Verticillium dahliae infection of strawberry (Fragaria x ananassa) is a major cause of disease-induced wilting in soil-grown strawberries across the world. To understand what components of the pathogen are affecting disease expression, the presence of the known effector VdAve1 was screened in a sample of Verticillium dahliae isolates. Isolates from strawberry were found to contain VdAve1 and were divided into two major clades, based upon their vegetative compatibility groups (VCG); no UK strawberry isolates contained VdAve1. VC clade was strongly related to their virulence levels. VdAve1-containing isolates pathogenic on strawberry were found in both clades, in contrast to some recently published findings. On strawberry, VdAve1-containing isolates had significantly higher virulence during early infection, which diminished in significance as the infection progressed. Transformation of a virulent non-VdAve1 containing isolate, with VdAve1 was found neither to increase nor decrease virulence when inoculated on a susceptible strawberry cultivar. There are therefore virulence factors that are epistatic to VdAve1 and potentially multiple independent routes to high virulence on strawberry in V. dahliae lineages. Genome sequencing a subset of isolates across the two VCGs revealed that isolates were differentiated at the whole genome level and contained multiple changes in putative effector content, indicating that different clonal VCGs may have evolved different strategies for infecting strawberry, leading to different virulence levels in pathogenicity tests. It is therefore important to consider both clonal lineage and effector complement as the adaptive potential of each lineage will differ, even if they contain the same race determining effector.
Citation: Fan R, Cockerton HM, Armitage AD, Bates H, Cascant-Lopez E, Antanaviciute L, et al. (2018) Vegetative compatibility groups partition variation in the virulence of Verticillium dahliae on strawberry. PLoS ONE 13(2): e0191824. https://doi.org/10.1371/journal.pone.0191824
Editor: Minou Nowrousian, Ruhr-Universitat Bochum, GERMANY
Received: October 5, 2017; Accepted: January 11, 2018; Published: February 16, 2018
Copyright: © 2018 Fan 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.
Data Availability: All raw data files are available from the SRA (https://www.ncbi.nlm.nih.gov/sra) using the following accession numbers: PRJNA344737 and PRJNA352681.
Funding: RF gratefully acknowledges the financial assistance of the China Scholarship Council (grant number: 201506300012). RJH was funded by BBSRC grants BB/K017071/1, BB/N006682/1, AA by BB/K017071/1, HC by innovate UK grant 101630. LA was funded by AHDB grant CP094.
Competing interests: The authors have declared that no competing interests exist.
V. dahliae hosts and control methods
Verticillium dahliae Kleb. is a soilborne plant pathogenic fungus which can cause wilting on over 200 plant species, including many agricultural and horticultural crops . This pathogen has a rapidly expanding host range, most notably on the crops lettuce and pepper where it has become a major pathogen within these industries [2–4]. V. dahliae cannot act as an effective saprophyte and thus must produce resting structures to persist within the soil between infection events. The resting structures are microsclerotia, these are robust melanised hyphae which are very difficult to eradicate and can persist in the soil for up to 20 years post-formation . Soil fumigation is the only effective control measure however it relies on the use of chemicals with harmful environmental effects which have either been banned or (in the EU/UK) are due to be phased out . Furthermore, crop rotation is largely ineffective as a management strategy and there are no available fungicides to combat the disease once established [6,7]. Consequently, breeding natural host disease resistance into crops remains the most effective strategy to control Verticillium wilt disease.
The Ve1 resistance gene is operational against ‘race 1’ isolates
Pathogens have evolved to deliver effector proteins in order to suppress host PTI (pathogen associated molecular pattern [PAMP]-triggered immunity). If successful, the pathogen will cause effector-triggered susceptibility (ETS) in the host and as such allow the pathogen to cause disease . The plant cell wall is the primary barrier preventing pathogen infection. In order to overcome this barrier, pathogens secrete cell wall-degrading enzymes (CDEs) which are able to catabolise plant cell wall polysaccharides . Although CWDEs are pathogenicity genes [9–11], some also function as PAMPs, and trigger host PTI independent of their depolymerisation activity [9–11].
Plants have an innate immune system, part of which is comprised of so called ‘resistance’ genes that encode receptors to recognize molecular patterns of pathogen proteins  and effectors . Dominant plant resistance genes that bring about resistance to V. dahliae ‘race 1’ isolates have been identified in cotton, lettuce, tomato, and sunflower [14–17] have been deployed successful in commercial tomato (Ve1) , and lettuce (Vr1)  cultivars. Ve1 is a cell surface like receptor operative against ‘race 1’ of the pathogen (defined by resistance in tomato), which carries the avirulence gene VdAve1 [19–20]. However, Ve1 on tomato and its orthologs in other crops are not effective against ‘race 2’ isolates which lack VdAve1 . Although attempts have been made to find tomato cultivars resistant to race 2 isolates, no high level resistance has been discovered .
Race structure in current crops
It was thought that genome rearrangement was responsible for the widespread success of V. dahliae as a pathogen across multiple host species [23–24], however, a similar pattern of rearrangement has been observed across the 10 species within the Verticillium genus, including those which are plant pathogens with a limited host range . It is now proposed that gene loss is more important in defining differences in speciation and phenotypic diversity within the Verticillium spp. .
V. dahliae has linage specific genome regions enriched in effectors which have arisen by segmental duplication to allow for functional divergence [24–26]. By contrast, the effector VdAve1 was most likely acquired by horizontal gene transfer from the host plant on a single occasion and this effector has undergone multiple independent gene loss events within different ‘race 2’ isolates of V. dahliae ; . Indeed, VdAve1 is flanked by several ‘young’ transposable elements in a lineage specific region of V. dahliae, which, it is believed provide a mechanism of gene loss ; [27,28]. Furthermore, it was suggested that ‘race 2’ isolates should be divided into discrete groups based on the sequences in the genomic region around VdAve1 gene  with these regions representing each VdAve1 loss event. Greater complication of the race structure has arisen, whereby ‘race 2’ isolates infecting tomato in Japan can be split into two further races based on their differential pathogenicity on resistant rootstocks , however further work is required to determine the causative effector(s).
A study investigating race distribution in Californian tomato exclusively isolated ‘race 2’ strains from tomato, despite isolating ‘race 1’ strains from other crop hosts. It was proposed that this is due to the predominant use of Ve1 containing tomato cultivars . Likewise, V. dahliae strains from cotton, mint, olive, and potato were exclusively found to be ‘race 2’ isolates whereas all isolates from artichoke were ‘race 1’ and isolates from lettuce, pepper and Californian strawberry contained both races . It is clear that there is a global overlapping distribution on ‘race 1’ and ‘race 2’ isolates however, the host determines the proportion of the races isolated .
