Battle of the bugs: how an oomycete pathogen shapes the microbiota of its host

Abstract

Plants harbor a complex community of prokaryotic and eukaryotic microbes in their above- and belowground tissues, collectively referred to as their ‘microbiota’, that plays a central role in their well-being by promoting growth, health, and resilience (Gross, 2022). Moreover, during pathogen attack, plants actively attract microbes into their microbiota from the surrounding environment that can mitigate disease either by directly antagonizing the pathogen, or through stimulation of plant immune responses (Teixeira et al., 2019). In addition to host-plant genetics and environmental cues, intermicrobial interactions greatly influence microbiota compositions (Singh et al., 2023). In an article published in this issue of New Phytologist, Gómez-Pérez et al. (2023, 2320–2334) describe the investigation of molecular mechanisms underlying intermicrobial interactions between a pathogenic oomycete and its host-plant microbiota (Gómez-Pérez et al., 2023). Using a novel approach, the authors show that oomycete pathogens release proteins with selective antimicrobial activity to restructure host microbial communities to promote host colonization (Fig. 1).

Oomycetes are filamentous eukaryotes that belong to the Stramenopiles lineage and are closely related to diatoms and brown algae. Some of the most devastating plant and animal pathogens are oomycetes, and these organisms cause significant economic losses and represent major threats to global food security and ecosystem stability. As with plant-associated fungi, the symbiotic relationships between plants and pathogenic oomycetes have mostly been studied in binary interactions, involving oomycetes with particular plant hosts. These studies identify many oomycete effectors that target plant regulatory networks to facilitate host colonization (Wang et al., 2019). However, they ignore the host-associated microbiota. Such oomycete effectors fit the ‘traditional’ definition as pathogen-specific small, cysteine-rich proteins that are secreted in planta to suppress host immunity. However, this dogma has been challenged by findings that effectors are also secreted by mutualistic, endophytic, and even saprophytic fungi, suggesting that they are better described as molecules that are secreted by microbes to beneficially manipulate their direct environment (Snelders et al., 2022). Taken together with the intimate association between plants and their microbiota, and the importance of the microbiota for plant health, this broader definition of effectors has led to the hypothesis that pathogens have evolved to target host-associated microbiota, in addition to the endogenous immune system of the host, using effectors to facilitate disease establishment (Snelders et al., 2018). Indeed, several fungal pathogens were recently shown to secrete effectors with antimicrobial activities that directly manipulate the microbiota of their hosts (Snelders et al., 2020, 2021, 2023; Chavarro-Carrero et al., 2023; Ökmen et al., 2023). Whether this also applied to oomycete plant pathogens was unknown until now.

In an earlier study, the authors showed that the plant pathogenic oomycete Albugo laibachii is a keystone member of the Arabidopsis thaliana phyllosphere microbiota that significantly affects epi- and endophytic bacterial colonization (Agler et al., 2016). However, how Albugo mechanistically affects phyllosphere microbiota compositions remained enigmatic. Based on analyses of environmental samples of wild A. thaliana throughout multiple years and locations, Gómez-Pérez and co-authors have now confirmed that this Albugo sp. is highly integrated in the phyllosphere microbiota, most particularly through negative associations. From this important conceptual observation, the authors continued to identify three A. candida proteins in the apoplast of infected plants that displayed selective antimicrobial activity as a means to target the A. thaliana phyllosphere microbiota by inhibiting microbes that determine community stability. This unequivocally demonstrates that oomycete plant pathogens manipulate host-associated microbiota through in planta-secreted antimicrobials (Fig. 1). Importantly, with this finding the authors show that microbiota manipulation is a general mechanism for niche establishment of plant pathogens across kingdoms.

Interestingly, despite their evolutionary distance, fungi and oomycetes display strikingly similar strategies in their interactions with plant hosts. For example, to prevent the release of immunogenic fragments from their cell walls by plant hydrolases, and the recognition of these molecules by plant receptor proteins, both oomycetes and fungi modify their cell wall components or secrete glycan-binding effector molecules (Rovenich et al., 2016). These effectors either form a protective layer on cell walls, inhibit host hydrolases, or sequester glycan fragments to prevent their recognition. Other types of effectors are released by oomycetes and fungi to interfere with conserved host immune components, including the disruption of reactive oxygen species (ROS) homeostasis, ethylene signaling, and autophagy (Wilson & McDowell, 2022). Considering that effectors which directly target the plant host, as well as the associated microbiota, exist in these distantly related organisms, and exert similar functions, the underlying mechanisms may have ancient evolutionary origins. Like fungi, oomycetes colonized microbe-rich aquatic and terrestrial habitats before land plants evolved. Hence, particular effectors with antimicrobial activity may have evolved from antimicrobial proteins that originally functioned in microbial competition before the appearance of land plants, and consequently the evolution of plant pathogenicity (Snelders et al., 2022). This has been suggested for the antifungal effector AMP3 of the soilborne pathogen Verticillium dahliae, which likely evolved from ancient antimicrobial proteins of terrestrial fungal ancestors (Snelders et al., 2021).

