RLPredictiOme, a Machine Learning-Derived Method for High-Throughput Prediction of Plant Receptor-like Proteins, Reveals Novel Classes of Transmembrane Receptors.

Cell surface receptors play essential roles in perceiving and processing external and internal signals at the cell surface of plants and animals. The receptor-like protein kinases (RLK) and receptor-like proteins (RLPs), two major classes of proteins with membrane receptor configuration, play a crucial role in plant development and disease defense. Although RLPs and RLKs share a similar single-pass transmembrane configuration, RLPs harbor short divergent C-terminal regions instead of the conserved kinase domain of RLKs. This RLP receptor structural design precludes sequence comparison algorithms from being used for high-throughput predictions of the RLP family in plant genomes, as has been extensively performed for RLK superfamily predictions. Here, we developed the RLPredictiOme, implemented with machine learning models in combination with Bayesian inference, capable of predicting RLP subfamilies in plant genomes. The ML models were simultaneously trained using six types of features, along with three stages to distinguish RLPs from non-RLPs (NRLPs), RLPs from RLKs, and classify new subfamilies of RLPs in plants. The ML models achieved high accuracy, precision, sensitivity, and specificity for predicting RLPs with relatively high probability ranging from 0.79 to 0.99. The prediction of the method was assessed with three datasets, two of which contained leucine-rich repeats (LRR)-RLPs from Arabidopsis and rice, and the last one consisted of the complete set of previously described Arabidopsis RLPs. In these validation tests, more than 90% of known RLPs were correctly predicted via RLPredictiOme. In addition to predicting previously characterized RLPs, RLPredictiOme uncovered new RLP subfamilies in the Arabidopsis genome. These include probable lipid transfer (PLT)-RLP, plastocyanin-like-RLP, ring finger-RLP, glycosyl-hydrolase-RLP, and glycerophosphoryldiester phosphodiesterase (GDPD, GDPDL)-RLP subfamilies, yet to be characterized. Compared to the only Arabidopsis GDPDL-RLK, molecular evolution studies confirmed that the ectodomain of GDPDL-RLPs might have undergone a purifying selection with a predominance of synonymous substitutions. Expression analyses revealed that predicted GDPGL-RLPs display a basal expression level and respond to developmental and biotic signals. The results of these biological assays indicate that these subfamily members have maintained functional domains during evolution and may play relevant roles in development and plant defense. Therefore, RLPredictiOme provides a framework for genome-wide surveys of the RLP superfamily as a foundation to rationalize functional studies of surface receptors and their relationships with different biological processes.

Fangorn Forest (F2): a machine learning approach to classify genes and genera in the family Geminiviridae.

BACKGROUND: Geminiviruses infect a broad range of cultivated and non-cultivated plants, causing significant economic losses worldwide. The studies of the diversity of species, taxonomy, mechanisms of evolution, geographic distribution, and mechanisms of interaction of these pathogens with the host have greatly increased in recent years. Furthermore, the use of rolling circle amplification (RCA) and advanced metagenomics approaches have enabled the elucidation of viromes and the identification of many viral agents in a large number of plant species. As a result, determining the nomenclature and taxonomically classifying geminiviruses turned into complex tasks. In addition, the gene responsible for viral replication (particularly, the viruses belonging to the genus Mastrevirus) may be spliced due to the use of the transcriptional/splicing machinery in the host cells. However, the current tools have limitations concerning the identification of introns. RESULTS: This study proposes a new method, designated Fangorn Forest (F2), based on machine learning approaches to classify genera using an ab initio approach, i.e., using only the genomic sequence, as well as to predict and classify genes in the family Geminiviridae. In this investigation, nine genera of the family Geminiviridae and their related satellite DNAs were selected. We obtained two training sets, one for genus classification, containing attributes extracted from the complete genome of geminiviruses, while the other was made up to classify geminivirus genes, containing attributes extracted from ORFs taken from the complete genomes cited above. Three ML algorithms were applied on those datasets to build the predictive models: support vector machines, using the sequential minimal optimization training approach, random forest (RF), and multilayer perceptron. RF demonstrated a very high predictive power, achieving 0.966, 0.964, and 0.995 of precision, recall, and area under the curve (AUC), respectively, for genus classification. For gene classification, RF could reach 0.983, 0.983, and 0.998 of precision, recall, and AUC, respectively. CONCLUSIONS: Therefore, Fangorn Forest is proven to be an efficient method for classifying genera of the family Geminiviridae with high precision and effective gene prediction and classification. The method is freely accessible at www.geminivirus.org:8080/geminivirusdw/discoveryGeminivirus.jsp .

