Role of Exopolygalacturonase-Related Genes in Potato-Verticillium dahliae Interaction

Verticillium dahliae is a hemibiotrophic pathogen responsible for great losses in dicot crop production. An ExoPG gene (VDAG_03463,) identified using subtractive hybridization/cDNA-AFLP, showed higher expression levels in highly aggressive than in weakly aggressive V. dahliae isolates. We used a vector-free split-marker recombination method with PEG-mediated protoplast to delete the ExoPG gene in V. dahliae. This is the first instance of using this method for V. dahliae transformation. Only two PCR steps and one transformation step were required, markedly reducing the necessary time for gene deletion. Six mutants were identified. ExoPG expressed more in the highly aggressive than in the weakly aggressive isolate in response to potato leaf and stem extracts. Its expression increased in both isolates during infection, with higher levels in the highly aggressive isolate at early infection stages. The disruption of ExoPG did not influence virulence, nor did it affect total exopolygalacturonase activity in V. dahliae. Full genome analysis showed 8 more genes related to polygalacturonase/pectinase activity in V. dahliae. Transcripts of PGA increased in the △exopg mutant in response to potato leaf extracts, compared to the wild type. The expression pattern of those eight genes showed similar trends in the △exopg mutant and in the weakly aggressive isolate in response to potato extracts, but without the increase of PGA in the weakly aggressive isolate to leaf extracts. This indicated that the △exopg mutant of V. dahliae compensated by the suppression of ExoPG by activating other related gene.


Introduction
Verticillium dahliae is a hemibiotrophic pathogen that causes wilt symptoms and results in great losses in a wide range of dicot hosts [1], including in important economic crops [2]. Traditional control management strategies, such as crop rotation and the use of green manure, are not very effective in preventing Verticillium wilt [3,4]. Soil fumigation results in effective Verticillium wilt reduction, but has a negative impact on the environment [5][6][7]. Genetic resistance has not been very useful so far, i.e., although tomato lines contain the Ve resistance gene, it only has resistance to V. dahliae race 1 [8,9], and cultivars with effective resistance have not been developed in the majority of the crops at risk.
The plant cell wall is a key barrier of protection from pathogen attacks. It consists primarily of cuticular wax, cutin, glycans, cellulose, pectic substances, and cell wall structural proteins [10]. However, many microbes produce various cell wall-degrading enzymes (CWDEs) to degrade these plant cell-wall components. CWDEs include cutinases, pectinases, cellulases, glucanases and proteases [10]. Glycosidic bonds of polysaccharides can be cut by enzymes belonging to the glycoside-hydrolases (GH) family [11]. Polygalacturonases belong to these pectin-degrading enzymes, which include both endopolygalacturonases (EndoPGs) and exopolygalacturonases (ExoPGs). EndoPGs hydrolyze the polysaccharide randomly to produce oligogalacturonides, while ExoPGs hydrolyze the polymer from the

ExoPG Expression against Potato Extracts
The ExoPG gene's response was measured under treatment with extracts from potato leaves, stems and roots. As shown in Figure 1, ExoPG responded more in the highly aggressive isolate Vd1396-9 to potato leaf and stem extracts, compared to in the weakly aggressive isolate Vs06-07. However, gene expression in both highly and weakly aggressive isolates were up-regulated in response to root extracts with no significant difference between the two. The expression of the ExoPG gene during infection of potato detached leaves w significantly higher in the highly aggressive isolate Vd1396-9 at 3 DAI, when compare to the weakly aggressive one (Figure 1).

