*Article* **Identification of Putative Virulence Genes by DNA Methylation Studies in the Cereal Pathogen** *Fusarium graminearum*

**Francesco Tini 1, Giovanni Beccari <sup>1</sup> , Gianpiero Marconi 1,\* , Andrea Porceddu <sup>2</sup> , Micheal Sulyok <sup>3</sup> , Donald M. Gardiner 4, Emidio Albertini <sup>1</sup> and Lorenzo Covarelli 1,†**


**Abstract:** DNA methylation mediates organisms' adaptations to environmental changes in a wide range of species. We investigated if a such a strategy is also adopted by *Fusarium graminearum* in regulating virulence toward its natural hosts. A virulent strain of this fungus was consecutively sub-cultured for 50 times (once a week) on potato dextrose agar. To assess the effect of subculturing on virulence, wheat seedlings and heads (cv. A416) were inoculated with subcultures (SC) 1, 23, and 50. SC50 was also used to re-infect (three times) wheat heads (SC50×3) to restore virulence. *In vitro* conidia production, colonies growth and secondary metabolites production were also determined for SC1, SC23, SC50, and SC50×3. Seedling stem base and head assays revealed a virulence decline of all subcultures, whereas virulence was restored in SC50×3. The same trend was observed in conidia production. The DNA isolated from SC50 and SC50×3 was subject to a methylation contentsensitive enzyme and double-digest, restriction-site-associated DNA technique (ddRAD-MCSeEd). DNA methylation analysis indicated 1024 genes, whose methylation levels changed in response to the inoculation on a healthy host after subculturing. Several of these genes are already known to be involved in virulence by functional analysis. These results demonstrate that the physiological shifts following sub-culturing have an impact on genomic DNA methylation levels and suggest that the ddRAD-MCSeEd approach can be an important tool for detecting genes potentially related to fungal virulence.

**Keywords:** DNA methylation; *Fusarium graminearum*; *in vitro* subcultures; virulence reduction; ddRAD-MCSeEd; virulence genes

### **1. Introduction**

Fusarium Head Blight (FHB) is one of the most widespread and damaging diseases of cereal crops, such as bread and durum wheat and barley, capable of strongly impairing not only yield but also quality, by contaminating grains with mycotoxins. The disease is caused by several members of the *Fusarium* species complex [1]. *F. graminearum* is globally considered the most dangerous FHB pathogen due to its aggressiveness and diffusion worldwide [2–4]. The pathogen is able to biosynthesize mycotoxins belonging to the type B trichothecenes, such as deoxynivalenol (DON) and nivalenol (NIV) [5] as well as the DON acetylated forms: 3-acetyl-deoxynivalenol (3-ADON) and 15-acetyl-deoxynivalenol (15-ADON) [4,6].

**Citation:** Tini, F.; Beccari, G.; Marconi, G.; Porceddu, A.; Sulyok, M.; Gardiner, D.M.; Albertini, E.; Covarelli, L. Identification of Putative Virulence Genes by DNA Methylation Studies in the Cereal Pathogen *Fusarium graminearum*. *Cells* **2021**, *10*, 1192. https://doi.org/10.3390/ cells10051192

Academic Editor: N. Louise Glass

Received: 19 March 2021 Accepted: 10 May 2021 Published: 13 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

All living organisms can adopt strategies to enable rapid adaptation to new environmental conditions, without changing the DNA sequence [7,8]. For example, to respond to the environmental changes or biotic and abiotic stresses, some organisms adjust physiological and development machinery by gene expression regulation. DNA methylation and demethylation of cytosine play key roles in such a strategy [9–12]. Cytosine methylation is conventionally classified in CG, CHG, and CHH sequence contexts, where H is adenine, cytosine, or thymine. DNA methylation involves the addition of a methyl group to cytosine to produce 5-methylcytosine (5mC). Furthermore, the addition of the same group to adenine has been recently explored (N6-methyladenine, 6 mA), [13,14]. Methylation changes on cytosine residues are important for transposon silencing, epigenetic regulation, and genome expression [15–17]. Generally, methylation is related to the silencing of genes and transposable elements, whereas demethylation is correlated to active transcription [18], even if the reverse has also been described [15]. DNA methylation is catalysed by a conserved set of proteins called DNA methyltransferases (MTases) [19]. DNA MTases in eukaryotes belong to five different groups based on their structure and functions [20–22]. DNA MTase homologs have been identified in many fungal pathogens, including *F. graminearum* [23]. In pioneering studies, the histone proteins DIM-2, DIM-5, and HP1 were defined to be essential for DNA methylation in *Neurospora* [24–26]. Homologues for these three proteins are present in all sequenced *Fusarium* species [27], demonstrating that the DNA methylation machinery is present in this genus [28].

