Gunther Döhlemann

Establishment of compatibility in biotrophic interactions

The basidiomycete fungi Ustilago maydis and Ustilago hordei parasitize their respective host plants maize and barley to cause smut disease. The pathogens establish biotrophic interactions, in which infected plant cells infected stay alive throughout the entire disease cycle. While U. hordei infections are systemic and symptoms are only produced in the inflorescences, plant tumors induced by U. maydis can appear at basically all aerial parts of the maize plant. In U. maydis infected maize plants, the metabolism is reprogrammed and carbohydrate fluxes are redirected towards the infected tissue in which massive proliferation of fungal hyphae occurs. Our research is focussed on the mechanisms of biotrophic pathogens to accommodate in the host tissue. Specifically, we are investigating i) how fungal effector proteins suppress host immunity and ii) which host factors are required to establish compatibility.

Functional analysis of U. maydis effector proteins

U. maydis does not form specialized intracellular infection structures such as the haustoria that are formed e.g. by plant pathogenic rust fungi. This suggests that the biotrophic interface, which surrounds the infectious hyphae is the major site of communication in this pathosystem.

The secreted effector protein Pep1 is specifically expressed during pathogenic development of U. maydis. Deletion mutants for pep1 show no defect during saprophytic growth but are arrested upon host penetration and elicit a hypersensitive response of the plant. Thus, a biotrophic interaction is not established and the infected plant tissue shows various defense responses, particularly the production of reactive oxygen species (ROS). We identified Pep1 as an efficient inhibitor of the plant oxidative burst response, which is elicited by so called pathogen-associated molecular patterns (PAMPs). This inhibitory effect of Pep1 on the early plant defense is explained by an inhibition of plant peroxidase activity through Pep1. Suppression of this conserved component of the plant ROS generating system by a pathogen effector therefore shows a new strategy of biotrophic pathogens to suppress innate plant immunity. Another secreted effector of U. maydis is Pit2. The pit2 gene is located in a small cluster of four genes (the pit-cluster) of which two encode proteins being essential for U. maydis virulence, the membrane protein Pit1 and the secreted effector Pit2. Similarly to Pep1, Pit2 is specifically produced during plant colonization and it is also secreted to the biotrophic interface. However, while pep1 mutants are blocked by early plant defenses, pit2 deletion mutants are able to establish a biotrophic interaction but fail to induce tumor formation and elicit host defense at later steps of infection. We are currently studying the molecular function of Pit2 to identify the host processes this effector is interfering with.

Organ specificity in the maize - U. maydis interaction

Unlike other closely related smut fungi, U. maydis can infect all aerial parts of its host plant. After active penetration of the respective tissues in the infected organs, various cell types such as epidermal and mesophyll cells as well as vascular tissue can be colonized. Such tissues and cell types differ in structure and physiology; a situation U. maydis must be prepared to deal with (Figure 1). In primordia of seedling leaves U. maydis mainly colonizes mesophyll and, in particular, vascular tissue/bundle sheath cells. Here, the fungal hyphae proliferate and grow along the leaf axis (Figure 1, left panel).

However, in anthers, there is only a single central vascular bundle; the locules, which comprise most of the anther, do not contain vascular tissues. Mature anther locules consist of five cell layers each composed of one cell type (Figure 1, middle panel) and nutrient/signal exchange in anthers is thought to happen primarily through plasmodesmata and apoplastic movement of materials from the vasculature through the connective tissue into the locules. Infection by U. maydis dramatically changes anther development: they are transformed into tumors and fail to develop meiocytes. In anther tumors, intracellular fungal hyphae are colonizing this densely packed conglomerate of undifferentiated tumor cells (Figure 1, right panel).

Fig. 1 | Biotrophic development of U. maydis in seedling leaf tissue and anthers. Left panel: Biotrophic U. maydis colonizing a seedling leaf (4 dpi). Fungal hyphae (WGA-AF488: green) grow both intra- and intercellularly, mostly around vascular bundles that are indicated by strong cell wall fluorescence (red). Middle panel: Maize anthers are four lobes surrounding the connective tissue; centrally there is a single vascular strand in the middle of the connective tissue. Architecture of maize anther locules with 5 distinct cell layers (numbered 1-5) as indicated schematically and in a confocal projection. Red: propidium iodide staining of plant DNA; grey: plant cell walls. C: U. maydis infected maize anther 8 dpi. Right panel: The defined structure of the anther is completely obscured, the anther tumor consists of undifferentiated, densely packed plant cells (upper right image). In the anther tumor cells, fungal hyphae (WGA-AF488: green) grow intracellularly (lower right image, arrows).