Mutant Production
Following the procedure showed in Figure 2, the 715 bp from the upstream regio and the 713 bp from the downstream region of the selected ExoPG, as well as the 789 b from the 5' segment and 909 bp from the 3' segment of Hygromycin resistant gene (Hp gene were both successful amplified by PCR ( Figure 3). Both 5' terminator of primers F AR and FS-BF were flanked with a 27 bp nucleotide sequence from 5' and 3' terminator the Hph gene sequence, respectively. The PCR product of the upstream region of the s lected ExoPG region and the 5' segment of the Hph gene were used as templates to use t overlapping PCR ( Figure 2; Table 1). During the first eight cycles of the overlapping PC these two DNA fragments used each other's overlapping regions as a "primer" to elonga a new DNA fragment. The primer (Primer: FS-AF/ HY-R; Table 1) can attach to both en of the new fragment, to enrich the new DNA fragment ( Figure 2; Table 1). The same pr cedure was performed with the PCR product of the downstream region of the selecte ExoPG region and 3' segment of Hph gene with primer YG-F/ FS-BR ( Figure 2; Table 1 an 2). As shown in Figure 3, the 1486 bp from DNA fusion of the upstream region of th selected ExoPG region and the 5' segment of Hph gene, the 1622 bp from DNA fusion the downstream region of the selected ExoPG region and the 3' segment of the Hph gen were both successfully amplified. After transforming both fused DNA from overlappin PCR into the protoplast by PEG-mediated transformation (Figure 2), seven positi knock-out transformants were obtained from 40 transformants identified by PCR (Figu 4). In addition, six single hygromycin gene replacement mutants of the seven tran formants were identified by Southern blot (Figure 4). The mutants △exopg-ko-18 and △e opg-ko-23 were randomly chosen for subsequent experiments. The expression of the ExoPG gene during infection of potato detached leaves was significantly higher in the highly aggressive isolate Vd1396-9 at 3 DAI, when compared to the weakly aggressive one (Figure 1).

Mutant Production
Following the procedure showed in Figure 2, the 715 bp from the upstream region and the 713 bp from the downstream region of the selected ExoPG, as well as the 789 bp from the 5 segment and 909 bp from the 3 segment of Hygromycin resistant gene (Hph) gene were both successful amplified by PCR ( Figure 3). Both 5 terminator of primers FS-AR and FS-BF were flanked with a 27 bp nucleotide sequence from 5 and 3 terminator of the Hph gene sequence, respectively. The PCR product of the upstream region of the selected ExoPG region and the 5 segment of the Hph gene were used as templates to use the overlapping PCR ( Figure 2; Table 1). During the first eight cycles of the overlapping PCR, these two DNA fragments used each other's overlapping regions as a "primer" to elongate a new DNA fragment. The primer (Primer: FS-AF/ HY-R; Table 1) can attach to both ends of the new fragment, to enrich the new DNA fragment ( Figure 2; Table 1). The same procedure was performed with the PCR product of the downstream region of the selected ExoPG region and 3 segment of Hph gene with primer YG-F/ FS-BR ( Figure 2; Tables 1 and 2). As shown in Figure 3, the 1486 bp from DNA fusion of the upstream region of the selected ExoPG region and the 5 segment of Hph gene, the 1622 bp from DNA fusion of the downstream region of the selected ExoPG region and the 3 segment of the Hph gene, were both successfully amplified. After transforming both fused DNA from overlapping PCR into the protoplast by PEG-mediated transformation (Figure 2), seven positive knock-out transformants were obtained from 40 transformants identified by PCR ( Figure 4). In addition, six single hygromycin gene replacement mutants of the seven transformants were identified by Southern blot (Figure 4). The mutants exopg-ko-18 and exopg-ko-23 were randomly chosen for subsequent experiments.        (10). 58 • C for 90 seconds (11). 72 • C for 120 s Repeat step 8 to 10 for 30 cycles (12). 72 • C for 10 min (13). 4 • C As shown in Figure 5, there were no significant differences in the growth, conidiation, or formation of microsclerotia between the exopg mutants ( exopg-ko-18, exopg-ko-23) and wildtype Vd1396-9.

Pathogenicity
To determine the difference in aggressiveness between the exopg mutants and wildtype ( Figure 6), all isolates were inoculated onto the susceptible potato cultivar Kennebec. There was no dramatic change in total area under disease progress curve (Total AUDPC) for infection or disease severity, or plant height and vascular discoloration measurements.  represents the wild type Vd1396-9.   Hph gene Hph-R [27] CTATTCCTTTGCCCTCGGACGAGT 55 Hph gene YG-F [28] GATGTAGGAGGGCGTGGATATGTCCT 55 Hph gene HY-R [28] GTATTGACCGATTCCTTGCGGTCCGAA 55 Hph gene Note: Hph gene: hygromycin resistant gene.

