P < 0.05, P < 0.01 and P < 0.001, respectively) according to a GLM test (Poisson distribution) on n = 15-20 technical replicates (individual seedlings). The experiment was repeated twice with similar results (Fig. S2). (B) Representative macroscopic infection phenotypes at 7 dpi.

Figure 6

The mlo2 mlo6 mlo12 triple mutants show an elevated bacterial titre upon challenge with P. syringae. Five-week-old Arabidopsis plants were infiltrated with P. syringae pv. maculicola lux (OD₆₀₀ = 0.0005) and luminescence (RLU cm⁻²; corresponding to bacterial titre) determined at 0 dpi (to ensure an equal start inoculum) and 3 dpi. (A) The boxplot shows data from one representative experiment based on n = 7 (0 dpi) and n = 11 plants (3 dpi) per genotype, with each plant value representing the mean of three leaves. Centre lines mark the medians, upper and lower box limits indicate the 25^th and 75^th percentiles, respectively; upper and lower whiskers extend 1.5 times the interquartile range from the 25^th and 75^th percentiles, respectively; and dots represent outliners. Letters indicate statistically different groups (at least P < 0.05) according to a GLM test (quasi-Poisson distribution) on n = 11 technical replicates (individual plants). The experiment was repeated four times with n = 7-13 plants per genotype in each experiment and in two different controlled environments (see Materials and Methods for details) with comparable results (Fig. S3). (B) Representative macroscopic infection phenotypes at 3 dpi. Yellow arrows indicate the inoculated leaves.

Figure 7

The mlo2 mlo6 mlo12 triple mutants show unaltered colonization by S. indica. Roots of 2-week-old plants were inoculated with S. indica (isolate DSM11827) and the rate of root colonization determined by quantifying the relative amount of fungal genomic DNA by qPCR analysis. Data show the mean ± SEM based on three experiments with 240 seedlings per genotype and experiment. There is no statistically significant difference between genotypes according to GLM (quasi-Poisson distribution) on n = 3 independent experimental replicates.