Some pathogenic fungi may partially lose or attenuate their virulence in response to environmental changes [29] as well as in response to other external factors, such as prolonged subculturing on artificial rich media [30]. Virulence may be restored if the attenuated strains are re-inoculated onto healthy host tissues. In the present work, we analysed whether virulence changes due to subculturing in different media/hosts was associated with DNA methylation changes and if these changes affected genes known to be involved in virulence regulation toward different hosts.

The loss of aggressiveness, following subculturing in artificial rich media, was evaluated by (i) execution of *in planta* virulence assays, (ii) characterization of secondary metabolites biosynthesis, and (iii) determination of *in vitro* fungal development and conidiation. Furthermore, the colony that had undergone consecutive transfers for one year on an artificial, rich, nutrient medium was used to infect healthy bread wheat heads. The DNA extracted from the last *in vitro* subculture was compared to the DNA isolated from mycelia sampled on infected wheat heads. Several genes affected by methylation level changes have already been demonstrated to be involved in virulence toward the host.

### **2. Materials and Methods**

### *2.1. Fungal Strain and Subculturing*

*F. graminearum* strain FG8 (15-ADON producer) from the fungal collection of the Department of Agricultural, Food, and Environmental Sciences (University of Perugia, Perugia, Italy) was used for all experiments. FG8 was isolated from durum wheat grain, molecularly identified and characterized for the *in vitro* mycotoxigenic profile [31]. The experimental design is shown in Figure 1.

**Figure 1.** Experimental design followed throughout the experiment.

To prepare the subcultures' inoculum, FG8 was cultured on potato dextrose agar (PDA, Biolife Italiana, Milan, Italy) for 50 weeks. Briefly, a piece of fungal mycelium was cultured on PDA in a 9-cm Petri dish at 22 ◦C. After one week, one mycelium plug (0.5 cm

of diameter) was used to inoculate a sterile PDA plate that was incubated for another week at 22 ◦C, whereas the rest of the mycelium was cut in small pieces and stored into 2-mL plastic tubes (Eppendorf, Hamburg, Germany) at −80 ◦C and represented the SC1 inoculum. The same actions were repeated every week for 50 weeks, obtaining a total of 50 subcultures stored at −80 ◦C (from SC1 to SC50). Sterile PDA plates were inoculated with mycelium plugs deriving from SC1, SC23, and SC50 samples for further analyses.

SC50×3 subcultures were obtained from sterile PDA plates inoculated with mycelium derived from three head-to-head passages, as described in Section 2.2.2.

### *2.2. Virulence Assays*

### 2.2.1. Crown Rot Assay

The virulence assay on the stem base of bread wheat was carried out following the method previously described [32–34]. The mycelium of SC1, SC23, and SC50 was cut in small squares and homogenised with 12 mL of sterile water with a Mixer Mill MM400 (Retsch, Haan, Germany) to obtain a gel for pipetting. Bread wheat seeds (cv. A416, an Italian cultivar with well-known susceptibility to FHB) were previously surface sterilized with a solution composed of 7% sodium hypochlorite (8% *v/v*), 98% ethanol (10% *v/v*), and sterile, deionised water (82% *v/v*) for 5 min and rinsed three times with sterile, deionised water. Surface-sterilized seeds were sown in 6 × 8 × 8 cm pots (10 seeds per pot), and filled with a sterile soil mix (50% sand and 50% peat). Pots were incubated at 22 ◦C with a 15/9 h day/night light cycle. A 3-cm-long PVC collar (3-mm internal diameter) was placed around the emerging coleoptiles. When the second leaf was fully expanded, plants were inoculated by injecting 700 μL of inoculum into the space between seedling and the PVC collar. PDA macerated with sterile water was used as a control treatment. Three replicates (corresponding to three different pots, 10 plants per pot/replicate) for each FG8 subculture and for the control were realized for a total of 12 pots (120 plants). The inoculated seedlings were covered by plastic bags for 3 days to keep the moisture high. Seedlings were maintained for 25 days at 22 ◦C with a 15/9 h day/night light cycle.

