Ustilago maydis Secreted Endo-Xylanases Are Involved in Fungal Filamentation and Proliferation on and Inside Plants

Plant pathogenic fungi must be able to degrade host cell walls in order to penetrate and invade plant tissues. Among the plant cell wall degrading enzymes (PCWDEs) produced, xylanases are of special interest since its degradation target, xylan, is one of the main structural polysaccharides in plant cell walls. In the biotrophic fungus Ustilago maydis, attempts to characterize PCWDEs required for virulence have been unsuccessful, most likely due to functional redundancy. In previous high-throughput screening, we found one xylanase to be important for U. maydis infection. Here, we characterize the entire U. maydis endo-xylanase family, comprising two enzymes from the glycoside hydrolase (GH) 10 family, Xyn1 and Xyn2, one from GH11, Xyn11A, and one from GH43, Xyn3. We show that all endo-xylanases except Xyn3 are secreted and involved in infection in a non-redundant manner, suggesting different roles for each xylanase in this process. Taking a closer look inside the plant during the pathogenic process, we observed that all secreted xylanases were necessary for fungal proliferation. Finally, we found that at least Xyn11A accumulated in the apoplast of the infected plant after three days, highlighting the role of these enzymes as important secreted proteins during fungal proliferation inside plant tissues.


Introduction
Plant smut diseases, which mainly affect grasses, are caused by a variety of fungal species collectively known as smut fungi. These pathogens are able to produce large amounts of teliospores, mainly in host floral organs, affecting their reproduction [1]. Smut diseases result in the loss of important crop cultivars, including wheat, barley, and maize, among others [2]. The biotrophic pathogen Ustilago maydis, currently one of the most wellstudied smut fungi, is an exception to the pattern of floral teliospore formation. U. maydis is able to colonize all aerial parts of its host, the maize plant, inducing plant tumors in the last stages of colonization, where teliospores develop [3,4]. The pathogenic cycle occurs when two sexually compatible strains come into contact on the maize plant surface and pheromones activate mutually compatible receptors. Then, they form a conjugation tube and mate, developing a filamentous cell cycle-arrested dikaryon [5,6]. These filaments sense physical and chemical plant signals, generating a morphogenetic structure called the appressoria, which mediates fungal penetration into plants [7]. After penetration, cell cycle arrest is released and the fungus expands into a branched filamentous form that induces hypertrophied plant cells, macroscopically visible as tumors [3,[8][9][10][11]. With the onset of plant tumor formation, diploid cells are formed after fusion of nuclei. Finally, hyphae subdivide as individual cells and undergo teliospore development.
The initial phases of fungal infection include the formation of infective structures in response to topographical (e.g., stomatal pores), chemical (e.g., epicuticular waxes), role of fungal xylanases in plant pathogenesis is still poorly understood. Although deletion of the gene encoding a GH11 xylanase in the necrotrophic fungi Botrytis cinerea has been shown to produce a drastic reduction in infection capability [34], and more recently a study on the necrotrophic fungus Rhizoctonia cerealis revealed a role in infection for a GH10 xylanase [35], deletion of different xylanase genes in the hemi-biotrophic fungi Fusarium oxisporum and Magnaporthe oryzae has shown no effect on plant infection [36][37][38], probably due to the redundancy between different enzymes in the family.
In a recent screen for U. maydis virulence factors, we found the xylanase Xyn1 to be important for infection [22]. In addition to Xyn1, we found three additional endo-xylanases in the genome of U. maydis. Taking advantage of this genetically tractable model organism, we wanted to address whether all other xylanases of this biotrophic fungus had a role in infection and possible redundancy. We observed that three out of the four U. maydis xylanases were secreted and that single, double, and triple deletions of the genes encoding for them reduced infection symptoms to a similar degree, discarding functional redundancy between different xylanases. To get better insight into the possible role of these secreted xylanases during infection, we studied the different steps of the infection process and found that secreted endo-xylanases are involved in proper filamentation and fungal progression on the plant surface and inside the plant, suggesting different roles for these enzymes during the pathogenic process in addition to plant penetration. Furthermore, we detected the presence of at least one xylanase in the apoplast of infected plants (three days after inoculation), supporting their importance as secreted proteins required for fungal progression inside plant tissues.
