Verticillium dahliae Vta3 promotes ELV1 virulence factor gene expression in xylem sap, but tames Mtf1-mediated late stages of fungus-plant interactions and microsclerotia formation

Verticillium transcription activator of adhesion 3 (Vta3) is required for plant root colonization and pathogenicity of the soil-borne vascular fungus Verticillium dahliae. RNA sequencing identified Vta3-dependent genetic networks required for growth in tomato xylem sap. Vta3 affects the expression of more than 1,000 transcripts, including candidates with predicted functions in virulence and morphogenesis such as Egh16-like virulence factor 1 (Elv1) and Master transcription factor 1 (Mtf1). The genes encoding Elv1 and Mtf1 were deleted and their functions in V. dahliae growth and virulence on tomato (Solanum lycopersicum) plants were investigated using genetics, plant infection experiments, gene expression studies and phytohormone analyses. Vta3 contributes to virulence by promoting ELV1 expression, which is dispensable for vegetative growth and conidiation. Vta3 decreases disease symptoms mediated by Mtf1 in advanced stages of tomato plant colonization, while Mtf1 induces the expression of fungal effector genes and tomato pathogenesis-related protein genes. The levels of pipecolic and salicylic acids functioning in tomato defense signaling against (hemi-) biotrophic pathogens depend on the presence of MTF1, which promotes the formation of resting structures at the end of the infection cycle. In summary, the presence of VTA3 alters gene expression of virulence factors and tames the Mtf1 genetic subnetwork for late stages of plant disease progression and subsequent survival of the fungus in the soil.

Introduction Pathogenic fungi pose a major threat to plants worldwide. Their control in the fields remains difficult, as resistant cultivars are unavailable for many crops. Verticillium dahliae causes vascular wilt disease in a variety of important crop plants. Control of this soil-borne ascomycete is challenging because pesticides lose their effectiveness once the fungus is inside the plant [1][2][3][4][5]. Melanized fungal dormant structures (microsclerotia) can remain viable in the soil during unfavorable environmental conditions or in the absence of hosts [1,6]. Microsclerotia germinate when sensing root exudates, and hyphae grow toward the plant. V. dahliae enters through root tips, lateral root hairs, or natural root wounds before colonizing the root cortex and central cylinder [1,2,7]. Within the plant, V. dahliae forms asexual conidiospores to systematically spread throughout the host's vasculature [7,8]. The above-ground plant tissue is supplied with important minerals and water via the xylem vessels [9][10][11]. However, germinating conidia can block this transport and colonize neighboring tissue, leading to plant defense responses and the occurrence of disease symptoms including wilting, stunting, chlorosis and early senescence [1,5].
The plant xylem sap is nutrient poor and unbalanced, containing plant defense proteins, hormones and low levels of amino acids and sugars [10][11][12]. V. dahliae requires the secretion of signaling molecules that impair the host's defenses to live under these special nutritional conditions, compete with host cells and establish systemic infections. Sensing and adaptive mechanisms allow the fungus to respond to environmental cues such as nutrient supply or host defense molecules by altering gene expression and transport [13]. Verticillium can tailor its secretion patterns to adapt to different environments. The fungus produces a specific secretome pattern in xylem sap with proteins for degradation of plant material and growth under that unique condition, which differs from other growth conditions [14].
Transcriptomic responses of V. dahliae toward susceptible and tolerant olive cultivars are significantly different. Fungal genes involved in niche-adaptation, virulence and microsclerotia development were induced preferentially in the susceptible cultivar [15]. During plant infection, different transcription factors sequentially regulate the expression of host interactionrelated genes. The Verticillium LisH domain transcription factor Som1 and the transcription AGL1 [27] as described [28]. V. dahliae was manipulated using A. tumefaciens-mediated transformation [17]. In-locus complemented strains were constructed in the same way to verify phenotypes of deletion strains. All strains used in this study are listed in S3 Table. Media and culture conditions Lysogeny broth [29] supplemented with 100 μg ml -1 kanamycin (AppliChem) was used to cultivate E. coli and A. tumefaciens at 37˚C and 25˚C, respectively. V. dahliae was incubated at 25˚C in or on simulated xylem medium (SXM), potato dextrose medium (PDM; Carl Roth) or Czapek-Dox medium (CDM) as described [30]. Conidiospores were harvested using sterile Miracloth (Calbiochem), washed and resuspended in sterile water, and counted using the Coulter Z2 Particle Count and Size Analyzer (Beckman Coulter) with Beckman Coulter Isoton II Diluent, or a Neubauer counting chamber.

Sample preparation and transcriptome analysis
Xylem sap was extracted from six-week-old, uninfected Moneymaker tomato plants (Solanum lycopersicum) as described [16]. We acquired approximately 1 l pure xylem sap from 325 plants and added sterile distilled water, which accounted for 10% of the final volume. Filtered xylem sap was kept at -80˚C until use. V. dahliae wild-type and the VTA3 deletion strain were grown shaking in liquid SXM at 25˚C to produce conidiospores. After harvest, 50 ml liquid SXM were inoculated with 5 × 10 7 spores. Following cultivation for five days, mycelia were transferred to extracted tomato xylem sap and further incubated with shaking at 25˚C for 8 h. For transfer, cultures were centrifuged at 2,500 rpm for 10 min and mycelia were washed with sterile water. This procedure was repeated once. Mycelia were harvested using Miracloth and ground to powder in liquid nitrogen. Fungal strains were cultivated in three independent replicates. RNA was extracted using the TRIzol/chloroform protocol [31]. DNase treatment was performed using the TURBO DNA-free Kit (Invitrogen Thermo Fisher Scientific). RNA concentrations and purity were checked with a Nanodrop spectrophotometer (Thermo Fisher Scientific).
Quality control, library preparation, sequencing and bioinformatic analysis were conducted by GENEWIZ (Azenta Life Sciences) on the HiSeq platform with 2 x 150 configuration as described [19]. Briefly, polyA selection for eukaryotic mRNA was performed. Raw reads were trimmed to remove possible adapter sequences and low-quality bases using Trimmomatic v.0.36 [32]. Trimmed reads were mapped to the V. dahliae JR2 genome obtained from Ensembl Fungi [24] using STAR aligner v.2.5.2b. Unique gene hit counts were calculated using featureCounts from the Subread package v.1.5.2. Comparison of gene expression between the three wild-type samples and the three samples of the VTA3 deletion strain was performed using DESeq2. Sample distances were measured using expression values from each sample to identify how similar the groups are (S1 Fig). Wald tests were used to generate P-values and log 2 fold changes (VTA3 deletion strain versus wild-type). Genes with an adjusted Pvalue greater than 0.05 and log 2 (fold change) < 1 or > -1 were excluded from further analysis.
All remaining genes were considered significantly differentially expressed and analyzed for functional enrichment (S4-S6 Tables) with the FungiFun2 web tool [33] using corresponding VdLs.17 identifiers. The web tool produces overlaps between categories thus some genes appear in more than one category. Transcripts with a log 2 (fold change) � -2 were regarded as most induced (S7 Table), those with a log 2 (fold change) � 2 as most reduced by Vta3 (S8 Table). These transcripts were further investigated separately ( Fig 1A) using Ensembl Fungi [24], InterPro [34], and BLAST with NCBI [35] and FungiDB [36].  The Verticillium dahliae transcription factor Vta3 affects the expression of 162 fungal genes in tomato xylem sap including the Egh16-like virulence factor-encoding gene ELV1 and the Master transcription factor-encoding gene MTF1. Equal numbers of spores of V. dahliae JR2 wild-type (WT) or VTA3 deletion strain (ΔVTA3) were used to inoculate liquid simulated xylem medium and incubated for five days with shaking. The mycelia were then transferred to extracted tomato xylem sap and incubated for a further 8 h. (a) The transcriptomes of three biological replicates per fungal strain were analyzed by RNA sequencing. After applying a log 2 (fold change) of � 2, transcript levels of 100 genes were significantly higher in the presence of VTA3 (induced; in blue), whereas expression of 62 genes was reduced (in orange). The products of the differentially expressed genes are involved in various processes including metabolism, redox processes, stress response, transcription, transport, virulence and unknown functions. Putative functions were assigned based on conserved domains identified by searches with InterPro and BLAST. Details are given in S7 Table  (Vta3-dependently induced candidates) and S8 Table (Vta3-dependently reduced candidates). (b, c) ELV1 and MTF1 expression levels were analyzed by reverse transcription-quantitative PCR. The transcript levels of references H2A and EIF2B were used for normalization and gene expression levels in wild-type were set to one. Mean values ± SE of the mean from three independent experiments are shown. Significance was calculated using t-tests. (b) The presence of VTA3 leads to a five-fold induction of ELV1 expression ( � , P < 0.05). (c) MTF1 expression was reduced six-fold ( � , P < 0.05) in the presence of VTA3.

