Figures
Abstract
Venturia inaequalis is the causal agent of apple scab, one of the most devastating diseases of apple. Due to several distinct features, it has emerged as a model fungal pathogen to study various aspects of hemibiotrophic plant pathogen interactions. The present study reports de novo assembling, annotation and characterization of the transcriptome of V. inaequalis. Venturia transcripts expressed during its growth on laboratory medium and that expressed during its biotrophic stage of infection on apple were sequenced using Illumina RNAseq technology. A total of 94,350,055 reads (50 bp read length) specific to Venturia were obtained after filtering. The reads were assembled into 62,061 contigs representing 24,571 unique genes. GO analysis suggested prevalence of genes associated with biological process categories like metabolism, transport and response to stimulus. Genes associated with molecular function like binding, catalytic activities and transferase activities were found in majority. EC and KEGG pathway analyses suggested prevalence of genes encoding kinases, proteases, glycoside hydrolases, cutinases, cytochrome P450 and transcription factors. The study has identified several putative pathogenicity determinants and candidate effectors in V. inaequalis. A large number of transcripts encoding membrane transporters were identified and comparative analysis revealed that the number of transporters encoded by Venturia is significantly more as compared to that encoded by several other important plant fungal pathogens. Phylogenomics analysis indicated that V. inaequalis is closely related to Pyrenophora tritici-repentis (the causal organism of tan spot of wheat). In conclusion, the findings from this study provide a better understanding of the biology of the apple scab pathogen and have identified candidate genes/functions required for its pathogenesis. This work lays the foundation for facilitating further research towards understanding this host-pathogen interaction.
Citation: Thakur K, Chawla V, Bhatti S, Swarnkar MK, Kaur J, Shankar R, et al. (2013) De Novo Transcriptome Sequencing and Analysis for Venturia inaequalis, the Devastating Apple Scab Pathogen. PLoS ONE 8(1): e53937. https://doi.org/10.1371/journal.pone.0053937
Editor: Jason E. Stajich, University of California Riverside, United States of America
Received: September 13, 2012; Accepted: December 4, 2012; Published: January 17, 2013
Copyright: © 2013 Thakur et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors acknowledge the financial support provided by Department of Biotechnology (Government of India) under sub-program I (Enhancing productivity of apple through understanding the molecular basis of host-pathogen interactions) of network project entitled “Improvement of Apple through Biotechnological Interventions” and in-house projects (MLP0031 and MLP0062) of Council of Scientific and Industrial Research, (CSIR) Government of India, that support the GJ lab research initiatives to unravel the molecular basis of fungal pathogenesis on apple. The computational part of this study was supported by MLP0037. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Venturia inaequalis (Cke.) is a phytopathogenic fungus that causes Apple scab, the black spot disease of apple [1], [2], [3]. It causes deformation in shape and size of the affected fruits, premature leaf/fruit fall and enhances susceptibility of apple tree to chilling and freezing injuries. Overall, it renders apple unsuitable for trade and causes up to 70% of yield reductions. Like other obligate parasites, it generally infects and lives in association with living host tissues. However, its ability to be cultured on laboratory medium, possibility of in vitro mating, existence of extensive population diversity, uninucleate conidia, genetically uniform progenies, stability of genotype and phenotype of the progeny even after multiple rounds of sub-culturing and availability of standardized protocol for genetic manipulation, etc. makes it a useful model to study the pathogenesis of hemibiotrophic fungal pathogens [1], [2], [3]. V. inaequalis has broad geographical distribution and an interesting growth pattern. It is mostly restricted in sub-cuticular space of the apple tissues, does not form haustoria, and no apparent mechanical pressure is observed during its penetration into host cuticle [1].
V. inaequalis is rapidly evolving, hyper variable and eight different races have been reported from different parts of the world [1]. Using multi locus microsatellite markers, Gladieux et al [4] had revealed the origin of V. inaequalis in central Asia (which is also centre of origin of apple), from where it seems to have traveled along with apple (domestication/introduction) using silk route to get into Europe and expanded into other continents. Practices like disease resistance breeding and protective fungicide spray are being routinely used to control apple scab disease [3], [5], [6]. Several disease resistance (R) genes have been identified in apple and one of them, namely Rvi6 (Vf) has been cloned, which is being explored to engineer scab resistant apples [7], [8]. But both of these commonly utilized methodologies to control apple scab, are facing big threats as the pathogen is rapidly evolving resistance against fungicides [9], [10], [11] and has breached several R gene mediated resistance [12], [13], [14]. So far, inadequate studies have been done to understand the molecular interaction between Venturia and apple, which could otherwise provide insights to develop alternative strategies to prevent scab disease.
V. inaequalis is a heterothallic fungus and contains seven chromosomes. Its genome size is estimated to be around 100 Mb [15] while most of the other ascomycetes genomes are mostly in the range of 40 Mb. The best available genetic map for V. inaequalis consists of eleven linkage groups with total map length of 1106 cM [16]. Efforts have been made to identify effectors which might determine the pathogenicity of this pathogen. Recently sixteen putative candidate effectors have been mined from Venturia ESTs, three of them were found to be differentially regulated during in planta growth [17]. Although the functions of these effectors has not been established as yet. In several other pathosystem, it has been demonstrated that pathogen uses effectors to suppress host immunity, while some of these effectors (better known as avr factors) induces host immunity upon being recognized by the host R gene(s). Hence, identifying the avr genes and those involved in suppressing host immunity would be useful in ensuring durable scab resistance in apple.
With the introduction of Next Generation Sequencing (NGS) platforms like Roche 454, Illumina Genome Analyzer and Applied Biosystems SOLiD, there is an exponential increase in both genome and transcriptome sequencing efforts [18], [19], [20], [21]. Advancement of various de novo sequence assembly algorithms has made it possible to assemble short reads obtained by these NGS platforms to generate draft genome/transcriptome sequences [22], [23], [24]. These tools are being utilized to understand the intricacies of host-pathogen interactions by unraveling the genomic and transcriptome sequences. Transcriptome sequencing is becoming an economically attractive alternative to whole genome sequencing, as one can study the expressed part of the genome and bring novel insights about the underlying biological processes. In the present study, Illumina Genome analyser GA IIx has been used to sequence the transcriptome of V. inaequalis and obtain genomic insights about this devastating pathogen. A total of 94,350,055 reads of V. inaequalis transcriptome, assembled into 62,061 contigs, representing 24,571 non-redundant unigenes, were functionally annotated. The present work also predicts the secretome, effectors and genes involved in host-pathogen interactions. Also the phylogenomic study of Venturia with respect to several ascomycetes fungal pathogens have been explored in this study.
Materials and Methods
Isolation, culturing and bioassay of Venturia inaequalis
Indian isolate of V. inaequalis was isolated from the diseased apple fruit sample collected in a sealed polyethylene biohazard bag, from Kullu district of Himachal Pradesh, India (No permission was required to collect the diseased apple fruit sample to carry out this research work). Dilution plating and repeated sub-culturing on PDA (39 g/L; Potato Dextrose Agar; HiMedia, Mumbai, India; plates incubated at 20°C) were used to obtain pure culture of V. inaequalis. The identity of the pathogen was confirmed through microscopic analysis, rDNA sequencing and by testing its ability to cause scab symptoms on susceptible cultivar Gala. In order to enrich the transcriptome, we explored the transcripts expressed during its growth on apple and on laboratory media. 10 µl mycelial suspension isolated from one month old culture of Venturia (grown on PDA plate; at 20°C) was inoculated on the detached leaves of susceptible apple cultivar Gala as described by Win et al. [25] and also in 2 ml of PDB (24 g/L; Potato Dextrose Broth; HiMedia, Mumbai, India) broth with proper shaking. To prevent microbial contaminations generally associated with field grown apple leaves (which might potentially create confusion during assembling Venturia specific transcripts), we used in vitro grown leaves of Gala in this study. The samples were collected from both the Venturia inoculated leaves and Venturia growing in PDB broth at 0, 2 and 5 dpi (day post inoculation).