Geminivirus data warehouse: a database enriched with machine learning approaches.

BACKGROUND: The Geminiviridae family encompasses a group of single-stranded DNA viruses with twinned and quasi-isometric virions, which infect a wide range of dicotyledonous and monocotyledonous plants and are responsible for significant economic losses worldwide. Geminiviruses are divided into nine genera, according to their insect vector, host range, genome organization, and phylogeny reconstruction. Using rolling-circle amplification approaches along with high-throughput sequencing technologies, thousands of full-length geminivirus and satellite genome sequences were amplified and have become available in public databases. As a consequence, many important challenges have emerged, namely, how to classify, store, and analyze massive datasets as well as how to extract information or new knowledge. Data mining approaches, mainly supported by machine learning (ML) techniques, are a natural means for high-throughput data analysis in the context of genomics, transcriptomics, proteomics, and metabolomics. RESULTS: Here, we describe the development of a data warehouse enriched with ML approaches, designated geminivirus.org. We implemented search modules, bioinformatics tools, and ML methods to retrieve high precision information, demarcate species, and create classifiers for genera and open reading frames (ORFs) of geminivirus genomes. CONCLUSIONS: The use of data mining techniques such as ETL (Extract, Transform, Load) to feed our database, as well as algorithms based on machine learning for knowledge extraction, allowed us to obtain a database with quality data and suitable tools for bioinformatics analysis. The Geminivirus Data Warehouse (geminivirus.org) offers a simple and user-friendly environment for information retrieval and knowledge discovery related to geminiviruses.

Functional and regulatory conservation of the soybean ER stress-induced DCD/NRP-mediated cell death signaling in plants.

BACKGROUND: The developmental and cell death domain (DCD)-containing asparagine-rich proteins (NRPs) were first identified in soybean (Glycine max) as transducers of a cell death signal derived from prolonged endoplasmic reticulum (ER) stress, osmotic stress, drought or developmentally-programmed leaf senescence via the GmNAC81/GmNAC30/GmVPE signaling module. In spite of the relevance of the DCD/NRP-mediated signaling as a versatile adaptive response to multiple stresses, mechanistic knowledge of the pathway is lacking and the extent to which this pathway may operate in the plant kingdom has not been investigated. RESULTS: Here, we demonstrated that the DCD/NRP-mediated signaling also propagates a stress-induced cell death signal in other plant species with features of a programmed cell death (PCD) response. In silico analysis revealed that several plant genomes harbor conserved sequences of the pathway components, which share functional analogy with their soybean counterparts. We showed that GmNRPs, GmNAC81and VPE orthologs from Arabidopsis, designated as AtNRP-1, AtNRP-2, ANAC036 and gVPE, respectively, induced cell death when transiently expressed in N. benthamiana leaves. In addition, loss of AtNRP1 and AtNRP2 function attenuated ER stress-induced cell death in Arabidopsis, which was in marked contrast with the enhanced cell death phenotype displayed by overexpressing lines as compared to Col-0. Furthermore, atnrp-1 knockout mutants displayed enhanced sensitivity to PEG-induced osmotic stress, a phenotype that could be complemented with ectopic expression of either GmNRP-A or GmNRP-B. In addition, AtNRPs, ANAC036 and gVPE were induced by osmotic and ER stress to an extent that was modulated by the ER-resident molecular chaperone binding protein (BiP) similarly as in soybean. Finally, as putative downstream components of the NRP-mediated cell death signaling, the stress induction of AtNRP2, ANAC036 and gVPE was dependent on the AtNRP1 function. BiP overexpression also conferred tolerance to water stress in Arabidopsis, most likely due to modulation of the drought-induced NRP-mediated cell death response. CONCLUSION: Our results indicated that the NRP-mediated cell death signaling operates in the plant kingdom with conserved regulatory mechanisms and hence may be target for engineering stress tolerance and adaptation in crops.