Pathogenicity
To determine the difference in aggressiveness between the △exopg mutants and wildtype ( Figure 6), all isolates were inoculated onto the susceptible potato cultivar Kennebec. There was no dramatic change in total area under disease progress curve (Total AUDPC) for infection or disease severity, or plant height and vascular discoloration measurements.  and colony phenotype (C) of wildtype (Vd1396-9) and mutant (△exopg-ko-18, △exopg-ko-23 and ExoPG-Ect-44) strains of V. dahliae. The bar graphs depict mean values (n = 8 for growth rate experiment, and n=5 for conidiation experiment) ± standard error. Error bars refer to standard error. For each parameter, mean values marked by the different letters are significantly different according to multiple comparison (Fisher's LSD) test (p < 0.05).

Total exopolygalacturonases Activity of △exopg Mutant
To determine the influence of ExoPG's disruption in total exopolygalacturonase activity, the supernatant from cultured isolates (△exopg-ko-18, △exopg-ko-23, weakly aggressive isolate Vs06-07 and wildtype Vd1396-9) were collected over time and assayed for enzyme activity. Pectin was added to the CDB medium to stimulate the production of Ex-oPG. As shown in Figure 7, the total exopolygalacturonase activity in all tested isolates under the elicitation of pectin was significantly higher than the control group, however,

Total Exopolygalacturonases Activity of exopg Mutant
To determine the influence of ExoPG's disruption in total exopolygalacturonase activity, the supernatant from cultured isolates ( exopg-ko-18, exopg-ko-23, weakly aggressive isolate Vs06-07 and wildtype Vd1396-9) were collected over time and assayed for enzyme activity. Pectin was added to the CDB medium to stimulate the production of ExoPG. As shown in Figure 7, the total exopolygalacturonase activity in all tested isolates under the elicitation of pectin was significantly higher than the control group, however, there was no significant differences between the exopg mutants and wildtype at almost all time points. On the contrary, the total exopolygalacturonase activity of weakly aggressive isolate Vs06-07 after treatment with pectin was significantly lower than wildtype Vd1396-9 ( Figure S1). there was no significant differences between the △exopg mutants and wildtype at almost all time points. On the contrary, the total exopolygalacturonase activity of weakly aggressive isolate Vs06-07 after treatment with pectin was significantly lower than wildtype Vd1396-9 ( Figure S1). Graph points represent mean values (n = 3, with three technical replicates for each biological replicate) ± standard error. For each strain, means labelled by different letters are significantly different between treatments according to multiple comparison (Fisher's LSD) test (p < 0.05). Error bars refer to standard error.

Identification of Other Polygalacturonase/Pecninase Coding Genes and Expression Pattern of these Genes
Polygalacturonase/pectinase-related genes included in the genome of V. dahliae were designated (i.e., Polygalacturonase A (PGA) to PGD and Pectinase 1 (PEC1) to PEC4). Since disruption of the ExoPG coding gene did not affect the virulence of V. dahliae, the expression levels of these genes (Accessions #: VDAG_07608, VDAG_02879, VDAG_08097, VDAG_05992, VDAG_09366, VDAG_08098, VDAG_01781, and VDAG_00768) were determined in wild type, weakly aggressive isolate (Vs06-07) and the mutant △exopg, with elicitation of Kennebec potato leaves or stems extracts. Most of the genes in water control treatment were down-regulated in the mutant △exopg compared to the wild type ( Figure 8). When responding to both or one of the potato extracts, the expression level of most of the genes, except PGA and PEC1, showed significant a decrease in both isolates, compared to that in wild type with water control treatment ( Figure 8).
However, the expressions of PGA showed a drastic increase in the mutant △exopg compared with the wild type, in response to potato leaf extracts ( Figure 8). The expression of PGA in response to stem extracts, and PEC1 to leaf extracts, both showed more significant increases in the wild type than in the mutant △exopg ( Figure 8). Even given that most genes expression in both wild type and mutant were down-regulated in response to potato extracts, the expression level of PGC, PGD, PEC2 and PEC3 show a higher level in the mutant △exopg than the wild type in response to one or both potato extracts (Figure 8). Interestingly, PGA and PGD showed a significantly lower level in the mutant △exopg than Figure 7. Total exopolygalacturonases activity of exopg mutants ( exopg-ko-18, exopg-ko-23) and wildtype (Vd1396-9). Graph points represent mean values (n = 3, with three technical replicates for each biological replicate) ± standard error. For each strain, means labelled by different letters are significantly different between treatments according to multiple comparison (Fisher's LSD) test (p < 0.05). Error bars refer to standard error.

