Methionine biosynthesis enzyme MoMet2 is required for rice blast fungus pathogenicity by promoting virulence gene expression via reducing 5mC modification

The emergence of fungicide resistance severely threatens crop production by limiting the availability and application of established fungicides. Therefore, it is urgent to identify new fungicidal targets for controlling plant diseases. Here, we characterized the function of a conserved homoserine O-acetyltransferase (HOA) from the rice blast fungus Magnaporthe oryzae that could serve as the candidate antifungal target. Deletion of the MoMET2 and MoCYS2 genes encoding HOAs perturbed the biosynthesis of methionine and S-adenyl methionine, a methyl group donor for epigenetic modifications, and severely attenuated the development and virulence of M. oryzae. The ∆Momet2 mutant is significantly increased in 5-methylcytosine (5mC) modification that represses the expression of genes required for pathogenicity, including MoGLIK and MoCDH-CYT. We further showed that host-induced gene silencing (HIGS) targeting MoMET2 and MoCYS2 effectively controls rice blasts. Our studies revealed the importance of HOA in the development and virulence of M. oryzae, which suggests the potential feasibility of HOA as new targets for novel anti-rice blast measurements.


Introduction
Rice blast caused by Magnaporthe oryzae results in over 10% yield loss worldwide [1,2].To combat this disease, several control measures have been developed, including the most widely used resistant cultivars and fungicides [3,4].However, resistance to disease can be easily lost following continuous cultivation due to fungal adaptability and rapid variation, as well as rapidly developed resistance to fungicides [4].Therefore, it is urgent to develop novel disease control strategies.
In eukaryotes, DNA methylation plays a key role in cellular activities, including genome regulation and development [5][6][7].This process is catalyzed by a family of DNA methyltransferases (DNMTs) that transfer a methyl group from S-adenyl methionine (SAM) to DNA, where it is established and maintained [8].DNA cytosine methylation (5-methylcytosine, 5mC), which occurs in diverse eukaryotic organisms, including animals, plants, and fungi [9], is the first discovered and most extensively studied DNA methylation.5mC is involved in the silencing of transposable elements (TEs) and repeat sequences, genome stability, development, and transcription regulation [10][11][12].
At present, genome-wide DNA methylation examined in several fungi using bisulfite sequencing indicates the global DNA methylation level is very low compared to the higher eukaryotic counterparts [13,14].DNA methylation patterns in fungi such as Beauveria bassiana, Cryphonectria parasitica, Mucor rouxii, Yarrowia lipolytica, and Ustilago maydis are closely related with the developmental stages [15,16].Methylomes assays showed that 0.22 and 1.5% cytosines in the genomes of M. oryzae and Neurospora crassa are methylated, respectively.This methylation contributes to gene expression, transposon silencing, and heterochromatin formation [17][18][19].However, fungal genome-wide DNA methylation during plantfungi interactions has not been reported, and there are few reports to date about methylated promoters mediating gene transcription in fungi.
Various DNMTs are involved in the establishment and maintenance of DNA methylation, but the distributions and numbers of enzymes involved are highly variable [20].DNMTs involve numerous pathways and methionine (Met) biosynthetic pathways are often focused on because they provide Met for generating SAM, an important methyl donor for cellular methylation including DNA methylation.Incidentally, the Met biosynthetic pathways are absent in mammals, making them attractive targets for antimicrobial compound discovery [21].
The canonical Met biosynthetic pathway consists of a sulfur uptake pathway and Met biosynthetic pathways A and B [22].Met can only be synthesized by pathway A in plants and some bacteria, whereas it is synthesized by both pathways A and B in fungi.This suggests that pathway B is specifically synthetic to fungi [23,24].Homoserine O-acetyltransferases (HOAs) catalyze the formation of l-O-acetyl-homoserine from l-homoserine through the transfer of an acetyl group from acetyl-CoA, and they function as the first committed step required for Met biosynthesis pathway B [25,26].Additionally, HOAs are crucial for the synthesis of Met and cysteine (Cys) in the yeast Saccharomyces cerevisiae, and the loss of HOA function results in Met auxotrophs in Candida albicans, Candida guilliermondii, and S. cerevisiae [27][28][29][30].Given the importance of HOA in cellular activities, small molecule inhibitors have been developed that inhibit HOA enzyme activities in human pathogenic fungi [31,32].Interestingly, the loss of HOA function decreases the pathogenicity of Gibberella zeae to wheat and maize [33].These studies implied that HOAs could be potential candidates for fungicide targets to control fungal diseases.To examine if HOAs could serve as potential fungicide targets, we characterized the function of HOAs encoded by MoMET2 and MoCYS2 in M. oryzae.Further, as hostinduced gene silencing (HIGS) and small molecule inhibitor screening are two recently developed approaches to target fungal virulence genes and have the potential to be utilized as the new approaches for managing diseases, we employed HIGS targeting MoMET2 and MoCYS2 that effectively controlled the rice blast.Our studies revealed the importance of HOA enzymes in M. oryzae and demonstrated their great potential as fungicidal targets.