Comprehensive analysis of the endoplasmic reticulum stress response in the soybean genome: conserved and plant-specific features.

BACKGROUND: Despite the relevance of the eukaryotic endoplasmic reticulum (ER)-stress response as an integrator of multiple stress signals into an adaptive response, knowledge about these ER-mediated cytoprotective pathways in soybean (Glycine max) is lacking. Here, we searched for genes involved in the highly conserved unfolded protein response (UPR) and ER stress-induced plant-specific cell death signaling pathways in the soybean genome. METHODS: Previously characterized Arabidopsis UPR genes were used as prototypes for the identification of the soybean orthologs and the in silico assembly of the UPR in soybean, using eggNOG v4.0 software. Functional studies were also conducted by analyzing the transcriptional activity of soybean UPR transducers. RESULTS: As a result of this search, we have provided a complete profile of soybean UPR genes with significant predicted protein similarities to A. thaliana UPR-associated proteins. Both arms of the plant UPR were further examined functionally, and evidence is presented that the soybean counterparts are true orthologs of previously characterized UPR transducers in Arabidopsis. The bZIP17/bZI28 orthologs (GmbZIP37 and GmbZIP38) and ZIP60 ortholog (GmbZIP68) from soybean have similar structural organizations as their Arabidopsis counterparts, were induced by ER stress and activated an ERSE- and UPRE-containing BiP promoter. Furthermore, the transcript of the putative substrate of GmIREs, GmbZIP68, harbors a canonical site for IRE1 endonuclease activity and was efficiently spliced under ER stress conditions. In a reverse approach, we also examined the Arabidopsis genome for components of a previously characterized ER stress-induced cell death signaling response in soybean. With the exception of GmERD15, which apparently does not possess an Arabidopsis ortholog, the Arabidopsis genome harbors conserved GmNRP, GmNAC81, GmNAC30 and GmVPE sequences that share significant structural and sequence similarities with their soybean counterparts. These results suggest that the NRP/GmNAC81 + GmNAC30/VPE regulatory circuit may transduce cell death signals in plant species other than soybean. CONCLUSIONS: Our in silico analyses, along with current and previous functional data, permitted generation of a comprehensive overview of the ER stress response in soybean as a framework for functional prediction of ER stress signaling components and their possible connections with multiple stress responses.

A long and stressful day: Photoperiod shapes aluminium tolerance in plants.

Aluminium (Al), a limiting factor for crop productivity in acidic soils (pH ≤ 5.5), imposes drastic constraints for food safety in developing countries. The major mechanisms that allow plants to cope with Al involve manipulations of organic acids metabolism and DNA-checkpoints. When assumed individually both approaches have been insufficient to overcome Al toxicity. On analysing the centre of origin of most cultivated plants, we hypothesised that day-length seems to be a pivotal agent modulating Al tolerance across distinct plant species. We observed that with increasing distance from the Equator, Al tolerance decreases, suggesting a relationship with the photoperiod. We verified that long-day (LD) species are generally more Al-sensitive than short-day (SD) species, whereas genetic conversion of tomato for SD growth habit boosts Al tolerance. Reduced Al tolerance correlates with DNA-checkpoint activation under LD. Furthermore, DNA-checkpoint-related genes are under positive selection in Arabidopsis accessions from regions with shorter days, suggesting that photoperiod act as a selective barrier for Al tolerance. A diel regulation and genetic diversity affect Al tolerance, suggesting that day-length orchestrates Al tolerance. Altogether, photoperiodic control of Al tolerance might contribute to solving the historical obstacle that imposes barriers for developing countries to reach a sustainable agriculture.