All U. maydis strains used in this study are listed in Table 1. As previously described [40], U. maydis pre-cultures and cultures were performed in YEPSL (0.4% bactopeptone, 1% yeast extract, and 0.4% saccharose) unless otherwise specified.
Cell wall integrity, ER, and oxidative stress assays were carried out with cultures grown at 28 • C to exponential phase in complete media (CM) supplemented with 2% D-glucose (CMD) and spotted at 0. 4  For conjugation tube formation assay, FB1 single xylanase deletion mutants were grown to exponential phase in liquid CMD. Then, cells were diluted to 0.5 OD 600 in 1 mL and incubated in a wheel at room temperature for 5 h with 1 µL (2.5 mg/mL) of pheromone a2. As a control, cells were also incubated with 1 µL of DMSO, since the pheromone is diluted in this organic solvent.
For mating assays, cells were grown in liquid YEPSL until exponential phase, washed twice with sterile bi-distilled water, spotted onto PD-charcoal plates, and grown for 24-48 h at 25-28 • C.
Maize (Zea mays) infection assays were performed as previously described [19], with minor modifications. Briefly, U. maydis cells were grown at 28 • C to exponential phase in liquid YEPSL and concentrated to an OD 600 of 3, washed twice in water, and injected into seven-day-old maize seedlings (Early Golden Bantam). Disease symptoms were quantified at 14 days post-infection. Statistical analyses were performed in GraphPad Prism 6 software.

Molecular Biology and Genetics Methods
Molecular biology techniques were used as described by Sambrook, Frisch, and Maniatis in Molecular Cloning: A Laboratory Manual [39]. U. maydis DNA isolation and transformation were carried out following the protocol described by Schulz et al. [44].
Xylanase deletion mutants were generated by homologous recombination as described previously [22]. Primers used in this study are listed in Supplementary Table S1.
For complementation of the ∆xyn1 mutant, the sequence upstream of xyn1 ORF until the next gene containing approximately 1 kb that would correspond to its promoter was amplified by PCR Q5 DNA polymerase (New England Biolabs, Ipswich, MA, USA) using primers Pxyn1_fwd/XmaI_Pxyn1_rev, adding a XmaI restriction site. After digestion with XmaI restriction enzyme (New England Biolabs, Ipswich, MA, USA), xyn1 promoter was cloned into p123 P otef :xyn1 (containing xyn1 ORF under the control of otef promoter) previously digested with XmaI and PvuII restriction enzymes, removing the otef promoter and generating the plasmid p123 Pxyn1:xyn1. This plasmid was SspI digested, purified with MEGAquick-spin™ Plus Total Fragment DNA Purification Kit (iNtRON Biotechnology, Seongnam, Gyeonggi, ROK), and integrated into the ip-locus of CL13 ∆xyn1 protoplasts.
For ∆xyn2 and ∆xyn11A complementation, the xyn2 and xyn11A genes were amplified by PCR using Q5 DNA polymerase (New England Biolabs, Ipswich, MA, USA) and primers designed in the NEBuilder assembly tool (Pxyn2_fwd/Pxyn2_rev for xyn2 and Pxyn11A_fwd/Pxyn11A_rev for xyn11A). All fragments were cloned into p123 plasmid previously digested with PvuII and NotI using NEBuilder ® HiFi DNA Assembly (New England Biolabs, Ipswich, MA, USA) and ligation transformed into E. coli competent cells. p123 P xyn2 :xyn2 and p123 P xyn11A :xyn11A derivative plasmids were digested with SalI and SspI, respectively. Both digested plasmids were purified with MEGAquick-spin™ Plus Total Fragment DNA Purification Kit (iNtRON Biotechnology, Seongnam, Gyeonggi, ROK) and integrated into the ip-locus of CL13 ∆xyn2 and CL13 ∆xyn11A protoplasts, respectively.
For GFP tagging of xyn2, xyn11A, and xyn3, gene ORF was amplified by PCR using Q5 DNA polymerase (New England Biolabs, Ipswich, MA, USA) and primers with BamHI and NcoI restriction sites to clone ORFs in frame with GFP into p123 GFP plasmid [45] previously digested with BamHI and NcoI. T4 DNA ligase was employed to ligate both fragments, and ligation transformed into E. coli competent cells. Every p123 derivative plasmid was SspI digested and transformed into U. maydis SG200 wild-type protoplasts.
