https://orcid.org/0000-0002-8908-820X
Figures
Abstract
The Snf2 chromatin remodeler, DECREASE IN DNA METHYLATION 1 (DDM1) facilitates DNA methylation. In flowering plants, DDM1 mediates methylation in heterochromatin, which is targeted primarily by MET1 and CMT methylases and is necessary for silencing transposons and for proper development. DNA methylation mechanisms evolved throughout plant evolution, whereas the role of DDM1 in early terrestrial plants remains elusive. Here, we studied the function of DDM1 in the moss, Physcomitrium (Physcomitrella) patens, which has robust DNA methylation that suppresses transposons and is mediated by a MET1, a CMT, and a DNMT3 methylases. To elucidate the role of DDM1 in P. patens, we have generated a knockout mutant and found DNA methylation to be strongly disrupted at any of its sequence contexts. Symmetric CG and CHG sequences were affected stronger than asymmetric CHH sites. Furthermore, despite their separate targeting mechanisms, CG (MET) and CHG (CMT) methylation were similarly depleted by about 75%. CHH (DNMT3) methylation was overall reduced by about 25%, with an evident hyper-methylation activity within lowly-methylated euchromatic transposon sequences. Despite the strong hypomethylation effect, only a minute number of transposons were transcriptionally activated in Ppddm1. Finally, Ppddm1 was found to develop normally throughout the plant life cycle. These results demonstrate that DNA methylation is strongly dependent on DDM1 in a non-flowering plant; that DDM1 is required for plant-DNMT3 (CHH) methylases, though to a lower extent than for MET1 and CMT enzymes; and that distinct and separate methylation pathways (e.g. MET1-CG and CMT-CHG), can be equally regulated by the chromatin and that DDM1 plays a role in it. Finally, our data suggest that the biological significance of DDM1 in terms of transposon regulation and plant development, is species dependent.
Citation: Griess O, Domb K, Katz A, Harris KD, Heskiau KG, Ohad N, et al. (2023) Knockout of DDM1 in Physcomitrium patens disrupts DNA methylation with a minute effect on transposon regulation and development. PLoS ONE 18(3): e0279688. https://doi.org/10.1371/journal.pone.0279688
Editor: Anton Wutz, Eidgenossische Technische Hochschule Zurich, SWITZERLAND
Received: July 24, 2022; Accepted: December 13, 2022; Published: March 8, 2023
Copyright: © 2023 Griess et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: BS-seq and RNA-seq data are available from the GEO database under the accession number GSE198693.
Funding: This work was supported by the Israeli Centers for Research Excellence Program of the Planning and Budgeting Committee, Israel Science Foundation (757/12), Israel Science Foundation (1636/15), and the European Research Council (679551) to A.Z, and Israel Science Foundation (767/09) to N.O. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
DNA methylation is catalyzed by DNA methyltransferases (DNMTs) with unique affinities and targeting mechanisms. In plants, DNA methylation is divided into three sequence contexts, CG, CHG, and CHH (H = A, C, or T) [1]. In flowering plants, CG methylation is maintained by MET1, CHG methylation by CMT3, and CHH methylation by CMT2 or DRM2 [2–6]. To access and methylate the DNA, DNMTs are assisted by Snf2 chromatin remodelers [7–10]. The most robust Snf2 remodeler for methylation in plants is DDM1 [11]. Mutations in DDM1 in flowering plants, such as in Arabidopsis, reduced methylation in all three sequence contexts suggesting its general importance for the methylation of distinct DNMTs [2,12]. DDM1 is especially important for DNA methylation of nucleosomal DNA of heterochromatic transposable elements (TEs), there it was suggested to counteract the chromatin compaction determined by the linker histone H1 or histone variant H2A.W [2,13–15]. To date, the roles of DDM1 in DNA methylation, TE regulation, and plant development were investigated in four flowering plants, namely Arabidopsis, tomato, rice, and maize [2,16–19]. In association with their methylation effect, ddm1 mutations in flowering plants were found to be crucial for the silencing of TEs as well as for proper development; thousands of TEs were activated in ddm1 mutants, which have shown pleiotropic to lethal developmental phenotypes [2,16,17,20,21].