Identification of HOA encoding genes in M. oryzae
To explore novel fungicidal targets in M. oryzae, we focused on key enzymes in the Met biosynthesis pathways.Using BlastP to search homologous protein sequences, we discovered potential HOA homologs that catalyze the first committed step in Met biosynthesis [26] and found that these Met synthetases are specific to fungi and not present in cereal crops (Fig 1).
In S. cerevisiae, the Met HOA encoding gene ScMET2 has been extensively studied previously [34].Using ScMET2 as a tracer to BLAST the genome of M. oryzae (http://fungidb.org/fungidb/), we identified two putative Met2 proteins, MGG_01469 and MGG_14202.Sequence alignments showed that MGG_01469 and MGG_14202 share 44.86% and 20.55% amino acid identity, respectively, with ScMet2 (S1A   ).Both ΔMomet2 and ΔΔMocys2Momet2 mutants failed to produce conidia on straw decoction and corn (SDC) medium, but the ΔMomet2 mutant could still produce conidia on complete medium (CM) (S2 Table ).The defect phenotypes were restored in the ΔMomet2 and ΔMocys2 mutants complemented with WT MoMET2 and MoCYS2, respectively.These results indicated that MoMet2 and MoCys2 are required for vegetative growth and conidiation.
To examine the role of MoMET2 and MoCYS2 in fungal pathogenicity, we inoculated conidial suspensions of WT, ΔMomet2, ΔMocys2, and complementation strains on the susceptible rice cultivar CO39 and barley seedling leaves.The ΔMomet2 and ΔMocys2 mutants were significantly reduced in the infection of rice (Fig 2A , 2B, 2C and 2D).Similar results were observed on the detached barley cultivar four-arris leaves (Fig 2E).As the ΔΔMocys2Momet2 double mutant could not produce conidia, mycelia of WT, ΔMomet2, ΔMocys2, and ΔΔMocys2Momet2 were used to inoculate the detached barley seedling leaves.The ΔMocys2 mutant exhibited very restricted lesions on unwounded leaves, while the ΔMomet2 and ΔΔMocys2Momet2 mutants displayed no obvious lesions (Fig 2E).These results demonstrated that both MoMet2 and MoCys2 are required for the virulence of M. oryzae, with MoMet2 having a stronger role.
To fully assess virulence attenuation, we performed an infection assay on rice sheath and observed invasive hyphae (IH) growth at 100 appressorium penetration sites at 24, 36, 48, and 72 h post-inoculation (hpi).We rated the hyphal growth as type I (no penetration), type II (with primary IH), type III (secondary IH does not extend to neighboring plant cells), and type IV (IH extended into neighboring plant cells) (Fig 2F and 2G) [35].WT and complementation strains exhibited approximately 90% successful appressorium penetration events, with more than 70% type III and IV IH at 36 hpi (Fig 2H).In contrast, only 78% ΔMocys2 appressoria had successful penetration with less than 40% of penetration sites showing type III and IV IH growth (Fig 2H).Strikingly, the ΔMomet2 mutant was restricted to types I, II, and III in IH growth at 24, 36, 48, and 72 hpi within one cell (Fig 2I).These results indicated that MoMet2 and MoCys2 play an important role in IH growth and host colonization.
To test whether MoMet2 and MoCys2 share the conserved HOA function, we expressed both proteins in S. cerevisiae HOA homolog mutant (ΔScmet2) lacking HOA [36], and found that the co-expression of MoMET2 and MoCYS2, but not individually, rescues the growth

Exogenous Met or Cys partially rescue mutants' defects in vegetative growth, conidiation, and virulence
The ΔMomet2 and ΔΔMocys2Momet2 mutants showed sparse mycelial growth on CM and SDC media but not at all on MM (S6A Fig and S2 Table ).However, growth was restored in

Deletion of MoMet2 and MoCys2 decreases Met and SAM biosynthesis, but only MoMet2 has an impact on 5mC
Since HOA is the key enzyme required for Met biosynthesis, we used UPLC-ESI-MS/MS to analyze Met and SAM levels in WT, ΔMomet2, ΔMocys2, and ΔΔMocys2Momet2.A SAM peak matching the retention time (0.37 minutes) of the SAM standard was detected in mycelium samples (Fig 3A).The ΔMomet2, ΔMocys2, and ΔΔMocys2Momet2 mutants had significantly lower levels of Met and SAM than WT (Figs 3E, 3F and S7).To test whether MoMet2 or MoCys2 affect DNA methylation, we examined 5mC levels in these strains by UPLC-ESI-MS/ MS.A peak matching the retention time of the 5mC standard was found in the DNA of M. oryzae (Fig 3C).Moreover, the same base fragment was detected in two samples by MS/MS of 242.1 (mass/charge ratio), which also matched the 5mC standard (Fig 3D).We estimated 5mC abundance and found the global methylation levels (mC/total C) in the ΔMomet2 mutant was significantly higher than that in WT (Fig 3G).No changes were found in the ΔMocys2 mutant (Fig 3H).These results indicated that disruption of the MoMET2 gene alters global methylation in M. oryzae.

The ΔMomet2 mutant has a higher level of 5mC at the gene body and the promoter
Since the ΔMomet2 mutant had a significantly higher level of 5mC than WT, we hypothesized that gene disruption of MoMET2 affects genome-wide DNA methylation patterns.To test this, we performed whole-genome bisulfite sequencing (WGBS) of mycelia from WT and ΔMomet2 (two biological replicates) and found that the overall DNA methylation level was low, with an average weighted methylation percentage (calculated as the number of reads supporting methylation over the number of cytosines sequenced) at CG dinucleotides of 1.14% (Table 1).In addition, DNA methylation in ΔMomet2 differed in the gene body and promoter regions, but not TE (Fig 4A).5mC profiling in methylated genes revealed that 5mC peaks tended to flank the transcriptional start site (TSS) with a clear depletion pattern near the TSS, especially in CG methylation sequences (Fig 4B).