Library preparation and transcriptome sequencing
Total RNA from each of the collected samples was isolated using iRIS [26]. The integrity and quantity of isolated RNA were assessed on Bioanalyser (Agilent; 2100). Further, 4 µg RNA from each of these samples was processed for RNA library preparation using TruSeq RNA sample Prep Kit (Illumina) as per manufacturer's instruction. Library was quantified using Qubit™ RNA assay kit for Qubit 2.0® Fluorometer (Life technologies) and 10 pM of the prepared library were loaded onto a flowcell for cluster generation. The following loading arrangement was used: Lane 2: Venturia (0 dpi), Lane 3: Gala leaves infected with Venturia (2 dpi), Lane 4: PhiX Control, Lane 5: PDB grown Venturia (2 dpi), Lane 6: Gala leaves infected with Venturia (5 dpi), Lane 7: PDB grown Venturia (5 dpi), Lane 8: mixture of all above mentioned libraries except PhiX control. As per our previous experience, we selected 10 pM concentration of prepared library for cluster generation in this study and obtained good number (∼470–540 k/mm2) of clusters. Library sequencing was performed on Illumina Genome Analyzer GAIIx as per manufacturer's instructions (Illumina Inc).
De novo assembly and sequence clustering
The raw sequencing data was transformed into Single End (SE) 72 bp reads, using GERALD base-calling (a CASAVA package tool provided by Illumina). Resulting sequence reads were stored in FASTQ format. The 3′-end read trimming (keeping first 50 bp) was performed using read filtering tool, filteR [27]. The reads for V. inaequalis were obtained from different lanes of the flowcell. Three lanes (lane 2, 5 and 7) were having only Venturia specific transcripts (as they were covering different stages of its growth on laboratory medium). Other three lanes (lane 3, 6 and 8) were containing mixed sample for Venturia and apple (as they were processed either from Venturia infected apple leaves: lane 3 and 6; or from the mixture: lane 8). To remove apple specific reads (from lane 3, 6 and 8), reads were mapped onto apple genome [28] using Bowtie [29] with default settings. The unmapped reads were considered to have come from V. inaequalis. De novo assembling of high quality reads was performed using SOAPdenovo, ABySS and Velvet and in combination with a series of stepwise strategies [22], [23], [24], [27], [30]. The best assembled transcripts from each assembler (SOAPdenovo, ABySS and Velvet) were selected; reflecting the best balance between the number of contigs produced, average coverage, N50 length value of total assembly and average sequence length attained. Redundant set of sequences were removed by filtering and merging similar sequence stretches with TGICL-CAP3 [31] and CD-HIT-EST [32] based sequential and hierarchical clustering.
Assembly validation and similarity search for assembled V. inaequalis transcripts
To assess the reliability of assembly, 155 experimentally validated nucleotide sequences for V. inaequalis available at GenBank were used (File S1). BLASTn analysis was performed for reported nucleotide sequences against the set of assembled transcripts sequences at E-value threshold of 1e−05. Primer pairs were designed to amplify the partial sequences of 15 randomly selected transcripts and the PCR amplicons were sequenced using dye Terminator 3.1 Cycle Sequencing Kit (Applied Biosystems Foster City,USA) on 3130 XL Genetic analyzer from Applied Biosystem. BLAST2 analyses were performed to determine the homology of the sequences obtained through Sanger sequencing with that obtained through de novo assembly of the transcriptome. For similarity search, the assembled and filtered transcript sequences obtained after hierarchical clustering were scanned against NR protein sequence database [33] using BLASTx [34] with E-value threshold of 1e−05. Further, these assembled transcripts were clustered on the basis of BLASTx hits against NR database. Several assembled transcript sequences, though display absolutely no similarity with each other, but map to the different parts of the same gene. Considering this, such transcripts were grouped into a single group representing unique genes. This reduces the inflated number of transcripts and gives better representation of total unique genes identified. The above strategy is known as Dissimilar Sequences (DS) clustering [27] and the same has been adopted in the present study to obtain more precise transcriptome representation for V. inaequalis. The non-redundant V. inaequalis transcripts obtained after DS clustering are referred to as Set A transcripts, while the transcripts with no homology in BLAST analysis have been referred to as Set B transcripts. The transcripts in Set A and Set B were translated as described in Figure 1.
A flow chart describing the strategies for prediction of RxLR effectors and open reading frame of the transcripts.
Sequence annotation and protein family classification
Assembled transcripts of V. inaequalis were searched against UniProt database [35] to assign associated GO [36], KEGG [37], [38] and EC [39] based annotation using Annot8r annotation tool [40], with E-value threshold of 1e−01. Protein families were classified by searching the assembled transcripts against Pfam [41] and InterProScan [42]. Pfam database was searched using HMMER3 [43] and the Conserved Domain database [44] was scanned using RPS-BLAST [45]. Protease families were identified using BLASTp (E-value <10−20) against MEROPS peptidase database release 9.6 [46]. Cytochromes (CYPs) were named according to classification details collected from BLASTp (E-value <10−5) against Fungal Cytochrome P450 database version 1.2 [47]. Transporters were classified based on BLASTp (E-value <10−10) against Transporter Classification Database [48]. KinBase database was scanned using BLASTp (E-value <10−10) to characterize sequences belonging to Kinase families [49]. Carbohydrate-degrading enzymes selected from InterProScan and Pfam analysis were classified according to GH (Glycoside hydrolase) family as classified in CAZy database [50]. PHI-base (pathogen-host interaction database) [51] was searched for different phenotypic categories like loss of pathogenicity, reduced virulence, effector, lethal, increased virulence etc. with BLASTp (E-value <10−5). As Set B transcripts did not show any homology while performing BLAST searches, only Set A transcripts of Venturia were used for comparative analysis. Proteases, carbohydrate-degrading enzymes and membrane transporters were predicted from Venturia transcriptome and compared with that encoded by Magnaporthe oryzae, Fusarium graminearum, Botrytis cinerea and Sclerotinia sclerotiorum. The dataset for proteases and membrane transporters were obtained from Zheng et al [52] while the carbohydrate-degrading enzymes were obtained from CAZy [50].
Secretome and RxLR effector identification
The amino acid sequences (Set A and Set B) of V. inaequalis transcriptome were further analyzed for prediction of secreted proteins. Sequences smaller than 70 amino acids, were not considered for further analysis (Figure 1). The remaining sequences with positive SignalP [53] prediction for signal peptide cleavage site at N-terminal region between 10–40 amino acids, without any transmembrane region as predicted by TMHMM [54], were selected as the candidate secreted proteins. RxLR pattern was searched using FuzzPro, a tool from EMBOSS package [55]. Sequences having RxLR patterns positioned between 30–60 amino acids and appearing after the observed signal cleavage site (30 amino acids from the start) were considered [56]. Also a less stringent approach was applied to predict RxLR effectors, wherein any protein with RxLR pattern and signal peptide cleavage site were considered as candidate RxLR effectors (Figure 1). The predicted secretome was also searched for similarity across known set of effectors (File S2, S3) using BLASTx (E-value <10−2). First set contained the sequences of known effectors of phytopathogenic fungi (n = 32) and the second set contained previously reported sequences of V. inaequalis predicted effectors (n = 16) [17].
Phylogenomic analysis
The reference protein sequences of nine species namely, Aspergillus fumigatus, Aspergillus nidulans, Candida albicans, Magnaporthe oryzae, Neurospora crassa, Sclerotinia sclerotiorum, Pyrenophora tritici-repentis, Gibberella zeae, Botryotinia fuckeliana, were downloaded from NCBI Protein RefSeq database [57]. In case of V. inaequalis, the amino acid sequences from transcripts with k-mer coverage [24] greater than or equal to 50 were used for ortholog detection and phylogenetic analysis. The sequences were analyzed using Hal, an automated pipeline for phylogenetic analysis of high throughput data [58]. Amino acid sequences of all ten species (A. fumigatus, A. nidulans, C. albicans, M. oryzae, N. crassa, S. sclerotiorum, P. tritici-repentis, G. zeae, B. fuckeliana and V. inaequalis) were imported in FASTA format and subjected to all vs all BLASTp search. MCL clustering algorithm [59] with a range of inflation parameters from 1.1 to 5 was used to group amino acid sequences into orthologous clusters. These clusters were screened for any possible redundancy. Redundancy could exist if clusters were found at more than one inflation parameter, clusters contained more than one amino acid sequence per genome, or the clusters containing amino acids sequences whose best reciprocal BLAST hit was found outside the cluster. Amino acid sequences for accepted orthologous clusters were extracted from their respective proteome dataset in FASTA format and subsequently aligned using ClustalW [60]. To counter uneven regions of alignments, three separate super-alignments were created: 1) by removing all gap-containing columns (remgaps), 2) by removing uneven regions of alignments based on the default conservative option (Gblocks-con) and 3) by liberal (Gblocks-lib) options of the Gblocks tool [61]. Best models of amino acid substitution for alignment were determined using ProtTest [62] and by the AIC criterion [63] with 10 models: Dayhoff, Blosum62, JTT, MtREV, WAG, RtREV, CpREV, VT, MtMM and GTR. For analysis, three phylogenetic trees were generated, representing three super-alignments (remgaps, Gblocks-con and Gblocks-lib). RAxML [64] was run on each of the super-alignments with the PROTCAT setting for the rate model with best-fitting model of amino acid substitution. Nodal support was estimated based on 1000 bootstrap replications using Candida albicans as an out-group. The trees were visualized using Dendroscope [65].