RNA extraction, cDNA synthesis and quantification of gene expression
Fungal strains were cultured as described above and RNAs were extracted from ground mycelia using TRIzol [31]. RNA from tomato hypocotyls was extracted in the same manner at 21 days post inoculation (dpi). QuantiTect Reverse Transcription Kit (Qiagen) was used for cDNA synthesis using 0.8 μg RNA. Transcript levels were analyzed in duplicate or triplicate (n = 1) using MESA GREEN qPCR MasterMix Plus for SYBR Assay (Eurogentec) in a CFX Connect Real Time PCR Detection System (Bio-Rad Laboratories). Primers used for quantitative PCR (qPCR) are listed in S9 Table. Expression levels of fungal genes were quantified relative to references histone H2A and EIF2B. For plant genes, tomato αTUB and elongation factor EF1α were used as references. The 2 À DDC T method [37] was applied and wild-type expression was set to one. Means of two or three independent experiments ± SE of the mean are presented.

Plasmid and strain construction of ELV1 deletion and complementation strains
The 1,277 bp ELV1 open reading frame (ORF) was replaced by a 2,194 bp nourseothricin resistance marker under control of a gpdA promoter and a trpC terminator amplified from pME4815 [14] using the primers ML8 and ML9. The primers IM32 and IM105 were used to amplify the 800 bp 5' flanking region from wild-type genomic DNA and the 1,000 bp 3' flanking region was amplified using the primers IM34 and IM35. All fragments were inserted into the EcoRV and StuI-treated 6,804 bp pME4564 [14] backbone resulting in pME5482. The V. dahliae wild-type was transformed with this plasmid resulting in the ELV1 deletion strains VGB670 and VGB671.
For the ELV1 complementation strain, the 5' flanking region, the backbone of pME4564 [14] and the 3' flanking region were amplified from the plasmid pME5482 using the primers IM34 and IM85. The ELV1 gene was obtained by amplification from wild-type genomic DNA using the primers IM86 and IM143. A 3,942 bp hygromycin B resistance marker under control of a gpdA promoter and a trpC terminator was amplified from the plasmid pPK2 [38] using the primers ML8 and RO3. All fragments were ligated and the resulting plasmid pME5483 was used to transform the ELV1 deletion strain VGB670. The ELV1 complementation strain was named VGB694. Correct integration of deletion and complementation constructs into V. dahliae was verified by Southern hybridization using SalI restriction and the 5' flanking region as probe (S2 Fig).

Plasmid and strain construction of MTF1 deletion and complementation strains
The 1,023 bp MTF1 ORF was replaced by homologous recombination with a 2,643 bp hygromycin B resistance marker cassette amplified from the plasmid pPK2 [38] using the primers ML8 and ML9. The primers IM40 and IM41 were used to amplify the 650 bp 5' flanking region from genomic DNA of the wild-type and the 1,300 bp 3' flanking region was amplified using primers IM42 and IM43. All fragments were inserted into the EcoRV and StuI-treated pME4564 [14] backbone. The resulting plasmid pME5480 was used to transform the V. dahliae JR2 wild-type by A. tumefaciens-mediated transformation resulting in the MTF1 deletion strains VGB575 and VGB576. For visualization of fungal hyphae on plant roots, the MTF1 deletion strain VGB575 was transformed with the plasmid pME4819 [39]. The μ-Slide 8 well microscopy chambers (Ibidi) were used to screen the transformants for green fluorescent hyphae. The MTF1 deletion strains harboring ectopically integrated GFP were named VGB625 and VGB626. For the MTF1 complementation strains, the 5' flanking region together with the MTF1 gene were amplified using the primers IM40 and IM84. This fragment was inserted into the EcoRV and StuI-treated pME4564 [14] backbone together with a nourseothricin resistance marker cassette and the 1,300 bp 3' flanking region. The resulting plasmid pME5481 was used for transformation of the MTF1 deletion strain VGB575 to obtain the MTF1 complementation strains VGB635 and VGB636. For the GFP-MTF1 strain, the nourseothricin resistance marker cassette, the 3' flanking region, the backbone of pME4564 [14] and the 5' flanking region were amplified from pME5481 with primers ML5 and IM92. GFP including a linker was amplified with primers ZQY10 and ZQY11 from pGreen2 [16]. Primers IM84 and IM91 were used to amplify the MTF1 gene from wild-type genomic DNA. These fragments were ligated to yield the plasmid pME5510 which was used to transform the MTF1 deletion strain VGB575. The resulting GFP-MTF1 expressing strain was named VGB650. Correct integration of deletion and complementation constructs into V. dahliae was verified by Southern hybridization using

Genomic DNA isolation and Southern hybridization
Genomic DNA extraction and Southern hybridization were performed as previously described [14,17]. Briefly, mycelia were harvested using Miracloth filters and homogenized into powder in liquid nitrogen. Genomic DNA was isolated with phenol from mycelium grown shaking at 25˚C in liquid PDM. Amersham AlkPhos Direct Labelling and CDP-Star detection reagents (GE Healthcare) were used for Southern hybridization.

Protein extraction and western experiments
For detection of GFP-fused proteins, 5 × 10 7 freshly harvested spores were inoculated into liquid PDM, SXM, CDM, extracted tomato xylem sap or onto 30 ml solid SXM plates covered with nylon membrane (GE Healthcare) and incubated for indicated times. For xylem sap, a five-day preculture was performed in SXM, after which the mycelia were transferred to xylem sap and incubated for an additional 8 h. After harvesting, the mycelium was ground to powder in liquid nitrogen. Protein extracts were obtained using B � buffer, and protein concentrations were determined as previously described [39]. Western experiments were performed with 80 to 160 μg protein extracts according to the protocol described [14,17]. Prior to blocking, Ponceau S staining was applied as a loading control [39]. Membranes were incubated with a monoclonal mouse α-GFP antibody (Santa Cruz Biotechnology) and a secondary horseradish peroxidase-coupled goat α-mouse antibody (Jackson ImmunoResearch). Signals were visualized on Amersham Hyperfilm ECL films (GE Healthcare) developed with Optimax film processor (Protec).