For the Xyn2 and Xyn11A mCherry-HA fused proteins, gene ORFs were PCR amplified with primers containing SacII and NcoI restriction sites and RSIATA sequence to clone them in frame into a p123 derivative plasmid harboring mCherry-HA under the control of pit2 promoter [43]. Both digested and purified p123 P pit2 :mCherry-HA and gene ORF were ligated in a 1:5 vector:fragment proportion with T4 DNA ligase and transformed in E. coli competent cells. Subsequent U. maydis transformation after SspI plasmid digestion was performed into the ip-locus of SG200 protoplasts.
Since the xyn1 ORF contains an NcoI restriction site inside its sequence, clonation into p123 P pit2 :mCherry-HA was performed following the same strategy described above, but using XbaI restriction enzyme instead of NcoI for gene ORF and plasmid.
For relative fungal biomass quantification, maize seedlings were infected with U. maydis, and 3-and 8-days post infection, leaves were ground in liquid nitrogen. Total DNA was isolated with DNeasy Plant Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Fungal biomass was then quantified by qRT-PCR using a Real-Time CFX Connect (Bio-Rad, Hercules, CA, USA) and SYBR ® Premix Ex Taq™ II (Tli RNase H Plus) (Takara Bio INC, Kusatsu, Japan) according to the manufacturer's protocol, measuring the signal of ppi1 fungal gene relative to the plant gene gapdh. For this process, 0.6 and 6 ng of total DNA was used as template for gapdh and ppi1 reaction, respectively.

Sequence Alignment and Phylogenetic Analysis
BlastP was used to search for xylanase sequences in other organisms. The alignments were obtained using MAFFT v7. Phylogenetic analysis of xylanases was inferred by using the maximum likelihood method based on the JTT matrix-based model [46]. The G-INSiterative strategy (recommended for <200 sequences with global homology) was used to generate GH11 and GH43 alignment and the L-INS-iterative strategy (recommended for <200 sequences with one conserved domain) to generate GH10 alignment. Refined datasets were applied to remove redundant sequences, and missing data were eliminated. MaxAlign was selected to maximizing the size of gap-free columns. Initial trees for the heuristic search were obtained automatically by applying the neighbor-joining algorithm. Trees were visualized and annotated using Interactive Tree of Life (iTOL v6, https://itol.embl.de/).

Infection Stage Analysis
For on plant filamentation and appressoria quantification, wild-type cells harboring cytoplasmic CFP and mutant cells harboring cytoplasmic RFP were grown in liquid YEPSL until exponential phase, washed twice with bi-distilled water, and concentrated to an OD 600 of 3. Cells from wild-type and mutant strains were mixed in equal proportions, centrifuged, and concentrated to an OD 600 of 3. Maize seedlings were infected, and leaves were recovered 18-20 h post-infection. Infected leaves were stained with calcofluor white (Sigma-Aldrich, St. Louis, MO, USA) to visualize U. maydis filaments and appressoria in a fluorescence microscope.
To analyze maize plant colonization capability, U. maydis cultures were grown at 28 • C to exponential phase in liquid YEPSL and concentrated to an OD 600 of 3, washed twice in water, and injected into seven-day-old maize seedlings. Three days post-infection, leaves were de-stained with ethanol for at least 24 h, treated at 60 • C with 10% KOH for 4 h, washed four times in phosphate buffer, and then stained with propidium iodide (PI) to visualize plant tissues in red and wheat germ agglutinin (WGA)-AF488 to visualize the fungus in green for 30 min in a vacuum pump with 5 min vacuum and 5 min rest cycles. At least four leaves from two independent experiments were stained and visualized by fluorescence microscopy (Leica SPE DM2500, Leica, Wetzlar, Germany).