Results and Discussion
V. inaequalis is one of the most important plant pathogen. Besides being havoc to apple industry, it has an interesting life style [1], [2]. Efforts are being made to understand the genomic structure and genome sequence of this deadly pathogen [16], [66], [67]. Understanding its transcriptome would provide insights about its pathogenicity mechanisms and arsenals being used to invade apple. With the advancement of sequencing tools and techniques, it is now possible to explore the entire transcriptome at a time. In the present study, we have applied Illumina Next Generation Sequencing platform to unravel the transcriptome of an Indian isolate of V. inaequalis. The present work reports transcriptome sequencing and de novo assembly of obtained short reads, comprehensive annotation of assembled sequences, protein-family classification, effector identification and exploration of phylogenomic relationship of this deadly pathogen with other ascomycetes fungi.
De novo sequence assembly of V. inaequalis transcriptome
De novo assembly of short reads without a reference genome still remains a challenge, in spite of recent development of several computational tools and approaches for data assembly and analysis. Single-end (SE) run of 72 cycles was performed on Illumina GAIIx for different growth/infectious stage of V. inaequalis. Quality check through filteR tool revealed comparatively declining read quality at 3′-end (Figure 2). It was found that the average read quality was acceptable up to 50th cycle. Therefore, trimming of reads was done to maintain read length of 50 bp. This way, a total of 147,780,763 SE reads were generated, out of which 129,766,417 reads passed quality filtering. After removing apple specific reads by mapping them onto apple genome, a total of 94,350,055 reads remained specific for V. inaequalis (Table 1). To obtain transcript sequences, the best assembled transcripts set from different available tools (Velvet, ABySS and SOAPdenovo) at different k-mers ranging from 19 to 47 were selected. The parameters considered were: transcripts having assembly length higher than 100 bp, average coverage, average transcript size, percentage of transcripts having length higher than 1000 bp, N50 value and highest transcript length. K-mer size of 29, in case of Velvet and SOAPdenovo and 27, in case of ABySS, emerged as the best choices for performing assembly. They displayed the best balance between transcripts number, coverage, maximum and average transcript length (File S4). At selected k-mer sizes, a total of 68,027 (with SOAPdenovo), 26,678 (with Velvet) and 45,805 (with ABySS) assembled transcripts were obtained. The average length for these transcript assemblies were 459 (with SOAPdenovo), 815 (with Velvet) and 568 (with ABySS) bases, having average coverage of 71, 170 and 108, respectively. Thus, the combined initial assembling of V. inaequalis transcriptome sequences yielded 140,510 transcripts.
The line diagram of the quality score of the reads obtained from different lanes.
Homology search and sequence clustering
Several sequences shared similarity between the different assembly sets (obtained from different assembling tools as described above) as well as exhibited overlap within each set, causing redundancy. In order to remove such redundancy and overrepresentation of transcripts, sequence similarity based hierarchical clustering was performed to merge such sequences After performing hierarchical clustering with TIGR Gene Indices clustering tools (TGICL), contig assembly program (CAP3) and Cluster database at high identity with tolerance (CD-HIT), number of assembled transcripts got reduced from a total of 140,510 to 62,061. A total of 29,750 such assembled sequences returned significant BLASTx hits while no hit was found for 32,311 sequences. Another clustering step was carried out for sequences with significant BLAST hits. Sequences with no apparent significant identity among themselves might belong to different parts of the same gene or may represent different isoforms. Counting them as separate transcripts would only inflate the number of unique genes [27]. Therefore, all such transcripts which returned hit to some common reference gene were assigned to a common cluster group, representing a unique gene. A set of in house scripts were used to scan for all those assembled transcript sequences that returned best hit with common reference but differed in their location. This step reduced the total number of transcripts with significant BLAST hits, from 29,750 to 24,571 (File S5). For the sake of uniformity, here onwards the non-redundant transcripts of V. inaequalis with BLAST hits (24,571) are referred to as Set A, while the transcripts with no homology in BLAST analysis (32,311) have been referred to as Set B. It is to be noted that Set B transcripts could also display the above mentioned clustering property and their total number might go lower than the observed value. Assembly statistics at both levels of clustering are given in Table 2, revealing average coverage being 71.63 and 127.42 for sequence similarity based clustering (with combination of TGICL and CD-HIT) and DS clustering, respectively. The sequences for Set A and Set B transcripts of V. inaequalis are provided in File S6 and File S7, respectively.
Validation of assembled sequences
The assembled transcriptome sequences were validated by BLASTn search against 155 publically available nucleotide sequences of V. inaequalis (File S1). Significant hits were observed for 131 sequences (84.51%), while no hit could be obtained for 24 previously reported nucleotide sequences of V. inaequalis. Most of the assembled transcript sequences were found correctly aligning in continuous manner, with an average identity of 98.79. The observed minimum coverage was 8.29% while maximum coverage was found to be 100%. Out of 131 nucleotide sequences, 105 sequences had at least 50% coverage, suggesting overall good assembly quality (File S8). Out of 24 NCBI sequences for which no hit was found across the assembled transcriptome of V. inaequalis, 15 were corresponding to 18S ribosomal RNA sequences and 6 were microsatellite sequences. The remaining three sequences for which no hit was found could be some stage specific transcripts. This is to be noted that even in finished genomes like A. thaliana around, 13% of the known nucleotide sequences could not be assigned to the final assembly [68]. In case of human only 64% of the reads could be mapped onto the RefSeq database of well annotated human genes [69].
To further validate the de novo transcriptome sequence assembly partial fragments of 15 randomly selected transcripts were PCR amplified and sequenced using Sanger dye termination based method. File S9 summarizes the BLAST2 analysis demonstrating high score, E-value and identity between the sequences obtained through Sanger sequencing with that of de novo assembled transcripts sequences.
Functional annotation and classification of V. inaequalis transcriptome
For functional annotation, the V. inaequalis transcripts were compared against amino acid sequences available at UniProt database using BLASTx algorithm. The associated hits were searched for their respective Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Enzyme Commission Codes (EC) for each query sequence and the highest bit score was selected. Annotation against GO database yielded significant hits for 18,431 out of 24,571 unigenes of V. inaequalis. Figure 3A and Figure 3B represent the distribution of V. inaequalis transcripts across the various GO categories associated with biological process and molecular function, respectively. Under biological process, categories like metabolic process, response to stimulus, nucleic acid metabolism, cellular process and transport etc were highly represented. For molecular function ontology, genes associated with binding, catalytic, transferase and oxidoreductase activities were found most abundant. This indicates occurrence of rapid growth and extensive metabolic activity for this pathogenic fungi.
Gene Ontology (GO) term assignments to V. inaequalis unigenes, based on significant GO slims, summarized into two main GO categories: biological process (A), and molecular function (B). Functional characterization and abundance of V. inaequalis transcriptome for enzyme classes, and KEGG pathways are represented in C and D respectively. Only top 25 most abundant EC and KEGG pathways are represented. Area under each pie represents the value in percent.
