effectR: An Expandable R Package to Predict Candidate RxLR and CRN Effectors in Oomycetes Using Motif Searches

Annotation of RxLR Effectors in Oomycete Genomes

Pathogens have evolved effector proteins to suppress host immunity and facilitate plant infections. RxLR effectors are small, secreted effector proteins with conserved RxLR and dEER amino acid motifs at the N terminus and highly variable C termini and are commonly found in oomycete species. We provide computational approaches to annotate RxLR candidate effector genes in a genome assembly in FASTA format with an available GFF file. Hidden Markov Modeling (HHM) is used in combination with regular expressions to search for RxLR and EER amino acid patterns.

Pangenome graph analysis reveals extensive effector copy-number variation in spinach downy mildew

Plant pathogens adapt at speeds that challenge contemporary disease management strategies like the deployment of disease resistance genes. The strong evolutionary pressure to adapt, shapes pathogens' genomes, and comparative genomics has been instrumental in characterizing this process. With the aim to capture genomic variation at high resolution and study the processes contributing to adaptation, we here leverage an innovative, multi-genome method to construct and annotate the first pangenome graph of an oomycete plant pathogen. We expand on this approach by analysing the graph and creating synteny based single-copy orthogroups for all genes. We generated telomere-to-telomere genome assemblies of six genetically diverse isolates of the oomycete pathogen Peronospora effusa, the economically most important disease in cultivated spinach worldwide. The pangenome graph demonstrates that P. effusa genomes are highly conserved, both in chromosomal structure and gene content, and revealed the continued activity of transposable elements which are directly responsible for 80% of the observed variation between the isolates. While most genes are generally conserved, virulence related genes are highly variable between the isolates. Most of the variation is found in large gene clusters resulting from extensive copy-number expansion. Pangenome graph-based discovery can thus be effectively used to capture genomic variation at exceptional resolution, thereby providing a framework to study the biology and evolution of plant pathogens.

Functional genomics identifies a small secreted protein that plays a role during the biotrophic to necrotrophic shift in the root rot pathogen Phytophthora medicaginis

Introduction

Hemibiotrophic Phytophthora are a group of agriculturally and ecologically important pathogenic oomycetes causing severe decline in plant growth and fitness. The lifestyle of these pathogens consists of an initial biotrophic phase followed by a switch to a necrotrophic phase in the latter stages of infection. Between these two phases is the biotrophic to necrotrophic switch (BNS) phase, the timing and controls of which are not well understood particularly in Phytophthora spp. where host resistance has a purely quantitative genetic basis.

Methods

To investigate this we sequenced and annotated the genome of Phytophthora medicaginis, causal agent of root rot and substantial yield losses to Fabaceae hosts. We analyzed the transcriptome of P. medicaginis across three phases of colonization of a susceptible chickpea host (Cicer arietinum) and performed co-regulatory analysis to identify putative small secreted protein (SSP) effectors that influence timing of the BNS in a quantitative pathosystem.

Results

The genome of P. medicaginis is ~78 Mb, comparable to P. fragariae and P. rubi which also cause root rot. Despite this, it encodes the second smallest number of RxLR (arginine-any amino acid-leucine-arginine) containing proteins of currently sequenced Phytophthora species. Only quantitative resistance is known in chickpea to P. medicaginis, however, we found that many RxLR, Crinkler (CRN), and Nep1-like protein (NLP) proteins and carbohydrate active enzymes (CAZymes) were regulated during infection. Characterization of one of these, Phytmed_10271, which encodes an RxLR effector demonstrates that it plays a role in the timing of the BNS phase and root cell death.

Discussion

These findings provide an important framework and resource for understanding the role of pathogenicity factors in purely quantitative Phytophthora pathosystems and their implications to the timing of the BNS phase.