Protein and Blotting Assays
For colony secretion assay, cells were grown in YEPSL to an OD 600 of 0.6-0.8, then were spotted onto a nitrocellulose filter secured to a rich media plate. After drop-drying, plates were sealed with parafilm and incubated face-up at 28 • C for 16 h. Cells growing on the nitrocellulose membrane were washed with distilled water, using a roller to eliminate all cells, and the membrane was incubated with mouse polyclonal anti-GFP antibody (Roche, Mannheim, BW, Germany) (1:1000). As secondary antibody, anti-mouse IgG-horseradish peroxidase conjugated antibody (1:5000; Sigma-Aldrich, St. Louis, MO, USA) was used. Immunoreactive dots were developed by SuperSignal™ West Femto Maximum Sensitivity substrate (ThermoFisher Scientific, Carlsbad, CA, USA). Image gel and membrane acquisition was carried out with ChemiDoc XRS (Bio-Rad, Hercules, CA, USA).
To determine whether xylanases were secreted to the maize apoplasts, cells harboring xylanases tagged with mCherry under the control of pit2 promoter [43] were grown in liquid YEPSL until exponential phase, washed twice in water, and concentrated to an OD 600 of 3. Around 200 maize seedlings for each strain were infected, and 3 dpi leaves were recovered to isolate apoplastic fluid by vacuum infiltration. For that purpose, 8 cm of leaf from 1 cm below the infection site was cut and coated with bi-distilled water in a big beaker with a steel sieve on top. Three cycles of 15 min vacuum at 60 mbar and 2.5 min atm pause were applied in a vacuum chamber while the water was continuously stirred with a magnetic stir bar to remove air bubbles. Then, leaves were carefully dried with paper towels, placed in syringes inside 50 mL tubes, and centrifuged at 3000 rpm for 15 min at 4 • C. Apoplastic fluid was pooled and stored at −80 • C or directly precipitated with chilled acetone and mCherry blotted using mouse polyclonal anti-mCherry antibody (Roche, Mannheim, BW, Germany) (1:1000). The same secondary antibody and blot development conditions described above were used here.

Microscopy
For visualization of DNA content and cellular morphology, cells were stained with DAPI and observed by differential interference contrast (DIC) and fluorescence microscopy using a DeltaVision microscopy system comprising an Olympus IX71 microscope (Olympus, Shinjuku, Tokyo, Japan) and CoolSnap HQ camera (Photometrics, Tucson, AZ, USA).
For in plant quantification of filament and appressoria formation in co-infection experiments with U. maydis CFP and RFP labelled strains, 20 hpi leaf samples were stained with calcofluor white (Sigma-Aldrich, St. Louis, MO, USA) to visualize fungal material and then checked for CFP or RFP fluorescence in the DeltaVision microscopy system.
To analyze the U. maydis progression inside the maize plant, leaf samples stained with PI and WGA-AF488 (described above) were examined using a Leica SPE (DM2500, Leica, Wetzlar, Germany) confocal microscope.
Image processing was carried out using Adobe Photoshop CS5 (Adobe, San Jose, CA, USA) and ImageJ (public domain).

Ustilago maydis Xylanases Are Secreted and Required for Full Virulence
We have previously shown that protein glycosylation is essential for U. maydis virulence, since mutants for genes in the two main glycosylation pathways are completely avirulent [47,48]. In a screen for U. maydis virulence factors related to glycosylation, we found UMAG_04422 (now called Xyn1) encoding an endo-1,4-β-xylanase (EC 3.2.1.8) in the GH10 family, which is involved in plant infection [22]. In order to identify other U. maydis xylanases, we performed a BlastP search against the U. maydis genome using the Xyn1 sequence as input, finding UMAG_03411 (now called Xyn2), which also belongs to the GH10 family. As xylanases are also found in the GH 5, 7, 8, 10, 11, and 43 families, we performed an analysis for known proteins from those families in the U. maydis genome by examining the MycoCosm database [49]. In this fungal genomics database, we found four annotated xylanases: the two GH10 xylanases previously shown (Xyn1 and Xyn2), the GH11 protein UMAG_06350 (UmXyn11A) annotated as an endo-1,4-beta xylanase, and the GH43 protein UMAG_04897 (now called Xyn3), an uncharacterized protein related to endo-1,4-beta-xylanase in U. trichophora. Among them, UmXyn11A has been previously identified and functionally characterized as an endo-1,4-β-xylanase capable of degrading xylan [50,51].