MoMet2 regulates global gene expression in M. oryzae
To examine the potential effect of MoMet2 on gene expression regulation, we performed a transcriptome analysis involving WT and the ΔMomet2 mutant (S8A Fig)   [37].Finally, we performed an integrative analysis of WGBS and RNA-seq on derepressed and repressed transcripts containing significantly altered 5mC modification (Fig 4C).The results showed that DNA methylation occurs in all sequence contexts: CG, CHG, and CHH (H = A, C, T), with CG methylation being predominant.
We next investigated the correlation between DNA methylation and gene expression in the ΔMomet2 mutant.Among 170 genes with hypermethylated promoters in the ΔMomet2 mutant, 11 were upregulated, and only 10 were downregulated.Furthermore, 2 genes were downregulated, and 3 genes were upregulated among 72 genes with hypermethylated gene bodies in the ΔMomet2 mutant (Fig 4D).Of 170 genes with increased 5mC in promoters, only 10 were differentially expressed in the ΔMomet2 mutant.Therefore, a conclusion that 5mC directly represses gene expression cannot be drawn.
Finally, to explore the biological processes MoMet2 might regulate, we performed a KEGG analysis on DEGs that were downregulated by 1.5-fold in the ΔMomet2 mutant.The analysis revealed that the "biosynthesis of secondary metabolites" was highly enriched (S8E Fig) .We therefore explored genes in this metabolic pathway with hypermethylation and low expression: MGG_00068 encoding a gliotoxin biosynthesis protein homolog, MoGliK.GliK has been demonstrated that plays critical roles in gliotoxin biosynthesis of Aspergillus fumigatus [38].Indeed, compared to WT, the ΔMomet2 mutant had significantly reduced MoGliK transcripts, accompanied by hypermethylation in the body and promoter region of the MoGLIK gene (Fig 4E and 4F).

Altered virulence gene expression contributes to MoMet2-associated pathogenicity
In plants, cytosine methylation primarily occurs in CG, CHG, and CHH (H = A, T, or C).We performed a bisulfite conversion reaction using a sequence starting 350 bp upstream of the transcription start site of MoGLIK to analyze the methylation status in its promoter region (Fig 5A ).Consistent with the omics data, the CG, CHG, and CHH methylation levels were higher in the ΔMomet2 mutant than in WT and complementation strains at the mycelium stage (Figs 5B, 5F, 5G, 5H and S10A), accompanied by lowered MoGLIK transcripts in the ΔMomet2 mutant (Fig 5E).To examine the methylation status in response to external stress or during infection, we calculated the methylation levels of WT, ΔMomet2 mutant, and complementation strains under 1 mM DTT stress treatment or after punch inoculation.The CG, CHG, and CHH methylation levels in the promoter region of MoGLIK in the ΔMomet2 mutant were higher than that in WT and complementation strains after DTT stress treatment ( To examine the role of MoGliK in pathogenicity, we inoculated conidial suspensions of WT, the ΔMoglik mutant, and the complemented strains on the susceptible rice cultivar CO39.The ΔMoglik mutant displayed significantly reduced virulence while the complemented strain  was similar to WT (Fig 5I and 5J).In the host penetration assay, the ΔMoglik mutant showed 17% type I, 20% type II, 54% type III, and 9% type IV IH growth while WT and the complemented strains had 8% type I, 16% type II, 42% type III, and 34% type IV IH growth (Fig 5K).These results indicated that MoGliK is required for pathogenicity and IH growth.Similar results were also observed for MGG_07949 encoding a protein MoCDH-cyt with cytochrome domain of cellobiose dehydrogenase which is an extracellular hemoflavoenzyme secreted by fungi to assist lignocellulolytic enzymes in biomass degradation [39], which was randomly selected in the united analysis with hypermethylation and low expression (S9 Fig) .Bisulfite sequencing analysis of methylated cytosine at MoCDH-CYT promoter showed that CG, CHG, and CHH methylation levels were higher in the ΔMomet2 mutant than that in WT at mycelium growth stages and during punch inoculation (S9B, S9C, S9F, S9G, S9H and S11 Taken together, these results indicated that MoMet2 regulates gene expression via 5mC methylation.

