Secreted fungal effector lipase releases free fatty acids to inhibit innate immunity-related callose formation during wheat head infection

Abstract

The deposition of the (1,3)-β-glucan cell wall polymer callose at sites of attempted penetration is a common plant defense response to intruding pathogens and part of the plant's innate immunity. Infection of the Fusarium graminearum disruption mutant Δfgl1, which lacks the effector lipase FGL1, is restricted to inoculated wheat (Triticum aestivum) spikelets, whereas the wild-type strain colonized the whole wheat spike. Our studies here were aimed at analyzing the role of FGL1 in establishing full F. graminearum virulence. Confocal laser-scanning microscopy revealed that the Δfgl1 mutant strongly induced the deposition of spot-like callose patches in vascular bundles of directly inoculated spikelets, while these callose deposits were not observed in infections by the wild type. Elevated concentrations of the polyunsaturated free fatty acids (FFAs) linoleic and α-linolenic acid, which we detected in F. graminearum wild type-infected wheat spike tissue compared with Δfgl1-infected tissue, provided clear evidence for a suggested function of FGL1 in suppressing callose biosynthesis. These FFAs not only inhibited plant callose biosynthesis in vitro and in planta but also partially restored virulence to the Δfgl1 mutant when applied during infection of wheat spikelets. Additional FFA analysis confirmed that the purified effector lipase FGL1 was sufficient to release linoleic and α-linolenic acids from wheat spike tissue. We concluded that these two FFAs have a major function in the suppression of the innate immunity-related callose biosynthesis and, hence, the progress of F. graminearum wheat infection.

Figure 1.

Callose deposition in wheat spikes after fungal infection. A, Sections of directly inoculated wheat spikelet comprising the transition zone of rachilla and rachis including vascular bundles, as indicated in B, at 7 dpi with F. graminearum wild type (F.g. wt), lipase-deficient strain Δfgl1-2 (F.g. Δfgl1-2), and P. teres wild type (P.t. wt). Untreated spikelets served as controls. Micrographs were taken by confocal laser-scanning microscopy. Green color was assigned to fungal hyphae stained with Alexa Fluor 488-conjugated wheat germ agglutinin, and grayscale was assigned to aniline blue-stained callose. Green and grayscale channels were merged for an overview of fungal colonization of spikelet tissue, and grayscale only (aniline blue) was used to highlight callose plugs (white arrows) in the phloem of vascular bundles of the transition zone. Micrographs are representative for infection at 7 dpi. Bars = 200 µm (left column) and 100 µm (middle column). The right column provides a macroscopic overview of the disease severity of directly inoculated spikelets at 14 dpi. B, Longitudinal section of wheat spikelets and rachis to highlight the transition zone of rachilla and rachis including vascular bundles (white arrows). The orange frame indicates the area of microscopic analysis in A. Bars = 1 mm. C, Number of callose plugs located in the phloem of vascular bundles of the transition zone of rachilla and rachis tissue from directly inoculated spikelets as analyzed in A at 1, 3, 7, and 14 dpi with F. graminearum wild type, lipase-deficient strain Δfgl1-2, and P. teres wild type. Untreated spikelets served as controls. Values represent means of eight biologically independent inoculation experiments for each fungal strain. The letters a, b, and c indicate P < 0.05 by Tukey’s test. Error bars represent se, and n = 8. D, High-resolution microscopy of callose deposition (fluorescence by aniline blue staining) in sieve elements (indicated by dotted lines) of the phloem in the transition zone of rachilla and rachis of control and infected wheat spikelets. Callose plugs in sieve elements of Δfgl1-2-infected spikelets are representative for P. teres-induced callose plugs. cd, Callose deposit; se, sieve element; sp, sieve plate. Bars = 20 µm.

Figure 2.