To get a better picture of the possible peculiarities of the xylanase family of U. maydis, we performed a phylogenetic analysis of all xylanases found in other selected fungal phytopathogens, including closely related Basidiomycetes such as U. hordei and Sporisorium reilianum, the more distant rust fungus Puccinia graminis, and the Ascomycetes Magnaporthe oryzae, Trichoderma reesei, Aspergillus nidulans, and Botrytis cinerea ( Figure 1A). We observed that Xyn1 from smut fungi is very distant from most of the other fungal GH10 xylanases, including smut fungus Xyn2 (Figures 1A and S1). The sequence analysis shows that Xyn1 contains a disordered region of 200 amino acids in its C-terminal domain ( Figure 1B), which is only present in the other two more divergent GH10 xylanases in this phylogenetic analysis: UHOR_06909 and sr15309 (from U. hordei and S. reilianum, respectively). This intrinsically disordered region (IDR) has been proposed to facilitate posttranslational modifications or protein-protein interactions or to regulate protein half-life during infection [52,53]. It is also remarkable that, in contrast to the high number of GH10 and GH11 xylanases, only a few GH43 xylanases are found in either Basidiomycetes or Ascomycetes (Figures 1A and S1).
We wanted to know which of the identified xylanases are secreted and thus could potentially have a role in plant infection. Xyn1 has been previously identified as a secreted glycoprotein downstream of glycosidase I (Gls1) [22]. We performed a colony secretion assay to analyze whether xylanases Xyn2, Xyn11A, and Xyn3 were also secreted. As expected, given the presence of signal peptides in their N-terminal regions, Xyn2 and Xyn11A, but not Xyn3, were found to be secreted ( Figure 1B,C). Thus, we excluded Xyn3 from further analysis since it would be expected to have a role in xylan-derivative carbohydrate metabolism inside fungal cells (under study) rather than in extracellular xylan degradation, as has been recently described in Neurospora crassa [54].  In order to decipher whether these secreted xylanases are expressed during the infection process, we analyzed their expression during maize plant infection at different time points using data from a high-throughput transcriptomic analysis of U. maydis pathogenesis [21]. These data show that xyn1 was the most highly expressed xylanase during infection, with maximum expression at 2 dpi. xyn2 had a much lower expression level, with a first peak during the first hours of infection and another one at a later stage (12 dpi). Finally, xyn11A was faintly expressed during the infection process, with higher expression at 0.5 dpi (Figure 2A). These data suggest that Xyn1 may be the main xylanase operating during biotrophic development, perhaps assisted by the other two xylanases at different stages of infection.

Xylanases Are Necessary to Assure Proper Fungal Filamentation and Progression inside the Plant
To discard general cellular defects in the deletion mutants, which could explain the observed defects in the infection process, we analyzed growth alterations, endoplasmic reticulum (ER), osmotic and oxidative stress resistance, and cell wall integrity under axenic conditions. DTT and tunicamycin were used as ER stressors [60], sorbitol and NaCl Mann-Whitney statistical test was performed for each mutant versus corresponding wild-type strain (* p-value < 0.05; *** p-value < 0.005; **** p-value < 0.001).
To examine this possibility, we analyzed the virulence capability of single, double, and triple secreted xylanase deletion mutants. Significantly, the loss of any of the three xylanases resulted in virulence defects. As can be observed in Figures 2B, S2 and S3, all single mutants showed reduced virulence in different genetic backgrounds. All mutants were checked by Southern blot analysis and complemented by re-introduction of the endogenous locus ( Figure S2). Surprisingly, the double ∆xyn1∆xyn2 and the triple ∆xyn1∆xyn2∆xyn11A mutants showed nearly the same effect as the single ones ( Figure 2B). This epistatic effect suggests that all xylanases have a non-redundant role during pathogenesis, and in agreement with the expression patterns, they may act at different stages of infection.
The lack of fully non-virulent PCWDE mutants has commonly been explained by redundancy and compensatory effects between PCWDEs of the same family [15,35,[55][56][57][58][59]. Although the quadruple deletion mutant, including the non-secreted Xyn3, was not created in this study, it is highly improbable that this mutant would result in an avirulent phenotype. Thus, our data suggest that such redundancy may also occur between PCWDEs with different catalytic activities, not just among xylanases themselves. Specifically, complete xylan hydrolysis requires additional enzymes. Although endo-1,4-β-D-xylanases are responsible for cleavage of the backbone chain, side group degradation is catalyzed by other enzymes, including α-L-arabinofuranosidases and α-D-glucuronidases [24]. Thus, it would be interesting to test the redundancy between cooperating enzymes for xylan degradation to further confirm their requirement for pathogenesis.

