Accurate plant pathogen effector protein classification ab initio with deepredeff: an ensemble of convolutional neural networks

Definition of the effector landscape across 13 phytoplasma proteomes with LEAPH and EffectorComb

'Candidatus Phytoplasma' genus, a group of fastidious phloem-restricted bacteria, can infect a wide variety of both ornamental and agro-economically important plants. Phytoplasmas secrete effector proteins responsible for the symptoms associated with the disease. Identifying and characterizing these proteins is of prime importance for expanding our knowledge of the molecular bases of the disease. We faced the challenge of identifying phytoplasma's effectors by developing LEAPH, a machine learning ensemble predictor composed of four models. LEAPH was trained on 479 proteins from 53 phytoplasma species, described by 30 features. LEAPH achieved 97.49% accuracy, 95.26% precision and 98.37% recall, ensuring a low false-positive rate and outperforming available state-of-the-art methods. The application of LEAPH to 13 phytoplasma proteomes yields a comprehensive landscape of 2089 putative pathogenicity proteins. We identified three classes according to different secretion models: 'classical', 'classical-like' and 'non-classical'. Importantly, LEAPH identified 15 out of 17 known experimentally validated effectors belonging to the three classes. Furthermore, to help the selection of novel candidates for biological validation, we applied the Self-Organizing Maps algorithm and developed a Shiny app called EffectorComb. LEAPH and the EffectorComb app can be used to boost the characterization of putative effectors at both computational and experimental levels, and can be employed in other phytopathological models.

Pyricularia oryzae: Lab star and field scourge

Pyricularia oryzae (syn. Magnaporthe oryzae), is a filamentous ascomycete that causes a major disease called blast on cereal crops, as well as on a wide variety of wild and cultivated grasses. Blast diseases have a tremendous impact worldwide particularly on rice and on wheat, where the disease emerged in South America in the 1980s, before spreading to Asia and Africa. Its economic importance, coupled with its amenability to molecular and genetic manipulation, have inspired extensive research efforts aiming at understanding its biology and evolution. In the past 40 years, this plant-pathogenic fungus has emerged as a major model in molecular plant-microbe interactions. In this review, we focus on the clarification of the taxonomy and genetic structure of the species and its host range determinants. We also discuss recent molecular studies deciphering its lifecycle.

TAXONOMY

Kingdom: Fungi, phylum: Ascomycota, sub-phylum: Pezizomycotina, class: Sordariomycetes, order: Magnaporthales, family: Pyriculariaceae, genus: Pyricularia.

HOST RANGE

P. oryzae has the ability to infect a wide range of Poaceae. It is structured into different host-specialized lineages that are each associated with a few host plant genera. The fungus is best known to cause tremendous damage to rice crops, but it can also attack other economically important crops such as wheat, maize, barley, and finger millet.

DISEASE SYMPTOMS

P. oryzae can cause necrotic lesions or bleaching on all aerial parts of its host plants, including leaf blades, sheaths, and inflorescences (panicles, spikes, and seeds). Characteristic symptoms on leaves are diamond-shaped silver lesions that often have a brown margin and whose appearance is influenced by numerous factors such as the plant genotype and environmental conditions. USEFUL WEBSITES Resources URL Genomic data repositories http://genome.jouy.inra.fr/gemo/ Genomic data repositories http://openriceblast.org/ Genomic data repositories http://openwheatblast.net/ Genome browser for fungi (including P. oryzae) http://fungi.ensembl.org/index.html Comparative genomics database https://mycocosm.jgi.doe.gov/mycocosm/home T-DNA mutant database http://atmt.snu.kr/ T-DNA mutant database http://www.phi-base.org/ SNP and expression data https://fungidb.org/fungidb/app/.

Effector-triggered susceptibility by the rice blast fungus Magnaporthe oryzae

Rice blast, the most destructive disease of cultivated rice world-wide, is caused by the filamentous fungus Magnaporthe oryzae. To cause disease in plants, M. oryzae secretes a diverse range of effector proteins to suppress plant defense responses, modulate cellular processes, and support pathogen growth. Some effectors can be secreted by appressoria even before host penetration, while others accumulate in the apoplast, or enter living plant cells where they target specific plant subcellular compartments. During plant infection, the blast fungus induces the formation of a specialized plant structure known as the biotrophic interfacial complex (BIC), which appears to be crucial for effector delivery into plant cells. Here, we review recent advances in the cell biology of M. oryzae-host interactions and show how new breakthroughs in disease control have stemmed from an increased understanding of effector proteins of M. oryzae are deployed and delivered into plant cells to enable pathogen invasion and host susceptibility.