Obligate biotrophic pathogens are specialized to colonize relatively narrow host ranges. Consequently, these pathogens are likely to encounter a smaller diversity of microbes than facultative or broad host-range pathogens. Thus, specialized pathogens may have evolved antimicrobial effectors to specifically manipulate a limited number of microbes to promote niche establishment on their plant hosts. In this study, Gómez-Pérez et al. demonstrate that the three identified antimicrobials secreted by A. candida display selective antimicrobial activity against Gram-positive bacteria. This contrasts with earlier findings for two other narrow host-range pathogens, whose specialized lifestyles seem to be contradictory to the broad activity spectra of the antimicrobial effectors they secrete. The foliar pathogen Zymoseptoria tritici and the smut fungus Ustilago maydis secrete the active ribonucleases Zt6 and Ribo1, respectively, which display potent antimicrobial activity (Kettles et al., 2018; Ökmen et al., 2023). One obvious distinction between the antimicrobial effectors of Z. tritici and U. maydis, and those of A. candida, are the different colonization stages during which they act. The infection cycles of all three pathogens begin on the plant-host surface (episphere), which is characterized by the presence of a complex microbiota. Subsequently, hyphae penetrate host tissue and colonize the endosphere. Such endosphere colonization typically requires specialization, enabling microbes to overcome adverse conditions in the plant episphere. Hence, endosphere microbial densities are typically lower. Thus, effectors with a broad activity spectrum, like Zt6 and Ribo1, may have evolved to target the large breadth of microbes in the highly competitive episphere, while proteins with highly selective antimicrobial activities, like those secreted by A. candida, may be required to manipulate specific, potentially antagonistic, microbes in the endosphere, and perhaps are the result of a long coevolutionary trajectory (Snelders et al., 2022).

The A. candida antimicrobials display selective activity against five bacterial isolates of the A. thaliana core microbiota, one of which determines microbial community stability as shown by dropout experiments. Interestingly, growth of this strain is suppressed by one of the antimicrobial effectors in planta. Considering that Albugo infection results in a reduced alpha diversity of the A. thaliana phyllosphere community (Agler et al., 2016), these results suggest that A. candida targets bacterial core taxa to cause dysbiosis.

Most of the antimicrobial proteins characterized to date are (predicted to be) small, cysteine-rich proteins with typical tight three-dimensional structures and recognized as toxin- or defensin-like folds. These criteria are often used in in silico pipelines to identify novel antimicrobials. In a novel approach, Gómez-Pérez et al. combined proteomics with antimicrobial activity prediction to identify antimicrobial protein candidates in apoplastic fluid isolated from A. candida-infected A. thaliana. To sidestep the limitation of protein size bias, which exists in most antimicrobial prediction tools, the authors used a weighted score resulting in candidates of 500 amino acids on average, which is considerably larger than ‘typical’ antimicrobials. Moreover, proteins with intrinsically disordered regions (IDRs) were highly overrepresented. Such regions lack secondary or tertiary structures under physiological conditions, but can fold in a stimulus-dependent manner (Marin et al., 2013). This structural flexibility may facilitate translocation, function, and evasion of recognition of bacterial effectors (Marin et al., 2013) and could have similar advantages for fungal and oomycete antimicrobials. In humans, so-called cationic intrinsically disordered antimicrobial peptides (CIAMPs), which are defined as any linear peptide with a high percentage of disorder-promoting amino acids and a positive net charge, display potent microbicidal activity depending on environmental conditions (Latendorf et al., 2019). Interestingly, the IDRs of the antimicrobials tested here were sufficient to inhibit bacterial growth in vitro and their activity positively correlated with their net charge. Thus, with their work, the authors provide new insights into structural characteristics of antimicrobial proteins that should be integrated into approaches for the identification of novel antimicrobial effectors, which are currently still heavily biased toward sequence and structural homology to ‘typical’ AMPs.