Transcriptome analysis of Phytophthora cactorum infecting strawberry identified RXLR effectors that induce cell death when transiently expressed in Nicotiana benthamiana

Phytophthora cactorum is a plant pathogenic oomycete that causes crown rot in strawberry leading to significant economic losses every year. To invade the host, P. cactorum secretes an arsenal of effectors that can manipulate host physiology and impair its defense system promoting infection. A transcriptome analysis was conducted on a susceptible wild strawberry genotype (Fragaria vesca) 48 hours post inoculation with P. cactorum to identify effectors expressed during the early infection stage. The analysis revealed 4,668 P. cactorum genes expressed during infection of F. vesca. A total of 539 secreted proteins encoded by transcripts were identified, including 120 carbohydrate-active enzymes, 40 RXLRs, 23 proteolytic enzymes, nine elicitins, seven cysteine rich proteins, seven necrosis inducing proteins and 216 hypothetical proteins with unknown function. Twenty of the 40 RXLR effector candidates were transiently expressed in Nicotiana benthamiana using agroinfiltration and five previously unreported RXLR effector genes (Pc741, Pc8318, Pc10890, Pc20813, and Pc22290) triggered cell death when transiently expressed. The identified cell death inducing RXLR effectors showed 31-66% identity to known RXLR effectors in different Phytophthora species having roles in pathogenicity including both activation and suppression of defense response in the host. Furthermore, homology analysis revealed that these cell death inducing RXLR effectors were highly conserved (82 - 100% identity) across 23 different strains of P. cactorum originating from apple or strawberry.

Genome-Enabled Insights into Downy Mildew Biology and Evolution

Oomycetes that cause downy mildew diseases are highly specialized, obligately biotrophic phytopathogens that can have major impacts on agriculture and natural ecosystems. Deciphering the genome sequence of these organisms provides foundational tools to study and deploy control strategies against downy mildew pathogens (DMPs). The recent telomere-to-telomere genome assembly of the DMP Peronospora effusa revealed high levels of synteny with distantly related DMPs, higher than expected repeat content, and previously undescribed architectures. This provides a road map for generating similar high-quality genome assemblies for other oomycetes. This review discusses biological insights made using this and other assemblies, including ancestral chromosome architecture, modes of sexual and asexual variation, the occurrence of heterokaryosis, candidate gene identification, functional validation, and population dynamics. We also discuss future avenues of research likely to be fruitful in studies of DMPs and highlight resources necessary for advancing our understanding and ability to forecast and control disease outbreaks.

Genome analysis of Phytophthora cactorum strains associated with crown- and leather-rot in strawberry

Phytophthora cactorum has two distinct pathotypes that cause crown rot and leather rot in strawberry (Fragaria × ananassa). Strains of the crown rot pathotype can infect both the rhizome (crown) and fruit tissues, while strains of the leather rot pathotype can only infect the fruits of strawberry. The genome of a highly virulent crown rot strain, a low virulent crown rot strain, and three leather rot strains were sequenced using PacBio high fidelity (HiFi) long read sequencing. The reads were de novo assembled to 66.4-67.6 megabases genomes in 178-204 contigs, with N50 values ranging from 892 to 1,036 kilobases. The total number of predicted complete genes in the five P. cactorum genomes ranged from 17,286 to 17,398. Orthology analysis identified a core secretome of 8,238 genes. Comparative genomic analysis revealed differences in the composition of potential virulence effectors, such as putative RxLR and Crinklers, between the crown rot and the leather rot pathotypes. Insertions, deletions, and amino acid substitutions were detected in genes encoding putative elicitors such as beta elicitin and cellulose-binding domain proteins from the leather rot strains compared to the highly virulent crown rot strain, suggesting a potential mechanism for the crown rot strain to escape host recognition during compatible interaction with strawberry. The results presented here highlight several effectors that may facilitate the tissue-specific colonization of P. cactorum in strawberry.

