DNA methylation and demethylation pathways regulate many aspects of plant development and stress responses through reversible, non-genetic modification of cytosine to 5-methylcytosine (5mC) (Fransz and De Jong, 2002; Henderson and Jacobsen, 2007; Zilberman et al., 2007; Zhang et al., 2010; Matzke and Mosher, 2014; Du et al., 2015; López Sánchez et al., 2016; Hewezi et al., 2018; Hewezi, 2020). In plants, this occurs in the CG, CHG, and CHH sequence contexts (where H is any nucleotide except G) (Lister et al., 2008). The de novo formation of 5mC is catalyzed by DNA methyltransferases DOMAINS REARRANGED METHYLTRANSFERASE2 (DRM2) and DRM3 through the RNA-directed DNA methylation (RdDM) pathway (Cao and Jacobsen, 2002a; Cao and Jacobsen, 2002b; Cao et al., 2003; Henderson et al., 2010). Following DNA replication and cell division, DNA methylation is maintained in a sequence context-specific manner (Saze et al., 2003). CG methylation is maintained mainly by METHYLTRANSFERASE1 (MET1). MET1 paralogs, MET2b and MET3, also contribute to CG maintenance but to a much lesser extent (Kankel et al., 2003; Quadrana et al., 2016). CHROMOMETHYLASE1 (CMT1), CMT2, CMT3, and the RdDM pathway are responsible for maintaining non-CG methylation (Henderson et al., 2010; Law and Jacobsen, 2010; Costa-Nunes et al., 2014; Stroud et al., 2014). 5mC can be actively removed and replaced by cytosine via the base excision repair process mediated by the paralogous DNA demethylases REPRESSOR OF SILENCING 1 (ROS1), DEMETER (DME), DME-LIKE2 (DML2) and DML3 (Choi et al., 2002; Gong et al., 2002; Agius et al., 2006; Morales-Ruiz et al., 2006; Penterman et al., 2007b; Gehring et al., 2009; Zhang and Zhu, 2012). Together, DNA methylation and demethylation fine-tune genome-wide methylation levels and subsequent transcriptional reprogramming (Gong et al., 2002; Penterman et al., 2007a; Penterman et al., 2007b; Zhu et al., 2007; Ortega-Galisteo et al., 2008; Lei et al., 2015; Bennett et al., 2022).

Recent studies have provided intriguing evidence for the involvement of DNA methylation and active demethylation pathways in modulating defense responses against various phytopathogens (Hewezi et al., 2018). Genome-wide methylation profiling of Arabidopsis leaves inoculated with Pseudomonas syringae pv. tomato DC3000 (Pst) revealed widespread changes in plant methylomes, particularly in gene-rich regions (Dowen et al., 2012). Pst-induced DNA methylation changes were associated with differential transcript abundances of a significant number of stress-responsive genes (Dowen et al., 2012). Similarly, infection of canola (Brassica napus) by the fungal pathogen Leptosphaeria maculans triggered differential DNA methylation in the promoters of thousands of genes both in resistant and susceptible cultivars, of which numerous genes are associated with defense responses (Tirnaz et al., 2020). The correlation between genome-wide DNA methylation patterns and gene expression levels were also reported in rice and tobacco in response to infection by rice Black Streaked Dwarf Virus and Cucumber Mosaic Virus, respectively (Wang et al., 2018; Li et al., 2020).

Changes in DNA methylation levels seem to impact plant interactions with fungal pathogens and oomycetes. For example, hypomethylated mutants defective in the establishment of DNA methylation increased plant resistance to Hyaloperonospora arabidopsidis (Hpa), whereas hypermethylated mutants impaired in active DNA demethylation increased plant susceptibility to both Hpa and Fusarium oxysporum (Le et al., 2014; Lopez Sanchez et al., 2016). Gene expression analyses of these mutants provided strong evidence for the involvement of DNA methylation in the regulation of numerous genes with defense- and stress-related functions. Notably, modulation of defense- and stress-responsive genes in hyper- or hypomethylated mutants showed opposite responses to various phytopathogens (López Sánchez et al., 2016).