A complex race structure in strawberry has been defined through a series of pathology experiments by Govorova and Govorov (1997), who identified six newly described races of Russian V. dahliae isolates based on resistance associated with five cultivars . This work did not seek to quantify race structure in terms of VdAve1 presence nor the effector complexes however there is support for a complex series of plant -pathogen interactions. Strawberry wilt resistance was identified at the QTL level, in wilt segregating progenies planted within a mixed inoculum field, resistance was not monogenic, but quantitative, thus providing further evidence for complex plant- pathogen interaction .
V. dahliae is an asexual organism comprised of several discrete clonal lineages. Vegetative compatibility groups (VCGs) determine whether V. dahliae strains can form hyphal anastomosis bridges, leading to genetic exchange through nuclear fusion and subsequent haploidisation. VCGs limit the transmission of genetic material between unrelated strains and thus promote the maintenance of distinct clonal lineages . Previous studies have described two VCGs present on UK strawberry: VCG 2A and VCG 4B . However recent work has redefined the clonal lineages based on over 26,000 SNPs and as such divided the population structure of V. dahliae into two major clades (I & II) with clade II dissected into two further sub clades (II-1 & II-2) .
Materials and methods
PCR detection of VdAve1 gene in V. dahliae isolates
DNA of V. dahliae isolates was extracted using Macherey–Nagel Plant DNA Extraction Kit (Germany) following the manufacturer's instruction. Primers targeting a partial region of the VdAve1 open reading frame, VdAve1-F/R (S1 Table), were designed using Geneious 7.0 and the V. dahliae specific primers Vd-F929-947/Vd-R1076-1094 (S1 Table) were used as a control to confirm DNA quality (Wei et al. 2015). The DNA of a VdAve1-carrying isolate (V. dahliae 12067) was used as a positive control. PCR conditions were 95°C for 5 min, followed by 35 amplification cycles of 30 s at 95°C, 30 s at 60°C and 30 s at 72°C. Amplicons were separated on a 2% agarose gel.
PCR detection of vegetative compatibility group
The vegetative compatibility group of the isolates used within the pathogenicity assay was determined through PCR amplification of a lineage specific region using the primers DB19 and DB22 [35,37]. The PCR was conducted using Q5 High-Fidelity DNA polymerase with conditions of 98 oC 30 s; 35 cycles of 98 oC 10 s, 55 oC 30 s, 72 oC 30 s and 72 oC at 2 min. Amplicon production was confirmed through gel electrophoresis on a 1% agarose gel before PCR clean up using the Machery-Nagel Nucleospin® Gel and PCR clean up kit and sequenced in two directions using Sanger sequencing.
Phylogeny of VC groups
Multiple sequence alignment of the region amplified by the DB19/22 primer set was performed using MUSCLE . A neighbour-joining phylogeny was constructed using the HKY85 substitution model, with support from 1000 Bootstrap replicates in the Geneious v10 software package (www.geneious.com) . The resulting dendrogram was visualised using FigTree (http://tree.bio.ed.ac.uk/software/figtree/).
Pathogenicity assay on strawberry
The wilt susceptible strawberry cv. Hapil was used to determine the pathogenicity of isolates. Conidial inoculum of thirteen isolates were prepared through transfer of five, 1 cm plugs of mycelia growing on PDA into 50 ml of PDB media incubated in the dark at 25°C and 150 rpm for one week. After filtration through 13 μm particle retention filter paper, the conidial density was adjusted to 1x106 spores.ml-1 using sterile distilled water. Three-week old strawberry runners were up-rooted from pots, excess soil was removed with water, and trimmed 2 cm from the root tips. Six replicate plants were inoculated per treatment, a single experiment was carried out per phenotyping event. Roots were dipped in spore suspensions for 5 min and re-planted in the pot. Plants were kept in a growth room with 16:8 hr light—dark cycles and at a constant temperature of 22°C. Disease severity was surveyed following five categories: 0—healthy shoots, all leaves green; 1—single leaf yellow-brown; 2 - <25% leaves yellow-brown; 3–26–50% leaves yellow-brown; 4–51–75% leaves yellow-brown; 5–76–100% leaves yellow-brown, the plants dead .
Wilt severity scores (Yi) [= 0, 1, 2, 3, 4 and 5] were analysed using a cumulative logit model in which Yi is assumed to follow a multinomial distribution. A cumulative logit model is a regression model for a cumulative logit : where xk represents the kth explanatory variate, the βk effect of xk, and αj the intercept for each cumulative logit. The larger the value of ∑βkxik, the higher the probability of Yi falling in a category at the upper end of the category scale. It is the extension of the common logistic model with only two possible outcomes (i.e. diseased or healthy). β estimates the change in the cumulative odds ratio (on loge scale) for one unit increase in the explanatory variate x. In the present study, there were three categorical explanatory variates: VCG, host and race group.
Generation of VdAve1 knock-in mutants and induction of VdAve1 expression
The known avirulence gene gene VdAve1 and its native promoter (called PAve1) were cloned from the ‘race 1’ isolate 12067 using the primer pair VdAve1C-F/R and NPAve1-F/R, respectively (S1 Table). The VdAve1 random insertion vector was constructed using USER-Brick vector system . Core USER Bricks, as well as the constitutive fungal promoter PgpdA and the selective marker Hygromycin, were amplified from the plasmid pRF-HU2-F/R and pRF-HUE-F/R using the primers specified in ). Two different insertion vectors were assembled; one containing the Ave1 native promoter, and another using the constitutive promoter PgpdA. Correct vector assembly was then verified by PCR. The valid vectors were transformed into Agrobacterium tumefaciens isolate GV3101. After A. tumefaciens-mediated transformation (ATMT) of V. dahliae, the hygromycin B-resistant transformants were purified through the single-spore purification process as described previously . Amplicon insertion in the VdAve1 knock-in isolates was confirmed through PCR with the primer pairs Hyg-F/R and VdAve1-F/R (S1 Table) and qRT-PCR detection with the housekeeping genes of elongation factor 1-α (EF-F/R) and ß-tubulin (btubulin-F/R), as well as the specific primers for VdAve1 (qAve1-F/R, S1 Table). Moreover, in order to induce VdAve1 expression, cultures of five VdAve1 knock-in mutants of each vector were incubated on PDA and Capek-Dox agar media for 5 days at 25°C before qRT-PCR quantification.