In-vitro protein pull-down and sample preparation for liquid chromatography/mass spectrometry
Liquid PDM (500 ml) was inoculated with 5 × 10 8 freshly harvested spores and cultured for five days at 25˚C with shaking. Proteins were extracted from the V. dahliae JR2 wild-type, the wild-type strain overexpressing ectopically integrated GFP and the VTA3-GFP expressing strain. The experiment was performed as previously described [31] with the following modifications: 15 μl of GFP-Trap Agarose beads were used. The three eluates were collected in one tube. Sediments were dissolved in 40 μl of sample buffer by shaking for 5 min and incubation in an ultrasonic bath for 3 min. The peptides were purified using StageTips [40,41]. Briefly, StageTips were equilibrated in four steps with (I) 100 μl methanol/0.1% formic acid, (II) 100 μl 70% acetonitrile/0.1% formic acid, (III) 100 μl water/0.1% formic acid and (IV) 100 μl water/ 0.1% formic acid. Samples were divided onto two StageTips and incubated on the columns for 5 min. This step was repeated after centrifugation at 4,000 rpm for 5 min. The columns were washed twice with 200 μl of water/0.1% formic acid. Elution was performed by adding 60 μl of 70% acetonitrile/0.1% formic acid, and the two parts of each sample were combined. Peptide samples were dried and liquid chromatography/mass spectrometry (LC/MS) analysis of the peptides was performed as described [31]. MaxQuant 1.6.10.43 [42] and Perseus 1.6.0.7 [43] were used to analyze MS raw data. Data were processed as described [31]. Missing values were replaced four times from the normal distribution. The VTA3-GFP expressing strain was used as the first group and the wild-type as the second group. Mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [44] partner repository with the dataset identifier PXD039123.

Phenotypical analyses
Freshly harvested spores were adjusted to 5 × 10 6 spores ml -1 and 10 μl were spotted onto indicated media plates. PDM, SXM, CDM, CDM with 3% cellulose as an alternative carbon source and CDM supplemented with 0.004% SDS as a stress inducing agent were used. Two independent transformants per genotype were compared with wild-type. After 10 days of incubation at 25˚C, growth and microsclerotia production of the strains were assessed by binocular (SZX12-ILLB2-200, illuminated with the KL1500-LCD light source, Olympus) and light microscopy (Axiolab, Zeiss) (equipped with SC30 cameras, Olympus) with cellSens Dimension software (Olympus).

Quantification of microsclerotia and conidia formation
The formation of microsclerotia in fungal colonies was assessed by measuring the brightness factor with the ROI Manager of ImageJ software [45] as previously described [31,39]. CDM plates containing cellulose were inoculated with 50,000 freshly harvested spores in triplicate (n = 1) and incubated at 25˚C for 10 days. In three independent experiments, two individual transformants of the deletion and complementation strains were compared with two independent wild-type cultures each (n = 6). Means ± SD are shown and statistical significance was calculated using t-tests.
Quantification of conidia production was performed as previously described [31,39] with the following modifications: Spores were incubated for five days and each strain was inoculated in three to four technical replicates (n = 1). In each experiment, two independent transformants of the deletion or complementation strains were compared with two independent wild-type cultures. Significance was calculated using t-tests.

Arabidopsis thaliana root infection
The assay for colonization of A. thaliana Col-0 roots by the indicated V. dahliae strains was performed as previously described [17,30]. Fluorescence micrographs were taken at two and five days post inoculation (dpi) with a 20×/0.5 air objective of the Axio Observer Z1 system (Zeiss) with Laser Lunch System (Model 3iL32, Intelligent Imaging Innovations), Quan-tEM:512SC camera (Photometrics) and the Slide Book 6.0 imaging software (Intelligent Imaging Innovations). Two independent experiments with strains constitutively expressing ectopically integrated GFP were performed. In each of them, two individual ΔMTF1 strains (ΔMTF1 GFP OE; VGB625, VGB626) and two individual ΔVTA3 strains (ΔVTA3 GFP OE; VGB184, VGB185) were compared with respective wild-type controls (WT GFP OE; VGB45, VGB392). Quantification of fungal root colonization was performed at 5 dpi. For each root, 10 to 15 stacks consisting of several individual pictures were acquired at randomly selected

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection sections. Roots of two plants per treatment were considered as one biological replicate. The green fluorescence of GFP-expressing fungal hyphae was quantified relative to the root area using the mask statistics feature of the Slidebook 6.0 software. The means of four biological replicates ± SD are presented and a t-test was used for statistical analysis.

Pathogenicity assay on tomato plants
Pathogenicity assays were performed as previously described [14,30]. In brief, S. lycopersicum ('Moneymaker', Kiepenkerl Bruno Nebelung) seeds were surface sterilized with 70% (v/v) Ethanol, 0.05% Tween 20. We wounded the roots of 10-day-old seedlings and inoculated them by incubation in 50 ml of 10 7 spores ml -1 or water (mock) for 40 min under constant agitation. Additionally, 3 × 10 7 spores or 3 ml water (mock) were added to the seedlings in pots containing a sand/soil mixture. Plants were incubated for 21 days in a BrightBoy GroBank (CLF Plant-Climatics). Plant weight, height and longest leaf length were measured and calculated into a disease score ranking relative to the means of mock-inoculated plants (set to 100%). Values above 80% were classified as 'healthy', 60-80% as 'mild symptoms', 40-60% as 'strong symptoms' and below 40% as 'very strong symptoms'. The disease scores of each plant are visualized in stack diagrams relative to the total amount of treated plants. Statistical significance was calculated using two-tailed Mann-Whitney U tests [46]. We also observed the discoloration of the tomato hypocotyls and tested the plants for fungal outgrowth from surface-sterilized stem sections.
V. dahliae DNA was quantified at 21 dpi in hypocotyls using primers OLG70/OLG71 [7] to amplify a specific rDNA fragment (parts of the 5.8S rRNA and ITS2) relative to the tomato elongation factor-encoding gene EF1α [47]. For this purpose, genomic DNA was isolated from 13 to 15 hypocotyls per treatment using the NucleoSpin Plant II Kit (Macherey-Nagel). We tested 10 ng DNA in technical triplicates using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories). The 2 À DDC T method [37] was applied and wild-type was used for normalization. Means of five to six biological replicates ± SE of the mean are presented and a ttest was used for statistical analysis.

Extraction of plant hormones and relative signal determination
Leaves from tomato plant infection experiments were harvested at 21 dpi, frozen and ground in liquid nitrogen. Phytohormone extraction with 80% methanol was performed on approximately 50 mg (fresh weight) ground material [48]. Pipecolic acid (Pip) and salicylic acid (SA) were reversed phase-separated using an ACQUITY UPLC system (Waters) and analyzed by nanoelectrospray ionization (TriVersa Nanomate, Advion BioSciences) coupled with an AB Sciex 4000 QTRAP tandem mass spectrometer (AB Sciex) employed in scheduled multiple reaction monitoring mode as described [49]. The reversed phase separation was achieved by UPLC using an ACQUITY UPLC HSS T3 column (100 mm x 1 mm, 1.8 μm; Waters). Solvent A and B were water and acetonitrile/water (90:10, v/v), respectively, both containing 0.3 mmol l -1 NH 4 HCOO (adjusted to pH 3.5 with formic acid). The flow rate was 0.16 ml min -1 and the separation temperature was constant at 40˚C. For Pip, the elution was performed isocratically for 1 min at 1% solution B, followed by a linear increase to 95% solution B in 4.5 min, this condition was held for 2.5 min. The column was re-equilibrated for start conditions in 6 min. For SA, the elution was performed isocratically for 0.5 min at 10% solution B, followed by a linear increase to 40% solution B in 1.5 min, this condition was held for 2 min, followed by a linear increase to 95% solution B in 1 min, this condition was held for 2.5 min. The column was reequilibrated for start conditions in 3 min. For nanoelectrospray ionization-tandem mass spectrometric analysis, Pip was ionized in positive mode and SA in negative mode. The mass transitions were as follows: 130/84 (declustering potential 90 V, entrance potential 8 V, collision energy 22 V) for Pip and 137/93 (declustering potential -25 V, entrance potential -6 V, collision energy -20 V) for SA. The relative signal area per fresh weight was calculated and wild-type was set to one. Means of 15 plants per treatment from two independent experiments ± SD are presented.