Stem-base infections were evaluated by measuring the length (cm) of the necrotic area on the first leaf and the presence/severity of necrosis across leaf sheaths with a 0–17 arbitrary scale (clean = 0; coleoptile = 1–2; 1st leaf = 3–4–5; 2nd leaf = 6–7–8; 3rd leaf = 9–10–11; 4th = 12–13–14; 5th leaf = 15–16–17).

The fungal subcultures virulence toward the bread wheat stem base was evaluated using crown rot disease index (DI). DI was calculated as the product between the average length of the necrotic area on the first leaf (cm) and the average value (0–17) of necrosis across leaf sheaths of 10 plants for 3 replicates.

### 2.2.2. Fusarium Head Blight Assay

Flasks containing 300 mL of mung bean broth were inoculated with a SC50 mycelium plug. Mung bean broth was prepared by boiling 1 L of sterile water and adding 40 g of mung beans for 10 min. Subsequently, beans were removed from the broth by filtering with cheesecloth and the broth was autoclaved. The inoculated flasks were shaken on an orbital shaker at 150 rpm for 10 days at room temperature and 12/12 h light/dark. The fungal broth was filtered through Miracloth (Millipore Corporation Billerica, MA, USA) and the conidia suspension concentration was adjusted to 1 × <sup>10</sup><sup>6</sup> conidia mL−<sup>1</sup> using a haemocytometer to count the cells. Sterilised seeds were incubated in Petri dishes for one day in the dark at 4 ◦C on water-soaked filter paper and three days in the dark at room temperature for germination. Germinated seeds were transplanted into 9 × 9 × 13 cm pots (one seed per pot) filled with peat and placed in a growth chamber at 23 ◦C with a photoperiod of 16 h. At mid-anthesis, wheat heads were point inoculated with macroconidia of SC50 by pipetting 10 μL of conidial suspension, containing approximately 10<sup>4</sup> conidia. The inoculum was injected between the glumes of a central spikelet. Heads were covered in plastic bags for 7 days to increase moisture content. Two weeks after inoculation, a little piece of mycelium was scraped from inoculated heads and used for inoculating other

healthy wheat heads. This was repeated for three consecutive direct head-to-head transfers to obtain a sample named SC50×3. A portion of scraped mycelium of SC50×3 was cultured on PDA and stored at −80 ◦C for DNA extraction or used to inoculate mung bean flasks to obtain SC50×3 conidia inoculum, as described for the FHB assay. At mid-anthesis, heads were point inoculated as mentioned above, with 10 μL of conidial suspension containing approximately 10<sup>4</sup> conidia. A total of 15 heads per subculture were inoculated (5 heads per replicate) for a total of 75 heads, including control (sterile-water inoculation). After inoculation, heads were covered for 72 h in plastic bags to maintain a high humidity level and promote the infection. Inoculated plants were placed into a growth chamber with a photoperiod of 16 h at 23 ◦C. Symptoms caused by the different subcultures were assessed at 14 days post-inoculation (dpi), determining the proportion of spikelets of each head that displays browning symptoms.

### *2.3. In Vitro Growth Rate Assay*

One mycelium plug of 0.5 cm diameter of SC1, SC23, and SC50 was taken from the edge of a 4-day-old colony and placed in the middle of the plates containing PDA. Six replicates per subculture were realized for a total of 18 plates. The growth rate was evaluated measuring the mycelial diameters in the two perpendicular directions, as previously described [35]. The diameters of the colonies were measured alongside the two axes. The growth value was calculated as the average of the measures taken from the two axes for three replicates per subculture and expressed in centimeters.

### *2.4. In Vitro Conidial Production*

Three PDA Petri dishes per sample (three replicates) were inoculated with one mycelium plug of each subculture (diameter of 0.5 cm) and incubated for 4 weeks at room temperature, under near-UV light for 12 h per day. At the end of the incubation period, 15 mL of sterile water were added with a pipette to each incubated plate and the mycelium was scraped and mixed to the added water with a sterile spatula. The conidia were separated from the mycelium by Miracloth filtration. Conidia concentration was estimated with a haemocytometer and conidia production was calculated as the average of three replicates per sample.