Best EC classification was obtained for a total of 9,731 unique genes, and KEGG classification was obtained for 10,821 unique genes. Figure 3C enlists the top 25 abundant enzyme classes observed for V. inaequalis unigenes. Interestingly, a large amount of transcripts belonged to nonspecific serine/threonine protein kinase enzyme class (12.82%). Figure 3D displays top 25 KEGG pathways represented by Venturia unigenes. Large proportion of such transcripts belonged to ribosome (4.84%), spliceosome (2.97%) and pathways involved in plant-pathogen interaction (2.90%). This was followed by RNA transport, starch and sucrose metabolism, plant hormone signal transduction, glycolysis/gluconeogenesis, protein processing in endoplasmic reticulum, ubiquitin mediated proteolysis and others. An InterproScan analysis of unigenes identified 5,418 conserved protein families in V. inaequalis, Pfam scan resulted in identification of 3,960 families. As Set B transcripts did not show any significant homology in BLAST analysis, we performed Pfam and InterProScan analysis to obtain insights about their putative functions. The analysis resulted in identification of 475 families with InterproScan and 352 families by Pfam scan. As shown in Figure 4, presence of conserved domain in several of Set B transcripts suggests that they might be encoding genes with important functions such as fibronectin attachment protein, topoisomerase, transcriptional regulator, DNA polymerase, pre-mRNA spicing factors etc.
Sequences showing no homology with nr database (Set B transcripts) were subjected to Conserved Domain Database (CDD) to predict the conserved domain. Top 15 highly represented functional conserved domains are shown by bars.
BLASTp analysis against KinBase database resulted into 240 hits across nine different kinase groups, namely, AGC (protein kinase A, G and C group), atypical kinases (Hisk, BRD, PDHK), CAMK (Calcium/Calmodulin regulated kinases), CK1 (Casein Kinase 1 group), CMGC (CDK, MAPK, GSK3 and CLK kinases), STE (MAP kinase cascade kinases), TK (Tyrosine kinase), TKL (Tyrosine kinase-like group) and others (Table 3). Highest number of transcripts was found to be associated with CMGC kinase group. Cytochrome P450s (CYPs) play an important role in physiology of fungi and are involved in biosynthesis of secondary metabolites and detoxification [70]. A total of 88 transcripts encoding CYP subfamily proteins were identified in V. inaequalis while selecting the unique hits from Pfam and InterproScan searches and BLASTp analysis against fungal Cytochrome P450 database (File S10). Overall, this study had identified transcripts that might be involved in several important molecular and cellular functions of V. inaequalis.
Secretome of V. inaequalis
Secreted pathogenic proteins and effectors, better known as secretome are crucial for establishing infection on the host plant [71], [72]. These secreted proteins may disable plant defense and sabotage cellular processes to suit the needs of invading pathogens. There are a number of computational tools available that predict whether a protein is likely to be secreted or not [73]. We used SignalP [53]; to predict the presence of signal peptides and TMHMM [54]; to predict the presence of transmembrane helices to define the secretome of V. inaequalis. Those proteins which contain signal peptides but lack transmembrane helices are considered as secreted proteins. Following such criteria, 463 Set A transcripts and 483 Set B transcripts were predicted to be secreted (Table 4). Interestingly, Venturia seems to harbor similar number of secreted proteins to that of majority of phytopathogenic fungi (Figure 5A).
Comparative analysis of secretome sizes of various filamentous fungi (A), showing that the secretome size of V. inaequalis is comparable to other fungi. Gene Ontology (GO) term assignments to V. inaequalis secretome, based on significant GO slims, summarized into two main GO categories: biological process (B), and molecular function (C). Area under each pie represents the value in percent. Nearly 50% of genes are involved in metabolic processes as shown in B, while peptidase activity and catalytic activity are predominant molecular functions (C).
To functionally annotate the secretome, we performed Gene Ontology and Pfam protein domains searches. GO biological process analysis revealed that nearly 50% of Set A transcripts are involved in metabolism (Figure 5B). The overrepresentation of genes associated with metabolic processes has also been observed in the secretome of Fusarium graminearum, the causal organism of Fusarium Ear Blight disease of small grain cereals [74], and in the secretome of Phytophthora infestans, the casual agent of tomato and potato late blight diseases [75]. Several genes associated with transport (protein, lipid etc), response to stress, pseudohyphal growth and killing of cells of other organisms were found present in the secretome. The secretome appeared enriched for peptidase activity (22.05%), catalytic activity (14.17%) and hydrolase activity (6.03%), as depicted by GO molecular function analysis (Figure 5C). Quite a few genes that might regulate signaling cascades were also observed in the identified secretome.
Several Pfam domains were noticed in amino acid sequences of the transcripts encoding predicted secreted proteins of V. inaequalis (File S11). The FAD binding domain, cutinase, eukaryotic aspartyl protease, hydrophobic surface binding protein A, glycosyl hydrolases families and those related to cell wall degradation were among the most frequently found Pfam domains in such transcripts. A domain associated with chitin recognition protein (PF00187) was one such domain. Chitin-binding proteins are thought to protect the fungal cell wall from chitinases that are produced by host plants [76]. Another important Pfam domain found in V. inaequalis is generally associated with isochorismatase family proteins. Conversion of isochorismate to 2, 3-dihydroxybenzoate and pyruvate has been reported to be catalyzed by such enzymes. Also, they are involved in synthesis of anti-microbial compounds such as phenazine [77] and siderophore and enterobactin [78]. Although isochorismatases are present in many filamentous ascomycetes, they are known to be secreted only in phytopathogenic fungi. As isochorismate is a precursor of SA (Salicyclic Acid, an important plant hormone involved in plant defense and systemic acquired resistance); the phytopathogens might use isochorismatases to sequester SA accumulation. This in turn might attenuate the plant defense [79]. Another important Pfam domain identified in Venturia secretome is associated with inosine-uridine preferring nucleoside hydrolase enzyme. This enzyme is important for parasitic organisms, which are deficient in de novo synthesis of purines and are dependent on salvaging the host purine nucleosides [80]. Interestingly the purine auxotrophs of V. inaequalis were compromised for pathogenesis on apple [81], suggesting the importance of inosine-uridine preferring nucleoside hydrolase enzyme during pathogenesis of Venturia on apple.
Cell wall degrading enzymes present in V. inaequalis
As mentioned above, several Pfam domains involved in cell wall degradation were present in the secretome. Phytopathogens have impressive arsenal of plant cell wall degrading enzymes (CWDE) which are secreted to depolymerize different constituents of plant cell wall and required for successful pathogenesis. Magnaporthe grisea, the casual agent of rice blast disease has 30 enzymes for cellulose degradation and 44 enzymes for hemicellulose degradation [82], [83]. The present study reports six cutinases (PF01083; methyl esterases that degrade cutin), two cellulases (glycosyl hydrolase family 5), three pectate lyases and two pectin acetyl esterases in the secretome. Also, a large number of β-glucosidase were found (n = 22) in the scab pathogen, followed by α-mannosidase (n = 25), polygalacturonase (n = 19). Four contigs with PF00295 domain (encoding Glycosyl hydrolase family 28) were predicted to be secreted. This domain is present in enzymes involved in pectin degradation. Ten genes belonging to family GH 61 were present in the transcriptome. It is interesting to note that members of this family protein act as factors that enhance the hydrolysis of lignocellulose [82]. Thus, several CWDEs encoding transcripts of Venturia were identified in this study and data is summarized in File S12. It would be interesting to explore their function during establishment of apple scab disease. Interestingly, CWDEs are thought to play critical role in Venturia-apple interactions and have been speculated to be required for host penetration and nutrient uptake from the plant [1].
Identification of V. inaequalis effectors
Beside CWDEs, small molecular weight secreted proteins of phytopathogenic fungi are known to reprogram the host metabolism and prevent the execution of plant defense responses [84], [85], [86], [87], [88]. These small molecular weight secreted proteins are well known as effectors. However, during million years of co-evolution, the host plants have evolved strategies to recognize pathogenic effectors and mount resistance (R) gene mediated disease resistance, which is potent enough to combat disease [89]. Such effectors are known as avirulence factors, as their presence render pathogen avirulent. Several avirulence factors were previously predicted from V. inaequalis and some of them are known to be race determinants [1], [2], [90]. However, till now no such Avr factors had been molecularly identified or cloned. In an attempt, Bowen et al [17] have identified 16 putative effector proteins of Venturia and demonstrated that three of them are induced during in planta infection. Presence of all 16 predicted effector proteins were noticed in the Venturia transcriptome sequenced in this study. Our transcriptome data could also reflect the presence of few Cladosporium fulvum (syn. Passalora fulva) (model ascomycetes fungi, extensively used to identify and characterize fungal effectors) effectors in Venturia (Table 5). Notably amongst them is a homolog of ECP6; the LysM domain containing effector of C. fulvum, which plays a critical role in its virulence [91] (in Set B; Table 5).