Bioinformatics methods
The Ensembl Fungi database [24] was used for annotation of V. dahliae genes. The FIMO web tool [50] was used for performing promoter analyses. Protein sequences were analyzed via the InterPro website [34]. The eventual presence of signal peptides in protein sequences was analyzed using SignalP 6.0 [51]. cNLS Mapper [52] was used to predict nuclear localization signals in the entire protein sequence. Protein sequences of other fungi mentioned in this study were obtained from NCBI [35]. Multiple sequence alignments were performed using the MegAlign Pro software (DNASTAR, version 16) with the MUSCLE algorithm for uncorrected pairwise distance alignments. Sequences of tomato genes were obtained from Ensembl Plants [24]. Functional enrichment analysis of differentially expressed genes was performed using the Fun-giFun2 web tool [33]. Significance (ns, not significant; � , P < 0.05; �� , P < 0.01; ��� , P < 0.001; ���� , P < 0.0001) was calculated by independent two-sample t-tests using the iCalcu website (https://www.icalcu.com/stat/two-sample-t-test-calculator.html). Student's t-test was used for equal variances, while Welch's t-test was applied in case of unequal variances. The two-tailed Mann-Whitney U test [46] was performed for statistical analysis of tomato plant infections.

Vta3 controls the expression of 1,179 V. dahliae genes including candidates involved in virulence and regulators of other genetic networks
The transcriptomes of V. dahliae wild-type and VTA3 deletion strains were compared to assess the Vta3 genetic network with emphasis on virulence-associated functions. Fungal strains were cultivated in liquid SXM before mycelia were shifted to extracted tomato xylem sap and further incubated for 8 h. Overall, the loss of VTA3 affected the expression of 1,179 genes. In the presence of VTA3, 650 transcripts were induced, while the transcription of 529 genes was reduced. The FungiFun2 web tool [33] was applied for functional enrichment analysis for all significant differentially expressed genes (log 2 (fold change) � 1 or � -1). Thereby, 209 of 1,179 genes were assigned to the five significantly enriched categories (i) heavy metal binding, (ii) heme binding, (iii) secondary metabolism, (iv) disease, virulence and defense and (v) virulence, disease factors (S4 Table). Among the 650 transcripts induced in the presence of VTA3, 15 were found in the significantly enriched category 'disease, virulence and defense' (S5 Table), including the SnodProt1-like proteins Cp1 and Cp2. Cerato-platanin family proteins, such as SnodProt1, are involved in parasitism, adhesion and development of fungi [53]. Whereas Cp1 contributes to virulence on cotton [54], Cp2 is dispensable for virulence on tomato plants [14]. 95 of the 529 transcripts reduced in the presence of VTA3 were classified in five significantly enriched categories. Besides secondary metabolism, the deduced proteins are attributed to DNA damage response, repair, recombination, and topology (S6 Table).
Additionally, the deduced proteins of the most strongly induced/reduced transcripts (log 2 (fold change) � -2 or � 2) were manually assigned functions based on BLAST search identification of encoded conserved domains. In the presence of VTA3, expression of 100 genes was induced with a log 2 (fold change) � -2 ( Fig 1A, S7 Table) and expression of 62 genes was reduced with a log 2 (fold change) � 2 ( Fig 1A, S8 Table). The identified candidates have

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection various functions, which are summarized in Fig 1A. For 42 induced and 31 reduced transcripts, no function could be assigned due to the absence of conserved domains or characterized homologs. Among these are two Vta3-induced and four reduced transcripts encoding potential effectors (� 200 aa, predicted signal peptide, � 2% cysteines).
Expression of three genes encoding putatively extracellular virulence-associated proteins is induced in the presence of VTA3, including an Egh16-like virulence factor domain protein (named Elv1). We further investigated Elv1 functions because Vta3 induces its transcript level 13-fold compared with only four to six-fold for the other two candidates. The increased expression of ELV1 in the presence of VTA3 was confirmed by qPCR ( Fig 1B). Elv1-related proteins presumably are important during early infection in the rice blast fungus Magnaporthe oryzae (formerly M. grisea) [55].
Expression of two potential transcription factor-encoding genes is reduced in the presence of an intact VTA3 regulatory gene suggesting a potential direct or indirect repressor function for other genetic networks (Fig 1A). One of the deduced proteins carries only a weak potential basic-leucine zipper motif, whereas the other protein is encoded by the bona fide MTF1 ortholog of Aspergillus nidulans mtfA (S8 Table). MtfA of Aspergillus spp. is involved in secondary metabolism and morphogenesis [56,57]. A possible virulence-associated role was suggested in the saprotrophic Aspergillus flavus when colonizing peanut seeds [58]. We confirmed by qPCR that Vta3 reduces MTF1 expression in xylem sap ( Fig 1C) as prerequisite for further investigation of possible Mtf1 functions in the pathogenic fungus V. dahliae during infection of tomato plants.

The Egh16-like virulence factor Elv1 contributes to V. dahliae virulence on tomato plants but is dispensable for vegetative growth and conidiation
Consistent with the RNA sequencing data, promoter analysis using the FIMO web tool [50] revealed that ELV1 has consensus sequences in its 5' regulatory region for possible binding of Crt1/RFX1 (DNA-binding motif discovered by [59]), an ortholog of Vta3 (S4A Fig). The V. dahliae ELV1 ORF consists of 1,277 bp, including three exons and two introns, encoding a 389 amino acid (aa) protein with a predicted molecular weight of 38.9 kDa (Fig 2A). An N-terminal signal peptide that directs the protein for secretion was predicted by SignalP 6.0 [51]. The protein harbors an Egh16-like virulence factor domain (IPR021476; 19-195 aa). Three further proteins with this domain are annotated in JR2 and share 60-68% identity with Elv1. A BLAST search revealed that Elv1 is homologous to Egh16 originally identified from a cDNA library of germinating conidia of the powdery mildew fungus Blumeria (formerly Erysiphe) graminis f. sp hordei [60], and to its homolog Egh16h1. Other related proteins with approximately 60% similarity are the appressoria-specific virulence factors Gas1 and Gas2 of M. oryzae, which are involved in lesion development in rice and barley [55]. Functionally uncharacterized putative homologs with similarities between 21-66% were found in the phytopathogens Colletotrichum graminicola and F. oxysporum, as well as in Neurospora crassa and A. fumigatus (S5A Fig). Predicted homologs exist only in filamentous fungi, but not in yeasts such as S. cerevisiae or Candida albicans.
We replaced the ELV1 ORF by a nourseothricin resistance marker. In-locus complemented strains with ELV1 alongside a hygromycin resistance marker were generated. Strains were verified by Southern hybridization (S2 Fig). A possible function of Elv1 in fungal growth outside the plant was tested by point-inoculating spores of wild-type, ELV1 deletion and complementation strains on solid media. Conidiation was analyzed in liquid SXM. We did not observe any defects of the ELV1 deletion strain in vegetative growth, microsclerotia or conidia formation (Fig 3A and 3B).