Plant immunity in soybean: progress, strategies, and perspectives

Soybean (Glycine max) is one of the most important commercial crops worldwide. Soybean hosts diverse microbes, including pathogens that may cause diseases and symbionts that contribute to nitrogen fixation. Study on soybean-microbe interactions to understand pathogenesis, immunity, and symbiosis represents an important research direction toward plant protection in soybean. In terms of immune mechanisms, current research in soybean lags far behind that in the model plants Arabidopsis and rice. In this review, we summarized the shared and unique mechanisms involved in the two-tiered plant immunity and the virulence function of pathogen effectors between soybean and Arabidopsis, providing a molecular roadmap for future research on soybean immunity. We also discussed disease resistance engineering and future perspectives in soybean.

The Phytophthora sojae nuclear effector PsAvh110 targets a host transcriptional complex to modulate plant immunity

Plants have evolved sophisticated immune networks to restrict pathogen colonization. In response, pathogens deploy numerous virulent effectors to circumvent plant immune responses. However, the molecular mechanisms by which pathogen-derived effectors suppress plant defenses remain elusive. Here, we report that the nucleus-localized RxLR effector PsAvh110 from the pathogen Phytophthora sojae, causing soybean (Glycine max) stem and root rot, modulates the activity of a transcriptional complex to suppress plant immunity. Soybean like-heterochromatin protein 1-2 (GmLHP1-2) and plant homeodomain finger protein 6 (GmPHD6) form a transcriptional complex with transcriptional activity that positively regulates plant immunity against Phytophthora infection. To suppress plant immunity, the nuclear effector PsAvh110 disrupts the assembly of the GmLHP1-2/GmPHD6 complex via specifically binding to GmLHP1-2, thus blocking its transcriptional activity. We further show that PsAvh110 represses the expression of a subset of immune-associated genes, including BRI1-associated receptor kinase 1-3 (GmBAK1-3) and pathogenesis-related protein 1 (GmPR1), via G-rich elements in gene promoters. Importantly, PsAvh110 is a conserved effector in different Phytophthora species, suggesting that the PsAvh110 regulatory mechanism might be widely utilized in the genus to manipulate plant immunity. Thus, our study reveals a regulatory mechanism by which pathogen effectors target a transcriptional complex to reprogram transcription.

RNA silencing proteins and small RNAs in oomycete plant pathogens and biocontrol agents

Introduction

Oomycetes cause several damaging diseases of plants and animals, and some species also act as biocontrol agents on insects, fungi, and other oomycetes. RNA silencing is increasingly being shown to play a role in the pathogenicity of Phytophthora species, either through trans-boundary movement of small RNAs (sRNAs) or through expression regulation of infection promoting effectors.

Methods

To gain a wider understanding of RNA silencing in oomycete species with more diverse hosts, we mined genome assemblies for Dicer-like (DCL), Argonaute (AGO), and RNA dependent RNA polymerase (RDRP) proteins from Phytophthora plurivora, Ph. cactorum, Ph. colocasiae, Pythium oligandrum, Py. periplocum, and Lagenidium giganteum. Moreover, we sequenced small RNAs from the mycelium stage in each of these species.

Results and discussion

Each of the species possessed a single DCL protein, but they differed in the number and sequence of AGOs and RDRPs. SRNAs of 21nt, 25nt, and 26nt were prevalent in all oomycetes analyzed, but the relative abundance and 5' base preference of these classes differed markedly between genera. Most sRNAs mapped to transposons and other repeats, signifying that the major role for RNA silencing in oomycetes is to limit the expansion of these elements. We also found that sRNAs may act to regulate the expression of duplicated genes. Other sRNAs mapped to several gene families, and this number was higher in Pythium spp., suggesting a role of RNA silencing in regulating gene expression. Genes for most effector classes were the source of sRNAs of variable size, but some gene families showed a preference for specific classes of sRNAs, such as 25/26 nt sRNAs targeting RxLR effector genes in Phytophthora species. Novel miRNA-like RNAs (milRNAs) were discovered in all species, and two were predicted to target transcripts for RxLR effectors in Ph. plurivora and Ph. cactorum, indicating a putative role in regulating infection. Moreover, milRNAs from the biocontrol Pythium species had matches in the predicted transcriptome of Phytophthora infestans and Botrytis cinerea, and L. giganteum milRNAs matched candidate genes in the mosquito Aedes aegypti. This suggests that trans-boundary RNA silencing may have a role in the biocontrol action of these oomycetes.