POOE: predicting oomycete effectors based on a pre-trained large protein language model

Oomycetes are fungus-like eukaryotic microorganisms which can cause catastrophic diseases in many plants. Successful infection of oomycetes depends highly on their effector proteins that are secreted into plant cells to subvert plant immunity. Thus, systematic identification of effectors from the oomycete proteomes remains an initial but crucial step in understanding plant-pathogen relationships. However, the number of experimentally identified oomycete effectors is still limited. Currently, only a few bioinformatics predictors exist to detect potential effectors, and their prediction performance needs to be improved. Here, we used the sequence embeddings from a pre-trained large protein language model (ProtTrans) as input and developed a support vector machine-based method called POOE for predicting oomycete effectors. POOE could achieve a highly accurate performance with an area under the precision-recall curve of 0.804 (area under the receiver operating characteristic curve = 0.893, accuracy = 0.874, precision = 0.777, recall = 0.684, and specificity = 0.936) in the fivefold cross-validation, considerably outperforming various combinations of popular machine learning algorithms and other commonly used sequence encoding schemes. A similar prediction performance was also observed in the independent test. Compared with the existing oomycete effector prediction methods, POOE provided very competitive and promising performance, suggesting that ProtTrans effectively captures rich protein semantic information and dramatically improves the prediction task. We anticipate that POOE can accelerate the identification of oomycete effectors and provide new hints to systematically understand the functional roles of effectors in plant-pathogen interactions. The web server of POOE is freely accessible at http://zzdlab.com/pooe/index.php. The corresponding source codes and data sets are also available at https://github.com/zzdlabzm/POOE.IMPORTANCEIn this work, we use the sequence representations from a pre-trained large protein language model (ProtTrans) as input and develop a Support Vector Machine-based method called POOE for predicting oomycete effectors. POOE could achieve a highly accurate performance in the independent test set, considerably outperforming existing oomycete effector prediction methods. We expect that this new bioinformatics tool will accelerate the identification of oomycete effectors and further guide the experimental efforts to interrogate the functional roles of effectors in plant-pathogen interaction.

Secretome analysis of the phytopathogen Macrophomina phaseolina cultivated in liquid medium supplemented with and without soybean leaf infusion

Macrophomina phaseolina (Tassi) Goid. is a fungal pathogen that causes root and stem rot in several economically important crops. However, most of disease control strategies have shown limited effectiveness. Despite its impact on agriculture, molecular mechanisms involved in the interaction with host plant remains poorly understood. Nevertheless, it has been proven that fungal pathogens secrete a variety of proteins and metabolites to successfully infect their host plants. In this study, a proteomic analysis of proteins secreted by M. phaseolina in culture media supplemented with soybean leaf infusion was performed. A total of 250 proteins were identified with a predominance of hydrolytic enzymes. Plant cell wall degrading enzymes together peptidases were found, probably involved in the infection process. Predicted effector proteins were also found that could induce plant cell death or suppress plant immune response. Some of the putative effectors presented similarities to known fungal virulence factors. Expression analysis of ten selected protein-coding genes showed that these genes are induced during host tissue infection and suggested their participation in the infection process. The identification of secreted proteins of M. phaseolina could be used to improve the understanding of the biology and pathogenesis of this fungus. Although leaf infusion was able to induce changes at the proteome level, it is necessary to study the changes induced under conditions that mimic the natural infection process of the soil-borne pathogen M. phaseolina to identify virulence factors.

The transcriptional landscape of plant infection by the rice blast fungus Magnaporthe oryzae reveals distinct families of temporally co-regulated and structurally conserved effectors

The rice blast fungus Magnaporthe oryzae causes a devastating disease that threatens global rice (Oryza sativa) production. Despite intense study, the biology of plant tissue invasion during blast disease remains poorly understood. Here we report a high-resolution transcriptional profiling study of the entire plant-associated development of the blast fungus. Our analysis revealed major temporal changes in fungal gene expression during plant infection. Pathogen gene expression could be classified into 10 modules of temporally co-expressed genes, providing evidence for the induction of pronounced shifts in primary and secondary metabolism, cell signaling, and transcriptional regulation. A set of 863 genes encoding secreted proteins are differentially expressed at specific stages of infection, and 546 genes named MEP (Magnaportheeffector protein) genes were predicted to encode effectors. Computational prediction of structurally related MEPs, including the MAX effector family, revealed their temporal co-regulation in the same co-expression modules. We characterized 32 MEP genes and demonstrate that Mep effectors are predominantly targeted to the cytoplasm of rice cells via the biotrophic interfacial complex and use a common unconventional secretory pathway. Taken together, our study reveals major changes in gene expression associated with blast disease and identifies a diverse repertoire of effectors critical for successful infection.

Effector Identification in Plant Pathogens

Effectors play a central role in determining the outcome of plant-pathogen interactions. As key virulence proteins, effectors are collectively indispensable for disease development. By understanding the virulence mechanisms of effectors, fundamental knowledge of microbial pathogenesis and disease resistance have been revealed. Effectors are also considered double-edged swords because some of them activate immunity in disease resistant plants after being recognized by specific immune receptors, which evolved to monitor pathogen presence or activity. Characterization of effector recognition by their cognate immune receptors and the downstream immune signaling pathways is instrumental in implementing resistance. Over the past decades, substantial research effort has focused on effector biology, especially concerning their interactions with virulence targets or immune receptors in plant cells. A foundation of this research is robust identification of the effector repertoire from a given pathogen, which depends heavily on bioinformatic prediction. In this review, we summarize methodologies that have been used for effector mining in various microbial pathogens which use different effector delivery mechanisms. We also discuss current limitations and provide perspectives on how recently developed analytic tools and technologies may facilitate effector identification and hence generation of a more complete vision of host-pathogen interactions. [Formula: see text] Copyright © 2023 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license.