Identification of Other Polygalacturonase/Pecninase Coding Genes and Expression Pattern of These Genes
Polygalacturonase/pectinase-related genes included in the genome of V. dahliae were designated (i.e., Polygalacturonase A (PGA) to PGD and Pectinase 1 (PEC1) to PEC4). Since disruption of the ExoPG coding gene did not affect the virulence of V. dahliae, the expression levels of these genes (Accessions #: VDAG_07608, VDAG_02879, VDAG_08097, VDAG_05992, VDAG_09366, VDAG_08098, VDAG_01781, and VDAG_00768) were determined in wild type, weakly aggressive isolate (Vs06-07) and the mutant exopg, with elicitation of Kennebec potato leaves or stems extracts. Most of the genes in water control treatment were down-regulated in the mutant exopg compared to the wild type ( Figure 8). When responding to both or one of the potato extracts, the expression level of most of the genes, except PGA and PEC1, showed significant a decrease in both isolates, compared to that in wild type with water control treatment ( Figure 8). However, the expressions of PGA showed a drastic increase in the mutant exopg compared with the wild type, in response to potato leaf extracts (Figure 8). The expression of PGA in response to stem extracts, and PEC1 to leaf extracts, both showed more significant increases in the wild type than in the mutant exopg (Figure 8). Even given that most genes expression in both wild type and mutant were down-regulated in response to potato extracts, the expression level of PGC, PGD, PEC2 and PEC3 show a higher level in the mutant exopg than the wild type in response to one or both potato extracts (Figure 8). Interestingly, PGA and PGD showed a significantly lower level in the mutant exopg than in the wild type under elicitation of stem extracts, but the opposite occurred after elicitation with leaf extracts (Figure 8). Our results indicate that different other polygalacturonase/pectinase genes may compensate that of ExoPG transcripts under different conditions, and among them, PGA might play a significant role for compensation ( Figure 8).
trends were seen in the mutant △exopg and weakly aggressive isolate in response to potato extracts (Figure 8 and 9). However, the PGA is interesting among all these genes, as it showed an obvious increase in the mutant △exopg exposed to the leaf extracts, but did not significantly change in weakly aggressive isolate, when both are compared to highly aggressive wild type (Figure 8 and 9). The expressions of PGC and PEC3 in weakly aggressive isolate were significantly increased, but this was not the case in the mutant △exopg (Figures 8 and 9). The expression of PGC and PEC3 showed a more significant increase in the weakly aggressive isolate than the highly aggressive wild type stain when exposed to stem extracts ( Figure 9). The number of transcripts of PGB, PGC, PGD and PEC3 in weakly aggressive strain were significantly higher than highly aggressive wild type when exposed to one or both extracts (Figure 9). On the other hand, the expression levels of PGA in response to stem extracts and PEC1 to leaf extracts both increased drastically in the highly aggressive wild type compared to in the weakly aggressive isolate (Figure 9).
When comparing the expression patter of all eight polygalacturonase/pectinase genes, between the highly aggressive wild type and the mutant exopg, as well as between the highly aggressive wild type and weakly aggressive isolate, in most cases, similar trends were seen in the mutant exopg and weakly aggressive isolate in response to potato extracts (Figures 8 and 9). However, the PGA is interesting among all these genes, as it showed an obvious increase in the mutant exopg exposed to the leaf extracts, but did not significantly change in weakly aggressive isolate, when both are compared to highly aggressive wild type (Figures 8 and 9). The expressions of PGC and PEC3 in weakly aggressive isolate were significantly increased, but this was not the case in the mutant exopg (Figures 8 and 9).
Gene disruption in fungi, by either Agrobacterium-mediated transformation or PEGmediated protoplast transformation, often requires vector construction [18,23,34,35]. In particular, Agrobacterium-mediated transformation requires a special vector in the procedure [33,36,37]. Constructing a vector is time-consuming and requires many steps. Here, we employed a vector-free split-marker recombination method for knocking out target genes in V. dahliae, as previously carried out in S. cerevisiae [20] and M. oryzae [19]. This method can be easily processed by only two steps of PCR, and therefore can typically be performed in one day. This method can substantially reduce the time and cost when compared to at least one-week requirement for constructing a vector for a targeted gene. Combining both vector-free split-marker recombination methods with PEG-mediated pro-toplast transformation reduces the total time for manipulation, including transformation, to as little as three days. This is drastically shorter than the time needed for vector-based Agrobacterium-mediated transformation. In addition to time consideration, this method could reduce the number of transformants that contain randomly inserted exotic DNA in the fungal genome. If each of the two fusion DNA fragments (flanking DNA with half marker gene) were separately inserted into the fungal genome, the transformants cannot grow on the selection media, because of the non-intact marker gene. Only when the two fusion DNA fragments are inserted in the proper place can the homologous recombination occur between overlapping regions of the two split-markers. The overlapping PCR in this study is a convenient and quick method to fuse different DNA fragments together, without finding or adding appropriate restriction enzyme sites in the edge of different DNA fragments.