Comparative Genomic Analysis of 31 Phytophthora Genomes Reveals Genome Plasticity and Horizontal Gene Transfer

Phytophthora species are oomycete plant pathogens that cause great economic and ecological impacts. The Phytophthora genus includes over 180 known species, infecting a wide range of plant hosts, including crops, trees, and ornamentals. We sequenced the genomes of 31 individual Phytophthora species and 24 individual transcriptomes to study genetic relationships across the genus. De novo genome assemblies revealed variation in genome sizes, numbers of predicted genes, and in repetitive element content across the Phytophthora genus. A genus-wide comparison evaluated orthologous groups of genes. Predicted effector gene counts varied across Phytophthora species by effector family, genome size, and plant host range. Predicted numbers of apoplastic effectors increased as the host range of Phytophthora species increased. Predicted numbers of cytoplasmic effectors also increased with host range but leveled off or decreased in Phytophthora species that have enormous host ranges. With extensive sequencing across the Phytophthora genus, we now have the genomic resources to evaluate horizontal gene transfer events across the oomycetes. Using a machine-learning approach to identify horizontally transferred genes with bacterial or fungal origin, we identified 44 candidates over 36 Phytophthora species genomes. Phylogenetic reconstruction indicates that the transfers of most of these 44 candidates happened in parallel to major advances in the evolution of the oomycetes and Phytophthora spp. We conclude that the 31 genomes presented here are essential for investigating genus-wide genomic associations in genus Phytophthora. [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.

An in-planta comparative study of Plasmopara viticola proteome reveals different infection strategies towards susceptible and Rpv3-mediated resistance hosts

Plasmopara viticola, an obligate biotrophic oomycete, is the causal agent of one of the most harmful grapevine diseases, downy mildew. Within this pathosystem, much information is gathered on the host, as characterization of pathogenicity and infection strategy of a biotrophic pathogen is quite challenging. Molecular insights into P. viticola development and pathogenicity are just beginning to be uncovered, mainly by transcriptomic studies. Plasmopara viticola proteome and secretome were only predicted based on transcriptome data. In this study, we have identified the in-planta proteome of P. viticola during infection of a susceptible ('Trincadeira') and a Rpv3-mediated resistance ('Regent') grapevine cultivar. Four hundred and twenty P. viticola proteins were identified on a label-free mass spectrometry-based approach of the apoplastic fluid of grapevine leaves. Overall, our study suggests that, in the compatible interaction, P. viticola manipulates salicylic-acid pathway and isoprenoid biosynthesis to enhance plant colonization. Furthermore, during the incompatible interaction, development-associated proteins increased while oxidoreductases protect P. viticola from ROS-associated plant defence mechanism. Up to our knowledge this is the first in-planta proteome characterization of this biotrophic pathogen, thus this study will open new insights into our understanding of this pathogen colonization strategy of both susceptible and Rpv3-mediated resistance grapevine genotypes.

WideEffHunter: An Algorithm to Predict Canonical and Non-Canonical Effectors in Fungi and Oomycetes

Newer effectorome prediction algorithms are considering effectors that may not comply with the canonical characteristics of small, secreted, cysteine-rich proteins. The use of effector-related motifs and domains is an emerging strategy for effector identification, but its use has been limited to individual species, whether oomycete or fungal, and certain domains and motifs have only been associated with one or the other. The use of these strategies is important for the identification of novel, non-canonical effectors (NCEs) which we have found to constitute approximately 90% of the effectoromes. We produced an algorithm in Bash called WideEffHunter that is founded on integrating three key characteristics: the presence of effector motifs, effector domains and homology to validated existing effectors. Interestingly, we found similar numbers of effectors with motifs and domains within two different taxonomic kingdoms: fungi and oomycetes, indicating that with respect to their effector content, the two organisms may be more similar than previously believed. WideEffHunter can identify the entire effectorome (non-canonical and canonical effectors) of oomycetes and fungi whether pathogenic or non-pathogenic, unifying effector prediction in these two kingdoms as well as the two different lifestyles. The elucidation of complete effectoromes is a crucial step towards advancing effectoromics and disease management in agriculture.