Genome assembly and annotation
Genome sequencing was performed on five V. dahliae isolates. For isolates, 12251, 12153, 12158 and 12161 DNA extraction was performed on freeze-dried mycelium using a GenElute plant DNA miniprep kit (Sigma-Aldrich). Paired-end genomic libraries were prepared using a Nextera sample preparation kit (Illumina), according to the manufacturer’s protocol. Analysis using a fragment analyzer (Advanced Analytical Technologies) confirmed that libraries had a high representation of DNA fragments 600–1,000 bp in length. Paired end libraries were sequenced using 250 bp reads on an Illumina MiSeq machine. Genomic DNA for isolate 12008 was prepared by extracting high molecular weight DNA from freeze-dried V. dahliae mycelium using the Macherey-Nagel Nucleospin Plant II kit (Fisher 11912262) for Illumina sequencing and QIAGEN Genomic-tip 500/G kit (Qiagen, Germany) for Pacbio sequencing, which was sequenced on an RSII machine. An Illumina library was also prepared for isolate 12008, with DNA sheared using the Covaris M220 with microTUBE-50 (Covaris 520166) and size selected using a Blue Pippin (Sage Science). The Illumina library was constructed with a PCR-free method using NEBNext End Repair (E6050S), NEBNext dA-tailing (E6053S) and Blunt T/A ligase (M0367S) New England Biolabs modules. The library was sequenced using a 2x 300 bp PE (MS-102-3003) kit on an Illumina Miseq v3.
Illumina data adaptor sequences and low-quality data were removed using fastqc-mcf (available from https://expressionanalysis.github.io/ea-utils/) and the output was assembled with SPAdes . RepeatMasker and TransposonPSI were used to identify repetitive sequence regions  . Long reads generated from Pacbio sequencing of isolate 12008 were assembled using Canu and polished using Illumina MiSeq reads in Pilon to correct erroneous SNPs and InDels [47,48]. The assembly was edited in accordance with recommendations from the NCBI contamination screen (run as part of submission to Genbank in November 2016 for isolate 12008 and July 2017 for illumina-only isolates) with contigs split, trimmed or excluded as required. Quast  was used to summarise assembly statistics and BUSCO  was used to assess completeness of gene space within the assembly. Gene prediction was performed using Braker1 and CodingQuarry [51,52] and functional annotations predicted for these proteins using InterproScan . Sequence data and annotated genomes were deposited at DDBJ/EMBL/GenBank within Bioprojects PRJNA344737 (isolate 12008) and PRJNA352681 (isolates 12151, 12153, 12158 and 12161). Accession numbers for these Whole Genome Shotgun projects were MPSH00000000, PHNU00000000, PHNV00000000, PHNW00000000, PHNX00000000 for isolates 12008, 12151, 12153, 12158 and 12161, respectively.
For newly sequenced genomes and for the existing JR2, VdLs17  and Va.Ms.102 , putative virulence factors and effector candidates were identified within predicted gene models. Secreted proteins were identified using SignalP 4.1 , removing those that were predicted to contain a transmembrane domain (TMHMM). Small secreted cysteine-rich proteins (SSCP) were identified from secretomes, detecting proteins shorter than 300 amino acid and having a cysteine content of greater than 3%. Secreted cell wall degrading enzymes were identified within from those proteins predicted as carbohydrate active enzymes (CAZYmes) from the CAZY database . Further effector candidates were identified in the predicted secretome using EffectorP . LysM proteins were identified using CAZY hmm model CBM50.hmm. NPP1 proteins were identified from Interproscan function annotations (IPR008701).
Generation of a genome-wide phylogenetic tree
In order to study the genomic region surrounding the VdAve1 locus, Illumina MiSeq reads of the five UK ‘race 2’ isolates were mapped to the reference genome of JR2 using Bowtie2 before being visually inspected using IGV . Depth of aligned reads were determined over the JR2 Ave1 locus and flanking regions (Chromosome 5: 0.5–1.1 Mb) using Samtools. Mean values were determined over 50 bp intervals and plotted for each isolate using R. The phylogenetic tree of these five ‘race 2’ isolates along with the isolates of VdLs17 and JR2 was generated using RealPhy  using V. dahliae isolate JR2 as a reference genome.
Gene expansion was investigated through clustering proteins into ortholog groups and then identifying the number of genes from each organism within that orthogroup. The proteomes of the five sequenced isolates and reference isolates JR2, VdLs17 and VaM102 were clustered using OrthoMCL  using a cut-off length of 50 amino acids and an inflation value of 5. Ortholog groups containing putative effectors were filtered if they were expanded in VC subclade II-1 or II-2. Expansion status was designated if all isolates of one clade had greater numbers of genes in the orthogroup than all isolates from the other clade. The number of genes observed in reference genomes was not considered at this step due to differences in gene prediction methodologies.
UK strawberry isolates do not contain VdAve1
In order to investigate whether there is any evidence for the presence of VdAve1 in V. dahliae in the UK, we screened a total of 38 isolates isolated from UK strawberry, hop, acer, raspberry, potato, phlox, cotinus, tomato and chrysanthemum by PCR using a pair of primers targeting a 351 bp region within the VdAve1 gene (see Table 1 and S1 Fig). In addition primers of VdF929947/ VdR10761094 targeting part of V. dahliae intergenic spacer regions, were used as a positive control for DNA quality and PCR success in all DNA samples, generating a band of 160 bp. In addition, four V. dahliae isolates from Californian strawberry populations (disparate to those used in the referenced work), were also included to confirm the previously reported presence of VdAve1 . In the UK isolates, the VdAve1 gene was present in hop, acer and cotinus isolates indicating the presence of ‘race 1’ V. dahliae within the UK (S1A Fig). No DNA amplification at the VdAve1 locus was observed in the 20 V. dahliae isolates from UK strawberry, (S1B Fig). We found that the Californian strawberries isolates contained both races, confirming previous studies . These results show that ‘race 1’ strawberry isolates do exist but are absent among isolates from UK strawberry, despite the presence of VdAve1-containing isolates from other hosts within the UK. The observed frequency of VdAve1 containing UK isolates across other hosts is 9/16 (S1A Fig).
VdAve1 is not associated with a single VC group
Further characterisation of a subset of 30 isolates from the initial screen for VdAve1 revealed variation in vegetative compatibility groups of the isolates in this study. Contrary to recent, but not all reports, VdAve1 was found in two major VC clades- VC subclade II-1 and VC subclade II-2 (Fig 1) and not just a single VCG. The two VC clades identified in this study contain the previously reported VC groups on UK strawberry: VGC4B (nested within Subclade II-1) and VGC2B824 (Subclade II-2) [35,36]. The presence of VdAve1 in a VC clade other than VCG2A (Subclade II-1) is in contrast to the recent finding of Jimenez-Dıaz but consistent with earlier studies by de Jonge [26,27].
UK strawberry isolates display a range of virulence on a common susceptible cultivar
In order to understand variation in virulence of VdAve1-containing isolates on strawberry, isolate pathogenicity was determined using a susceptible cultivar of Fragaria x ananassa. Seven ‘race one’ VdAve1-containing isolates (two from UK hop, two from UK acer, three from Californian strawberry) and six ‘race two’ isolates (one from Californian strawberry and six from UK strawberry) were tested on the V. dahliae susceptible strawberry cv. Hapil. A wide range of symptoms were observed (Fig 2A) ranging from full plant collapse to mild stunting. No clear association was observed between the presence of VdAve1 and the virulence of the isolate by the end of the experiment, at six weeks post-inoculation (Fig 2B).