Cyst nematodes (Globodera and Heterodera spp.) and root-knot nematodes (Meloidogyne spp.) are sedentary obligate biotrophs that infect the root systems of a broad-spectrum of host plants, including the model plant Arabidopsis. Nematode parasitism of host plants is characterized by the formation of syncytia and giant-cells as permanent feeding sites for cyst and root-knot nematodes, respectively. The formation of giant-cells by root-knot nematodes is accompanied by the formation of galls at the site of infection as a result of increasing cell division of neighboring cells. The parasitic second-stage juveniles (J2s) feed from these metabolically hyperactive feeding cells and develop into adult females and complete the life cycle. Several experimental evidences indicate that epigenetic mechanisms, particularly DNA methylation, play important roles in plant-nematode interactions (Hewezi, 2020). For example, mutants partially impaired in non-GC DNA methylation exhibited reduced susceptibility to cyst and root-knot nematodes (Hewezi et al, 2018; Ruiz-Ferrer et al., 2018; Atighi et al., 2021). Genome-wide DNA methylation analysis of soybean roots exposed to soybean cyst nematode (SCN, Heterodera glycines) revealed widespread DNA hypomethylation in the promoter and transcribed regions of a substantial number of genes (Rambani et al., 2015) that were previously reported to be differentially expressed in the syncytium (Ithal et al., 2007; Klink et al., 2009, Klink et al., 2010; Kandoth et al., 2011). Similarly, beet cyst nematode (BCN, Heterodera schachtii) was found to induce considerable increases in hypomethylation levels of protein-coding genes and transposable elements (TEs) (Hewezi et al., 2017; Piya et al., 2017). Further analysis indicated that BCN-induced DNA methylation changes may directly impact the transcript abundance of more than one-fourth of differentially expressed genes in syncytium (Hewezi et al., 2017). Methylome analysis of two near isogenic lines of soybean differing in their response to SCN revealed distinct and specific methylation profiles over protein-coding and miRNA genes as well as TEs (Rambani et al., 2020a; Rambani et al., 2020b). While the susceptible line exhibited global hypomethylation patterns, the resistant line showed global hypermethylation (Rambani et al., 2020a; Rambani et al., 2020b). Global DNA hypomethylation patterns, predominantly in the CHH context, were also observed in early developing galls induced by the root-knot nematode Meloidogyne graminicola in rice (Atighi et al., 2021). In contrast, it has been recently shown that early developing galls formed by Meloidogyne javanica have undergone hypermethylation particularly in the CHG context (Silva et al., 2022).

The biological significance of DNA methylation changes induced by parasitic nematodes was established using various functional assays of overexpression and mutant analyses (Hewezi et al., 2017; Rambani et al., 2020a; Rambani et al., 2020b; Atighi et al., 2021; Silva et al., 2022). However, the transcriptional activity of key genes involved in DNA methylation and active demethylation in nematode feeding sites remains mostly unknown. We addressed this issue by profiling the spatio-temporal expression patterns of genes involved in maintenance and de novo establishment of DNA methylation as well as those involved in active DNA demethylation in the syncytia and galls induced respectively by H. schachtii and M. incognita in Arabidopsis roots. Also, mutants of these genes were assayed for susceptibility to H. schachtii and M. incognita. Our analyses revealed a key role of DNA methylation and active demethylation in mediating plant susceptibility to two evolutionary distant nematode species.