Conserved RxLR motifs, located within N-terminal 60 amino acids downstream of signal peptide cleavage sites are present in many oomycetes effector proteins [92]. The RxLR motif is thought to be required for host translocation of effectors. Recently, presence of RxLR motifs in M. grisea and some other ascomycetes fungi has also been reported. In order to explore whether V. inaequalis also possess such RxLR effectors, amino acid sequences of secretome were subjected to motif analysis. A total of 41 (28 in Set A and 13 in Set B) potential RxLR effectors were identified through these searches (Table 4). However, under stringent condition, we could only get two RxLR effectors. This suggests that the RxLR motif might not be a useful motif for identifying candidate effectors from Venturia. This could also reflect that oomycetes and fungi, separated by several hundred million years of evolution, might use different mechanism for effector delivery. However, availability of genome sequence could provide better representation of RxLR effectors in Venturia. In order to obtain functional insights about the predicted secreted effectors with RxLR motifs, BLASTx analysis and GO annotations were performed for each of them and the results are summarized in File S13. Interestingly, BLAST analysis revealed that several of putative RxLR containing sequences demonstrate homology with cell wall degrading enzymes, such as beta-glucosidase/beta-1,4-glucosidase, extracellular exo-polygalacturonase, pectate lyase A etc. Functional characterization of these effectors could help us in understanding how they interfere with host machinery and assist in scab disease establishment. Also, it would be interesting to explore whether any of these predicted effectors may function as avirulence factor(s) that govern race specificity on differential cultivars of apple. Identification of such factors would be really helpful in resistant breeding programs for ensuring durable resistance.
Pathogenicity associated genes of V. inaequalis
To predict potential pathogenicity genes of V. inaequalis, whole transcriptome BLASTp analysis against the pathogen-host interaction gene database (PHI database) was performed. PHI database is a collection of experimentally verified pathogenicity, virulence, lethal and effector genes from fungi, oomycetes and bacteria. The analysis resulted in 2,159 hits with E-value cut off of 1e−05. A total of 482 unique hits corresponding to known pathogenicity determinants of various pathogenic fungi (File S14), 102 hits corresponding to genes associated with loss of pathogenicity, 282 hits for reduced virulence and three for pathogenic effectors were observed in PHI database (File S14). GO biological process analysis revealed that majority of Venturia transcripts were orthologous to PHI genes involved in metabolic processes, followed by those responsible for transport of proteins/lipids (Figure 6A, File S15). GO analysis further reflected that majority of transcripts orthologous to PHI genes associated with loss of pathogenicity (Figure 7A) and reduced virulence (Figure 7B), might also be involved in metabolic process. Surprisingly, holomogs for known effectors in PHI database largely correspond to metabolic process (File S15); suggesting Venturia effectors might be targeting host metabolic processes for successful colonization. The protein kinase activity, followed by ATP binding and peptidase activity were predominant GO molecular function (Figure 6B, File S15). Similar trend was also observed when transcripts with loss of pathogenicity (Figure 7C) and reduced virulence (Figure 7D) were analyzed for GO molecular function. This highlights the importance of energy dependent signaling cascades during establishment of apple scab disease.
GO term assignments to V. inaequalis unigenes, with homology to PHI genes. A: represents biological process and B: represents the molecular function. PHI analysis revealed that the metabolic processes are overrepresented in PHI genes ortholog of V. inaequalis, thus highlighting the potential role of these processes in pathogenicity.
GO term assignments to V. inaequalis unigenes with homology to PHI genes associated with loss of pathogenicity (A, C) and reduced virulence (B, D). A, B and C, D represents the biological process and molecular function, respectively. This analysis revealed that the majority of PHI genes associated with loss of pathogenicity and reduced virulence are involved in metabolic processes, thus highlighting the potential role of these processes in pathogenicity.
Apart from transcripts having homologies within PHI database, the orthologs of other genes that are known pathogenicity determinants were also identified in this study. Notably, two contigs showing homology with hydrophobins were identified. Considering the fact that hydrophobins are required for conidial development, viability and pathogenic development of M. grisea [93], [94] as well as are involved in mediating fungal interaction with hydrophobic surfaces [95], it is encouraging to speculate that they might play role during apple scab pathogenesis.
Comparative analysis of protein families of V. inaequalis
The size of various protein families of V. inaequalis were compared with that of four plant pathogens; F. graminearum, M. oryzae, B. cinerea and S. sclerotiorum [52] (File S16, Figure 8). The size of several protein families, such as cutinases (n = 12), glycoside hydrolase (n = 345) and cytochrome P450 (n = 171) were quite comparable across all these pathogens. However, the transposases (n = 3) were underrepresented while zinc finger transcription factors (n = 434) and protein kinases (n = 481) were found to be overrepresented in V. inaequalis. In general, overall number of glycoside hydrolases (GH) possessed by V. inaequalis (n = 238) is similar to that encoded by M. oryzae (n = 236), B. fuckeliana (n = 237), A. nidulans (n = 246), A. oryzae (n = 296), and N. crassa (n = 176) (File S17). Also the percentage representation of individual members of different GH family was quite comparable across different pathogens analyzed. However, overrepresentation of a few GH families was observed in Venturia. Notably, amongst them are GH1, GH17, GH31, GH38, GH47 and GH63 (File S17). Interestingly, a few GH families (GH13 and GH18) were underrepresented in Venturia as compared to other fungal pathogens. Also, few GH family members, generally present in other fungal pathogens, were found to be absent in V. inaequalis transcripts. We identified a few members of GH77 in Venturia which are otherwise found to be absent in other fungal pathogens analyzed in this study. The differences between protein families might point towards the underlying variability in pathogenicity mechanisms of these phytopathogens. The availability of Venturia genome sequence and multi-stage transcriptome data may facilitate further validation of this comparative study.
The histogram reflects the number of selected protein families across different phytopathogens namely F. graminearum, M. oryzae, B. cinerea and S. sclerotiorum.
Search against MEROPS peptidase database led to identification of eight major categories of peptidases in Venturia transcriptome. Figure 9 provides the list of peptidase family present in Venturia and along with their comparative account across different fungal pathogens. Serine peptidase, metallopeptidases and cysteine peptidases constitute the majority of peptidase family present in different fungal pathogens analyzed, including V. inaequalis. Interestingly though, the asparagine peptide lyase, inhibitors family and peptidases of unknown catalytic type were exclusively present in V. inaequalis (File S18). We further attempted to identify the transcripts encoding various transporters in Venturia. Table 6 represents the list of seven superfamilies of transporters present and also provides comparative enlisting of different subclasses present in different fungal pathogens. It is noteworthy that all these seven transporters superfamily involved in various physiological activities are overrepresented in Venturia, as compared to other fungal pathogens. The finding that transporters are overrepresented in Venturia, suggests that some of the transporters might be facilitating Venturia to obtain nutrients from the host during in-planta growth and might also assist it to tackle the altered pH homeostasis, generally observed during execution of plant defense. Also, the transporters might be assisting the apple scab pathogen to generate resistance against commonly used fungicides, as seen in case of P. tritici-repentis a wheat pathogen, wherein the efflux transporters are known to impart fungicide resistance [96].
V. inaequalis transcripts encoding peptidase families were predicted using MEROPS database and the respective percent sizes were compared with that of F. graminearum, M. oryzae, B. cinerea and S. sclerotiorum.
Phylogenomic analysis of V. inaequalis
The reference amino acid sequences of nine organisms (File S19), Aspergillus fumigatus, Aspergillus nidulans, Candida albicans, Magnaporthe oryzae, Neurospora crassa, Sclerotinia sclerotiorum, Pyrenophora tritici-repentis, Gibberella zeae, Botryotinia fuckeliana and assembled sequences of V. inaequalis obtained in this study, were analyzed using Hal, an automated pipeline for phylogenetic analysis of genomic protein sequence data [58]. The Hal pipeline generated three separate sets of alignments, one removing all gap-containing columns (remgaps) and two removing problematic regions of alignments based on the default conservative (Gblocks-con) and liberal (Gblocks-lib) options of the program Gblocks [61]. Three phylogenetic trees representing three super-alignments (remgaps, Gblocks-con and Gblocks-lib) were generated and they were found to be well resolved and robust with high bootstrap values for different clades (Figure 10). As expected, different clades were observed in the phylogenetic trees and relationship between most of the pathogens analyzed were similar to that observed by Gao et al [97]. Interestingly, in every phylogentic tree, each generated by different methods, V. inaequalis was placed closer to P. tritici-repentis (Figure 10). As V. inaequalis and P. tritici-repentis are placed in Pleosporaceae family, our study strengthens the classification of V. inaequalis into this family and establishes its closeness to that of other members of Pleosporaceae family at genome wide scale.