Disease scores six weeks post-inoculation (B) of the strawberry cultivar ‘Hapil’ inoculated with 13 isolates of Verticillium dahliae. ‘Race 1’ isolates are unshaded while ‘Race 2‘ isolates are shaded grey, subclade II-1 and II-2 isolates are denoted by thin and thick borders respectively. Error bars are standard errors.
VdAve1 isolates, when infected on strawberry have higher virulence early in the infection process and VC group is significantly associated with overall virulence level
A cumulative logit model was used to assess the severity of disease over time. From the results presented in Table 2, a clear pattern is apparent. VdAve1 containing isolates across the VC groups display higher virulence early in the infection process, with the most pronounced effects at three weeks post inoculation. This variation is less visible as infection progresses and is undetectable at six weeks post-inoculation. Variation in pathogenicity associated with VC subclade grows in importance over time, with isolates in subclade II-2 displaying far higher pathogenicity, than isolates in subclade II-1. Isolate 12008 was found to be the most pathogenic isolate, with the highest level of disease recorded six weeks post inoculation (Fig 2B). This isolate lacks VdAve1 but is within subclade II-2.
The presence of VdAve1 alone is neither correlated with gain or loss of virulence on strawberry in artificial transformation of a highly virulent isolate
In order to examine whether the virulence of 12008 could be increased by the action of VdAve1 we cloned the VdAve1 gene from a ‘race 1’ isolate 12067 from tomato and then transformed the VdAve1 expression vector into isolate 12008, a highly virulent isolate from UK strawberry using Agrobacterium tumefaciens mediated transformation. Vectors were constructed with either the inclusion of the native promoter region or use of a constitutive fungal promoter. In both cases five independent transformants (with the construct integrated at a random location in the genome) were selected for analysis. Strawberry plants inoculated with VdAve1 knock-in conidia cultured in PDA media showed comparable wilt disease severity to the wild type (WT) isolate 12008 irrespective of whether the native promoter or constitutive promoter was used. Furthermore, even though qRT-PCR results confirmed that VdAve1 knock-in mutant of PA4 with the native promoter had a similar VdAve1 expression level with that of the ‘race 1’ isolate 12067 in PDA, no significant difference was found between disease scores of the two mutants compared with the wild type isolate 12008 (Fig 3A; S3 and S4 Tables). qRT-PCR results with cultures of VdAve1 knock-in mutants incubated on PDA or Czapek-Dox plates showed that VdAve1 expression levels were induced (to differing degrees) on both PDA media and Czapek-Dox media (Fig 3B). In conclusion the presence of VdAve1 did not increase the virulence of V. dahliae on strawberry, when transformed into a virulent isolate.
(A) on the susceptible strawberry cv. Hapil. Mutants labelled with ‘PA’ carry the VdAve1 native promoter cloned from isolate 12067, whereas ‘PG’ mutants contained a constitutive promoter PgpdA from Aspergillus nidulans. The relative expression of VdAve1 to the housekeeping genes elongation factor 1-α and beta tubulin in PDA and Czapek Dox agar cultures (B).
Genome sequencing and assembly of selected isolates
To further characterise the molecular mechanisms of V. dahliae virulence, isolates from subclade II-1 and subclade II-2 were selected for genome sequencing. Long read PacBio and short read Illumina sequencing technologies were used to generate a reference genome for VC subclade II-1 isolate 12008. A further four isolates were sequenced using Illumina technology only, comprising two isolates from VC subclade II-2 (12158 and 12161) and two from VC subclade II-1 (12251, 12253).
High coverage was obtained for the isolate 12008 (174X), and 44-80x for the four Illumina-only isolates. A highly contiguous assembly was generated for isolate 12008 using 107-fold coverage of PacBio reads. De novo assembly using canu resulted in a 33.4 Mb assembly in 103 contigs (> 500 bp), with an N50 metric of 747 kb and a largest scaffold of 2.4 Mb (Table 3). RepeatModeler and RepeatMasker were used to identify repetitive and low-complexity regions within the 12008 assembly, masking 9.62% (3.37 Mb) of the assembly. BUSCO was used to assess gene space, identifying that 3689 of 3725 (99%) core Sordariomycete genes were present in the assembly. Gene prediction was carried out using Braker1 and CodingQuary, which predicted a total of 10486 genes in the unmasked assembly.
De novo assembly of the four Illumina-only genomes resulted in assemblies of a similar size (32.4–32.9 Mb) and gene-space (98.9–99.2% BUSCO genes) but more fragmented, with assemblies in 1154–1483 contigs (Table 3). Repetitive regions of the V. dahliae genome were not as complete in Illumina-only assemblies as the PacBio assembly, with 2.67–5.25% of these assemblies masked. A total of 9925–10330 genes were predicted in these assemblies (Table 3).
Putative secreted effector proteins were identified using EffectorP and through searches for SSCPs. Results of effector prediction were broadly similar between the five sequenced isolates, with 186–196 EffectorP candidates and 125–136 SSCPs (Table 3). This led to a total set of 226–243 putative effectors predicted from each of the sequenced isolates. The number of secreted CAZymes in the five genomes ranged from 298–306. Seven chitin-binding lysin motif (LysM) proteins were identified in each proteome, apart from LS17 where eight were identified. Three LysM proteins were predicted in each of the five sequenced genomes. However, despite a similar total number of LysM proteins, reference proteomes carried different numbers of secreted LysM effectors (Table 3). Putative NPP1 (necrosis-inducing phytophthora protein) effectors were also identified, with isolate 12161 carrying an additional NPP1 protein that was not present in the other four isolates (Table 3).
Vegetative compatibility groups have different effector complements and patterns of VdAve1 loss differ between VC groups
Even though the five isolates of 12008, 12251, 12253, 12158 and 12161 from UK strawberry were ‘race 2’ isolates, their virulence on strawberry was divergent as 12008 and 12251 can cause severe wilt disease while the isolate 12161 had very low virulence (Fig 3A). After sequencing, short reads of each genome were mapped to the reference genome of ‘race 1’ isolate JR2. The pattern of read coverage in the regions flanking VdAve1 gene was distinct in 12008, 12251 and 12253 from that observed in 12158 and 12161 (Fig 4). Moreover, some reads from isolates 12158 and 12161 showed alignments at around 672 kb of the JR2 contig while no read coverage was observed over this region in the other three isolates (Fig 4, panel A), indicating that these five ‘race 2’ isolates could be divided into two distinct groups.