A Begomovirus Nuclear Shuttle Protein-Interacting Immune Hub: Hijacking Host Transport Activities and Suppressing Incompatible Functions.

Begomoviruses (Geminiviridae family) represent a severe constraint to agriculture worldwide. As ssDNA viruses that replicate in the nuclei of infected cells, the nascent viral DNA has to move to the cytoplasm and then to the adjacent cell to cause disease. The begomovirus nuclear shuttle protein (NSP) assists the intracellular transport of viral DNA from the nucleus to the cytoplasm and cooperates with the movement protein (MP) for the cell-to-cell translocation of viral DNA to uninfected cells. As a facilitator of intra- and intercellular transport of viral DNA, NSP is predicted to associate with host proteins from the nuclear export machinery, the intracytoplasmic active transport system, and the cell-to-cell transport complex. Furthermore, NSP functions as a virulence factor that suppresses antiviral immunity against begomoviruses. In this review, we focus on the protein-protein network that converges on NSP with a high degree of centrality and forms an immune hub against begomoviruses. We also describe the compatible host functions hijacked by NSP to promote the nucleocytoplasmic and intracytoplasmic movement of viral DNA. Finally, we discuss the NSP virulence function as a suppressor of the recently described NSP-interacting kinase 1 (NIK1)-mediated antiviral immunity. Understanding the NSP-host protein-protein interaction (PPI) network will probably pave the way for strategies to generate more durable resistance against begomoviruses.

Machine learning approaches and their current application in plant molecular biology: A systematic review.

Machine learning (ML) is a field of artificial intelligence that has rapidly emerged in molecular biology, thus allowing the exploitation of Big Data concepts in plant genomics. In this context, the main challenges are given in terms of how to analyze massive datasets and extract new knowledge in all levels of cellular systems research. In summary, ML techniques allow complex interactions to be inferred in several biological systems. Despite its potential, ML has been underused due to complex computational algorithms and definition terms. Therefore, a systematic review to disentangle ML approaches is relevant for plant scientists and has been considered in this study. We presented the main steps for ML development (from data selection to evaluation of classification/prediction models) with a respective discussion approaching functional genomics mainly in terms of pathogen effector genes in plant immunity. Additionally, we also considered how to access public source databases under an ML framework towards advancing plant molecular biology and introduced novel powerful tools, such as deep learning.

Comparative genomics of Staphylococcus aureus associated with subclinical and clinical bovine mastitis.

Many efforts have been made to understand the pathogenesis of bovine mastitis to reduce losses and promote animal welfare. Staphylococcus aureus may cause bovine clinical mastitis, but it is mainly associated with subclinical infection, which is usually persistent and can easily reoccur. Here, we conducted a comparative genomic analysis between strains of S. aureus causing subclinical infection (Sau170, 302, 1269, 1364), previously sequenced by our group, and two well-characterized strains causing clinical mastitis (N305 and RF122) to find differences that could be linked to mastitis outcome. A total of 146 virulence-associated genes were compared and no appreciable differences were found between the bacteria. However, several nonsynonymous single nucleotide polymorphisms (SNPs) were identified in genes present in the subclinical strains when compared to RF122 and N305, especially in genes encoding host immune evasion and surface proteins. The secreted and surface proteins predicted by in silico tools were compared through multidimensional scaling analysis (MDS), revealing a high degree of similarity among the strains. The comparison of orthologous genes by OrthoMCL identified a membrane transporter and a lipoprotein as exclusive of bacteria belonging to the subclinical and clinical groups, respectively. No hit was found in RF122 and N305 for the membrane transporter using BLAST algorithm. For the lipoprotein, sequences of Sau170, 302, 1269, and 1364 with identities between 68-73% were found in the MDS dataset. A conserved region found only in the lipoprotein genes of RF122 and N305 was used for primer design. Although the polymerase chain reaction (PCR) on field isolates of S. aureus did not validate the findings for the transporter, the lipoprotein was able to separate the clinical from the subclinical isolates. These results show that sequence variation among bovine S. aureus, and not only the presence/absence of virulence factors, is an important aspect to consider when comparing isolates causing different mastitis outcomes.