Maximum Likelihood based phylogenetic analysis using RAxML tool, on three different super-alignments of V. inaequalis and that of other nine selected fungal pathogens, were used to establish phylogenomic relationship. A, B, C represents phylogenetic tree from super-alignment constructed by Gblocks-con, Gblocks-lib and remgaps, respectively using Hal pipeline.
Conclusion
In this study, we report de novo assembly of the transcriptome of V. inaequalis, the apple scab pathogen, and its in-depth computational analysis to explore genome wide insights about this deadly pathogen. In total, we obtained 94,350,055 SE reads specific to Venturia, assembled them into 62,061 contigs. Out of which 29,750 transcripts demonstrated significant similarity to sequences in other species by BLAST analysis, representing a total of 24,571 unique genes (Set A). No significant homolog was found for 32,311 assembled transcript sequences (Set B). Functional annotations using GO, EC, KEGG, InterproScan, Pfam analysis and searches against various databases identified genes that might be participating in several important biological and metabolic pathways. A total of 463 Set A transcripts and 483 Set B transcripts were found encoding putative secreted proteins. Metabolism was the predominant biological process ontology category while peptidase, catalytic and hydrolase activities were amongst predominant molecular function categories. Also the present work identified several cell wall degrading enzymes and Pfam domains participating in cell wall degradation from Venturia transcriptome. Furthermore, a few host translocated putative RxLR effectors, orthologs of several known candidate effectors and those showing homologies to known pathogenicity determinants of various pathogenic fungi as per PHI database entries were identified in this study. A large number of transcripts encoding membrane transporters were identified and the comparative analysis unraveled that the number of transporters encoded by Venturia are significantly more as compared to that encoded by several other important plant fungal pathogens. Overall, the transcriptome analysis of V. inaequalis provided wealth of information which would facilitate further research to understand the biology and pathogenicity mechanism of this pathogen, which in turn would make possible to evolve novel strategies for engineering disease resistance in apple.
Supporting Information
File S1.
155 experimentally validated nucleotide sequences for V. inaequalis available at GenBank.
https://doi.org/10.1371/journal.pone.0053937.s001
(TXT)
File S2.
Sequences of known effectors of various pathogenic fungi.
https://doi.org/10.1371/journal.pone.0053937.s002
(TXT)
File S3.
Sequences of known V. inaequalis effectors.
https://doi.org/10.1371/journal.pone.0053937.s003
(TXT)
File S4.
Effect of k-mer size on assembling performance of transcriptome using Velvet, SOAPdenovo and ABySS.
https://doi.org/10.1371/journal.pone.0053937.s004
(XLS)
File S5.
Dissimilar sequence groupings for assembled V. inaequalis transcriptome sequences.
https://doi.org/10.1371/journal.pone.0053937.s005
(XLS)
File S8.
Assembly validation with known V. inaequalis nucleotide sequences.
https://doi.org/10.1371/journal.pone.0053937.s008
(XLS)
File S9.
Assembly validation through Sanger Sequencing.
https://doi.org/10.1371/journal.pone.0053937.s009
(DOC)
File S10.
Blast analysis result against Fungal Cytochrome P450 database.
https://doi.org/10.1371/journal.pone.0053937.s010
(XLS)
File S11.
Pfam domains predicted to be present in the secretome of V. inaequalis.
https://doi.org/10.1371/journal.pone.0053937.s011
(XLS)
File S12.
Cell wall degrading enzymes predicted in transcriptome of V. inaequalis.
https://doi.org/10.1371/journal.pone.0053937.s012
(XLS)
File S14.
Summary of PHI database gene orthologs in V. inaequalis.
https://doi.org/10.1371/journal.pone.0053937.s014
(DOC)
File S15.
GO annotation of PHI gene orthologs of V. inaequalis.
https://doi.org/10.1371/journal.pone.0053937.s015
(XLS)
File S16.
Sizes of selected protein families in V. inaequalis and other plant pathogens.
https://doi.org/10.1371/journal.pone.0053937.s016
(XLS)
File S17.
Percentage of carbohydrate-degrading enzymes in ascomycetes fungal pathogens, arranged by GH family.
https://doi.org/10.1371/journal.pone.0053937.s017
(XLS)
File S18.
Protease genes in different fungal genomes, arranged by MEROPS family.
https://doi.org/10.1371/journal.pone.0053937.s018
(XLS)
File S19.
Number of reference protein sequences of selected phytopathogenic fungi used for Phylogenomics.
https://doi.org/10.1371/journal.pone.0053937.s019
(DOC)
Acknowledgments
KT and SB acknowledge the senior research fellowships from CSIR and UGC respectively, while VC is thankful to DST for INSPIRE-JRF fellowship. KT is registered for PhD at Department of Biotechnology, Panjab University, Chandigarh, India. We are also thankful to Ms. Surbhi Sood for help during patching and maintaining the Indian isolate of V. inaequalis. This manuscript has IHBT publication no: 2402.
Author Contributions
Provided intellectual input in compilation of this manuscript: JK. Supervised the computational part of this study: RS. Carried out wet lab experiments associated with work: KT. Prepared cDNA libraries for Illumina sequencing: KT. Significantly contributed in acquisition and interpretation of data: KT. Performed reads generation, quality filtering, assembling, annotations, phylogenomics and associated computational study in this work: VC. Performed bioassays: SB. Assisted in cDNA library preparation for Illumina sequencing: SB. Performed Illumina sequencing run: MKS. Assisted in Illumina sample preparation work flow: MKS. Conceived and designed the experiments: GJ. Analyzed the data: KT VC GJ RS. Contributed reagents/materials/analysis tools: GJ. Wrote the paper: KT VC GJ.
References
- 1. Jha G, Thakur K, Thakur P (2009) The Venturia apple pathosystem: pathogenicity mechanisms and plant defense responses. J Biomed Biotech .
- 2. Bowen JK, Mesarich CH, Bus VGM, Beresford RM, Plummer KM, et al. (2011) Venturia inaequalis: the casual agent of apple scab. Mol Plant Pathol 12: 105–122.
- 3.
MacHardy WE (1996) Apple Scab, Biology, Epidemiology, and Management. St. Paul, MinnUSA: APS Press.
- 4. Gladieux P, Zhang X-G, Afoufa-Bastien D, Sanhueza R-MV, Sbaghi M, et al. (2008) On the origin and spread of the scab disease of apple: out of central Asia. PLoS ONE 3: e1455 Gladieux. doi:10.1371/journal.pone.0001455.
- 5. Gupta GK (1985) Recent trends in forecasting and control of apple scab (Venturia inaequalis). Pesticides 19: 19–31.
- 6. Gessler C, Patocchi A, Sansavini S, Tartarini S, Gianfranceschi L (2006) Venturia inaequalis resistance in apple. Crit Rev Plant Sci 25: 473–503.
- 7. Malnoy M, Jin Q, Borejsza-Wysocka EE, He SY, Aldwinckle HS (2007) Overexpression of the apple MpNPR1 gene confers increased disease resistance in Malus x domestica. Mol Plant Microbe Interact 20: 1568–1580.
- 8. Joshi SG, Schaart JG, Groenwold R, Jacobsen E, Schouten HJ, et al. (2011) Functional analysis and expression profiling of HcrVf1 and HcrVf2 for development of scab resistant cisgenic and intragenic apples. Plant Mol Biol 75: 579–591.
- 9. Palani PV, Lalithakumari D (1999) Resistance of Venturia inaequalis to the sterol biosynthesis inhibiting fungicide, penconazole [1-(2-(2,4-dichlorophenyl) pentyl)-1H-1,2,4-triazole]. Mycol Res 103: 1157–1164.
- 10. Köller W, Parker DM, Turechek WW, Avila-Adame C, Cronshaw K (2004) A two-phase resistance response of Venturia inaequalis populations to the QoI fungicides Kresoxim-Methyl and Trifloxystrobin. Plant Dis 88: 537–544.
- 11. Marine SC, Schmale DG III, Yoder KS (2007) Resistance to myclobutanil in populations of Venturia inaequalis in Winchester, Virginia. Plant Health Prog .