Alignment of short reads to the VdAve1 region of the JR2 genome highlights different patterns of gene loss around the VdAve1 region which are dependent upon VC subclade; Subclade II-2 isolates (12008,12251,12253) show similar patterns of read alignments to each other and subclade II-1 isolates (12158,12161) again show similarity within VC subclade but not between groups.
Phylogenetic analysis of RealPhy using whole-genome sequencing reads of these 5 ‘race 2’ isolates and the reference genome JR2 showed that V. dahliae isolates from UK strawberry should be divided into two clusters, consistent with the coverage mapping result. In the phylogenetic tree, the isolates of 12158 and 12161 were clustered together in a separate clade from the other isolates (Fig 5). Interestingly, the isolates of 12008, 12251, 12253, JR2 and VdLs17 were clustered together (Fig 5), which is concordant with the VC group analysis (Fig 1).
Effector expansion and losses in VC groups
To further identify those genes associated with differences in VC group, orthology analysis was performed on the proteomes of the five UK isolates, V. dahliae reference isolates JR2 and VdLs17 and the V. alfalfae isolate VaMs102. In total, 80552 proteins clustered into 9697 orthogroups and a further 3109 proteins were considered orphan proteins, not clustering using OrthoMCL . Results showed that the majority of proteins were conserved between V. dahliae and V. alfalfae, with 7286 orthogroups common to all isolates representing 81% of the proteins (8581) from 12008. A further 498 orthogroups were conserved between all V. dahliae isolates, representing a further 5% of proteins from 12008. A total of 579 orthogroups contained proteins predicted as effectors. Expansion of effector genes was investigated between VC subclades II-1 and II-2; 37 orthogroups were identified that showed patterns of gain or loss congruent with VC subclade, when compared between the five sequenced isolates (Table 4). Ten of these ortholog groups contained secreted carbohydrate active enzymes, indicating differences, with five expanded in each VC subclade, indicating differences in cell-wall degrading enzymes may be responsible for aggressive performance. Nevertheless, many of these orthogroups represented secreted effector genes of unannotated functions, indicating that VC subclades have unique or differentiated compliments of effector proteins to modulate host defence (Table 4, S2 Table).
Our results clearly demonstrate that defining the clonal structure of Verticillium is key to understanding variability in pathogenicity and patterns of effector gain and loss. We conclude that VC subclade II-2 is associated with high virulence on the host plant strawberry, while VC subclade II-1, on the whole, contains isolates of lower virulence. To our knowledge this is the first time that variation in virulence on strawberry has been linked to VC group. However, comparable studies on the hosts artichoke and watermelon have also described an association of VC subclade II-2 with high virulence [61–63] indicating that this may be a common phenomena. High virulence was associated with VCG 1A on olive and cotton hosts, which may be explained by the presence of the high defoliating pathotype (D) within this lineage ; . In contrast, no significant association was found between any VCG and virulence of V. dahliae on inoculated woody ornamentals, including Acer species . We found no significant association between host of origin and variation in virulence on strawberry in our set of isolates (Table 2), which may suggest that the variation in virulence observed on some hosts (e.g. strawberry) that is associated with VC group may not be replicated in other hosts. This requires further study over a much larger number of isolates before a robust conclusion can be drawn.
Phylogenetic analysis of whole genome assemblies showed that the five sequenced isolates of Verticillium dahliae from UK strawberry clustered into two groups: the higher virulence isolates 12008, 12251 and 12253 and the lower virulence isolates, 12158 and 12161 in a separate clade adding support to the single locus separation seen when genotyping using the DB19/22 primers. This result suggests that the higher (VC subclade II-2) and lower virulence (VC subclade II-1) isolates are quite different across the whole genome. Even though these five isolates are defined as ‘race 2’ isolates, the group of isolates 12008, 12251 and 12253 were closely clustered with the ‘race 1’ isolate JR2, which indicated genetic diversity among ‘race 2’ isolates. It was recently reported that ten ‘race 2’ isolates can be divided into three subgroups depending upon the exact location of the read coverage drop when mapping the genomic reads to the ‘race 1’ isolate JR2 assembly . Following this method, we found that the read depth of 12158 and 12161, surrounding the VdAve1 gene dropped at the same positions when aligned to JR2, but this differed from the pattern of read dropout 12008, 12251 and 12253. It is therefore parsimonious to conclude that VdAve1 was conserved before the split of Vd into VC subclade II-1 and VGC 2B and that there have been multiple independent loss events of VdAve1 in these non-recombining, clonal lineages.
The plant cell wall provides the first barrier to prevent pathogen infection, and pathogens secrete numerous cell wall-degrading enzymes (CWDEs) to depolymerize the polysaccharide components of the plant cell wall . Enzymes that are capable of hydrolytically cleaving glycosidic bonds in oligo- or polysaccharides (including cellulose and hemicellulose) are generally summarized under the term glycoside hydrolases [67,68]. Glycoside hydrolases are important plant cell wall degradation enzymes. In the current study, analysis of secreted and effector gene presence in orthogroups showed that genes containing domains of glycoside hydrolase were expanded in the high virulence isolate group of 12008 and 12251. Specifically, studies in the fungal genus Fusarium showed that plant galactose-containing hemicellulosic compounds may represent natural substrates of galactose oxidase . The enriched domains of cell wall-degrading enzymes in isolates 12008, 12251 and 12253 may be responsible for greater degradation of plant cell wall composition and subsequent high virulence in V. dahliae isolates when infecting strawberry. Not all identified secreted CAZymes were identified as targeting the plant cell wall, with orthogroup18 containing genes with chitinase activity. As chitin is a well characterised MAMP, expansion of these genes may lead to decreased recognition in these more virulent isolates. This requires further testing to determine the causative elements responsible for the differences in virulence seen between the two VCG subclades. It also remains to be seen whether there is any difference in the resistance of cultivars to the two different clonal lineages of V. dahliae. Previous QTL mapping experiments in strawberry have been conducted under field conditions in highly infested plots which were artificially inoculated with many different culture of V. dahliae . Further work using inoculations with single isolates is needed to understand whether resistance responses vary between isolates that differ in VCG.
A recent study showed that ‘race 2’ isolates are obtained with high frequency in many crops with the exception of several crops including strawberry in California, where ‘race 1’ isolates are found at a higher frequency . This suggests that an ortholog of Ve1 is either absent or non-functional in the octoploid strawberry. This study supports the existence of ‘race 1’ strawberry USA isolates, however, when screening 20 V. dahliae isolates from UK strawberry, all of them were found to be non-VdAve1-containing ‘race 2’ isolates. Inoculation of common UK-grown wilt-susceptible cultivars of strawberry plants with ‘race 1’ isolates from other origins can cause wilt disease, indicating the lack of ‘race 1’ isolates on strawberry in the UK is not due to differences in host response between Californian and UK strawberry material. This indicates that Ve1-mediated resistance is not operating in the octoploid strawberry [20,70,71].