FFA and 2-DDG impact on callose biosynthesis and disease phenotypes of F. graminearum-infected wheat spikes. A, Callose synthase activity of membranes isolated from untreated wheat spikes at Zadoks stages 7.5 to 7.9 (Zadoks et al., 1974). Ethanol-dissolved FFAs and the callose synthesis inhibitor 2-DDG were added to membrane preparations at the final concentrations indicated. FFAs were palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), linoleic acid (18:2), and α-linolenic acid (18:3). B, Callose synthase activity of membranes isolated from spikelets and the adjacent rachis region 24 h after FFA and 2-DDG application. Four microliters of ethanol-dissolved FFAs or 2-DDG (concentration, 3.2 mm) and pure ethanol (EtOH) was added to wheat spikelets at Zadoks stages 7.02 to 7.05 (Zadoks et al., 1974). For A and B, values represent means of two biologically independent experiments. The letters a to d indicate P < 0.05 by Tukey’s test. Error bars represent se, and n = 6. C, Inoculation of two central spikelets with 10 µL of water (control [c]) and F. graminearum strains wt-GFP-1 and Δfgl1-GFP-1. Four microliters of ethanol-dissolved FFAs and 2-DDG (concentration, 3.2 mm) and pure ethanol was added to each of the previously inoculated spikelets at 3 dpi. Arrows indicate the inoculation sites. Pathogenicity tests were repeated 10 times for strain wt-GFP-1 and 20 times for strain Δfgl1-GFP-1. Spikes are representative for infection severity at 21 dpi. wt-GFP-1 inoculation is representative for infection with or without the addition of FFAs or 2-DDG (Table I). D, Micrographs of isolated stigma and pistil from a floret at 1 dpi. Hyphal growth was directed from pollen on stigmas toward the ovary. No differences were seen between wt-GFP-1 and Δfgl1-GFP-1 growth at 1 dpi. Bars = 40 µm. E, Number of callose plugs located in phloem of the vascular bundles of the transition zone of rachilla and rachis tissue from spikelets directly inoculated with Δfgl1-2 supplemented (as described in C) with either α-linolenic acid (F.g. Δfgl1-2 + 18:3) or 2-DDG (F.g. Δfgl1-2 + 2-DDG) at 7 dpi, including data from an analysis of F. graminearum wild type (F.g. wt), lipase-deficient strain Δfgl1-2 without supplementation, and the control (Fig. 1C). Values represent means of eight biologically independent inoculation experiments for each fungal strain. The letters a and b indicate P < 0.05 by Tukey’s test. Error bars represent se, and n = 8. F, Section of a directly inoculated wheat spikelet comprising the transition zone of rachilla and rachis including vascular bundles at 7 dpi with Δfgl1-2 supplemented with α-linolenic acid. Micrographs were taken by confocal laser-scanning microscopy. Green color was assigned to fungal hyphae stained with Alexa Fluor 488-conjugated wheat germ agglutinin, and grayscale was assigned to aniline blue-stained callose. Micrographs are representative for infection at 7 dpi. Bars = 200 µm (left) and 100 µm (right).

Figure 3.

FFA and 2-DDG impact on the progress of F. graminearum infection and mycotoxin production in wheat spikes. A, Sections of directly inoculated spikelets comprising the transition zone of rachilla and rachis (section A), the transition zone of rachilla and rachis of spikelets above inoculated spikelets (section B), and the rachis region below inoculated spikelets (section C). Inoculation was with F. graminearum wild type (F.g. wt), lipase-deficient strain Δfgl1-2 (F.g. Δfgl1-2), and Δfgl1-2 supplemented with α-linolenic acid (F.g. Δfgl1-2 + 18:3). Supplementation was by the addition of 4 µL of ethanol-dissolved 18:3 (concentration, 3.2 mm) to each of the previously inoculated spikelets at 3 dpi. Micrographs were taken by confocal laser-scanning microscopy at 14 dpi. Green color was assigned to fungal hyphae stained with Alexa Fluor 488-conjugated wheat germ agglutinin, and grayscale was assigned to aniline blue-stained callose. Micrographs are representative for the infection progress in spikelets inoculated with F. graminearum wild type (no differences with or without the addition of FFAs or 2-DDG) and lipase-deficient strain Δfgl1-2. Changes in infection progress due to the addition of unsaturated FFAs or 2-DDG to Δfgl1-2-infected spikelets are represented by images of infection after the addition of 18:3. Bars = 200 µm. B, DON concentrations of wheat spike tissue from sections A, B, and C (compare the schematic overview in A) at 7 and 14 dpi with F. graminearum wild type, lipase-deficient strain Δfgl1-2, and Δfgl1-2 supplemented with α-linolenic acid or 2-DDG. Values represent means of two biologically independent experiments. The letters a, b, and c indicate P < 0.05 by Tukey’s test. Error bars represent se, and n = 3.

Figure 4.

FFA concentrations in wheat spikes after lipase application and F. graminearum infection. A, FFA concentrations in wheat spikelets after application of the purified F. graminearum effector lipase FGL1. Untreated spikelets served as controls. Samples were taken at 0, 0.5, 3, and 16 h after application. FFAs were palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), linoleic acid (18:2), and α-linolenic acid (18:3). B, FFA concentrations in wheat spikelets inoculated with F. graminearum wild type (F.g. wt) and the lipase-deficient mutant Δfgl1-2 (F.g. Δfgl1-2). Uninfected spikelets served as controls. Samples were taken at 3, 5, and 7 dpi. Values represent means of two biologically independent experiments. *Statistically different from the control; ⁺statistically different from Δfgl1-2 at P < 0.05 by Tukey’s test. Error bars represent se, and n = 6.

Figure 5.

Effector lipase-dependent type II resistance to fungal infection in wheat. The model shows the determination of type II resistance to fungal infection in wheat based on our results of pathogen-caused suppression of the plant defense-related callose formation. Left, susceptible wheat characterized by callose synthases that are strongly inhibited by polyunsaturated FFAs due to pathogen-derived effector lipase activity; middle, application of polyunsaturated FFAs (partially) restores the virulence of lipase-deficient fungal pathogen unable to release FFAs due to missing lipase; right, lipase-deficient fungal pathogen (or incompatible pathogen) without release of FFAs due to missing lipase. Pathogen-induced callose formation is not inhibited and promotes type II resistance in planta.