Pathogens can overcome plant cell wall protection by producing cell wall-degrading enzymes (CWDEs) [10]. In A. citri, an EndoPG gene is required for pathogenicity [15]. El-Bebany et al. [16] found that when V. dahliae responds to root extracts from both moderately resistant and susceptible potato cultivars, the expression of ExoPG (VDAG_03463) was up-regulated in the highly aggressive V. dahliae isolate (Vd1396-9), but down-regulated in a weakly aggressive one (Vs06-14) [16]. Based on these results, we wanted to investigate if the function of this ExoPG was involved in pathogenicity or interaction with the host. The degree of up or down-regulation in V. dahliae isolates with differential aggressiveness levels were not provided by El-Bebany et al. [16], since the experiment was conducted using a subtractive hybridization cDNA-AFLP method.
In this study, we determined the fold-change of ExoPG expression in differentially aggressive V. dahliae isolates in response to potato leaf, stem, or root extracts, and during plant tissue infection. ExoPG responded more in the highly aggressive isolate Vd1396-9 to potato leaf and stem extracts, compared to in the weakly aggressive one. ExoPG was up-regulated in both isolates elicited with root extracts. However, ExoPG was up-regulated more in the highly aggressive isolate than in the weakly aggressive one at early infection stages. This may indicate that this gene may be activated by potato extracts in both isolates, but at a higher level in the highly aggressive one, and therefore may not play a primary role in infection. Subsequent comparison of the exopg mutants and wildtype proved that ExoPG is not indispensable for V. dahliae's pathogenicity. The total exopolygalacturonase activity in V. dahliae did not change due to the disruption of the ExoPG gene. However, when compared to the weakly aggressive strain vd06-07 of V. dahliae, wildtype isolate showed much higher ExoPG activity. So, it seems that other genes or pathways rather than the knocked out gene might help with the pathogenicity of highly aggressive isolate of V. dahliae.
Comparison of the expression patterns of the other polygalacturonases/pectinases indicates that most of those genes showed no difference or decreased in the mutant exopg, compared to wild type without elicitation, but PGA showed a more significant increase in the mutant exopg than the wild type under elicitation with potato leaf extracts. Even though the expression of most of the genes in response to potato extracts showed downregulation in both mutant and wild type, most genes still showed relatively higher levels in the mutant compared to the wild type. Similar trends in most genes were observed between weakly and highly aggressive isolates, except for PGA which does not show the same increase as that in mutant in response to leaf extracts. This indicates that these polygalacturonases/pectinases coding genes, PGA in particular, may compensate for the function of the knocked out gene in the exopg mutant during interaction with potato, and this particular ExoPG is not a crucial virulence factor in V. dahliae. The increased activity of PGA in the mutant could also explain the reason why disruption of ExoPG had no effect on the virulence of transformants. Interestingly, although most of the polygalacturonases/pectinases showed a similar response to potato leaf and stem extracts in the mutant exopg, activity of PGA showed an opposite trend with potato leaf or stem extracts in the mutant, indicating that those polygalacturonases/pectinases may also have various interactions in response to infection of different host tissues. In Fusarium oxysporum, three polygalacuronase PG1, PG5, PGX4 were identified, and none of these had an effect on virulence [38][39][40]. In Aspergillus niger, deletion of an exopolygalacturonase gene, PGXB, caused partial reduction in virulence on apple fruits, and six other polygalacturonase genes showed a higher level of expression in ∆pgxB mutant than in wild type [41]. In Ralstonia solanacearum, disruption of polygalacturonases, endoPG PehA or ExoPG PehB, could both reduce virulence on eggplant and tomato [42]. Both PehA and PehB function quantitatively, and play important roles in rapid colonization in the host's vascular system [42]. In rice pathogen Burkholderia glumae, knocking out of polygalacturonase genes PehA or PehB could not affect the virulence on rice [43]. These studies showed species-dependent results by polygalacturonases. It seems that some polygalacturonases may play the most important roles in pathogenicity, while others do not. V. dahliae possesses many polygalacturonases, but obviously, the ExoPG tested in this study is not important for virulence, as other polygalacturonases, especially PGA, could compensate for its function. It will be interesting to investigate the functions of the PGA in V. dahliae, because it showed critically higher expression in exopg than the wild type under elicitation with potato leaf extracts, and this did not exhibit in weakly aggressive isolate. One or more of these polygalacturonases may play a more central role for virulence under other conditions.