In a previous study on U. maydis mutants with deletions for clusters encoding secreted effectors, more than 50% of the mutants showed no phenotype or only marginal virulence defects (Kämper et al. 2006). This raised the question, why U. maydis has retained these effector genes. In this context, we hypothesized that those effectors might allow U. maydis to tailor its weaponry to the specific conditions characteristic for different host organs. Parallel transcriptome profiling of both U. maydis and Z. mays genes in seedlings, adult leaves and tassels, validated this hypothesis and revealed dramatic differences in gene expression of both the host and the pathogen in the different organs. To functionally test the organ-specific role of U. maydis effectors, the cluster deletion mutants were inoculated on adult leaves and tassels as well as seedlings. Strikingly, five of the U. maydis mutants showed organ-specific virulence phenotypes. This finding of organ specific virulence factors sets a new paradigm for virulence of fungal plant pathogens.

We are currently investigating the organ-specific functions of individual effector proteins to elucidate how U. maydis adapts to different host environments and deploys specific effector proteins that comprise cell-type specific host targets.

Suppression of host immunity

In transcriptome analyses of U. maydis infected maize plants, a large number of maize genes was regulated during distinct steps of fungal infection. To functionally analyze host being relevant for compatibility with U. maydis, a virus induced gene silencing (VIGS) protocol was established that allowed systemic silencing of maize genes during fungal infection. Using this approach, a maize terpene synthase (TPS6) was found to restrict U. maydis infection. In contrast, expression of the conserved suppressor of programmed cell death, Bax-Inhibitor-1 (BI-1) was required for fungal virulence, i.e. silencing of BI-1 led to a reduction of host colonization by U. maydis and caused increased plant cell death at sites of infection. Recently, we identified an apoplastic maize cystatin to be highly induced at the early biotrophic interaction with U. maydis wild type but not upon infection of the incompatible pep1 deletion mutant. VIGS of the cystatin revealed that it is essential for U. maydis infection. In cystatin silenced plants, U. maydis infection is terminated by plant cell death. As trigger for this plant defense response we identified a set of apoplastic cysteine proteases (ACPs) that can be activated by salicylic acid (SA) treatment and all these proteases can be inhibited by the cystatin. Consequently, the cystatin was found to block SA-triggered defense including cell death when being present in the apoplast. We now aim to elucidate, which roles the individual ACPs have in maize defense. To identify the signals that trigger SA-associated plant defense responses we are particularly interested in the targets of these proteases.

Fig. | 2 Induction of maize defense by apoplastic cysteine proteases (ACPs) and their inhibition by a maize cystatins (CC). SA-associated defense induction is activated via ACPs and finally leads to plant cell death. In a biotrophic interaction, JA signaling activates cystatin expression. This inhibits activation of ACPs and thereby blocks plant defense. JA: Jasmonic Acid; SA: Salicylic Acid; ROS: Reactive Oxygen Species;?: ACP targets are unknown.

Cell death regulation in barley - Ustilago interactions

In plants, programmed cell death is not only a fundamental process during all vegetative and reproductive stages of plant development, but also an essential defense mechanism during pathogen attack. Therefore, the prevention of cell death is essential to biotrophic plant pathogens such as Ustilago hordei. In the compatible interaction of U. hordei and barley, cell death is fully prevented and no macroscopic symptoms of infection are visible. In contrast, U. hordei deletion mutants of the secreted effector protein Pep1 cause death of the infected host cells. Using transgenic barley lines overexpressing BI-1, we are investigating the impact of programmed cell death on non-host resistance to U. maydis as well as host- resistance to the U. hordei pep1 deletion mutant. We are combining confocal microscopy, gene expression analyses and enzyme activity assays to dissect the individual pathways that are involved in host cell death during these incompatible interactions. In parallel, we are studying U. hordei gene expression on the whole genome level to identify effector proteins that are involved in suppression of host cell death