Host-induced gene silencing (HIGS) in MoMET2 and MoCYS2 improves rice blast resistance
HIGS is a powerful strategy for developing transgenic rice cultivars to control fungal diseases.Since MoMet2 and MoCys2 contribute to M. oryzae pathogenesis and there were no MoMET2 and MoCYS2 homologs in rice (Fig 1 ), they became the ideal targets for HIGS.After removing the conserved sequence, a 264-bp and a 129-bp DNA sequence remained in the MoCys2-and MoMet2-coding regions, respectively.To design a chimeric hairpin RNAi construct that could silence every single gene or simultaneously two target genes (Fig 6A ), the selected fragments (Fig 6B ) were used to generate a dsRNA sequence with a hairpin structure.The chimeric RNAi construct was first transformed into the WT strain, and the correspondingly silenced target genes were analyzed (S12A, S13A and S14A Figs).The transformants exhibited phenotypes similar to deletional mutants in vegetative growth and virulence (S12B, S12C, S12D, S13B, S13C, S13D, S14B, S14C and S14D Figs).These results indicated that the selected sequence fragments effectively silence target genes in M. oryzae, and, thereby are suitable for HIGS.
Using the same DNA sequence, we generated a siRNA construct and introduced it into Zhonghua11 (Zh11) for HIGS (HIGS-MoCYS2-MET2).We also generated the rice cultivar Zh11 expressing the empty vector as a negative control (CK).We then inoculated Guy11 spore suspensions onto CK and HIGS-MoCYS2-MET2 transgenic plants.At 7d post-inoculation (dpi), the HIGS-MoCYS2-MET2 plants produced significantly smaller lesions than CK (Fig 6C and 6D).To verify whether resistance to M. oryzae in the transgenic rice lines was caused by in planta-derived silencing of the two target genes, we quantified the transcript levels of represent significant differences.(B) 5mC profiles in gene regions.The 2 kb upstream and downstream flanking coding regions were aligned for all genes.TSS, transcription start site; TES, transcription end site.(C) Plot showing integrative analysis of WGBS and RNA-seq to analyze derepressed and repressed transcripts containing significantly altered 5mC modification.The genes associated with hypomethylation and high expression and genes associated with hypermethylation and low expression, red dots.The genes associated with hypomethylation and low expression and genes associated with hypermethylation and high expression, blue dots.(D) Venn diagram showing the overlaps between expressed genes and 5mC-marked genes.a, DMR Hypergenes, b, DEG upgenes, c, DMR Hypogenes, d, DEG downgenes.(E, F) The DNA methylation levels surrounding promoters and gene body regions, and the relative gene expression levels, were shown with screenshots of the Integrative Genomics Viewer (IGV) display.Guy11, blue.ΔMomet2, pink.https://doi.org/10.1371/journal.pgen.1010927.g004The qPCR analysis showed that the transcripts of MoMET2 and MoCYS2 were reduced by approximately 30%-55% and 72%-88%, respectively, in four transgenic lines, compared to CK (Fig 6E).We further inoculated HIGS-MoCYS2-MET2 transgenic plants with Guy11 and expressing the Scmet2 in Guy11 (Scmet2-Guy11), respectively.At 7 dpi, the HIGS-MoCYS2-MET2 plants infected with Guy11 produced significantly smaller lesions than that infected with Scmet2-Guy11 (Fig 6F and 6G).Moreover, there was no significant difference in the transcripts of MoMET2 and MoCYS2 (Fig 6H).These findings indicated that HIGS-MoCYS2-MET2 improves rice resistance against M. oryzae.