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection ELV1 expression was strongly induced in the presence of VTA3 in xylem sap (Fig 1A and  1B). Therefore, we investigated whether V. dahliae Elv1 is involved in tomato plant colonization. Disease symptoms were assessed at 21 dpi of 10-day-old tomato seedlings treated with spores of wild-type, ELV1 deletion or complementation strains or water (mock). A general disease index was calculated from the measured plant height, weight and longest leaf length [14,30]. The diagram in Fig 4A displays the number of plants classified as healthy or with weak, strong or very strong symptoms compared with mock-inoculated plants. Deletion of ELV1 resulted in approximately 25% more plants with no or only weak symptoms compared with wild-type-infected plants. Hypocotyl discoloration as a sign of fungal infection was observed in cross-sections of plants treated with wild-type, ELV1 deletion or complementation strains, but not in mock-inoculated plants (Fig 4B). Fungal DNA (Fig 4C) and PR gene expression (S6 Fig) were quantified to address possible alterations in plant colonization or immune responses, respectively. Both were similar to the wild-type in plants infected by ELV1 deletion and complementation strains. Taken together, these data suggest that Elv1 is dispensable for growth, conidiation and initial plant colonization but contributes to late and full disease development in tomato plants.

The C 2 H 2 -type zinc-finger transcriptional regulator Mtf1 is present during vegetative growth of V. dahliae
Vta3-dependent control of the virulence factor Elv1-encoding gene supports a temporal Vta3 function in late V. dahliae plant xylem colonization. Therefore, we examined whether the

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection reduced expression of the Vta3-dependent transcriptional regulator Mtf1 is also associated with late stages of fungus-induced tomato plant disease and the subsequent formation of survival structures. Like ELV1, MTF1 also carries consensus sequences for putative binding of Vta3 as the Crt1/RFX1 counterpart of V. dahliae in its 5' regulatory region (S4B Fig). The A. nidulans counterpart MtfA was first identified in a mutant strain that had regained the ability to produce sterigmatocystin protecting the overwintering fruiting bodies of this fungus, which is lost in veA deletion strains [56]. Both VeA and MtfA are master transcriptional regulators of secondary metabolism and development in A. nidulans and A. fumigatus, although different underlying genetic pathways are involved [61]. In V. dahliae, the ortholog MTF1 is comprised of a 1,023 bp ORF including two exons and one intron. The deduced protein consists of 321 aa and has a molecular weight of approximately 34.9 kDa (Fig 2B). The Mtf1 sequence harbors a putative nuclear localization signal (258-292 aa) and a C 2 H 2 -type zinc finger domain (IPR013087) with high conservation among ascomycetes. A comparison of the V. dahliae Mtf1 sequence with putative orthologs in other fungi, including the phytopathogens C. graminicola and M. oryzae, and the model organism N. crassa, revealed similarities of 49-69% (S5B Fig). An ortholog was not found in either S. cerevisiae or humans.
We examined the production of a functional Mtf1 protein during growth and development of V. dahliae using a strain expressing MTF1 fused to GFP and compared it with the production of Vta3 fused to GFP (VTA3-GFP strain from [17]). Phenotypic examination of the GFP-MTF1-expressing strain revealed that it resembles the wild-type, suggesting that the fusion protein is functional (S7 Fig). Vta3-GFP (115 kDa) was detected under condition favoring vegetative growth and conidia production as well as in minimal medium and natural plant xylem sap. On SXM plates, where increasing numbers of microsclerotia are formed over time, Vta3-GFP was visible after two days, whereas at later time points only free GFP (27 kDa) was detectable. In contrast, the GFP-Mtf1 fusion protein with a size of about 62 kDa was detectable only when V. dahliae grew predominantly vegetatively, and a weak signal corresponding to GFP-Mtf1 was visible in minimal medium (S7 Fig). Consistent with the Vta3-dependent downregulation of MTF1 transcription (Fig 1A and 1C), strong signals corresponding to Vta3-GFP were detected when V. dahliae grew in xylem sap, whereas GFP-Mtf1 was not found (S7 Fig). These results indicate that Vta3 is produced in the plant xylem and underline a possible function of Vta3 in reducing Mtf1 levels in the plant.

Mtf1 enhances microsclerotia formation in V. dahliae
Deletion strains were constructed carrying a hygromycin resistance marker in place of the MTF1 ORF to investigate Mtf1 functions. In-locus complemented strains were generated by inserting MTF1 into the deletion strain under its native promoter with a nourseothricin resistance marker. All strains were verified by Southern hybridization (S3 Fig). Microsclerotia are specialized melanized cells that enable V. dahliae to survive in the soil for several years [1,62]. Although GFP-fused Mtf1 was not detected under conditions favoring microsclerotia production (S7 Fig), the transcriptional regulator may still be involved in the induction of microsclerotia formation. The ability to form microsclerotia was compared between wild-type, MTF1 deletion and complementation strains by point-inoculation of spores on different media. Melanization of fungal colonies decreased in the absence of MTF1 (Fig 5A). Microscopic observations of fungal material from colony centers implied that MTF1 is required for microsclerotia abundance but not for their size or shape. MTF1 deletion did not result in any growth defect or other phenotypic differences compared with wild-type.
The melanization of fungal colonies grown on cellulose-containing CDM plates was quantified as a measure for the number of microsclerotia formed. There was a 40% reduction in Melanization of fungal colonies grown on CDM + cellulose was quantified as a measure of the number of microsclerotia formed. Pictures were taken from top view scans of the colonies after removal of aerial mycelium. The brightness factor of three colonies per transformant (n = 1) was calculated using the ROI Manager of ImageJ software. Three independent experiments were performed. Means ± SD are shown. Significant differences from wild-type were determined using t-tests. Melanization of ΔMTF1 was reduced to approximately 60% of wild-type levels ( ���� ,

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection melanization in the MTF1 deletion strain compared with wild-type (Fig 5B). The complementation strain was able to form microsclerotia at levels comparable to the wild-type. The transcription factors Vta1 and Cmr1 are produced during microsclerotia development and are required for melanization [30,63]. VTA1 expression was reduced by 40% in the MTF1 deletion strain compared with wild-type (Fig 5C), while CMR1 expression was not affected by MTF1 deletion (S8 Fig). These results suggest that Mtf1 promotes microsclerotia formation and VTA1 expression to favor melanization independent of Cmr1, which may act in a parallel signaling pathway.

Colonization of Arabidopsis roots by V. dahliae is independent of MTF1
Possible Mtf1 functions in different stages of plant infection were addressed. Roots are the entry points to the host for V. dahliae as a soil-borne pathogen [1]. Arabidopsis thaliana roots were treated with spores of strains expressing ectopically integrated GFP under control of a gpdA promoter. Root colonization of wild-type and MTF1 and VTA3 deletion strains was compared. Transcription factor-encoding VTA3 is required for fungal propagation on A. thaliana roots [17]. By fluorescence microscopy, we show that roots were similarly colonized by wild-type and MTF1 deletion strains (Fig 6A). Initial colonization of the root surface was observed by 2 dpi, and whole roots were covered with hyphae of wild-type or MTF1 deletion strains by 5 dpi. In contrast, the VTA3 deletion strain was significantly impaired in propagation on A. thaliana roots (P < 0.05; Fig 6B). These results indicate that MTF1 is not necessary for attachment and initial growth on root surface and that the regulation of root colonization by Vta3 is an Mtf1-independent process.