Revisiting the Soybean GmNAC Superfamily.

The NAC (NAM, ATAF, and CUC) genes encode transcription factors involved with the control of plant morph-physiology and stress responses. The release of the last soybean (Glycine max) genome assembly (Wm82.a2.v1) raised the possibility that new NAC genes would be present in the soybean genome. Here, we interrogated the last version of the soybean genome against a conserved NAC domain structure. Our analysis identified 32 putative novel NAC genes, updating the superfamily to 180 gene members. We also organized the genes in 15 phylogenetic subfamilies, which showed a perfect correlation among sequence conservation, expression profile, and function of orthologous Arabidopsis thaliana genes and NAC soybean genes. To validate our in silico analyses, we monitored the stress-mediated gene expression profiles of eight new NAC-genes by qRT-PCR and monitored the GmNAC senescence-associated genes by RNA-seq. Among ER stress, osmotic stress and salicylic acid treatment, all the novel tested GmNAC genes responded to at least one type of stress, displaying a complex expression profile under different kinetics and extension of the response. Furthermore, we showed that 40% of the GmNACs were differentially regulated by natural leaf senescence, including eight (8) newly identified GmNACs. The developmental and stress-responsive expression profiles of the novel NAC genes fitted perfectly with their phylogenetic subfamily. Finally, we examined two uncharacterized senescence-associated proteins, GmNAC065 and GmNAC085, and a novel, previously unidentified, NAC protein, GmNAC177, and showed that they are nuclear localized, and except for GmNAC065, they display transactivation activity in yeast. Consistent with a role in leaf senescence, transient expression of GmNAC065 and GmNAC085 induces the appearance of hallmarks of leaf senescence, including chlorophyll loss, leaf yellowing, lipid peroxidation and accumulation of H2O2. GmNAC177 was clustered to an uncharacterized subfamily but in close proximity to the TIP subfamily. Accordingly, it was rapidly induced by ER stress and by salicylic acid under late kinetic response and promoted cell death in planta. Collectively, our data further substantiated the notion that the GmNAC genes display functional and expression profiles consistent with their phylogenetic relatedness and established a complete framework of the soybean NAC superfamily as a foundation for future analyses.

A WW Domain-Containing Protein Forms Immune Nuclear Bodies against Begomoviruses.

The bipartite begomoviruses (Geminiviridae family), which are DNA viruses that replicate in the nucleus of infected cells, encode the nuclear shuttle protein (NSP) to facilitate the translocation of viral DNA from the nucleus to the cytoplasm via nuclear pores. This intracellular trafficking of NSP-DNA complexes is accessorized by the NSP-interacting guanosine triphosphatase (NIG) at the cytosolic side. Here, we report the nuclear redistribution of NIG by AtWWP1, a WW domain-containing protein that forms immune nuclear bodies (NBs) against begomoviruses. We demonstrated that AtWWP1 relocates NIG from the cytoplasm to the nucleus where it is confined to AtWWP1-NBs, suggesting that the NIG-AtWWP1 interaction may interfere with the NIG pro-viral function associated with its cytosolic localization. Consistent with this assumption, loss of AtWWP1 function cuased plants more susceptible to begomovirus infection, whereas overexpression of AtWWP1 enhanced plant resistance to begomovirus. Furthermore, we found that a mutant version of AtWWP1 defective for NB formation was no longer capable of interacting with and relocating NIG to the nucleus and lost its immune function against begomovirus. The antiviral function of AtWWP1-NBs, however, could be antagonized by viral infection that induced either the disruption or a decrease in the number of AtWWP1-NBs. Collectively, these results led us to propose that AtWWP1 organizes nuclear structures into nuclear foci, which provide intrinsic immunity against begomovirus infection.