Chromosome-level assembly of the Phytophthora agathidicida genome reveals adaptation in effector gene families

Phytophthora species are notorious plant pathogens, with some causing devastating tree diseases that threaten the survival of their host species. One such example is Phytophthora agathidicida, the causal agent of kauri dieback - a root and trunk rot disease that kills the ancient, iconic and culturally significant tree species, Agathis australis (New Zealand kauri). A deeper understanding of how Phytophthora pathogens infect their hosts and cause disease is critical for the development of effective treatments. Such an understanding can be gained by interrogating pathogen genomes for effector genes, which are involved in virulence or pathogenicity. Although genome sequencing has become more affordable, the complete assembly of Phytophthora genomes has been problematic, particularly for those with a high abundance of repetitive sequences. Therefore, effector genes located in repetitive regions could be truncated or missed in a fragmented genome assembly. Using a combination of long-read PacBio sequences, chromatin conformation capture (Hi-C) and Illumina short reads, we assembled the P. agathidicida genome into ten complete chromosomes, with a genome size of 57 Mb including 34% repeats. This is the first Phytophthora genome assembled to chromosome level and it reveals a high level of syntenic conservation with the complete genome of Peronospora effusa, the only other completely assembled genome sequence of an oomycete. All P. agathidicida chromosomes have clearly defined centromeres and contain candidate effector genes such as RXLRs and CRNs, but in different proportions, reflecting the presence of gene family clusters. Candidate effector genes are predominantly found in gene-poor, repeat-rich regions of the genome, and in some cases showed a high degree of duplication. Analysis of candidate RXLR effector genes that occur in multicopy gene families indicated half of them were not expressed in planta. Candidate CRN effector gene families showed evidence of transposon-mediated recombination leading to new combinations of protein domains, both within and between chromosomes. Further analysis of this complete genome assembly will help inform new methods of disease control against P. agathidicida and other Phytophthora species, ultimately helping decipher how Phytophthora pathogens have evolved to shape their effector repertoires and how they might adapt in the future.

Karyotype variation, spontaneous genome rearrangements affecting chemical insensitivity, and expression level polymorphisms in the plant pathogen Phytophthora infestans revealed using its first chromosome-scale assembly

Natural isolates of the potato and tomato pathogen Phytophthora infestans exhibit substantial variation in virulence, chemical sensitivity, ploidy, and other traits. A chromosome-scale assembly was developed to expand genomic resources for this oomyceteous microbe, and used to explore the basis of variation. Using PacBio and Illumina data, a long-range linking library, and an optical map, an assembly was created and coalesced into 15 pseudochromosomes spanning 219 Mb using SNP-based genetic linkage data. De novo gene prediction combined with transcript evidence identified 19,981 protein-coding genes, plus about eight thousand tRNA genes. The chromosomes were comprised of a mosaic of gene-rich and gene-sparse regions plus very long centromeres. Genes exhibited a biased distribution across chromosomes, especially members of families encoding RXLR and CRN effectors which clustered on certain chromosomes. Strikingly, half of F1 progeny of diploid parents were polyploid or aneuploid. Substantial expression level polymorphisms between strains were identified, much of which could be attributed to differences in chromosome dosage, transposable element insertions, and adjacency to repetitive DNA. QTL analysis identified a locus on the right arm of chromosome 3 governing sensitivity to the crop protection chemical metalaxyl. Strains heterozygous for resistance often experienced megabase-sized deletions of that part of the chromosome when cultured on metalaxyl, increasing resistance due to loss of the sensitive allele. This study sheds light on diverse phenomena affecting variation in P. infestans and relatives, helps explain the prevalence of polyploidy in natural populations, and provides a new foundation for biologic and genetic investigations.