Mtf1 increases the virulence of V. dahliae on tomato plants
The functions of Mtf1 in later stages of plant colonization and symptom induction were examined by pathogenicity assays with the host plant tomato. Roots of 10-day-old tomato seedlings were treated with spores of wild-type, MTF1 deletion or complementation strains and disease symptoms were assessed at 21 dpi. The general disease scoring revealed that more than 80% of plants displayed strong and very strong symptoms when infected by either wild-type or MTF1 complementation strains, while less than 20% remained healthy or showed only weak symptoms. In contrast, this disease rate was halved to approximately 40% of plants displaying strong and very strong symptoms when infected by the MTF1 deletion strain. Nearly 60% of plants treated with the MTF1 deletion strain were either healthy or displayed only a mild stunting phenotype (Fig 7A). This suggests a strong contribution of Mtf1 to V. dahliae virulence.
Fungal infection leads to discoloration of the plant hypocotyl visible in cross-sections ( Fig  7B). Brownish streaks were observed in 95% of wild-type-infected plants, in 91% of MTF1 complementation strain-infected plants and in 85% of MTF1 deletion strain-treated plants. No discoloration was evident in hypocotyls of mock-inoculated plants. The presence of V. dahliae inside the plant was analyzed by re-isolating mycelium from stem sections on PDM plates and by quantifying fungal DNA in hypocotyls. No fungus grew from stems of mock-inoculated plants. Fungal material was successfully re-isolated from 75% of wild-type-infected stem P < 0.0001), which is rescued in MTF1-C (ns, not significant). (c) Expression levels of VTA1, which encodes the microsclerotia melanization-controlling transcription factor Vta1, were compared between wild-type, MTF1 deletion and complementation strains. RNA was extracted from mycelia grown in liquid SXM for five days and subsequently in xylem sap for 8 h. Means of three biological replicates normalized to transcript levels of wild-type and references H2A and EIF2B with error bars representing the SE of the means are shown. VTA1 expression was reduced to 60% in ΔMTF1 ( � , P < 0.05) and wild-type-like in MTF1-C (calculated using t-tests). https://doi.org/10.1371/journal.ppat.1011100.g005

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection

Fig 6. MTF1 is dispensable for Arabidopsis thaliana root colonization by Verticillium dahliae.
A. thaliana seedlings (three-weekold) were treated by root-dipping with spores of V. dahliae JR2 wild-type (WT GFP OE), ΔMTF1 (ΔMTF1 GFP OE) or ΔVTA3 (ΔVTA3 GFP OE) strains constitutively expressing (OE) ectopically integrated GFP. Fungal colonization of the roots was observed at two and five days post inoculation (dpi). Root cells were stained with 0.05% propidium iodide/0.01% silwet solution. Two independent experiments were performed. (a) Representative fluorescence micrographs depict similar initial (2 dpi) and complete root colonization (5 dpi) by V. dahliae wild-type or MTF1 deletion strains compared with the ΔVTA3 control, which did not propagate to the same extent on the root surface (scale = 10 μm). (b) Quantification of fungal root colonization was performed at 5 dpi. For each root, 10 to 15 images were taken at randomly selected sections. Roots of two plants per treatment were considered as one biological replicate. The green fluorescence of GFP-expressing fungal hyphae was quantified relative to the root area. Shown are the means of four biological replicates ± SD. A t-test was used for statistical analysis. No significant difference was detected between plant roots colonized by the wild-type or the MTF1 deletion strain (ns, not significant), in contrast to the ΔVTA3 control strain, which is impaired in its ability to colonize plant roots ( � , P < 0.05).

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection sections and from 74% of plants infected by the MTF1 complementation strain. In plants inoculated with the MTF1 deletion strain, the fungus was re-isolated from 62% of stem sections (Fig 7C). Fungal DNA in hypocotyls was analyzed by qPCR relative to plant DNA. Although less fungal DNA was measured in MTF1 deletion strain-inoculated plants compared with wild-type-inoculated plants, the reduction was not significant (Fig 7D).
Coupled with the observations that the MTF1 deletion strain is not impaired in hyphal growth (Figs 5 and 6) or conidiospore production (S9 Fig), our data indicate that Mtf1 is involved in virulence and increases symptom severity in tomato plants by acting at later stages of plant colonization.

Mtf1 acts in the sequential Vta3-Vta2 network required for plant disease
Transcriptional profiling of V. dahliae strains cultivated in xylem sap revealed that the transcriptional regulator Vta3 reduces MTF1 expression either as a repressor or indirectly through a yet elusive molecular mechanism (Fig 1A and 1C). A possible interplay of MTF1 expression with other prominent transcription factors involved in pathogenicity (Som1 and Vta2) was examined in corresponding deletion strains using qPCR and brought into context with the Vta3-dependently reduced expression of MTF1 ( Fig 8A). As shown previously in Fig 1C, MTF1 expression was reduced six-fold in the presence of VTA3. Som1 and Vta3 control genetic networks for sequential steps of plant root penetration and infection prior to Vta2-controlled genes [16,17]. MTF1 expression was induced eight-fold in the VTA2 deletion strain in comparison with the wild-type, whereas no MTF1 expression was measured in the MTF1 deletion strain. MTF1 transcript levels were not significantly elevated in the SOM1 deletion strain. Notably, the transcription of SOM1, VTA3 and VTA2 does not change significantly when an intact MTF1 gene is present or not (S10B and S10C Fig).
Thus, MTF1 expression is affected by the concerted and presumably sequential action of Vta3 and Vta2, a second regulatory protein controlling adhesion and plant root colonization. This reflects a complex network in which Som1 induces VTA3 and VTA2 expression. Vta3 in turn induces VTA2 expression [17] (Fig 8B). Mtf1 is therefore integrated into a regulatory network required for fungal development and virulence.