### *2.5. Determination of Secondary Metabolites Biosynthesized In Vitro by F. graminearum Subcultures*

### 2.5.1. *F. graminearum* Subcultures Preparation

Ten milliliters of deionized, sterile water were added to 20 g of rice kernels and placed into 100-mL glass flasks, autoclaved three times on alternate days, and inoculated with one mycelium plug per sample. Flasks were incubated for 4 weeks at 22 ◦C in the dark and the developed cultures were milled with mortar and pestle and stored at −80 ◦C. Three replicates per sample were realized. Three non-inoculated flasks with rice kernels were used as controls.

### 2.5.2. Extraction and Analysis of Secondary Metabolites

Five grams of each ground sample were extracted using 20 mL of extraction solvent (acetonitrile-water-acetic acid, 79:20:1, *v/v/v*) followed bya1+1 dilution using acetonitrilewater-acetic acid (20:79:1, *v/v/v*) and direct injection of 5 μL of diluted extract. Concentrations exceeding the linear range of the detector were quantified by reanalysis of the extracts after further dilution steps (1:50 and 1:1000, respectively). LC-MS/MS screening of target fungal metabolites was performed with a QTrap 5500 LC-MS/MS System (Applied Biosystems, Foster City, CA, USA) equipped with a Turbo Ion Spray electrospray ionization (ESI) source and a 1290 Series HPLC System (Agilent, Waldbronn, Germany). Chromatographic separation was performed at 25 ◦C on a Gemini® C18-column, <sup>150</sup> × 4.6 mm i.d., 5-μ<sup>m</sup> particle size, equipped with a C18 4 × 3 mm i.d. security guard cartridge (all from Phenomenex, Torrance, CA, USA). The chromatographic method as well as chromatographic

and mass spectrometric parameters are described in Reference [36]. Confirmation of positive analyte identification was obtained by the acquisition of two MRMs per analyte (with the exception of MON and three nitro propionic acids that exhibit only one fragment ion), which yielded 4.0 identification points according to the commission decision (Commission Decision, 2002). In addition, the liquid chromatography retention time and the intensity ratio of the two MRM transitions agreed with the related values of an authentic standard within 0.03 min and 30% rel., respectively. Quantification was performed via external calibration using serial dilutions of a multi-analyte stock solution. Results were corrected for apparent recoveries obtained for wheat [36]. The accuracy of the method is verified on a continuous basis by regular participation in proficiency testing schemes.

### *2.6. DNA Extraction*

To proceed with methylation analysis, DNA from SC50 and SC50×3 was extracted. In detail, SC50 was grown in Petri dishes containing PDA. Fungal mycelium was scraped using a spatula, and placed into 2-mL sterile plastic tubes with a steel bead at −80 ◦C. The SC50×3 mycelium developed on host tissues after three head-to-head consecutive transfers (Section 2.2.2) was collected with tweezers and stored in 2-mL plastic tubes with one steel bead at −80 ◦C. Mycelium samples (SC50 and SC50×3) were freeze-dried (Heto Power Dry LL3000) and reduced to a fine powder whit using a Mixer Mill MM400 (Retsch). Genomic DNA of the four samples was obtained using a PureLink™ Plant Total DNA Purification Kit (Thermo Fisher Scientific, Walthman, MA, USA) according to the manufacturer's instruction. Extracted DNA concentration was quantified with a Qubit® 3.0 Fluorometer (Thermo Fisher Scientific), using a dsDNA High Sensitivity (HS) Assay (Thermo Fisher Scientific) kit, following the manufacturer's protocol.

### *2.7. DNA Methylation Analysis*

The library set-up protocol was performed according to Reference [37]. Three specific enzyme combinations were chosen to infer the CG (*Aci*I/MseI), CHG (*Sex*AI/*Mse*I), and CHH (*Eco*T22I/*Mse*I), methylation contexts, respectively. Briefly, for each library, 150 ng DNA were double-digested with one of these enzyme combinations following the protocol previously described [37,38]. The libraries were then pooled, purified using magnetic beads (Agencourt AMPure XP, Beckman Coulter, MA, USA), size selected by gel electrophoresis, and purified using QIAquick Gel Extraction kits (Qiagen, Hilden, Germany) for fragments ranging from 250 bp to 600 bp. Size-selected libraries were quantified using a Qubit® 3.0 Fluorometer (Thermo Fisher Scientific), and a normalized DNA amount (15 ng) was amplified with a primer that introduced an Illumina index (at the Y common adapter site) for demultiplexing. Following PCR with uniquely indexed primers, multiple samples were pooled and subjected to PCR-enrichment, as previously described [37]. The grouped libraries were pooled in an equimolar fashion, and the final library was Illumina-sequenced using 150-bp single-end chemistry. Raw reads from the Illumina sequencing of the CG, CHG, and CHH libraries were analyzed following the protocol and the pipeline previously described [37].