Xylanases Are Necessary to Assure Proper Fungal Filamentation and Progression Inside the Plant
To discard general cellular defects in the deletion mutants, which could explain the observed defects in the infection process, we analyzed growth alterations, endoplasmic reticulum (ER), osmotic and oxidative stress resistance, and cell wall integrity under axenic conditions. DTT and tunicamycin were used as ER stressors [60], sorbitol and NaCl as osmotic stressors, H 2 O 2 as oxidant [61], calcofluor white (CFW) and Congo red as cell wall integrity sensors [62], and SDS as membrane-perturbing drug [62]. We found no significant differences in any of the conditions assayed, except for a subtle reduction in growth of xyn1 deletion mutant on all of the compounds tested but tunicamycin, where the opposite effect was observed (Figures 3 and S4). Strikingly, the lack of xyn1 resulted in longer cells ( Figure 3A,B), which was consistent among independent mutants with single ∆xyn1 cassette integrations in two different genetic backgrounds ( Figure S5). It is notable that this phenotype indicates a role for Xyn1 in the fungal cell itself, which suggests that this secreted xylanase may also have xylanase-independent activity, as fungal cells lack xylan. Despite the cells being longer and slightly more susceptible to some stressors, the doubling time of this mutant was similar to the wild-type strain ( Figure 3D). All together, these data point to a specific role for these enzymes in plan pathogenesis rather than a pleiotropic effect indirectly affecting infection.
To determine which step of pathogenic development is affected by the loss of xylanases, we first examined the mating capability between sexually compatible cells. We observed similar mating efficiency and conjugation tube formation in all xylanase deletion mutants compared to the wild-type strains ( Figure 4A,B). Therefore, pathogenic defects that affect xylanase mutant infection probably occur during host interactions.
To address whether fungal colonization inside plant tissue was impaired, we visualized infected maize leaves at three days post-infection by microscopy and quantified fungal biomass inside the plant by performing qPCR of DNA samples in infected maize leaves at three-and eight-days post-inoculation (dpi). Although all xylanase mutants were able to perform cell-to-cell progression inside the plant with normal morphology (Figure 5A), we observed defective proliferation for all mutants inside the plant, with a significant reduction in detected biomass at 8 dpi but not at 3 dpi ( Figure 5B). table that this phenotype indicates a role for Xyn1 in the fungal cell itself, which suggests that this secreted xylanase may also have xylanase-independent activity, as fungal cells lack xylan. Despite the cells being longer and slightly more susceptible to some stressors, the doubling time of this mutant was similar to the wild-type strain ( Figure 3D). All together, these data point to a specific role for these enzymes in plan pathogenesis rather than a pleiotropic effect indirectly affecting infection.  To determine which step of pathogenic development is affected by the loss of xylanases, we first examined the mating capability between sexually compatible cells. We observed similar mating efficiency and conjugation tube formation in all xylanase deletion mutants compared to the wild-type strains ( Figure 4A,B). Therefore, pathogenic defects that affect xylanase mutant infection probably occur during host interactions. To address whether fungal colonization inside plant tissue was impaired, we visualized infected maize leaves at three days post-infection by microscopy and quantified fungal biomass inside the plant by performing qPCR of DNA samples in infected maize leaves at three-and eight-days post-inoculation (dpi). Although all xylanase mutants were able to perform cell-to-cell progression inside the plant with normal morphology ( Figure  5A), we observed defective proliferation for all mutants inside the plant, with a significant reduction in detected biomass at 8 dpi but not at 3 dpi ( Figure 5B). (B) Conjugation tube formation for FB1 wild-type and xylanase deletion mutants was quantified after 5 h incubation with a2 pheromone in axenic conditions. Scale bar represents 20 µm. Total number of cells, corresponding to four biological replicates, is indicated above each column. T-test statistical analysis was performed (ns, not statistically significant).