Mtf1 induces expression of fungal effectors and tomato defense signaling
Fungal pathogens secrete effector proteins for successful colonization and manipulation of plant defense responses. The Ave1 effector is recognized by the cell-surface receptor-like protein Ve1 but acts as a virulence factor on tomato plants lacking Ve1 [64]. AVE1 transcript levels were compared between wild-type, MTF1 deletion and complementation strains cultured in xylem sap. AVE1 expression was reduced to 25% in the absence of MTF1 (Fig 9A). This Mtf1-mediated activation of Ave1 might explain the reduced aggressiveness of the MTF1 deletion strain on tomato cultivars that lack Ve1 (Fig 7).
PR gene expression was examined in tomato hypocotyls by qPCR at 21 dpi to investigate variations in host responses to colonization by V. dahliae. Mock-inoculated plants served as controls. SA-inducible PR1a expression was previously used as a marker of defense signals in tomato plants [65,66]. Expression of three other putative PR genes (PR1, PR3 and PR5) is known to be elevated in symptomatic lettuce leaves after V. dahliae infection [67]. Transcript levels of all tested PR genes were strongly induced in plants infected by the wild type strain of V. dahliae, suggesting that fungal infection activates SA signaling in tomato. PR gene expression in plants infected by the MTF1 complementation strain was comparable to wild-type. In plants inoculated with the MTF1 deletion strain, PR1, PR1a and PR5 expression was similar to mock-inoculated plants and significantly reduced (P < 0.0001) compared with wild-type-infected plants. PR3 expression was decreased to 60% of the wild-type level in plants treated with the MTF1 deletion strain (Fig 9B).
The small, secreted, cysteine-containing protein Scp7 from V. dahliae induces expression of the SA-marker genes PR1 and PR2 in Nicotiana benthamiana, suggesting activation of the SA signaling pathway [68]. We examined SCP7 transcript levels during xylem sap cultivation in the MTF1 deletion strain, which cannot fully induce PR gene expression in tomato plants. In the absence of MTF1, SCP7 transcript levels were reduced to 40% of wild-type (Fig 9C). SCP7 expression was similar to the wild-type in the MTF1 complementation strain. These results suggest that Mtf1 induces gene expression of the Scp7 effector, leading to the induction of PR gene expression in tomato plants.
An additional phytohormone analysis was performed on tomato leaves after root treatment with water or V. dahliae spores to emphasize the analysis of PR gene expression. Wild-type, MTF1 deletion and complementation strains were used and the relative signal area per fresh weight was calculated. We analyzed levels of Pip, which induces resistance of tomato plants to Botrytis cinerea and Pseudomonas syringae pv. tomato [69] and levels of SA, which functions in tomato defense against various pathogens [70]. Plant infection by V. dahliae wild-type or MTF1 complementation strains resulted in elevated Pip and SA levels in systemic tissue compared with mock-inoculated plants. In contrast, after treatment with spores of the MTF1 deletion strain, Pip and SA levels were comparable to water treatment and significantly lower complementation (MTF1-C), MTF1 deletion (ΔMTF1) and other gene deletion strains, which were cultured in extracted tomato xylem sap for 8 h following preculture in simulated xylem medium. MTF1 expression was normalized to the transcript levels of reference genes H2A and EIF2B and gene expression levels in wild-type were set to one. Significant differences from wild-type were determined using t-tests. Means of three independent experiments are shown with error bars representing the SE of the means. MTF1 expression was not measured in ΔMTF1 ( ���� , P < 0.0001) but was wild-type-like in MTF1-C (ns, not significant). MTF1 transcript levels were not significantly increased in the SOM1 deletion strain (ΔSOM1). In the VTA2 deletion strain (ΔVTA2), expression of MTF1 was induced eight-fold, and six-fold in the VTA3 deletion strain (ΔVTA3) ( � , P < 0.05). (b) The schematic displays the control network of pathogenesis-related transcription factors in V. dahliae. Connections indicate direct or indirect control. Som1 induces expression of VTA3 and VTA2, whereas Vta3 induces VTA2 expression [17]. MTF1 expression is reduced in the presence of VTA3 (Figs 1A, 1C and 8A) and VTA2 (Fig 8A), respectively. https://doi.org/10.1371/journal.ppat.1011100.g008

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection

Fig 9. Expression of fungal effector genes and pathogenesis-related protein (PR) genes in tomato plants is increased in the presence of
Verticillium dahliae MTF1. Transcript levels were analyzed by reverse transcription-quantitative PCR and gene expression levels in wild-type were set to one. Significant differences from wild-type (WT) were determined using t-tests. Means of three independent experiments are shown with error bars representing the SE of the means. (a,c) Strains were cultured in extracted tomato xylem sap for 8 h following preculture in simulated xylem medium. Means of three biological replicates normalized to the transcript levels of reference genes H2A and EIF2B are shown. (a) In the absence of MTF1 (ΔMTF1), expression of the virulence factor Ave1-encoding gene was reduced by almost 80% compared with wild-type and the MTF1 complementation strain (MTF1-C) ( � , P < 0.05; ns, not significant). (b) Expression levels of tomato PR genes were quantified in hypocotyls 21 days after inoculation with V. dahliae JR2 wild-type, ΔMTF1 or MTF1-C spores. Water-treated plants (mock) served as controls. Hypocotyls from 13 to 15 plants per treatment were pooled (n = 1). Means of six biological replicates normalized to the transcript levels of reference genes EF1α and αTUB are shown. The expression levels of PR1, PR1a, PR3 and PR5 did not differ significantly between wild-type and MTF1-C-infected plants. In plants treated with water or ΔMTF1, the expression levels of PR genes were significantly lower ( �� , P < 0.01; ��� , P < 0.001; ���� , P < 0.0001). (c) The transcript level of the Scp7 effector-encoding gene was significantly reduced in ΔMTF1 to 40% of the wild-type level. https://doi.org/10.1371/journal.ppat.1011100.g009

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection compared with wild-type (P < 0.001) or complementation strains (Fig 10). Taken together, these data demonstrate that infection by V. dahliae requires an intact MTF1 gene to activate defense responses in tomato plants.

Discussion
Vta3 exhibits a dual regulatory function important for late stages of the V. dahliae life cycle, namely plant xylem colonization and resting structure formation for survival in the soil. Fungal pathogens require signaling pathways for successful plant colonization allowing adaptation to changing environmental stimuli, such as host immune responses [71]. V. dahliae invades through the roots and colonizes the host's xylem vessels [1,2]. The fungus senses differences in its environment and responds specifically to plant xylem sap by adjusting its secretion [14]. Transcription factors are important downstream elements in the signaling cascades that control fungal cellular responses to the plant environment through expression of genes required for adaptation [71]. Characterization of transcriptional regulators in phytopathogens and identification of their targets therefore contributes to a better understanding of fungal virulence and interactions with host plants.
In this work, we studied Vta3 genetic networks during xylem sap cultivation of V. dahliae. Vta3 induces expression of FLO1 and FLO11 encoding adhesins, thereby rescuing adhesion in a non-adherent yeast strain. In V. dahliae, VTA3 is required for root colonization and tomato plant infection. The transcriptional regulator Som1 promotes adhesion, is required earlier during root penetration and induces VTA3 expression. Som1 and Vta3 both promote conidiation and vegetative growth [17]. Since growth and spore release are crucial for plant colonization Phytohormones were analyzed in leaves 21 days after root treatment with water (mock) or spores of indicated strains. The relative signal area per fresh weight (FW) was calculated and normalized to wild-type (set to one). Means of 15 plants per treatment from two independent experiments are shown with error bars representing SD. Significant differences from wild-type (WT) were determined using t-tests. Plant infection by wild-type or MTF1 complementation (MTF1-C) strains led to the induction of pipecolic acid and salicylic acid levels compared with mock, whereas treatment with spores of the MTF1 deletion strain (ΔMTF1) resulted in mock-like phytohormone levels that were significantly lower compared with wild-type (ns, not significant; �� , P < 0.01; ��� , P < 0.001; ���� , P < 0.0001). https://doi.org/10.1371/journal.ppat.1011100.g010