It is possible that geographical isolation plays a role in the frequency of VdAve1 in the population. For example, all the artichoke isolates tested from California were ‘race 1’, though race 2 artichoke isolates from Tunisia have been reported, indicating that although the global migration of genotypes in V. dahliae has likely occurred and continues to play an important role in the diversification of fungal populations, within-field variation may be low and may be dominated by widespread clonal types. Thus, race structure in specific crops and commodities may exist, depending on the geography. Geography alone though does not fully explain the abundance of non-VdAve1 containing isolates sampled from UK strawberry fields, as it is clear that VdAve1-containing isolates are present in the UK, as was revealed through extended sampling (S1 Fig). It is of course possible that VdAve1- containing isolates have remained associated with their respective hosts and have not spread to infect strawberry growing regions of the UK. More detailed and systematic sampling would allow the spatial and host-associated distributions of VdAve1- containing isolates to be determined.
The results of pathogenicity assays using ‘race 1’ isolates from Californian strawberries revealed that VdAve1 does enhance early development of pathogenic symptoms, the significance of which is attenuated as the disease progresses (Table 2). Although VdAve1 was found to increase early symptom development in strawberry, whether present in VC subclade II-1 or VC subclade II-2, transformation of VdAve1 either under its native promoter or under the A. nidulans promoter PgpdA failed to increase the virulence of the isolate 12008. Isolate 12008 is the most virulent isolate, based upon the pathogenicity test that we carried out (Fig 2) and therefore we conclude that other factors present in the 12008 genome provide an alternative route to high virulence, such that the effect of VdAve1 is masked.
Our work highlights that it is important that the nomenclature of V. dahliae is altered to move the classification of ‘race’ based on resistance responses in a single plant species, to one that is guided both by clonal lineage and effector complement as the adaptive potential of each lineage will differ, even if they contain the same effector.
S1 Fig. PCR detection of VdAve1 gene in Verticillium dahliae isolates.
Isolates from UK hop, acer, raspberry, potato, phlox, cotinus, and chrysanthemum (A) and strawberries (B) from UK and California. Primers target for V. dahliae intergenic spacer regions were used as control for DNA quality. The isolate of 12067 is a ‘race 1’ isolate as positive control and NTC was negative control of sterilized water.
S2 Table. Orthology tables from newly sequenced V. dahliae strains and existing genomes.
S3 Table. Dunnett post hoc test on ANOVA to determine differences between disease scores of Hapil plants 6 weeks post inoculation with isolate 12008 and each transformant or ‘race 1’ isolate 12067.
SE is standard error.
RF gratefully acknowledges the financial assistance of the China Scholarship Council (grant number: 201506300012). RJH was funded by BBSRC grants BB/K017071/1, BB/N006682/1, AA by BB/K017071/1, HC by innovate UK grant 101630. LA was funded by AHDB grant CP094. We would like to thank Vance Whitaker and Natalie Perez for Verticillium isolates. We also thank other members of the NIAB EMR (Andy Passey, Michelle Hulin, Beth Greenfield and Charlotte Nellist) for assistance throughout this project. We thank Bart Thomma and Luigi Faino for information about high-molecular weight DNA extraction.
RF and HC designed all experiments in collaboration with RJH. RF carried out all pathogenicity tests and transformations. HC carried out the VC group analysis. ECL assisted with genome assembly and annotation. AA provided support and supervision to RF for all informatics analysis, LA and HB generated Illumina genome sequences, XX provided statistical advice and assisted RJH with the development of the likelihood ratio tests. XH supervised RF and provided useful insights throughout the project. RF, HC, XX, AA and RJH wrote the manuscript. All authors approved the manuscript. The authors have no conflicts of interest to declare.
- 1. Pegg GF, Brady BL (2002) Verticillium wilts. CABI, 2002 ISBN 1845933222, 9781845933227.
- 2. Atallah ZK, Hayes RJ, Subbarao KV (2011) Fifteen years of Verticillium wilt of lettuce in America’s salad bowl: A tale of immigration, subjugation, and abatement. Plant Dis 95: 784–792.
- 3. Bhat RG, Subbarao KV (1999) Host Range Specificity in Verticillium dahliae. Phytopathology 89: 1218–1225. pmid:18944648
- 4. Bhat RG, Smith RF, Koike ST, Wu BM, Subbarao KV (2003) Characterization of Verticillium dahliae Isolates and Wilt Epidemics of Pepper. Plant Dis 87: 789–797.
- 5. European Food Safety Authority (2011) Conclusion on the peer review of the pesticide risk assessment of the active substance chloropicrin. EFSA Journal 9: 2084.
- 6. Wilhelm S (1955) Longevity of the Verticillium wilt fungus in the laboratory and field. Phytopathology 45: 180–181.
- 7. Fradin EF, Thomma BPHJ (2006) Physiology and molecular aspects of Verticillium wilt diseases caused by V. dahliae and V. albo-atrum. Mol Plant Pathol 7: 71–86. pmid:20507429
- 8. Jones JDG, Dangl JL (2006) The plant immune system. Nature 444: 323–329. pmid:17108957
- 9. Kubicek CP, Starr TL, Glass NL (2014) Plant cell wall-degrading enzymes and their secretion in plant-pathogenic fungi. Annu Rev Phytopathol 52: 427–451. pmid:25001456
- 10. Brito N, Espino JJ, González C (2006) The endo-beta-1,4-xylanase xyn11A is required for virulence in Botrytis cinerea. Mol Plant Microbe Interact 19: 25–32. pmid:16404950
- 11. Tzima AK, Paplomatas EJ, Rauyaree P, Ospina-Giraldo MD, Kang S (2011) VdSNF1, the sucrose nonfermenting protein kinase gene of Verticillium dahliae, is required for virulence and expression of genes involved in cell-wall degradation. Mol Plant Microbe Interact 24: 129–142. pmid:20839958
- 12. De Wit PJGM (1997) Pathogen avirulence and plant resistance: a key role for recognition. Trends Plant Sci 2: 452–458.
- 13. Boller T, He SY (2009) Innate immunity in plants: an arms race between pattern recognition receptors in plants and effectors in microbial pathogens. Science 324: 742–744. pmid:19423812
- 14. Fick GN, Zimmer DE (1974) Monogenic resistance to Verticillium wilt in sunflowers. Crop science 14: 895–896.
- 15. Hayes RJ, McHale LK, Vallad GE, Truco MJ, Michelmore RW, Klosterman SJ, et al. (2011) The inheritance of resistance to Verticillium wilt caused by race 1 isolates of Verticillium dahliae in the lettuce cultivar La Brillante. Theor Appl Genet 123: 509–517. pmid:21567237
- 16. Schaible L, Cannon OS, Waddoups V (1951) Inheritance of resistance to Verticillium wilt in a Tomato cross. Phytopathology 41: 986–990.