Discussion
Although available antifungal agents are effective in combating rice blasts, the emergence of resistance to these agents facilitates a new understanding of fungal pathogens, plant-fungi interactions, and virulence factors.The study of new antifungal drug targets and a better understanding of fungal pathogenesis are both anticipated benefits of characterizing virulence factors.In our study, we identified and characterized the functions of MoMet2 and MoCys2 from M. oryzae.We found that MoMet2 regulates gene expression via 5mC modification during infection and that MoGLIK is one of the key target genes (Fig 6I).In WT, hypomethylation of the MoGLIK promoter and gene body leads to its high transcription, which contributes to M. oryzae virulence.In contrast, the MoGLIK promoter and gene body are highly methylated in the ΔMomet2 mutant, which impedes MoGLIK transcription to inhibit the pathogenicity of M. oryzae.Of note, HIGS targeting MoMET2 and MoCYS2 effectively controls rice blasts.Our studies revealed the molecular mechanisms of MoMET2 mediating gene expression via 5mC modification in M. oryzae and provided a candidate target that can be used as both a novel fungicide target and in HIGS to effectively control the rice blast.
In the last few years, the stable plant transformation-based HIGS system has been proven to efficiently control certain pests, nematodes, filamentous pathogens, and parasitic plants [40][41][42][43][44]. Target gene selection is the most important prerequisite for HIGS.In general, genes that play a vital role in pathogen growth, development, and pathogenicity are considered HIGS target candidates.However, the selected gene fragment, size, and sequence specificity, as well as positions of dsRNA, all need to be considered for HIGS, and thus few effective HIGS targets have been uncovered [45][46][47].Met biosynthetic pathways are often focused because they are absent from mammals, making biosynthetic enzymes attractive targets for antimicrobial discoveries.Met is widely involved in vegetative growth, asexual development, multiple stress resistance, and the pathogenicity of various fungal pathogens.HOA participates in the first committed step of Met metabolism, and loss of HOA function causes defects in fungal development, and virulence [33].In Cryptococcus neoformans, a small-molecule screen identified the first inhibitor of fungal HOA [31].In Candida albicans, L-and D-penicillamine exhibit antifungal activities by inhibiting the function of HOA [32].These findings provide support for the ability to target HOAs with antifungal compounds.Consistent with these findings, transgenic rice produced in our HIGS experiments exhibited strong resistance to M. oryzae (Fig 6C -6E) due to the downregulated expression of MoCYS2 and MoMET2 by HIGS (Fig 6F -6H).It would be interesting to test whether existing HOA inhibitors are effective against M. oryzae, and to develop novel fungicides with stronger effects targeting HOA.Met is not only an essential amino acid for various biological processes but also can be converted to SAM, donating methyl groups to more than 80 biological reactions [48].Some Metrelated genes have been previously reported to be associated with the pathogenicity of M. oryzae [49][50][51][52], but the relationship between them and DNA methylation is unclear.A positive correlation between SAM contents and DNA methylation has been hypothesized and confirmed in some studies [53].However, a previous study suggested that the SAM content of female mice liver was reduced by methyl deficiency while their 5mC content was rather increased [54].Additionally, 12 h cultures of Met-starved met-1 mutant and Cys-starved cys-3 mutant in Neurospora crassa displayed ~3-fold less SAM than did WT cultures, whereas their methylation levels appeared comparable with WT [55].These suggested that SAM contents is not always linear to DNA methylation in the genome.Subtle reduced SAM would have drastic cellular effects including induced DNA methylation, when SAM is reduced to a certain level, SAM primarily provides methyl groups for the DNA methylation reaction [56], maybe a reason of resulting in significantly increased DNA methylation in the ΔMomet2 mutant (Fig 3G).The ΔMocys2 mutant showed no effects on the methylation (Fig 3H), indicating that other methyl donors [57] may substitute SAM for its function on DNA methylation in fungi.Moreover, the DNA methylation process showed its complexity among organisms, even in the case of methyltransferase knockout, not all changes occur in a singular direction [58].
Not only that, the fluctuation of Met content alter the methylation levels of mammalian including some locus-specific DNA and histone methylation, which is sensitive to Met [59,60].It has been demonstrated that hypermethylation in the reelin promote in the cortex of mice upon Met supplementation [61] and administration of Met led to prominent hypermethylation of the brain-derived neurotrophic factor (Bdnf) DNA [62].The methylation of MoGLIK was similar to Guy11 in the complemented strains (Figs 5 and S10), indicating that Met2 is associated with MoGLIK methylation in direct or indirect manner.Based on these results, we considered that maybe MoGliK was sensitive to DNA methylation which leads to the only gene was evaluated among this gli gene cluster.
DNA methylation links genetic components and environmental changes.The global DNA methylation level is very low in fungi when compared to other eukaryotes [63].However, it plays an important role in gene regulation and development [64].In M. oryzae, the existence of DNA methylation was examined by restriction enzyme-based work on a retrotransposon, MAGGY [19].A comparative methylome analysis showed that mycelia are the tissue with the highest number of methylated cytosine and that significantly proportion of these methylation sites disappears in conidia and appressoria of M. oryzae [18], suggesting that DNA methylation patterns change according to developmental stage.In contrast, our studies revealed a difference in DNA methylation in the promoter and gene body between WT and the ΔMomet2 mutant especially during infection.Our findings thus highlight the importance of virulence gene expression regulated by DNA methylation in M. oryzae.
Finally, during the characterization of genome-wide methylation in M. oryzae, we attempted to inject fungi spores into the bundle sheath cells and collected IH by infection by Guy11 at 7 dpi.(E) Relative mRNA expression of two target genes of M. oryzae during Guy11 infection of transgenic lines and CK at 7 dpi.(F) Disease phenotypes of CK and HIGS-MoCYS2-MET2 transgenic plants at 7 dpi by the strain expression the S. cerevisiae ortholog in M. oryzae Guy11 (Scmet2-Guy11).Rice cultivar CK and HIGS-MoCYS2-MET2 were punch-inoculated with Scmet2-Guy11 (1 × 10 5 spores ml -1 ).CK, the rice cultivar Zh11 expressing empty vector.(G) Quantitative analysis on lesion length of rice after infection by and Guy11 and Scmet2-Guy11 at 7 dpi.(H) Relative mRNA expression of two target genes of M. oryzae during Scmet2-Guy11 and Guy11 infection of transgenic lines at 7 dpi.G, transgenic lines inoculating with Guy11; S, transgenic lines inoculating with Scmet2-Guy11.(I) Proposed model for the roles of methylation of the MoGLIK promoter in regulating pathogenicity.Error bars represent the SD and different letters indicate significant differences (P < 0.05) tested by one-way ANOVA with Duncan's post hoc test for (D and E).Significant differences were tested by Student's t test for (G and H): *P < 0.05, **P < 0.01.https://doi.org/10.1371/journal.pgen.1010927.g006microdissection.However, the bundle sheath cells were too thin to be fixed, and the amount of IH was insufficient for sequencing analysis, despite separating bundle sheath-infected cells by single-cell flow cytometer.In resequencing to examine the ratio of plant genome to pathogen genome during infection, M. oryzae accounted for only 1.044% of the detected genomes, with the rest belonging to the plant (S15 Fig) .Therefore, it is difficult to distinguish fungi genomewide DNA methylation during plant-fungi interactions.To circumvent the difficulty in distinguishing fungal DNA methylation during infection, we examined the role of methylation during stress and infection at the individual level.Stress treatment with DTT (1 mM) and punching resulted in the same trend as the mycelial stage, whereby the DNA methylation level of target genes is higher in the ΔMomet2 mutant than in WT.Comparison of the dynamic of DNA methylation under different development and infection stages may suggest the critical role of DNA methylation in altering pathogen strategies to infect its host plants.Our studies revealed differences in methylation levels from promoter and gene body regions at the mycelial stage, and this difference is most drastic during infection.Therefore, monitoring the fungus genome-wide DNA methylation status in the plant-fungi interaction with deficient methylation levels is expected to develop more accurate or targeted methods.
To test the homologies of MoMET2 and MoCYS2 to ScMET2, the full-length cDNA of MoMET2 and MoCYS2 were individually cloned into the vector pYES2 (Invitrogen).Subsequent to PCR sequencing and screening on SD medium without uracil, the MoMet2-pYES2, MoCys2-pYES2 constructs and MoMet2-pYES2, along with MoCys2-pYES2 combined constructs, were each transformed into the yeast ΔScmet2 mutant (BY4741 loss of YNL277W).Yeast colonies were cultured on YPD medium (10 g Yeast Extract, 20 g Peptone, 20 g D-Glucose) and adjusted to an OD600 value of 0.1.Following this, 10-fold serial dilutions were cultured on SD-Met (galactose) plates at 30˚C for 4 d and subsequently [65].