The reduced fungal progression of xylanase mutants inside the plant together with the induction of expression of these genes during the biotrophic establishment stage suggest that U. maydis proliferation inside plant tissues requires degradation of xylan in cell walls. However, we wanted to further investigate a possible additional role for xylanases in earlier stages of infection, when they start to be expressed, such as filamentation on the plant surface and appressoria formation. Although appressoria formation was not affected, we observed a significant reduction in filamentation capability on the plant surface for both ∆xyn1 and ∆xyn2 mutants versus the wild-type strain ( Figure 6A,B,D,E), with longer filaments formed ( Figure 6G). These defects were not observed in the ∆xyn11A mutant ( Figure 6C,F), suggesting a more specific role of this xylanase in later stages of infection.
Many pathogenic fungi use the combination of turgor pressure and secreted cell wall degrading enzymes to penetrate plant tissues [12,13,[63][64][65]. Once inside the plant, the fungi must continue degrading the cell walls in order to progress from cell to cell. In this scenario, we can postulate that U. maydis uses xylanases to degrade plant cell walls and proliferate inside the plant. However, the alteration in the number and length of filaments on the plant surface observed for ∆xyn1 and ∆xyn2 mutants is surprising. One possibility is that U. maydis secretes xylanases during the filamentation process and the residue resulting from xylan degradation acts as a signal for fungal penetration. In this way, the fungus might be able to identify regions where plant cell walls are more exposed as potential penetration sites. Once inside the plant, U. maydis would continue producing xylanases to degrade the cell walls, allowing proper proliferation of hyphae. In our studies we have not been able to identify different stages of the infection process affected by all xylanases, thus further investigations are required to know why these xylanases show non-redundant roles. The reduced fungal progression of xylanase mutants inside the plant together with the induction of expression of these genes during the biotrophic establishment stage suggest that U. maydis proliferation inside plant tissues requires degradation of xylan in cell walls. However, we wanted to further investigate a possible additional role for xylanases in earlier stages of infection, when they start to be expressed, such as filamentation on the plant surface and appressoria formation. Although appressoria formation was not affected, we observed a significant reduction in filamentation capability on the plant surface for both ∆xyn1 and ∆xyn2 mutants versus the wild-type strain ( Figure 6A,B,D,E), with longer filaments formed ( Figure 6G). These defects were not observed in the ∆xyn11A mutant ( Figure 6C,F), suggesting a more specific role of this xylanase in later stages of infection. Many pathogenic fungi use the combination of turgor pressure and secreted cell wall degrading enzymes to penetrate plant tissues [12,13,[63][64][65]. Once inside the plant, the fungi must continue degrading the cell walls in order to progress from cell to cell. In this scenario, we can postulate that U. maydis uses xylanases to degrade plant cell walls and proliferate inside the plant. However, the alteration in the number and length of filaments on the plant surface observed for ∆xyn1 and ∆xyn2 mutants is surprising. One possibility is that U. maydis secretes xylanases during the filamentation process and the residue resulting from xylan degradation acts as a signal for fungal penetration. In this way, the fungus might be able to identify regions where plant cell walls are more exposed as potential penetration sites. Once inside the plant, U. maydis would continue producing xylanases to degrade the cell walls, allowing proper proliferation of hyphae. In our studies we have not been able to identify different stages of the infection process affected by all xylanases, thus further investigations are required to know why these xylanases show non-redundant roles.

Xyn11A Is Secreted to the Apoplast during Fungal Progression inside the Plant
Given that xylanases are secreted in vitro and are required for fungal progression inside the maize plant, we wanted to determine if they are secreted into the maize apoplast during infection. To test this, we infected plants with SG200 strains harboring Xyn1-, Xyn2-, or Xyn11A:mCherry-HA fusion proteins under the control of pit2 promoter, which is induced during the first days of infection [43], and purified apoplastic fluid at three days post-infection. The results obtained by anti-mCherry Western blot showed a band near 55 KDa corresponding to Xyn11A:mCherry-HA (51 KDa) and a second band between 25 KDa and 35 KDa equivalent to free mCherry (28.8 KDa) (Figure 7). To our knowledge, this is the first time this enzyme has been identified in the apoplast of infected Length of filaments with appressoria is quantified in (G). White arrows indicate appressoria. Scale bar represents 10 µm. At least two leaves from two independent experiments were analyzed. T-test statistical analysis was performed (ns, not statistically significant; * p-value < 0.05; ** p-value < 0.01).