- 17. Zhang Y, Wang X, Yang S, Chi J, Zhang G, Ma Z (2011) Cloning and characterization of a Verticillium wilt resistance gene from Gossypium barbadense and functional analysis in Arabidopsis thaliana. Plant Cell Rep 30: 2085–2096. pmid:21739145
- 18. Hayes RJ, Vallad GE, Klosterman SJ, Subbarao KV (2011) Selection for Resistance to Verticillium Wilt Caused by Race 2 Isolates of Verticillium dahliae in Accessions of Lettuce (Lactuca sativa L.). HortScience 46: 201–206.
- 19. Kawchuk LM, Hachey J, Lynch DR, Kulcsar F, van Rooijen G, Waterer DR, et al. (2001) Tomato Ve disease resistance genes encode cell surface-like receptors. Proc Natl Acad Sci U S A 98: 6511–6515. pmid:11331751
- 20. de Jonge R, van Esse HP, Maruthachalam K, Bolton MD, Santhanam P, Keykha Saber M, et al. (2012) Tomato immune receptor Ve1 recognizes effector of multiple fungal pathogens uncovered by genome and RNA sequencing. Proc Natl Acad Sci U S A 109: 5110–5115. pmid:22416119
- 21. Alexander LJ (1962) Susceptibility of certain Verticillium-resistant tomato varieties to an Ohio isolate of the pathogen. Phytopathology 52: 998–1000.
- 22. Baergen KD, Hewitt JD, Clair DS (1993) Resistance of Tomato Genotypes to Four Isolates of Verticillium dahliae Race 2. HortScience 28: 833–836.
- 23. Seidl MF, Thomma BPHJ (2014) Sex or no sex: evolutionary adaptation occurs regardless. Bioessays 36: 335–345. pmid:24531982
- 24. Faino L, Seidl MF, Shi-Kunne X, Pauper M, van den Berg GCM, Wittenberg AH et al. (2016) Transposons passively and actively contribute to evolution of the two-speed genome of a fungal pathogen. Genome Res 26: 1091–1100. pmid:27325116
- 25. Shi X, Faino L, van den Berg G, Thomma B, Seidl M (2017) Evolution within the fungal genus Verticillium is characterized by chromosomal rearrangement and gene loss. BioRxiv.
- 26. de Jonge R, Bolton MD, Kombrink A, van den Berg GCM, Yadeta KA, Thomma BP (2013) Extensive chromosomal reshuffling drives evolution of virulence in an asexual pathogen. Genome Res 23: 1271–1282. pmid:23685541
- 27. Jiménez-Díaz RM, Olivares-García C, Trapero-Casas JL, Jiménez-Gasco MM, Navas-Cortés JA, Landa BB, et al. (2017) Variation of pathotypes and races and their correlations with clonal lineages in Verticillium dahliae. Plant Pathol 66: 651–666.
- 28. Klosterman SJ, Subbarao KV, Kang S, Veronese P, Gold SE, Thomma BP, et al. (2011) Comparative genomics yields insights into niche adaptation of plant vascular wilt pathogens. PLoS Pathog 7: e1002137. pmid:21829347
- 29. Usami T, Momma N, Kikuchi S, Watanabe H, Hayashi A, Mizukawa M, et al. (2017) Race 2 of Verticillium dahliae infecting tomato in Japan can be split into two races with differential pathogenicity on resistant rootstocks. Plant Pathol 66: 230–238.
- 30. Maruthachalam K, Atallah ZK, Vallad GE, Klosterman SJ, Hayes RJ, Davis RM, et al. (2010) Molecular variation among isolates of Verticillium dahliae and polymerase chain reaction-based differentiation of races. Phytopathology 100: 1222–1230. pmid:20698756
- 31. Short DPG, Gurung S, Maruthachalam K, Atallah ZK, Subbarao KV (2014) Verticillium dahliae race 2-specific PCR reveals a high frequency of race 2 strains in commercial spinach seed lots and delineates race structure. Phytopathology 104: 779–785. pmid:24502204
- 32. Govorova GF, Govorov DN (1997) Methodological Bases of Strawberry Breeding for Resistance to Verticillium Wilt. In: Tjamos EC, Rowe RC, Heale JB, Favel DR, editors. Cape Sounion, Athens, Greece. Advances in Verticillium Research and Disease Management,. Greece: NA, Vol. VII. p. 58.
- 33. Antanaviciute L, Šurbanovski N, Harrison N, McLeary K, Simpson DW, Wilson F, et al. (2015) Mapping QTL associated with Verticillium dahliae resistance in the cultivated strawberry Fragaria × ananassa. Horticulture Research 2: 15009. pmid:26504565
- 34. Glass NL, Kuldau GA (1992) Mating type and vegetative incompatibility in filamentous ascomycetes. Annu Rev Phytopathol 30: 201–224. pmid:18643772
- 35. Collado-Romero M, Mercado-Blanco J, Olivares-García C, Jiménez-Díaz RM (2008) Phylogenetic analysis of Verticillium dahliae vegetative compatibility groups. Phytopathology 98: 1019–1028. pmid:18943740
- 36. Milgroom MG, Jiménez-Gasco M del M, Olivares García C, Drott MT, Jiménez-Díaz RM (2014) Recombination between clonal lineages of the asexual fungus Verticillium dahliae detected by genotyping by sequencing. PLoS ONE 9: e106740. pmid:25181515
- 37. Collado-Romero M, Berbegal M, Jiménez-Díaz RM, Armengol J, Mercado-Blanco J (2009) A PCR-based ‘molecular tool box’ for in planta differential detection of Verticillium dahliae vegetative compatibility groups infecting artichoke. Plant Pathol 58: 515–526.
- 38. Edgar RC (2004) MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5: 113. pmid:15318951
- 39. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, et al. (2012) Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28: 1647–1649. pmid:22543367
- 40. Sowik I, Bielenin A, Michalczuk L (2001) In vitro testing of strawberry resistance to Verticillium dahliae and Phytophthora cactorum. Sci Hortic 88: 31–40.
- 41. Schabenberger O, Pierce F (2001) Contemporary statistical models for the plant and soil sciences. CRC Press. https://doi.org/10.1201/9781420040197
- 42. Sørensen LQ, Lysøe E, Larsen JE, Khorsand-Jamal P, Nielsen KF, Normand Frandsen RJ (2014) Genetic transformation of Fusarium avenaceum by Agrobacterium tumefaciens mediated transformation and the development of a USER-Brick vector construction system. BMC Mol Biol 15: 15. pmid:25048842
- 43. Dobinson KF, Grant SJ, Kang S (2004) Cloning and targeted disruption, via Agrobacterium tumefaciens-mediated transformation, of a trypsin protease gene from the vascular wilt fungus Verticillium dahliae. Curr Genet 45: 104–110. pmid:14618375
- 44. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, et al. (2012) SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol 19: 455–477. pmid:22506599
- 45. Smit, AFA, Hubley, R & Green P (2015) RepeatMasker Open-4.0. Available: http://www.repeatmasker.org.