Strains and cultural conditions
The wild type (WT) was Guy11 strain in this study.For vegetative growth, small agar blocks (2 × 2 mm) were cut from the edge of 4-d-old cultures and placed onto complete medium (CM), minimal medium (MM) or straw decoction and corn (SDC) for culture in the dark at 28˚C for 7 d.The diameter growth was measured after incubation for 7 d.Liquid CM medium was used to prepare mycelia for DNA and RNA extraction.For conidia production, mycelia were grown in the dark on SDC or CM medium at 28˚C for 7 d, followed by constant illumination for 3 d under fluorescent light [66].

Targeted gene deletion and complementation
The gene deletion mutants were generated using the standard one-step gene replacement strategy.First, two 1.0 kb sequences flanking the targeted genes were PCR amplified with primer pairs (S1 Table ).Then, the resulting PCR products of MoMET2 were digested with restriction endonucleases and ligated with the HPH cassette released from pCX62.Finally, the completed inserts were sequenced.The 3.4 kb fragments, which contain the flanking sequences and hygromycin resistance cassette, were amplified and transformed into protoplasts of WT.Putative mutants were screened by PCR and confirmed by Southern blotting analysis.The complement fragments, which contain the entire MoMET2 gene with its native promoter regions, were amplified by PCR (Phanta Super-Fidelity DNA Polymerase, Vazyme Biotech Co., Nanjing, China) with primers (S1 Table ) and inserted into pYF11 (bleomycin resistance) to complement the respective mutant strains.Other gene deletion mutants and corresponding complemented strains were obtained using the same strategy.

Virulence and invasive growth assays
For infection, conidia were harvested from 10-d-old SDC agar cultures, filtered, and resuspended to a concentration of 5 × 10 4 spores ml -1 in a 0.2% (w/v) gelatin solution.The conidia were sprayed onto two-week-old seedlings of rice (Oryza sativa cv.CO39) [67].For rice leaves, 5 ml of a conidial suspension of each treatment was sprayed.For the detached barley inoculation assay, conidia were suspended to a concentration of 1 × 10 5 spores ml -1 in a 0.2% (w/v) gelatin solution, and 20 μl conidia were inoculated onto leaves of 7-day-old barley seedlings [68].Inoculated plants were kept in a growth chamber at 25˚C with 90% humidity and in the dark for the first 24 h, followed by a 12 h /12 h light /dark cycle.Lesion formation was observed daily and recorded by photography 7 dpi.The degree of disease lesions was analyzed by ImageJ [69].
For observation of the penetration and invasive growth in rice cells, conidial suspensions (1 × 10 5 spores ml -1 ) of WT, mutants and complement strains were injected into rice leaf sheath.The inner epidermises of infected sheaths were harvested at different hours post-inoculated by WT and mutant strains and observed under a microscope.

UPLC-ESI-MS/MS analysis for Met, SAM, and 5mC
A HPLC-ESI-MS/MS system, consisting of electrospray-time-of-flight mass spectrometry (Triple TOF 5600+, AB Sciex) and liquid chromatography (LC-20ADXR HPLC, Shimadzu), was used for 5mC detection.Data acquisition and processing were performed using PeakView version 2.0 (AB Sciex).The HPLC separation was performed on a reversed-phase column (C18, 2.1 mm × 100 mm, 2.6 μm; Kinetex) with a flow rate of 0.2 ml/min at 40˚C.Formic acid (FA) in water (0.1%, v/v, solvent A) and FA in methanol (0.1%, v/v, solvent B) were employed as the mobile phase.A gradient of 0.5 min 2% B, 2 min 10% B, 3 min 90% B, 3.5 min 2% B and 5 min 2% B was used.The mass spectrometry detection was performed under ESI positive mode with a DuoSpray dualion source.The retention time of 5mC is around 0.74 min.The mass transition for 5mC was set as m/z 242.11 > 126.1 [70].
Su et al. developed a UHPLC-ESI-MS/MS method to simultaneously determine the levels of 14 metabolites, including 4 Met metabolism metabolites (Met, homocysteine, SAM and Sadenosylhomocysteine) [71].Based on the sample preparation and experimental methods, we discovered the ideal test suitable for M. oryzae.The ACQUITY UPLC BEH Amide column (2.1 × 100 mm, 1.7 μm, Waters, USA) was used for metabolite separation at 30˚C.The mobile phases were water (A) and Acetonitrile (ACN) (B) both containing 0.1% FA.The LC gradient program was optimized as follows: 0.5 min 15% B, 3 min 85% B, 4 min 85% B, 4.01 min 15% B, 5 min 15% B. The injection volume was 1.0 μl.