Xyn11A Is Secreted to the Apoplast during Fungal Progression Inside the Plant
Given that xylanases are secreted in vitro and are required for fungal progression inside the maize plant, we wanted to determine if they are secreted into the maize apoplast during infection. To test this, we infected plants with SG200 strains harboring Xyn1-, Xyn2-, or Xyn11A:mCherry-HA fusion proteins under the control of pit2 promoter, which is induced during the first days of infection [43], and purified apoplastic fluid at three days post-infection. The results obtained by anti-mCherry Western blot showed a band near 55 KDa corresponding to Xyn11A:mCherry-HA (51 KDa) and a second band between 25 KDa and 35 KDa equivalent to free mCherry (28.8 KDa) (Figure 7). To our knowledge, this is the first time this enzyme has been identified in the apoplast of infected plants. This observation strongly supports our previous results showing a role for xylanases in fungal progression inside the plant. In contrast, western blots for Xyn1 and Xyn2 in apoplast extracts resulted in the detection of bands corresponding to unexpectedly high molecular weights. Thus, it remains to be determined whether or not Xyn1 and Xyn2 are also secreted into the maize apoplast during U. maydis infection.
plants. This observation strongly supports our previous results showing a role for xylanases in fungal progression inside the plant. In contrast, western blots for Xyn1 and Xyn2 in apoplast extracts resulted in the detection of bands corresponding to unexpectedly high molecular weights. Thus, it remains to be determined whether or not Xyn1 and Xyn2 are also secreted into the maize apoplast during U. maydis infection.

Conclusions
Here, we show that xylanases secreted by U. maydis GH10 and GH11 are involved in filament formation on the plant surface as well as during hyphae progression inside maize plant cells. Our results indicate that at least xylanase 11A is secreted into the maize apoplast during infection, which suggests that it has an important role in fungal progression inside the plant. Interestingly, the epistatic effect observed here for the different secreted xylanase mutants suggests that there is not such a high degree of redundancy in U. maydis, as has been observed for other fungi. Instead, each secreted xylanase should have an independent role during infection, and further investigations are required to identify these roles. In addition, the lack of completely non-virulent phenotypes in the triple xylanase deletion mutant suggests that these enzymes may act in conjunction with other types of PCWDEs to ensure successful xylan degradation and full pathogenic development. In the future, it would be of interest to determine what other PCWDEs might cooperate with the xylanases in the infection process or, conversely, whether U. maydis infection does not in fact primarily depend on PCWDEs.

Conclusions
Here, we show that xylanases secreted by U. maydis GH10 and GH11 are involved in filament formation on the plant surface as well as during hyphae progression inside maize plant cells. Our results indicate that at least xylanase 11A is secreted into the maize apoplast during infection, which suggests that it has an important role in fungal progression inside the plant. Interestingly, the epistatic effect observed here for the different secreted xylanase mutants suggests that there is not such a high degree of redundancy in U. maydis, as has been observed for other fungi. Instead, each secreted xylanase should have an independent role during infection, and further investigations are required to identify these roles. In addition, the lack of completely non-virulent phenotypes in the triple xylanase deletion mutant suggests that these enzymes may act in conjunction with other types of PCWDEs to ensure successful xylan degradation and full pathogenic development. In the future, it would be of interest to determine what other PCWDEs might cooperate with the xylanases in the infection process or, conversely, whether U. maydis infection does not in fact primarily depend on PCWDEs.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/jof7121081/s1. Figure S1: Fungal xylanase phylogenetic tree; Figure S2: Xylanase mutants contain a single cassette integration, leading to similar virulence defects, and phenotype is complemented by re-introduction of wild-type allele; Figure S3: Infection assay of xylanase mutants in FB1 × FB2 strains; Figure S4: Stress and cell wall integrity assays for xylanase defective mutants; Figure S5: Lack of xyn1 affects cell length in both CL13 and SG200 backgrounds. Table S1: List of primers used in this study.