Progress and Challenges in Elucidating the Functional Role of Effectors in the Soybean-Phytophthora sojae Interaction

Phytophthora sojae, the agent responsible for stem and root rot, is one of the most damaging plant pathogens of soybean. To establish a compatible-interaction, P. sojae secretes a wide array of effector proteins into the host cell. These effectors have been shown to act either in the apoplastic area or the cytoplasm of the cell to manipulate the host cellular processes in favor of the development of the pathogen. Deciphering effector-plant interactions is important for understanding the role of P. sojae effectors in disease progression and developing approaches to prevent infection. Here, we review the subcellular localization, the host proteins, and the processes associated with P. sojae effectors. We also discuss the emerging topic of effectors in the context of effector-resistance genes interaction, as well as model systems and recent developments in resources and techniques that may provide a better understanding of the soybean-P. sojae interaction.

EffectorP 3.0: Prediction of Apoplastic and Cytoplasmic Effectors in Fungi and Oomycetes

Many fungi and oomycete species are devasting plant pathogens. These eukaryotic filamentous pathogens secrete effector proteins to facilitate plant infection. Fungi and oomycete pathogens have diverse infection strategies and their effectors generally do not share sequence homology. However, they occupy similar host environments, either the plant apoplast or plant cytoplasm, and, therefore, may share some unifying properties based on the requirements of these host compartments. Here, we exploit these biological signals and present the first classifier (EffectorP 3.0) that uses two machine-learning models: one trained on apoplastic effectors and one trained on cytoplasmic effectors. EffectorP 3.0 accurately predicts known apoplastic and cytoplasmic effectors in fungal and oomycete secretomes with low estimated false-positive rates of 3 and 8%, respectively. Cytoplasmic effectors have a higher proportion of positively charged amino acids, whereas apoplastic effectors are enriched for cysteine residues. The combination of fungal and oomycete effectors in training leads to a higher number of predicted cytoplasmic effectors in biotrophic fungi. EffectorP 3.0 expands predicted effector repertoires beyond small, cysteine-rich secreted proteins in fungi and RxLR-motif containing secreted proteins in oomycetes. We show that signal peptide prediction is essential for accurate effector prediction, because EffectorP 3.0 recognizes a cytoplasmic signal also in intracellular, nonsecreted proteins.[Formula: see text] Copyright © 2022 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license.

Exploiting Structural Modelling Tools to Explore Host-Translocated Effector Proteins

Oomycete and fungal interactions with plants can be neutral, symbiotic or pathogenic with different impact on plant health and fitness. Both fungi and oomycetes can generate so-called effector proteins in order to successfully colonize the host plant. These proteins modify stress pathways, developmental processes and the innate immune system to the microbes' benefit, with a very different outcome for the plant. Investigating the biological and functional roles of effectors during plant-microbe interactions are accessible through bioinformatics and experimental approaches. The next generation protein modeling software RoseTTafold and AlphaFold2 have made significant progress in defining the 3D-structure of proteins by utilizing novel machine-learning algorithms using amino acid sequences as their only input. As these two methods rely on super computers, Google Colabfold alternatives have received significant attention, making the approaches more accessible to users. Here, we focus on current structural biology, sequence motif and domain knowledge of effector proteins from filamentous microbes and discuss the broader use of novel modelling strategies, namely AlphaFold2 and RoseTTafold, in the field of effector biology. Finally, we compare the original programs and their Colab versions to assess current strengths, ease of access, limitations and future applications.

PHI-base in 2022: a multi-species phenotype database for Pathogen-Host Interactions

Since 2005, the Pathogen–Host Interactions Database (PHI-base) has manually curated experimentally verified pathogenicity, virulence and effector genes from fungal, bacterial and protist pathogens, which infect animal, plant, fish, insect and/or fungal hosts. PHI-base (www.phi-base.org) is devoted to the identification and presentation of phenotype information on pathogenicity and effector genes and their host interactions. Specific gene alterations that did not alter the in host interaction phenotype are also presented. PHI-base is invaluable for comparative analyses and for the discovery of candidate targets in medically and agronomically important species for intervention. Version 4.12 (September 2021) contains 4387 references, and provides information on 8411 genes from 279 pathogens, tested on 228 hosts in 18, 190 interactions. This provides a 24% increase in gene content since Version 4.8 (September 2019). Bacterial and fungal pathogens represent the majority of the interaction data, with a 54:46 split of entries, whilst protists, protozoa, nematodes and insects represent 3.6% of entries. Host species consist of approximately 54% plants and 46% others of medical, veterinary and/or environmental importance. PHI-base data is disseminated to UniProtKB, FungiDB and Ensembl Genomes. PHI-base will migrate to a new gene-centric version (version 5.0) in early 2022. This major development is briefly described.