Our analysis revealed an important role of DNA methylation and active demethylation pathways in mediating Arabidopsis interactions with H. schachtii and M. incognita. Promoter activity and mutant analyses of 12 key genes involved in DNA methylation or active demethylation unveiled important similarity and distinct differences between these two pathosystems in term of expression patterns and plant susceptibility. For example, the promoter activity of MET1 and its homolog MET3 were both detected in the syncytia and galls during the early stage of infection. In contrast, the expression of MET2b was detected only in galls at 4 dpi ( Figures 1A , 3A ), a finding that is consistent with increased susceptibility of met2b mutant plants (SALK_093835C) to M. incognita but not H. schachtii ( Figures 2B , 4B ). Strikingly, no such activity of MET3 promoter was detected in the root tissues under non-infected conditions (Bennett et al., 2021), but this promoter was activated specifically in the syncytia and galls only at 3 and 4 dpi, respectively ( Figures 1A , 3A ). Consistent with this finding, mutant alleles of MET3 were more susceptible to both H. schachtii and M. incognita ( Figures 2B , 4B ). The finding that MET3 expression is restricted to the developing embryo (Jullien et al., 2012; Bennett et al., 2021) suggests a unique role of the encoding enzyme in reprogramming CG methylation in the syncytia and galls in a way similar to the developing embryos, which undergo extensive DNA methylation changes (Xiao et al., 2006; Gehring et al., 2009; Jullien et al., 2012; Wang et al., 2015; Satyaki and Gehring, 2017).

Notably, none of the examined mutants of the CMT genes were found to alter plant susceptibility to H. schachtii despite the strong activation of CMT1 and CMT3 in the syncytia during both early and late stages of infection. While this finding can be attributed to a possible functional redundancy between CMT1 and CMT3, whose expression was detected in the syncytia ( Figure 1B ), this explanation doesn’t support the results showing that single mutants of this gene family were altered in their susceptibility to M. incognita ( Figures 4B, C ). One possible explanation is that global loss of CHG methylation in cmt1 and cmt3 (Stroud et al., 2013) mutants have little or no effect on syncytium formation and function. Alternatively, global decrease in CHG methylation can be compensated with an increase of CG and CHH methylation in the syncytium as recently reported (Rambani et al., 2020b). In support of this hypothesis, the expression of MET1 and MET2b as well as DRM2 and DRM3, which establish CHH methylation de novo, are strongly induced in the syncytia (Figures 1A-B). Although DRMs are established as CHH methyltransferases, they regulate CHG methylation as well (Cao and Jacobsen, 2002b; Stroud et al., 2013).

The intense GUS staining of the DRM2 and DRM3 promoters in the syncytia and galls ( Figures 1B , 3B ) along with the results showing that mutant alleles of DRM2 and DRM3 were more susceptible to both M. incognita and H. schachtii point to the importance of establishing DNA methylation de novo in syncytial and gall cells as a defense mechanism to control cyst and root-knot nematode infection. Similarly, RdDM-defective mutants were found to be more susceptible to Botrytis cinerea and Plectosphaerella cucumerina (López et al., 2011). More recently, it was shown that tomato DRM5 expression is significantly upregulated in the roots of a resistant tomato cultivar upon infection with a virulent root-knot nematode field population (Leonetti and Molinari, 2020). However, various rice mutants deficient in the RdDM pathway were reported less susceptible to M. graminicola despite the fact that more than 99% of the identified differentially methylated regions in the galls were CHH-hypomethylated (Atighi et al., 2020). Similarly, it has been recently shown that M. javanica reproduction on drm1/drm2 double mutant was greatly reduced (Silva et al., 2022). Together, these findings suggest that the RdDM pathway regulates plant resistance against various phytopathogens (Erdmann and Picard, 2020) including cyst and root-knot nematodes.

It may be important to mention that the differences between the two alleles of MET3 (SALK_099592C and SALK_024049C), CMT1 (SALK_138685C and SALK_030404C), and DRM3 (SALK_136439C and SALK_024820C) in response to infection by H. schachtii or M. incognita are most likely due to difference in gene expression as shown in Supplementary Figure 1B . Similarly, of the two mutant lines of MET2b (SALK_102231 and SALK_093835C) only SALK_093835C exhibited increased susceptibility to M. incognita despite the fact that MET2b is expressed at a similar level in both mutants. The differences between the two mutant alleles could be related to different T-DNA insertion sites ( Supplementary Figure 1A ), which may impact the function of truncated transcripts.