Fungi Isolates and Plant Material
Highly aggressive Vd1396-9 and weakly aggressive Vs06-07 V. dahliae isolates, and susceptible potato cultivar Kennebec were used in this study [44,45]. Kennebec seedlings (germinated for 10 days) were planted in a soil, sand, and peat moss mixture with a ratio of 12:4:1, with day/night temperatures of 22/18 • C and 16/8 h photoperiod in growth cabinet.

ExoPG Expression in Response to Potato Tissue Extracts and on Inoculated Detached Leaves
The expression of ExoPG in the differentially aggressive V. dahliae isolates Vd1396-9 and Vs06-07 in response to elicitation with potato leaf, stem, and root extracts were performed as described by Zhu et al. [18]. Briefly, the potato leaf, stem, or root extracts were added into 7-day cultured V. dahliae isolates Vd1396-9 and Vs06-07, respectively, and after an additional one-week culturing, the fungal material was collected for RNA extraction.
The expression of ExoPG in the same differentially aggressive V. dahliae isolates inoculated onto potato detached leaves was also assessed as described by Zhu et al. [18]. In brief, V. dahliae isolates Vd1396-9 and Vs06-07 were cultured on potato dextrose agar (PDA) for 3 weeks prior to harvesting the conidia by water flooding. Four-week-old detached leaves were inoculated by immersing their petioles into 1 mL of conidial suspensions of Vd1396-9 or Vs06-07, and sterilized water was used as a control. Four to six individual detached leaves from different plants were combined as one sample, and three samples were collected for each treatment at each selected time-point (3,12, and 17 days after inoculation (DAI)). RNA extraction and Quantitative Real-Time RT-PCR were used here with primers (ExoPG-QRT-F/R; Table 2) following the protocol from Omega Fungal RNA kit (Omega Bio-Tek, Inc., Norcross, GA, USA) and SsoFast EvaGreen Super mix (Bio-Rad Lab, Philadelphia, PA, USA). The cycle parameters were as follows: initial denaturation at 95 • C for 30 s, followed by 40 cycles including denaturation at 95 • C for 5 s, then heating up to 60 • C for 30 s (annealing/extension); after that, a melting curve was generated by heating from 65 • C to 95 • C with increases of 0.5 • C and 5 s dwell time, as well as a plate read at each temperature.