Recent advances in understanding of fungal and oomycete effectors

Fungal and oomycete pathogens secrete complex arrays of proteins and small RNAs to interface with plant-host targets and manipulate plant regulatory networks to the microbes' advantage. Research on these important virulence factors has been accelerated by improved genome sequences, refined bioinformatic prediction tools, and exploitation of efficient platforms for understanding effector gene expression and function. Recent studies have validated the expectation that oomycetes and fungi target many of the same sectors in immune signaling networks, but the specific host plant targets and modes of action are diverse. Effector research has also contributed to deeper understanding of the mechanisms of effector-triggered immunity.

The Minichromosome Maintenance Complex Component 2 (MjMCM2) of Meloidogyne javanica is a potential effector regulating the cell cycle in nematode-induced galls

Root-knot nematodes Meloidogyne spp. induce enlarged multinucleate feeding cells—galls—in host plant roots. Although core cell-cycle components in galls follow a conserved track, they can also be usurped and manipulated by nematodes. We identified a candidate effector in Meloidogyne javanica that is directly involved in cell-cycle manipulation—Minichromosome Maintenance Complex Component 2 (MCM2), part of MCM complex licensing factor involved in DNA replication. MjMCM2, which is induced by plant oxilipin 9-HOT, was expressed in nematode esophageal glands, upregulated during parasitic stages, and was localized to plant cell nucleus and plasma membrane. Infected tomato hairy roots overexpressing MjMCM2 showed significantly more galls and egg-mass-producing females than wild-type roots, and feeding cells showed more nuclei. Phylogenetic analysis suggested seven homologues of MjMCM2 with unknown association to parasitism. Sequence mining revealed two RxLR-like motifs followed by SEED domains in all Meloidogyne spp. MCM2 protein sequences. The unique second RxLR-like motif was absent in other Tylenchida species. Molecular homology modeling of MjMCM2 suggested that second RxLR2-like domain is positioned on a surface loop structure, supporting its function in polar interactions. Our findings reveal a first candidate cell-cycle gene effector in M. javanica—MjMCM2—that is likely secreted into plant host to mimic function of endogenous MCM2.

Genomics and Informatics, Conjoined Tools Vital for Understanding and Protecting Plant Health

Genomics' impact on crop production continuously expands. The number of sequenced plant and microbial species and strains representing diverse populations of individual species rapidly increases thanks to the advent of next-generation sequencing technologies. Their genomic blueprints revealed candidate genes involved in various functions and processes crucial for crop health and helped in understanding how the sequenced organisms have evolved at the genome level. Functional genomics quickly translates these blueprints into a detailed mechanistic understanding of how such functions and processes work and are regulated; this understanding guides and empowers efforts to protect crops from diverse biotic and abiotic threats. Metagenome analyses help identify candidate microbes crucial for crop health and uncover how microbial communities associated with crop production respond to environmental conditions and cultural practices, presenting opportunities to enhance crop health by judiciously configuring microbial communities. Efficient conversion of disparate types of massive genomics data into actionable knowledge requires a robust informatics infrastructure supporting data preservation, analysis, and sharing. This review starts with an overview of how genomics came about and has quickly transformed life science. We illuminate how genomics and informatics can be applied to investigate various crop health-related problems using selected studies. We end the review by noting why community empowerment via crowdsourcing is crucial to harnessing genomics to protect global food and nutrition security without continuously expanding the environmental footprint of crop production.