Remarkably, the promoter activity of ROS1, DME, DML2, and DML3 exhibited opposite expression patterns in the syncytia and galls ( Figures 1C , 3C ). With the exception of DML2 upregulation in the syncytium at 3 and 7 dpi, the remaining three DNA demethylase genes were either downregulated or not expressed at a detectable level in the syncytia ( Figure 1C ). In contrast, strong GUS activity was observed for all four demethylase genes in young galls at 4 dpi ( Figure 3C ). Consistent with the contrasting expression patterns, the mutant alleles of ROS1, DML2, and DML3 showed opposite responses to infection by cyst and root-knot nematodes. Indeed, the DNA demethylase mutants exhibited increased susceptibility to H. schachtii and decreased susceptibility to M. incognita ( Figures 2C , 3C ). The involvement of active DNA demethylases in disease resistance has been reported in few studies. For example, a loss-of-function mutation in the Arabidopsis ROS1 gene enhanced plant susceptibility to the bacterial pathogen Pseudomonas syringae pv. tomato DC3000 (Pto DC3000) (Yu et al, 2013). In addition, Arabidopsis ros1/dml2/dml3 (rdd) triple mutant displayed enhanced disease susceptibility to the fungal pathogen Fusarium oxysporum (Le et al., 2014). The opposite response of Arabidopsis hypermethylated mutants to various plant pathogens has also been reported. For instance, the hypermethylated mutant ros1 was less resistant to Hpa, but more resistant to the necrotrophic fungi Alternaria brassicicola and Plectosphaerella cucumerina as well as M. javanica (López Sánchez et al., 2016; Silva et al., 2022).

Our gene expression analysis of PR1, PR2, and PDF1.2 pointed to opposite regulation of defense-related genes in the DNA demethylase mutants in response to infection by M. incognita and H. schachtii ( Figures 5B, C ). This may partially explain the opposite susceptibility phenotypes of these mutants in response to infection by H. schachtii and M. incognita ( Figures 2C , 4C ), and highlights the complexity of plant defense pathways against plant-parasitic nematodes and their interactions with active demethylation pathway. Our results are consistent with the finding that active DNA demethylases function mainly in regulating the expression of defense- and stress-related genes (Yu et al., 2013 ; Le et al., 2014 ; Lopez Sanchez et al., 2016) and point to distinct methylation-dependent mechanisms regulating defense gene expression upon infection by cyst and root-knot nematodes. Basal defense response does not seem to be altered in the demethylase mutants under non-infected conditions because PR1, PR2, and PDF1.2 are similarly expressed in these mutants and wild-type plants ( Figures 5A–C ). Additionally, it is unlikely that changes in the expression of PR1, PR5, and PDF1.2 are due to changes in methylation status in cis at the promoters of these genes upon nematode infection because nematode-induced potential hypermethylation of these PR genes in four demethylated mutants argues against upregulation of these genes in response to M. incognita infection ( Figures 5G–I ). Furthermore, PR1, PR2, and PDF1.2 are not subjected to DNA methylation changes upon H. schachtii infection in Col-0 plants (Hewezi et al., 2017). Additionally, it has been shown that the majority of defense-related genes in ros1 mutant are indirectly regulated by DNA methylation (López Sánchez et al., 2016; Halter et al., 2021). Taken together, it is conceivable to suggest that changes in DNA methylation patterns of trans-acting factors directly or indirectly controlling the expression of defense- and stress-related genes in DNA demethylase mutants are the causal factors contributing towards opposite regulation of PR1, PR2, and PDF1.2 upon infection by H. schachtii and M. incognita. Taken together, our results set the stage for future studies to determine the exact mechanism through which DNA methylation and active demethylation of trans-acting factors modulate plant immunity and basal defense response upon infection with parasitic nematodes. This may eventually lead to the development of novel approaches to improve plant resistance to nematode infection.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

TH designed the original research plans. MB performed the majority of the experimental work, analyzed the data, and wrote the first draft. TEH, VL-C, NA, and JR helped MB in data collection and analysis. TH supervised the experimental work, analyzed the data, and wrote the final version. All authors contributed to the article and approved the submitted version.