Protoplast Preparation and Transformation
Preparation of the protoplast cells of V. dahliae (highly aggressive isolate Vd1396-9) was conducted following the description of Dobinson [25] and Yelton et al. [46], with some improvements. In brief, V. dahliae was cultured in PDB broth for 4 days at 24 • C without shaking, and the mycelium collected and finely ground using sterile mortar and pestle under a laminar flow hood. The ground mycelium was then re-cultured in fresh PDB broth containing 0.001% thiamine for an additional 14 hours in the shaker at 120 rpm. Mycelia were collected by filtration on miracloth and washed with mycelia buffer (10 mM NaPO 4 pH 7.5; 10 mM EDTA pH 8.0; 1 mM dithiothreitol) 2-3 times. The mycelium was then re-suspended in 30 mL mycelia buffer and shaken for 2 h at 24 • C at 60 rpm. The culture was then centrifuged at 1900 g for 10 minutes to collect the mycelium, decanted the supernatant, then incubated in 15 mL OM buffer (1.2 M MgSO 4 , 10 mM NaPO 4 pH 5.8) with 10 mg/mL lysing enzymes from Trichoderma harzianum (Sigma-Aldrich Canada Co., Oakville, ON, Canada) and shaken overnight at 30 • C at 65 rpm. The suspension was finally filtered by miracloth and transferred into a 50 mL centrifuge tube, overlayed with 10 mL ST buffer (0.6 M sorbitol; 100 mM Tric-HCl pH 7.0), then centrifuged at 4 • C at 4000× g for 20 min. A glass Pasteur pipet was used to retrieve the 5-10 mL of protoplast cells in the interface between the OM and ST buffers. Two to 4 volumes of STC buffer (1.2 M sorbitol; 10 mM Tric-HCl pH 7.5; 10 mM CaCl 2 ) were added, centrifuged at 4 • C, 4000× g for 20 min. The precipitate was washed 2 times with STC buffer, and finally re-suspended in 0.5 mL STC buffer with 8% PEG (PEG3350) or 8% DMSO, yielding a final protoplast concentration ranging from 3 to 5 × 10 7 protoplast/mL.
We used the transformation protocol of Dobinson [25], with some modifications. An amount of 2 µg of DNA was added to 200 µL protoplast cell and incubated on ice for 20 min, followed by gentle addition of 0.625 mL PTC buffer (40% PEG 3350 in TSTC buffer (1 M sucrose; 50 mM Tris.HCl pH 8.0; 50 mM CaCl 2 )). The solution was mixed well following buffer addition, then incubated at room temperature for 20 min, followed by the addition of 5 mL complete media (CM) [47], containing 1 M sucrose, and shaking at 90 rpm, overnight at 24 • C.

Vector-Free Split-Marker Recombination Method for Knocking Out of ExoPG in V. dahliae
Vector-free split-marker recombination method requiring only two PCR steps followed by one transformation step was used in the present study [19,20]. Our aim was to knock out the front 1319 bp fragment of the ExoPG open reading frame (ORF) (1836 bp), but due to the proximity of the 3 DNA sequence of ExoPG to another gene, we designed the downstream homologous recombination region so as not to interfere with the ORF of the other gene, but to cover a short part of that region with the ExoPG ORF. Briefly, as shown in Figure 2 and 3, the first step was to amplify the flanked upstream and downstream regions of the selected gene from the V. dahliae genome with the primers FS-AF/R and FS-BF/R (Table 2), then amplify the 5 and 3 segments of the Hph from pSK846 vector [30] with the primers Hph-F/HY-R, YG-F/Hph-R (Table 2), respectively. Both primer FS-AR and FS-BF were fused with a short sequence of 5 terminator and 3 terminator of the Hph gene ( Figure 2). The amplified PCR product of the upstream region of the selected gene from V. dahliae and the 5 segment of the Hph gene was used as a template for overlapping PCR (Table 1), in order to obtain the DNA fragment of the upstream region fused with the 5 segment of Hph. Note the annealing temperature for the first 8 cycles of "self-primer" should be low in order to promote this combination. The downstream region of the selected gene from V. dahliae and the 3 segment of the Hph gene was used as a template for overlapping PCR to obtain the DNA fragment of the downstream region fused with the 3 segment of Hph (Figure 2). Finally, both overlapping PCR products of the upstream region fused with 5 segment of Hph and the downstream region fused with the 3 segment of Hph were transformed into protoplast cells of V. dahliae wild type Vd1396-9 ( Figure 2). Only the homolog recombination which arose from both upstream and downstream regions of the selected gene, as well as the overlapping 5 segment and 3 segment of the Hph gene could produce positive transformants ( Figure 2). All the transformants were selected on PDA plates with hygromycin B following the description of Zhu et al. [18]. The knocked out transformants were screened by PCR with primers ExoPG-ORF-F/R, ExoPG-UAF/Hph-TR (Table 2). Positive transformants were confirmed by Southern blot following Zhu et al. [18]. The virulence of exopg mutants were tested on potato cv. Kennebec following the description of Zhu et al. [18]. Briefly, Kennebec plants were grown in LA4 soil mix (SunGro Horticulture, Agawam, MA 01001, USA). After 21 days, plants were up-rooted, roots were washed and 1 cm was trimmed from the root tips, then they were inoculated with 10 6 conidia/mL suspension of exopg mutants ( exopg-ko-18, exopg-ko-23), wild type, Vd1396-9 and ectopic control, ExoPG-Ect-13 (fragment randomly inserted in V. dahliae genome without replacing the original ExoPG ORF) for 60 Sec. Plants were transplanted into 6-inch pots containing a mixture of pasteurized sand, soil and peat moss (16:4:1) and placed back into the controlled growth area for 2 weeks. Total area under the disease progress curve (total AUDPC) of percentage infection and disease severity was calculated weekly according to Zhu et al. [18]. Plant height measurements and vascular discoloration ratings were conducted at 5 weeks post-infection according to Alkher et al. [44].