An oomycete effector targets a plant RNA helicase involved in root development and defense

Oomycete plant pathogens secrete effector proteins to promote disease. The damaging soilborne legume pathogen Aphanomyces euteiches harbors a specific repertoire of Small Secreted Protein effectors (AeSSPs), but their biological functions remain unknown. Here we characterize AeSSP1256. The function of AeSSP1256 is investigated by physiological and molecular characterization of Medicago truncatula roots expressing the effector. A potential protein target of AeSSP1256 is identified by yeast-two hybrid, co-immunoprecipitation, and fluorescent resonance energy transfer-fluorescence lifetime imaging microscopy (FRET-FLIM) assays, as well as promoter studies and mutant characterization. AeSSP1256 impairs M. truncatula root development and promotes pathogen infection. The effector is localized to the nucleoli rim, triggers nucleoli enlargement and downregulates expression of M. truncatula ribosome-related genes. AeSSP1256 interacts with a functional nucleocytoplasmic plant RNA helicase (MtRH10). AeSSP1256 relocates MtRH10 to the perinucleolar space and hinders its binding to plant RNA. MtRH10 is associated with ribosome-related genes, root development and defense. This work reveals that an oomycete effector targets a plant RNA helicase, possibly to trigger nucleolar stress and thereby promote pathogen infection.

Comparative Genomic Analysis Reveals Genetic Variation and Adaptive Evolution in the Pathogenicity-Related Genes of Phytophthora capsici

Phytophthora capsici is an oomycete pathogen responsible for damping off, root rot, fruit rot, and foliar blight in popular vegetable and legume crops. The existence of distinct aggressiveness levels and physiological races among the P. capsici population is a major constraint to developing resistant varieties of host crops. In the present study, we compared the genomes of three P. capsici isolates with different aggressiveness levels to reveal their genomic differences. We obtained genome sequences using short-read and long-read technologies, which yielded an average genome size of 76 Mbp comprising 514 contigs and 15,076 predicted genes. A comparative genomic analysis uncovered the signatures of accelerated evolution, gene family expansions in the pathogenicity-related genes among the three isolates. Resequencing two additional P. capsici isolates enabled the identification of average 1,023,437 SNPs, revealing the frequent accumulation of non-synonymous substitutions in pathogenicity-related gene families. Furthermore, pathogenicity-related gene families, cytoplasmic effectors and ATP binding cassette (ABC) transporters, showed expansion signals in the more aggressive isolates, with a greater number of non-synonymous SNPs. This genomic information explains the plasticity, difference in aggressiveness levels, and genome structural variation among the P. capsici isolates, providing insight into the genomic features related to the evolution and pathogenicity of this oomycete pathogen.

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

Background

Plant pathogens cause billions of dollars of crop loss every year and are a major threat to global food security. Effector proteins are the tools such pathogens use to infect the cell, predicting effectors de novo from sequence is difficult because of the heterogeneity of the sequences. We hypothesised that deep learning classifiers based on Convolutional Neural Networks would be able to identify effectors and deliver new insights.

Results

We created a training set of manually curated effector sequences from PHI-Base and used these to train a range of model architectures for classifying bacteria, fungal and oomycete sequences. The best performing classifiers had accuracies from 93 to 84%. The models were tested against popular effector detection software on our own test data and data provided with those models. We observed better performance from our models. Specifically our models showed greater accuracy and lower tendencies to call false positives on a secreted protein negative test set and a greater generalisability. We used GRAD-CAM activation map analysis to identify the sequences that activated our CNN-LSTM models and found short but distinct N-terminal regions in each taxon that was indicative of effector sequences. No motifs could be observed in these regions but an analysis of amino acid types indicated differing patterns of enrichment and depletion that varied between taxa.

Conclusions

Small training sets can be used effectively to train highly accurate and sensitive deep learning models without need for the operator to know anything other than sequence and without arbitrary decisions made about what sequence features or physico-chemical properties are important. Biological insight on subsequences important for classification can be achieved by examining the activations in the model

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04293-3.