- 12. Parisi L, Lespinasse Y, Guillaumes J, Krüger J (1993) A new race of Venturia inaequalis virulent to apples with resistance due to the Vf gene. Phytopathol 83: 533–537.
- 13. Guérin F, Le Cam B (2004) Breakdown of the scab resistance gene Vf in apple leads to a founder effect in populations of the fungal pathogen Venturia inaequalis. Phytopathol 94: 364–369.
- 14. Bénaouf G, Parisi L (2000) Genetics of host-pathogen relationships between Venturia inaequalis races 6 and 7 and Malus species. Phytopathol 90: 236–242.
- 15. Broggini GAL, Le Cam B, Parisi L, Wu C, Zhang H-B, et al. (2007) Construction of a contig of BAC clones spanning the region of the apple scab avirulence gene AvrVg. Fungal Genet Biol 44: 44–51.
- 16. Xu X, Roberts T, Barbara D, Harvey NG, Gao L, et al. (2009) A genetic linkage map of Venturia inaequalis, the causal agent of apple scab. BMC Res Notes 2: 163 .
- 17. Bowen JK, Mesarich CH, Rees-George J, Cui W, Fitzgerald A, et al. (2009) Candidate effector gene identification in the ascomycete fungal phytopathogen Venturia inaequalis by expressed sequence tag analysis. Mol Plant Pathol 10: 431–448.
- 18. The Bovine Genome Sequencing and Analysis Consortium (2009) Elsik C, Tellam R, Worley K (2009) The genome sequence of taurine cattle: A window to ruminant biology and evolution. Science 324: 522–528 .
- 19. Ming R, Hou S, Feng Y, Yu Q, Dionne-Laporte A, et al. (2008) The draft genome of the transgenic tropical fruit tree papaya (Carica papaya Linnaeus). Nature 452: 991–996.
- 20. Huang S, Li R, Zhang Z, Li L, Gu X, et al. (2009) The genome of the cucumber, Cucumis sativus L. Nat Genet 41: 1275–1281.
- 21. Kim Y, Nandakumar MP, Marten MR (2007) Proteomics of filamentous fungi. Trends Biotechnol 25: 395–400 .
- 22. Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, et al. (2009) ABySS: a parallel assembler for short read sequence data. Genome Res 9: 1117–1123.
- 23. Birol I, Jackman SD, Nielsen CB, Qian JQ, Varhol R, et al. (2009) De novo transcriptome assembly with ABySS. Bioinformatics 25: 2872–2877.
- 24. Zerbino DR, Birney E (2008) Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome Res 18: 821–829.
- 25. Win J, Greenwood DR, Plummer KM (2003) Characterization of a protein from Venturia inaequalis that induces necrosis in Malus carrying the Vm resistance gene. Physiol Mol Plant Pathol 62: 193–202.
- 26. Ghawana S, Paul A, Kumar H, Kumar A, Singh H, et al. (2011) An RNA isolation system for plant tissues rich in secondary metabolites. BMC Res Notes 4: 85.
- 27. Gahlan P, Singh HR, Shankar R, Sharma N, Kumari A, et al. (2012) De novo sequencing and characterization of Picrorhiza kurrooa transcriptome at two temperatures showed major transcriptome adjustments. BMC Genomics 13: 126.
- 28. Velasco R, Zharkikh A, Affourtit J, Dhingra A, Cestaro A, et al. (2010) The genome of the domesticated apple (Malus×domestica Borkh). Nat Genet 42: 833–839.
- 29. Langmead B, Trapnell C, Pop M, Salzberg SL (2009) Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol 10: R25.
- 30. Li R, Zhu H, Ruan J, Qian W, Fang X, et al. (2010) De novo assembly of human genomes with massively parallel short read sequencing. Genome Res 20: 265–272.
- 31. Pertea G, Huang X, Liang F, Antonescu V, Sultana R, et al. (2003) TIGR Gene Indices clustering tools (TGICL): a software system for fast clustering of large EST datasets. Bioinformatics 19: 651–652.
- 32. Li W, Godzik A (2006) Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics 22: 1658–1659.
- 33.
Non redundant protein database Available: ftp://ftp.ncbi.nih.gov/blast/db/FASTA/nr.gz. Accessed 2012 Apr 4.
- 34. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215: 403–410.
- 35.
UniProt database Available: http://www.uniprot.org/downloads. Accessed 2012 May 20.
- 36. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, et al. (2000) Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25: 25–29.
- 37. Kanehisa M, Goto S, Sato Y, Furumichi M, Tanabe M (2012) KEGG for integration and interpretation of large-scale molecular datasets. Nucleic Acids Res 40: D109–D114.
- 38. Kanehisa M, Goto S (2000) KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res 28: 27–30.
- 39. Bairoch A (2000) The ENZYME database. Nucleic Acids Res 28: 304–305.
- 40.
Annot8r program Available: http://www.nematodes.org/bioinformatics/annot8r/index.shtml. Accessed 2012 Mar 30.
- 41. Punta M, Coggill PC, Eberhardt RY, Mistry J, Tate J, et al. (2012) The Pfam protein families database. Nucleic Acids Res 40: D290–D301.
- 42. Quevillon E, Silventoinen V, Pillai S, Harte N, Mulder N, et al. (2005) InterProScan: protein domains identifier. Nucleic Acids Res 33 (Web Server issue) W116–W120.
- 43.
HMMER3 Available: http://hmmer.org/. Accessed 2012 Jun 1.
- 44. Marchler-Bauer A, Lu S, Anderson JB, Chitsaz F, Derbyshire MK, et al. (2011) CDD: a Conserved Domain Database for the functional annotation of proteins. Nucleic Acids Res 39 (Database issue) D225–229.
- 45. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402.
- 46. Rawlings ND, Barrett AJ, Bateman A (2012) MEROPS: the database of proteolytic enzymes, their substrates and inhibitors. Nucleic Acids Res 40: D343–D350.
- 47. Park J, Lee S, Choi J, Ahn K, Park B, et al. (2008) Fungal cytochrome P450 database. BMC Genomics 9: 402.
- 48. Saier MH Jr, Yen MR, Noto K, Tamang DG, Elkan C (2009) The Transporter Classification Database: recent advances. Nucleic Acids Res 37: D274–278.
- 49.
Schomburg D, Schomburg I (2010) Enzyme databases. In: Carugo O, Eisenhaber F, editors. Data Mining Techniques for the Life Sciences. Totowa, NJ: Humana Press, Vol. 609. pp. 113–128.
- 50. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, et al. (2009) The Carbohydrate-Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res 37 (Database issue) D233–8.
- 51. Winnenburg R, Baldwin TK, Urban M, Rawlings C, Köhler J, et al. (2006) PHI-base: a new database for pathogen host interactions. Nucleic Acids Res 34: D459–D464.
- 52. Zheng P, Xia Y, Xiao G, Xiong C, Hu X, et al. (2011) Genome sequence of the insect pathogenic fungus Cordyceps militaris, a valued traditional Chinese medicine. Genome Biol 12: R116.
- 53. Petersen TN, Brunak S, von Heijne G, Nielsen H (2011) SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods 8: 785–786.
- 54. Krogh A, Larsson B, von Heijne G, Sonnhammer ELL (2001) Predicting transmembrane protein topology with a hidden markov model: application to complete genomes. J Mol Biol 305: 567–580.
- 55. Rice P, Longden I, Bleasby A (2000) EMBOSS: The European Molecular Biology Open Software Suite. Trends Genet 16: 276–277.
- 56. Win J, Morgan W, Bos J, Krasileva KV, Cano LM, et al. (2007) Adaptive evolution has targeted the C-terminal domain of the RXLR effectors of plant pathogenic oomycetes. Plant Cell 19: 2349–2369.
- 57.
NCBI Available: http://www.ncbi.nlm.nih.gov/. Accessed 2012 Apr 12.
- 58. Robbertse B, Yoder RJ, Boyd A, Reeves JB, Spatafora JW (2011) Hal: an automated pipeline for phylogenetic analyses of genomic data. PLoS Curr .
- 59.
Dongen SM van (2000) Graph clustering by flow simulation. Ph.D. Thesis, University of Utrecht, The Netherlands. Available: http://igitur-archive.library.uu.nl/dissertations/1895620/inhoud.htm. Accessed 2012 Aug 15.