Bioinformatics Analysis of the Receptor-Like Kinase (RLK) Superfamily.

Receptor-like kinases (RLKs) play key roles during development and in responses to the environment. In plant immunity, some members of RLKs function as pattern recognition receptors (PRRs), which, upon recognition of pathogen-associated molecular patterns (PAMP), are recruited into active complexes to induce pathogen-triggered immunity (PTI). In this chapter, we describe the bioinformatics tools and procedures for the identification and phylogenetic classification of RLKs from different plant species as a framework for understanding RLK function in signal transduction and immunity.

Rama: a machine learning approach for ribosomal protein prediction in plants.

Ribosomal proteins (RPs) play a fundamental role within all type of cells, as they are major components of ribosomes, which are essential for translation of mRNAs. Furthermore, these proteins are involved in various physiological and pathological processes. The intrinsic biological relevance of RPs motivated advanced studies for the identification of unrevealed RPs. In this work, we propose a new computational method, termed Rama, for the prediction of RPs, based on machine learning techniques, with a particular interest in plants. To perform an effective classification, Rama uses a set of fundamental attributes of the amino acid side chains and applies a two-step procedure to classify proteins with unknown function as RPs. The evaluation of the resultant predictive models showed that Rama could achieve mean sensitivity, precision, and specificity of 0.91, 0.91, and 0.82, respectively. Furthermore, a list of proteins that have no annotation in Phytozome v.10, and are annotated as RPs in Phytozome v.12, were correctly classified by our models. Additional computational experiments have also shown that Rama presents high accuracy to differentiate ribosomal proteins from RNA-binding proteins. Finally, two novel proteins of Arabidopsis thaliana were validated in biological experiments. Rama is freely available at http://inctipp.bioagro.ufv.br:8080/Rama .

The molecular chaperone binding protein BiP prevents leaf dehydration-induced cellular homeostasis disruption.

BiP overexpression improves leaf water relations during droughts and delays drought-induced leaf senescence. However, whether BiP controls cellular homeostasis under drought conditions or simply delays dehydration-induced leaf senescence as the primary cause for water stress tolerance remains to be determined. To address this issue, we examined the drought-induced transcriptomes of BiP-overexpressing lines and wild-type (WT) lines under similar leaf water potential (ψw) values. In the WT leaves, a ψw reduction of -1.0 resulted in 1339 up-regulated and 2710 down-regulated genes; in the BiP-overexpressing line 35S::BiP-4, only 334 and 420 genes were induced and repressed, respectively, at a similar leaf ψw = -1.0 MPa. This level of leaf dehydration was low enough to induce a repertory of typical drought-responsive genes in WT leaves but not in 35S::BiP-4 dehydrated leaves. The responders included hormone-related genes, functional and regulatory genes involved in drought protection and senescence-associated genes. The number of differentially expressed genes in the 35S::BiP-4 line approached the wild type number at a leaf ψw = -1.6 MPa. However, N-rich protein (NRP)- mediated cell death signaling genes and unfolded protein response (UPR) genes were induced to a much lower extent in the 35S::BiP-4 line than in the WT even at ψw = -1.6 MPa. The heatmaps for UPR, ERAD (ER-associated degradation protein system), drought-responsive and cell death-associated genes revealed that the leaf transcriptome of 35S::BiP-4 at ψw = -1.0 MPa clustered together with the transcriptome of well-watered leaves and they diverged considerably from the drought-induced transcriptome of the WT (ψw = -1.0, -1.7 and -2.0 MPa) and 35S::BiP-4 leaves at ψw = -1.6 MPa. Taken together, our data revealed that BiP-overexpressing lines requires a much higher level of stress (ψw = -1.6 MPa) to respond to drought than that of WT (ψw = -1.0). Therefore, BiP overexpression maintains cellular homeostasis under water stress conditions and thus ameliorates endogenous osmotic stress.