Growth Rate and Conidiation of exopg Mutants
exopg mutants ( exopg-ko-18, exopg-ko-23), wild type Vd1396-9 and ectopic control ExoPG-Ect-13 were grown on PDA. The colony diameter was measured at 7 days and 16 days to calculate the growth of colony per day. Growth rate per day = (16 days colony diameter-7days diameter)/9 days. After 4 weeks, colony morphology photos were taken and conidia production was assessed according to Zhu et al. [18].

Identification of Other V. dahliae Polygalacturonase Related Genes and Their Expression Pattern
We hypothesized that the suppression of ExoPG may be compensated by the activity of other polygalacturonase/pectinase coding genes. To test this hypothesis, using tBLASTn searches from the CLC genomics workbench (CLCBio, Aarhus Denmark), we could find 8 contigs possibly with polygalacturonase/pectinase genes within the available full genome sequence of V. dahliae [51]. The open reading frames (ORFs) encoded on these contigs (containing putative polygalacturonase/pectinase) were then put in Web AUGUSTUS (http://bioinf.uni-greifswald.de/augustus/, accessed on 1 May 2021) [52] for prediction. For further confirmation, identified ORFs were checked for signature domains of the target polygalacturonase/pectinase genes using InterProScan [53].
To determine the expression profile of the genes involved in polygalacturonase activity, samples were taken from seven-day-old cultures of Vd1396-9 (wildtype), Vd06-07 (weakly aggressive isolate) and exopg under elicitation of Kennebec potato leaves or stems or water (control) extracts following Zhu et al. [18]. RNA was extracted from the freezedried samples and analyzed by quantitative RT-PCR (qRT-PCR) using CFX96 Thermal Cycler (BioRad, Hercules, CA, USA). The expression pattern of 8 different genes with polygalacturonase/pectinase activity (including PGA: VDAG_07608, PGB: VDAG_02879, PGC: VDAG_08097, PGD: VDAG_05992, PEC1: VDAG_09366, PEC2: VDAG_08098, PEC3: VDAG_01781 and PEC4: VDAG_00768) related to polygalacturonase activity was evaluated, using specific primer sets (Table 3) and Syber green qRT-PCR Master Mix kit (Bio-Rad Lab, Irvine, CA, USA) according to the manufacturer protocol. The cycle parameters were as follows: initial denaturation at 96 • C for 4 min, followed by 40 cycles including denaturation at 96 • C for 5 s, followed by heating up to 60 • C for 30 s (annealing/extension). In all applications, the Histone H3 and Actin genes were used as the housekeeping genes ( Table 2). All PCR reactions were carried out in triplicate. The analysis of data was performed using the 2 −∆∆Ct method [54].

Statistical Analysis
PROC MIXED program processed by SAS Statistical Analysis Software (SAS Institute, Cary, NC, USA; release 9.1 for Windows) was used for data analysis in the current study. Some sets of data were applied with Log 10 transformation for statistical analysis when necessary. The PROC UNIVARIATE was used to test the normality, all data qualified for normal distribution with Shapiro-Wilk test >0.9. The test for homogeneity was determined by residual comparison with studentized residuals' critical values [55]. Mean values of all data were separated by least squared means and classified by the macro PDMIX800.sas [56] with α = 0.05 into a bunched letters result. Significant differences between different treatments were shown with totally different letters (p < 0.05).