- 60. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, et al. (2007) ClustalW and ClustalX version 2.0. Bioinformatics 23: 2947–2948.
- 61. Castresana J (2000) Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol Biol Evol 17: 540–552.
- 62. Abascal F, Zardoya R, Posada D (2005) ProtTest: selection of best-fit models of protein evolution. Bioinformatics 21: 2104–2105.
- 63.
Leeuw Jan De (1973) Information theory and extension of the maximum likelihood principle by Hirotogu Akaike. In: Proceedings of the 2nd International Symposium on Information Theory, Budapest, Hungary; 267–281.
- 64. Stamatakis A (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688–2690.
- 65. Huson DH, Richter DC, Rausch C, Dezulian T, Franz M, et al. (2007) Dendroscope: An interactive viewer for large phylogenetic trees. BMC Bioinformatics 22: 460.
- 66.
Rees DJG, Husselmann LHH, Celton J-M (2009) de novo Genome Sequencing of The Apple Scab (Venturia inaequalis) Genome, Using Illumina Sequencing Technology. Plant & Animal Genomes XVII Conference Abstract P013. Available: http://www.intl-pag.org/17/abstracts/P01_PAGXVII_013.html. Accessed: 2012 May 25.
- 67. Celton J-M, Christoffels A, Sargent DJ, Xu X, Rees DJ (2010) Genome-wide SNP identification by high-throughput sequencing and selective mapping allows sequence assembly positioning using a framework genetic linkage map. BMC Biol 8: 155.
- 68. Weber APM, Weber KL, Carr K, Wilkerson C, Ohlrogge JB (2007) Sampling the Arabidopsis transcriptome with massively parallel pyrosequencing. Plant Physiol 144: 32–42.
- 69. Mane SP, Evans C, Cooper KL, Crasta OR, Folkerts O, et al. (2009) Transcriptome sequencing of the Microarray Quality Control (MAQC) RNA reference samples using next generation sequencing. BMC Genomics 10: 264.
- 70. Crešnar B, Petrič S (2011) Cytochrome P450 enzymes in the fungal kingdom. Biochim Biophys Acta 1814: 29–35.
- 71. Mueller O, Kahmann R, Aquilar G, Trejo-Aquilar B, Wu A, et al. (2008) The secretome of the maize pathogen Ustilago maydis. Fungal Genet Biol 45: S63–S70.
- 72. Doehlemann G, van der Linde K, Assmann D, Schwammbach D, Hof A, et al. (2009) Pep1, a secreted effectorprotein of Ustilago maydis, is required for successful invasion of plant cells. PLoS Pathog 5: e1000290.
- 73. Soanes DM, Alam I, Cornell M, Wong HM, Hedeler C, et al. (2008) Comparative genome analysis of filamentous fungi reveals gene family expansions associated with fungal pathogenesis. PLoS ONE 3: e2300 .
- 74. Brown NA, Antoniw J, Hammond-Kosack KE (2012) The predicted secretome of the plant pathogenic fungus Fusarium graminearum: A refined comparative analysis. PLoS ONE 7: e33731 .
- 75. Raffaele S, Win J, Cano LM, Kamoun S (2010) Analyses of genome architecture and gene expression reveal novel candidate virulence factors in the secretome of Phytophthora infestans. BMC Genomics 11: 637.
- 76. van den Burg HA, Harrison SJ, Joosten MH, Vervoort J, de Wit PJGM (2006) Cladosporium fulvum Avr4 protects fungal cell walls against hydrolysis by plant chitinases accumulating during infection. Mol Plant Microbe Interact 19: 1420–1430.
- 77. Parsons JF, Calabrese K, Eisenstein E, Ladner JE (2003) Structure and mechanism of Pseudomonas aeruginosa PhzD, an isochorismatase from the phenazine biosynthetic pathway. Biochemistry 42: 5684–5693.
- 78. Gehring AM, Bradley KA, Walsh CT (1997) Enterobactin biosynthesis in Escherichia coli: isochorismate lyase (EntB) is a bifunctional enzyme that is phosphopantetheinylated by EntD and then acylated by EntE using ATP and 2,3-dihydroxybenzoate. Biochemistry 36: 8495–5803.
- 79. Wildermuth MC, Dewdney J, Wu G, Ausubel FM (2001) Isochorismate synthase is required to synthesize salicylic acid for plant defence. Nature 414: 562–565.
- 80. Leuthner B, Aichinger C, Oehmen E, Koopmann E, Müller O, et al. (2005) A H2O2-producing glyoxal oxidase is required for filamentous growth and pathogenicity in Ustilago maydis. Mol Genet Genomics 272: 639–650.
- 81. Day PR, Boone DM, Keitt GW (1956) Venturia inaequalis (Cke.)Wint. XI. The chromosome number. Am J Bot 43: 835–838.
- 82. King BC, Waxman KD, Nenni NV, Walker LP, Bergstrom GC, et al. (2011) Arsenal of plant cell wall degrading enzymes reflects host preference among plant pathogenic fungi. Biotechnol Biofuels 4: 4.
- 83. Wu S-C, Halley JE, Luttig C, Fernekes LM, Gutiérrez-Sanchez G, et al. (2006) Identification of an endo-β-1,4-D-Xylanase from Magnaporthe grisea by gene knockout analysis, purification and heterologous expression. Appl Environ Microbiol 72: 986–993 .
- 84. Bouwmeester K, Meijer HJG, Govers F (2011) At the frontier; RXLR effectors crossing the Phytophthora-host interface. Front Plant Sci .
- 85. Oliva R, Win J, Raffaele S, Boutemy L, Bozkurt TO, et al. (2010) Recent developments in effector biology of filamentous plant pathogens. Cell Microbiol 12: 705–715.
- 86. Stassen JHM, Van den Ackerveken G (2011) How do oomycete effectors interfere with plant life? Curr Opin Plant Biol 14: 407–414.
- 87. de Jonge R, Bolton MD, Thomma BPHJ (2011) How filamentous pathogens co-opt plants: the ins and outs of fungal effectors. Curr Opin Plant Biol 14: 400–406.
- 88. Schornack S, van Damme M, Bozkurt TO, Cano LM, Smoker M, et al. (2010) Ancient class of translocated oomycete effectors targets the host nucleus. PNAS 107: 17421–17426 .
- 89. Jones JDG, Dangl JL (2006) The plant immune system. Nature 444: 323–329.
- 90. Bus VGM, Rikkerink EHA, Caffier V, Durel C-E, Plummer KM (2011) Revision of the nomenclature of the differential host-pathogen interactions of Venturia inaequalis and Malus. Annu Rev Phtyopathol 49: 391–413.
- 91. de Jonge R, van Esse HP, Kombrink A, Shinya T, Desaki Y, et al. (2010) Conserved fungal LysM effector Ecp6 prevents chitin-triggered immunity in plants. Science 329: 953–955 .
- 92. Win J, Krasileva KV, Kamoun S, Shirasu K, Staskawicz BJ, et al. (2012) Sequence Divergent RXLR Effextors Share a Structural Fold Conserved across Plant Pathogenic Oomycete Species. PLoS Pathog 8: e1002400 .
- 93. Kim S, Ahn II-P, Rho H-S, Lee Y-H (2005) MHP1, a Magnaporthe grisea hydrophobin gene, is required for fungal development and plant colonization. Mol Microbiol 57: 1224–1237.
- 94. Talbot NJ, Kershaw MJ, Wakley GE, de Vries O, Wessels J, et al. (1996) MPG1 encodes a fungal hydrophobin involved in surface interactions during infection-related development of Magnaporthe grisea. Plant Cell 8: 985–999.
- 95. Elliot MA, Talbot NJ (2004) Building filaments in the air: aerial morphogenesis in bacteria and fungi. Curr Opin Microbiol 7: 594–601.
- 96. Reimann S, Deising HB (2005) Inhibition of Efflux Transporter-Mediated Fungicide Resistance in Pyrenophora tritici-repentis by a Derivative of a 4′-Hydroxyflavone and Enhancement of Fungicide Activity. Appl Environ Microbiol 71: 3269–3275 .
- 97. Gao Q, Jin K, Ying S-H, Zhang Y, Xiao G, et al. (2011) Genome sequencing and comparative transcriptomics of the model entomopathogenic fungi Metarhizium anisopliae and M. acridum. PLoS Genet e1001264 .