RNA extraction, RNA-seq, and qRT-PCR
Total RNA as isolated from 2-d-old hyphae using Omega Total RNA Kit I according to the manufacturer's manual.RNA quantity and quality was measured using Nanodrop ND-1000 and 1% agarose gel electrophoresis.RNA-seq service was provided by Novogene Corporation (Beijing, China).First-strand of cDNA was synthesized from 1 μg total RNA using HiScript III RT SuperMix (R323, Vazyme, China).ChamQ Universal SYBR qPCR Master Mix (Q711, Vazyme, China) was used for quantitative RT-PCR (qRT-PCR) reactions.QRT-PCR was performed using the ABIPRISM 7500 Fast Real-Time PCR System.The expression of each gene was normalized to the expression of internal control ACTIN, and primers used for qRT-PCR were listed in S1 Table.

Whole-Genome bisulfite sequencing and analysis
Genomic DNA of 2-d-old Guy11 and the ΔΜοmet2 mutant mycelium were isolated using TIANGEN DNAquick Plant Systerm (DP321) according to the manufacturer's manual.Genomic DNA degradation and contamination was validated by agarose gels.DNA purity was checked using NanoPhotometer spectrophotometer (IMPLEN, CA, USA).DNA concentration was measured using Qubit DNA Assay Kit in Qubit 2.0 Flurometer (Life Technologies, CA, USA).
A total amount of 100 ng genomic DNA spiked with 0.5 ng lambda DNA was fragmented by sonication to 200-300 bp with Covaris S220.These DNA fragments were treated with bisulfite using EZ DNA Methylation-Gold Kit (Zymo Research), and the library was constructed by Novogene Corporation (Beijing, China).Subsequently, pair-end sequencing of the sample was performed on Illumina platform (Illumina, CA, USA).Library quality was assessed on the Agilent Bioanalyzer 2100 system.
The library was sequenced on Illumina Novaseq platform.Image analysis and base calling were performed with Illumina CASAVA pipeline, and finally generated 150bp paired-end reads.
We used FastQC (fastqc_v0.11.5) to perform basic statistics on the quality of the raw reads.Then, those read sequences produced by the Illumina pipeline in FASTQ format were pre-processed through fastp (fastp 0.20.0).The remaining reads that passed all the filtering steps was counted as clean reads, and all subsequent analyses were based on this.Finally, we used FastQC to perform basic statistics on the quality of the clean data reads.

Bisulfite sequencing
Genomic DNA was extracted from WT and the ΔΜοmet2 mutant mycelium.Bisulfite treatment was conducted using the EZ DNA Methylation-Gold Kit (D5005, Zymo Research, USA).Based on MGG_00068, MGG_07949 promoter sequences, the bisulfite primers were designed through Meth Primer software (http://www.urogene.org/cgi-bin/methprimer/methprimer.cgi) to test the methylation status of the target regions.The bisulfite-treated DNA was used for PCR amplification with ZymoTaq DNA Polymerase (E2001, Zymo Research, USA).The PCR products were cloned into the T-Vector pMD 19 (3271, TaKaRa, Japan) and 12 independent clones were sequenced for each variety.For each sample, at least 12 individual clones were sequenced, and the sequencing data were analyzed using Kismeth software (http://katahdin.mssm.edu/kismeth/revpage.pl) and CyMATE software (http://www.cymate.org/).The primer sequences used for bisulfite sequencing are listed in S1 Table.

Construction of gene silencing vectors
The HIGS experiments mainly followed the method reported previously [72,73].The silencing vector pSilent-1 was constructed by inserting a PCR-amplified fragment of the M. oryzae CUT intron with multiple cloning sites.The 5' multiple cloning site includes unique recognition sites for XhoI, SnaBI, and HindIII, and the 3'multi-cloning site includes those for BglII, SphI, StuI, KpnI, and ApaI [74].We identified the protein's conserved domains and thus presumed its putative function via the CD-Search interface (https://www.ncbi.nlm.nih.gov/Structure/cdd/docs/cdd_search.html).In order to silence the MoCYS2 and MoMET2 in M. oryzae, a 264-bp variable domain fragment of MoCYS2 and a 129-bp variable domain fragment of MoMET2 were synthesized, and were subsequently inserted into pSilent-1 in sense and antisense orientation, respectively.The constructed pSilent-1-MoCYS2, pSilent-1-MoMET2 and pSilent-1-MoCYS2-MoMET2 were co-transformed into the protoplasts of the WT strain.Then the fragment containing the gene-specific DNA fragments in two opposite orientations and the GUS linker were amplified and cloned into the pU2301 vector and the recombinant plasmids were used for rice transformation.Then the expression of target genes was analyzed by qRT-PCR.