- 46. Haas BJ, Zeng Q, Pearson MD, Cuomo CA, Wortman JR (2011) Approaches to Fungal Genome Annotation. Mycology 2: 118–141. pmid:22059117
- 47. Walker BJ, Abeel T, Shea T, Priest M, Abouelliel A, Sakthikumar S, et al. (2014) Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS ONE 9: e112963. pmid:25409509
- 48. Koren S, Walenz BP, Berlin K, Miller JR, Phillippy AM (2016) Canu: scalable and accurate long-read assembly via adaptive k-mer weighting and repeat separation. bioRxiv.
- 49. Gurevich A, Saveliev V, Vyahhi N, Tesler G (2013) QUAST: quality assessment tool for genome assemblies. Bioinformatics 29: 1072–1075. pmid:23422339
- 50. Simão FA, Waterhouse RM, Ioannidis P, Kriventseva EV, Zdobnov EM (2015) BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics 31: 3210–3212. pmid:26059717
- 51. Hoff KJ, Lange S, Lomsadze A, Borodovsky M, Stanke M (2016) BRAKER1: Unsupervised RNA-Seq-Based Genome Annotation with GeneMark-ET and AUGUSTUS. Bioinformatics 32: 767–769. pmid:26559507
- 52. Testa AC, Hane JK, Ellwood SR, Oliver RP (2015) CodingQuarry: highly accurate hidden Markov model gene prediction in fungal genomes using RNA-seq transcripts. BMC Genomics 16: 170. pmid:25887563
- 53. Jones P, Binns D, Chang H-Y, Fraser M, Li W, McAnulla C, et al. (2014) InterProScan 5: genome-scale protein function classification. Bioinformatics 30: 1236–1240. pmid:24451626
- 54. Faino L, Seidl MF, Datema E, van den Berg GCM, Janssen A, Wittenberg AHJ et al. (2015) Single-Molecule Real-Time Sequencing Combined with Optical Mapping Yields Completely Finished Fungal Genome. MBio 6. pmid:26286689
- 55. Petersen TN, Brunak S, von Heijne G, Nielsen H (2011) SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Meth 8: 785–786.
- 56. Lombard V, Golaconda Ramulu H, Drula E, Coutinho PM, Henrissat B (2014) The carbohydrate-active enzymes database (CAZy) in 2013. Nucleic Acids Res 42: D490–5. pmid:24270786
- 57. Sperschneider J, Gardiner DM, Dodds PN, Tini F, Covarelli L, Singh KB, et al. (2016) EffectorP: predicting fungal effector proteins from secretomes using machine learning. New Phytol 210: 743–761. pmid:26680733
- 58. Robinson JT, Thorvaldsdóttir H, Winckler W, Guttman M, Lander ES, Getz G, et al. (2011) Integrative genomics viewer. Nat Biotechnol 29: 24–26. pmid:21221095
- 59. Bertels F, Silander OK, Pachkov M, Rainey PB, van Nimwegen E (2014) Automated reconstruction of whole-genome phylogenies from short-sequence reads. Mol Biol Evol 31: 1077–1088. pmid:24600054
- 60. Li L, Stoeckert CJJ, Roos DS (2003) OrthoMCL: Identification of Ortholog Groups for Eukaryotic Genomes. Genome Research 13: 2178–2189. pmid:12952885
- 61. Jiménez-Díaz RM, Mercado-Blanco J, Olivares-García C, Collado-Romero M, Bejarano-Alcázar J, Rodriguez-Jurado D, et al. (2006) Genetic and Virulence Diversity in Verticillium dahliae Populations Infecting Artichoke in Eastern-Central Spain. Phytopathology 96: 288–298. pmid:18944444
- 62. Dervis S, Bicici M (2005) Vegetative compatibility groups in Verticillium dahliae isolates from cotton in Turkey. Phytoparasitica 33: 157–168.
- 63. Dervis S, Yetisir H, Yıldırım H, Tok FM, Kurt S, Karaca F (2009) Genetic and pathogenic characterization of Verticillium dahliae isolates from eggplant in Turkey. Phytoparasitica 37: 467–476.
- 64. Göre ME, Erdoğan O, Caner ÖK, Aydın MH, Berk S (2014) VCG diversity and virulence of Verticillium dahliae from commercially available cotton seed lots in Turkey. Eur J Plant Pathol 140: 689–699.
- 65. Dervis S, Mercado-Blanco J, Erten L, Valverde-Corredor A, Pérez-Artés E (2010) Verticillium wilt of olive in Turkey: a survey on disease importance, pathogen diversity and susceptibility of relevant olive cultivars. Eur J Plant Pathol 127: 287–301.
- 66. Goud JC, Termorshuizen AJ (2002) Pathogenicity and Virulence of the Two Dutch VCGs of Verticillium dahliae to Woody Ornamentals. European Journal of Plant Pathology 108: 771–782.
- 67. Gui Y-J, Chen J-Y, Zhang D-D, Li N-Y, Li T-G, Zhang WQ, et al. (2017) Verticillium dahliae manipulates plant immunity by glycoside hydrolase 12 proteins in conjunction with carbohydrate-binding module 1. Environ Microbiol 19: 1914–1932. pmid:28205292
- 68. Zhao Z, Liu H, Wang C, Xu J-R (2013) Comparative analysis of fungal genomes reveals different plant cell wall degrading capacity in fungi. BMC Genomics 14: 274. pmid:23617724
- 69. Ögel ZB, Brayford D, McPherson MJ (1994) Cellulose-triggered sporulation in the galactose oxidase-producing fungus Cladobotryum (Dactylium) dendroides NRRL 2903 and its re-identification as a species of Fusarium. Mycol Res 98: 474–480.
- 70. Fradin EF, Abd-El-Haliem A, Masini L, van den Berg GCM, Joosten MHAJ, Thomma BP (2011) Interfamily transfer of tomato Ve1 mediates Verticillium resistance in Arabidopsis. Plant Physiol 156: 2255–2265. pmid:21617027
- 71. Song Y, Zhang Z, Seidl MF, Majer A, Jakse J, Javornik B, et al. (2017) Broad taxonomic characterization of Verticillium wilt resistance genes reveals an ancient origin of the tomato Ve1 immune receptor. Mol Plant Pathol 18: 195–209. pmid:26946045