Genomic Variability of Phytophthora palmivora Isolates from Different Oil Palm Cultivation Regions in Colombia

Palm oil is the most consumed vegetable oil globally, and Colombia is the largest palm oil producer in South America and fourth worldwide. However, oil palm plantations in Colombia are affected by bud rot disease caused by the oomycete Phytophthora palmivora, leading to significant economic losses. Infection processes by plant pathogens involve the secretion of effector molecules, which alter the functioning or structure of host cells. Current long-read sequencing technologies provide the information needed to produce high-quality genome assemblies, enabling a comprehensive annotation of effectors. Here, we describe the development of genomic resources for P. palmivora, including a high-quality genome assembly based on long and short-read sequencing data, intraspecies variability for 12 isolates from different oil palm cultivation regions in Colombia, and a catalog of over 1,000 candidate effector proteins. A total of 45,416 genes were annotated from the new genome assembled in 2,322 contigs adding to 165.5 Mbp, which represents an improvement of two times more gene models, 33 times better contiguity, and 11 times less fragmentation compared with currently available genomic resources for the species. Analysis of nucleotide evolution in paralogs suggests a recent whole-genome duplication event. Genetic differences were identified among isolates showing variable virulence levels. We expect that these novel genomic resources contribute to the characterization of the species and the understanding of the interaction of P. palmivora with oil palm and could be further exploited as tools for the development of effective strategies for disease control.

Improved Genome Sequence and Gene Annotation Resource for the Potato Late Blight Pathogen Phytophthora infestans

Phytophthora infestans is a devastating pathogen causing potato late blight (Solanum tuberosum). Here we report the sequencing, assembly and genome annotation for two Phytophthora infestans isolates sampled in Republic of Korea. Genome sequencing was carried out using long read (Oxford Nanopore) and short read (Illumina Nextseq) sequencing technologies that significantly improved the contiguity and quality of P. infestans genome assembly. Our resources would help researchers better understand the molecular mechanisms by which P. infestans causes late blight disease in the future.

Effector Repertoire of Phytophthora betacei: In Search of Possible Virulence Factors Responsible for Its Host Specificity

Phytophthora betacei is an oomycete plant pathogen closely related to Phytophthora infestans. It infects tree tomato (Solanum betaceum) in northern South America, but is, under natural conditions, unable to infect potatoes or tomatoes, the main hosts of its sister species P. infestans. We characterized, and compared the effector repertoires of P. betacei and other Phytophthora species. To this end, we used in silico approaches to predict and describe the repertoire of secreted proteins in Phytophthora species and determine unique and core effectors. P. betacei has the largest proteome and secretome of all Phytophthora species evaluated. We identified between 450 and 1933 candidate effector genes in Phytophthora ramorum, Phytophthora sojae, Phytophthora cactorum, Phytophthora parasitica, Phytophthora palmivora, P. infestans, and P. betacei genomes. The P. betacei predicted secretome contains 5653 proteins, 1126 of which are apoplastic effectors and 807cytoplasmic effectors. Genes encoding cytoplasmic effectors include 791 genes with an RxLR domain (the largest number known so far in a Phytophthora species) and 16 with a Crinkler (CRN) domain. We detected homologs of previously described avirulence gene (Avr) present in Phytophthora spp., such as Avr1, Avr3b, Avr4, and Avrblb1, suggesting a high level of effector gene conservation among Phytophthora species. Nonetheless, fewer CRN effectors were obtained in P. betacei compared to all other Phytophthora species analyzed. The comparison between P. infestans and P. betacei effector profiles shows unique features in P. betacei that might be involved in pathogenesis and host preference. Indeed, 402 unique predicted effector genes were detected in P. betacei, corresponding to 197 apoplastic effector genes, 203 RxLR cytoplasmic effector genes, and 2 effector genes with CRN domain. This is the first characterization of the effector profile of P. betacei and the broadest comparison of predicted effector repertoires in the genus Phytophthora following a standardized prediction pipeline. The resultant P. betacei putative effector repertoire provides a reasonable set of proteins whose experimental validation could lead to understand the specific virulence factors responsible for the host specificity of this species.