Statistical analysis
Results are presented as the mean ± standard deviation (SD) of at least three replicates.Student's t tests were performed using Excel.One-way analysis of variation (ANOVA) followed by Duncan's new-multiple range tests was performed using SPSS 2.0.Statistical tests are described in the figure legends.
MoMet2 and MoCys2 are required for vegetative growth, conidiation, and virulence of M. oryzaeTo examine the functions of MoMet2 and MoCys2, we generated individual ΔMomet2 and ΔMocys2, and double ΔΔMocys2Momet2 mutant strains (S3 Fig).The individual ΔMomet2 and ΔMocys2 mutants displayed apparent defects in vegetative growth and conidiation in comparison to the wild-type (WT) strain (S2 Table

Fig 2 .
Fig 2. MoMet2 and MoCys2 are important for virulence in M. oryzae.(A, B) Pathogenicity assay.Four milliliters of conidial suspension (5 × 10 4 spores ml -1 ) of each strain were used for spraying on Oryza sativa cv.CO39 and photographed 7 dpi.(C, D) Diseased leaf area analysis.Data are presented as a bar chart showing the percentage of lesion area analyzed by ImageJ.Error bars represent the SD of three biological replicates and different letters indicate significant differences (P < 0.05) tested by one-way ANOVA with Duncan's post hoc test.(E) Pathogenicity test on barley leaves.Unwounded barley leaves were incubated with mycelial pellets of different strains.Photographs were taken at 5 dpi.(F, G, H, I) Statistical analysis for each type of invasive hyphae (IH) shape.Type I, no penetration; type II, only with a single IH without branches; type III, with 1-3 branches but restricted in one cell; type IV, more than three branches and extended to neighboring plant cells.For each tested strain, 100 infecting hyphae (n = 100) were counted per replicate, and the experiment was repeated three times with similar results.Bar = 10 μm.https://doi.org/10.1371/journal.pgen.1010927.g002 Pearson correlation analysis revealed a distinct pattern of the global transcriptome between the two strains (S8B Fig).Additionally, differentially expressed gene (DEG) analysis (|fold change|>1.5)uncovered that 693 genes were significantly upregulated while 627 genes were downregulated in the ΔMomet2 mutant (S8C Fig).The dendrogram suggested a high repeatability of WT and the

Fig 3 .
Fig 3. MoMET2 is involved in 5mC DNA methylation in M. oryzae.(A) M. oryzae mycelium SAM was detected by UPLC.The selective multiple reaction monitoring (MRM) transitions for SAM were set as m/z 399.1-250.1.Peak diagram of M. oryzae mycelium.(B) M. oryzae SAM are detected by UPLC-ESI-MS/MS.The parent ion SAM (m/z near 399.1) and base fragment (m/z near 250.1)highlighted by arrows from samples also match standard SAM.(C) M. oryzae mycelium 5mC was detected by UPLC.The selective multiple reaction monitoring (MRM) transitions for 5mC were set as m/z 242.1-126.1.Peak diagram of M. oryzae gDNA samples.(D) M. oryzae 5mC is detected by UPLC-ESI-MS/MS.The parent ion 5mC (m/z near 242.1) and base fragment (m/z near 126.1)highlighted by arrows from samples also match standard 5mC.(E, F, G, H) Quantification of Met, SAM, and 5mC levels in M. oryzae mycelium samples.The 5mC concentrations are listed as 5mC per billion dC.Error bars represent the SD and different letters indicate significant differences (P < 0.05) tested by oneway ANOVA with Duncan's post hoc test for (E and F) and significant differences were tested by Student's t test for (G and H) (*P < 0.05).https://doi.org/10.1371/journal.pgen.1010927.g003 Figs 5C, 5F, 5G, 5H and S10B) or punch inoculation (Figs 5D, 5F, 5G, 5H and S10C), accompanied by lowered MoGLIK transcripts in the ΔMomet2 mutant (Fig 5E).

Fig 4 .
Fig 4. Integrative analysis of WGBS and RNA-seq identified transcripts affected by MoMET2.(A) Violin plot of the distribution of DNA methylation levels quantified as weighted methylation over genes, promoters, and transposable elements (TEs).Cytosine methylation was analyzed in the CG, CHG, and CHH sequence context.Methylation was measured in WT and ΔMomet2 mutant strains.By averaging the weighted values of comparable methylation in the same original of the repeated samples, the Wilcox test-based p values for the number on the figures were determined.P < 0.05 Figs), accompanied by lower MoCDH-CYT transcripts in the ΔMomet2 mutant (S9D Fig).Moreover, the ΔMocdh-cyt mutant displayed a significantly reduced virulence in comparison to the WT and complemented strains (S9E Fig).

Fig 5 .
Fig 5. MoMet2 represses MoGLIK expression via DNA methylation on HY, stress, and infection.(A, B, C, D) Distribution of cytosine DNA methylation in three contexts in 350 bp of the MoGLIK promoter region in the Guy11, ΔMomet2 and the complemented mutant ΔMoMET2/MoMET2 HY. (B) HY added 1mM DTT (C) and punched rice (D) as measured by bisulfite sequencing.Sequencing data were analyzed using Kismeth software.CpG, lilac; CpHpG, blue; CpHpHp, green.At least 12 clones were sequenced per sample.(E) The relative expression of MoGLIK in different treatments.(F, G, H) Methylation levels of CpG, CpHpG, and CpHpHp in Guy11, ΔMomet2 and the