The relative methylation levels at each site were calculated following a described procedure [37] and the DMPs (Differentially Methylated Positions) were called following the methyl kit's manual best practices [39]. The mapping of the DMPs in the same scaffold and as closer than a given threshold provided their clustering together to identify the DMRs (Differentially Methylated Regions), as previously reported [37].

### Synteny Block and Statistical Analysis

Synteny block analysis was performed with MCSCANX with default settings. *F. graminearum* proteins were used as a query against a database of *F. verticillioides* proteins for BLASTP homology searches. The BLASTP results were exported in a tabular format (m 8). The criteria for synteny block analysis were: match score 50, match size >5, gap\_penalty of −1, and max gaps of 25. The chromosomes of *F. graminearum* were partitioned in an

adjacent window of 20 kb and, for each of these regions, the proportion of mapping genes collinear to *F. verticillioides* was calculated. Chromosomal windows with a portion of collinear genes below 50% were considered to be non-conserved. The relative abundance of DMPs and DMR mapping in conserved and not conserved (NC) regions were compared with a permutation test.

The effect of subculturing on pathogen virulence, mycelium growth, conidia production, and metabolite biosynthesis were tested by one-way ANOVA and Duncan's multiple comparison tests, as implemented in the program DSAASTAT [40].

### **3. Results**

### *3.1. Subculturing Reduces Fungal Virulence, but Passaging Can Rescue These Defects*

To assess the effect of continuous subculturing on virulence, three subcultures of the FG8 strain (SC1, SC23, and SC50) were selected for virulence assays toward bread wheat with inoculations performed on stem bases.

A different aggressiveness between SC1, SC23, and SC50 was observed (Figure 2A). All three subcultures showed the ability to cause the typical necrotic lesions on the first leaves and across leaf sheaths of soft wheat plants. In detail, crown rot disease index (DI), calculated as the product between the average length of the necrotic area on the first leaf (cm) and the average value (0–17), was 53.8, 24.5, and 17.4 in plants inoculated with SC1, SC23, and SC50, respectively. The DI decrease was significant (*p* < 0.05) and followed the gradient SC1 > SC23 > SC50.

**Figure 2.** (**A**) Stem base Disease Index (DI) of soft wheat inoculated with subcultures SC1, SC23, and SC50 of *F. graminearum* strain FG8. Columns represent the average of three replicates (± SE) with each composed of 10 plants. Values with different letters are significantly different based on Duncan's multiple comparison tests (*p* < 0.05). (**B**) Average percentage of symptomatic spikelets (%) of heads point-inoculated with the FG8 different subcultures. Columns represent the average of three replicates (±SE), in which each is composed of five heads. Values with different letters are significantly different based on Duncan's multiple comparison tests (*p* < 0.05).

The aggressiveness of the same subcultures was also evaluated toward bread wheat heads. All three subcultures were able to induce the typical FHB bleached spikelets with aggressiveness decreasing by the subculturing time (Figure 2B). These subcultures were compared with SC50×3, which was obtained from mycelia derived from three head-tohead passages, as described in MM. The aggressive average levels described a trend with SC50×3 > SC1 ≥ SC23 ≥ SC50 (Figure 2B). In detail, the initial subculture (SC1) caused 18.5% of symptomatic spikelets whereas the virulence of SC50 was significantly (*p* < 0.05) reduced when compared to the first one, with only 10% of spikelets showing symptoms. SC23, with an average of 16% infected spikelets, showed an intermediate aggressiveness in comparison to SC1 and SC50. These results showed that continuous subculturing on PDA caused a pathogen virulence decrease as a consequence of its adaptation to a nutrient-rich medium while three transfers of SC50 on wheat heads fully restored virulence (38% of bleached spikelets) to a degree even higher than that observed for SC1 (*p* < 0.05).