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection [1], these could influence virulence. Nonetheless, Som1 and Vta3 induce gene expression of putative virulence factors, such as Cap20, Nlp2 and the Pr1-like protein Pry1. Vta3 and Som1 also induce gene expression of the transcriptional regulator Vta2, which positively affects conidiation and systemic infection via plant roots [17]. Vta2 affects gene expression of putative secreted proteins and adhesins potentially involved in early plant infection [16]. This study revealed 1,179 genes attributed to various cellular functions that depend on Vta3 for expression (Fig 1A). Vta3 induces gene expression of uncharacterized as well as known candidates, such as the pigment biosynthesis protein Ayg1 [72], the transcription factor Sge1 [73] and the SnodProt1-like protein Cp1 [54], all involved in virulence.
None of the Vta3-dependently reduced transcripts could be assigned a virulence function. However, we identified the transcription factor Mtf1 as a new member reduced in its expression dependent on Vta3 within a network important for disease induction in tomato plants. MTF1 is dispensable for early colonization and penetration of plant roots but is important in xylem sap for full disease development (Figs 6 and 7). The MTF1 deletion strain can grow in tomato plants but appears to be less virulent than wild-type. We found no growth or conidiation defect of MTF1 deletion strains (Figs 5-7 and S9), suggesting that the less severe disease symptoms in tomato plants are due to a specific virulence effect of Mtf1, possibly through regulation of effector gene expression. In this regard, we showed that AVE1 and SCP7 expression is induced in xylem sap in the presence of MTF1 (Fig 9A and 9C). Ave1 increases V. dahliae aggressiveness on plants lacking the Ve1 resistance gene. In tomato plants carrying Ve1, the receptor-like cell-surface protein Ve1 recognizes Ave1 and triggers a hypersensitive response resulting in resistance. It was hypothesized that Ave1 may increase water flux in the xylem, potentially facilitating host colonization by V. dahliae [64]. Scp7 is secreted by V. dahliae, targets the plant nucleus and induces immunity via SA and jasmonate signaling. The authors hypothesize that there may be a plant R protein that recognizes Scp7, which triggers an immune response [68]. Infection by biotrophic or hemibiotrophic pathogens such as V. dahliae usually induces SA signaling as part of plant immunity [74]. A few other V. dahliae effectors capable of modulating plant immune responses have been described. For example, the glycoside hydrolase 11 family member 424Y, like Scp7, localizes to the host nucleus, where it activates SA and jasmonate signaling [75]. The apoplastic effector Cut11 induces defense responses in N. benthamiana, cotton and tomato [76]. Scp41 and the fungal isochorismatase Ics1 can suppress plant immunity during host colonization [74,77]. Although Mtf1 is not involved in expression of these effector genes (S10A Fig), plants colonized by the MTF1 deletion strain show no defense reactions as PR gene expression is not fully induced ( Fig 9B) and levels of the plant defense hormones Pip and SA are comparable to uninfected plants (Fig 10). In tomato, lettuce or A. thaliana, PR1, PR1a, PR3 and PR5 expression levels are elevated after pathogen infection and a defense function is assumed [65][66][67]78]. According to the 'growthdefense tradeoff' hypothesis, plants treated with the MTF1 deletion strain may also grow taller because they do not invest in defense [79]. MTF1 transcription is reduced in a Vta3-dependent manner, consequently VTA3 deletion strains overexpress MTF1 (Figs 1C and 8A). Plants treated with the VTA3 deletion strain exhibit no defense reactions (PR gene expression is not induced, S11 Fig), as deletion of VTA3 leads to a growth defect of these strains on plant roots, resulting in failure already at early infection stages [17].
To our knowledge, a role of Mtf1 in virulence of phytopathogenic fungi has not previously been described. In Aspergillus spp., MtfA is known for its functions in morphogenesis and secondary metabolism [56][57][58]. The cellular functions of Mtf1 in the soil-borne plant pathogen V. dahliae and its targets are summarized in Fig 11. MTF1 belongs to a genetic network with the transcriptional regulators Vta3 and Vta2, as the presence of both VTA3 and VTA2 reduces MTF1 transcription (Fig 8). Mtf1 probably acts genetically downstream of Som1, Vta3 and Vta2, whose gene expression is in turn not controlled by Mtf1 (S10B and S10C Fig). Vta3 and Mtf1 also control their own distinct targets. For example, AVE1 and SCP7 expression is induced in the presence of MTF1, whereas Vta3 does not affect their transcript levels in xylem sap (S12 Fig). Although Mtf1 and Vta3 are both involved in microsclerotia formation (Fig 5) and plant disease (Fig 7), the pathways differ as many factors involved in these processes are affected by Vta3 [17] but not by Mtf1 (S10B and S10C Fig). For example, VTA3 presence induces gene expression of the Egh16-like virulence factor Elv1 (Fig 1A and 1B), which does not change significantly in the absence or presence of MTF1 (S10B Fig). Egh16-like virulence factors are found in many pathogenic filamentous fungi (S5A Fig) and are thought to play a role in host interaction. Homologs are present in entomopathogenic fungi [80,81], in M. oryzae [55] and in the obligate biotrophic fungus B. graminis f. sp. hordei, where their gene expression is induced upon penetration into barley leaves [60,82]. In powdery mildew fungi, different Egh16-like virulence factors are involved to varying degrees in fungal growth, the infection process and host defense. Chitinase activity was suggested as possible mode of action suppressing chitin perception and plant immunity [83]. We found that Elv1, as one of several

PLOS PATHOGENS
Elv1 virulence factor and Mtf1 regulator functions in plant infection factors whose gene expression is Vta3-dependently induced in V. dahliae, enhances tomato plant disease development (Fig 4) without affecting fungal morphogenesis (Fig 3) and PR gene expression (S6 Fig). The global regulator Vta3 is involved in the control of several processes. VTA3 deletion strains already fail during initial colonization of plant roots [17]. Overexpression of ELV1 is presumably not sufficient to overcome the virulence deficiency of VTA3 deletion strains, because an intact VTA3 gene is required for hyphae to reach the plant xylem sap, where ELV1 expression is induced.
In summary, we describe ELV1 and MTF1 as novel targets of the developmental and pathogenicity regulator Vta3 in the vascular plant pathogen V. dahliae. Further experiments could help to clarify the complex regulatory networks and the mode of action of the regulatory proteins. For example, possible direct binding of Vta3 or Mtf1 in the promoter region of potential target genes could be investigated using chromatin immunoprecipitation assays. Protein pulldowns were performed with GFP-fused Vta3 to investigate whether there is evidence for a direct interaction of the regulators Vta3 and Mtf1. V. dahliae was cultured in liquid PDM for five days when both proteins were visible in western experiments (S7 Fig). Ribosomal proteins predominantly co-enriched with Vta3, and the data set did not include Mtf1 as interaction partner of Vta3 (S13 Fig, S10 Table).
The Egh16-like virulence factor Elv1 contributes to virulence, while the regulatory protein Mtf1 is required for late stages of plant infection and host immune responses. Furthermore, Mtf1 promotes microsclerotia formation enabling fungal survival in the soil and re-infection of host plants. Our data indicate that Vta3 is particularly important as a transient factor that initiates processes and terminates others during fungal growth in plant xylem sap and in preparing for survival in the soil after plant death. In future studies, it will be interesting to compare the expression patterns of ELV1, MTF1 and VTA3 during infection of tomato plants by V. dahliae. This will require more sensitive approaches, as a first set of pilot experiments with different plant parts and at different time-points after inoculation of the plants with fungal spores resulted in such low levels of fungal RNA compared with plant RNAs that expression of ELV1, MTF1 and VTA3 could not be detected. Fungal transcription factors are not encoded by highly transcribed genes and can be highly instable [84]. Furthermore, it will be important to determine whether inhibiting Vta3-controlled components, such as Elv1 and Mtf1, leads to development of new strategies reducing both fungal survival in the plant and its decades-long survival in the soil, thereby better mitigating the currently increasing Verticillium-caused agricultural damage.