While DNA methylation is found in all land plants, its genomic profile and enzymology can be different among species [22,23]. For example, in comparison to Arabidopsis and other flowering plants, in the moss P. patens non-CG methylation is dependent on a CMT and DNMT3 rather than on CMT2, CMT3, and DRM2 [24,25]. DNMT3 is an ortholog of animal DNMT3 that has been lost during the evolution of flowering plants, and its CHH methylation activity at heterochromatic sequences was replaced by the flowering specific CMT, CMT2 [25]. DRMs are plant-specific homologs of DNMT3 and function together with the RNA directed DNA methylation (RdDM) pathway in methylating euchromatic TE sequences [2,3]. DRMs exist in P. patens, however, have only a trivial role in DNA methylation [25]. Essentially, in Arabidopsis, CMT2 and DRM2 are responsible for 70% and 30% of CHH methylation, respectively [2,3], whereas 95% of CHH methylation is mediated by DNMT3 in P. patens [25]. Another unique difference between P. patens and many flowering plants is the robust CHG methylation in the former, which is found at a similar level as its CG methylation. For example, CG and CHG methylation in Arabidopsis TEs is approximately 80% and 50%, respectively, whereas, in P. patens it is about 80% for both contexts.
DNA methylation in P. patens regulates TE expression and development [26]. Symmetric and asymmetric methylation function redundantly in silencing TEs [26]. Additionally, CHG methylation in P. patens has a stronger silencing effect than CG methylation, and a null mutant for CHG and CHH methylation develops abnormally [26]. While CG and CHG methylation can be separately eliminated in P. patens (including that of CHH methylation), efforts to generate a null P. patens mutant for both CG and CHG methylation failed to be accomplished, possibly due to lethality [26].
Accordingly, by studying the function of DDM1 in P. patens, we would elucidate its role in an early terrestrial plant, i.e., a non-flowering one. More specifically, we will discover the effect of DDM1 on DNA methylation mediated by a distinct DNMT (i.e., DNMT3), its contribution to a methylome mediated primarily by heterochromatic methylases (i.e., with a trivial RdDM activity), its relative influence on CG versus CHG methylation (that are equally methylated in P. patens), as well as the impact of symmetric methylation on TE control and P. patens development (assuming CG and CHG will be significantly reduced).
To study the role of DDM1 in P. patens, we first knocked out its single DDM1 gene by homologous recombination, which proved to be viable. We then profiled Ppddm1 methylome, transcriptome, and morphology. Our data found PpDDM1 to be crucial for the three DNA methylation sequence contexts. Specifically, both symmetrical methylation contexts (CG and CHG) were affected similarly and significantly more than asymmetric CHH methylation (reduction of 75% vs 25%, respectively). Despite the methylation effect, TE expression and development of Ppddm1 were hardly altered. Overall, our data substantiate the role of DDM1 in DNA methylation in plants evolution, show the dependency of a plant DNMT3 on DDM1, demonstrate the relative effect of DDM1 on CG and CHG methylation, as well as find that a massive disruption of CG and CHG methylation can be insufficient to activate TEs nor to alter plant development.
Results
Physcomitrium patens DDM1 knockout develops normally
To investigate the biological role of DDM1 in early terrestrial plants, we aimed to mutate it in Physcomitrium patens. A DDM1 gene has been previously identified in P. patens, named PpDDM1 (Pp1s65_183v6.1; [27]). To verify that PpDDM1 is the only DDM1 gene in P. patens (some plant species encode for multiple genes [16,17]), we generated a phylogenetic tree containing all P. patens Snf2 ATPases together with previously identified functional DDM1 orthologs from plants and mammals. This analysis located only PpDDM1 in the monophyletic DDM1 clade (Fig 1A). We confirmed this finding by generating a second phylogenetic tree containing Arabidopsis Snf2 remodelers (CHRs), together with PpDDM1 and its two closest homologs in P. patens (Pp3C11_18g and Pp3c73340). In this tree, PpDDM1 was placed inside the DDM1 clade, whereas Pp3C11_18g and Pp3c73340 were located together with the Arabidopsis CHR11 and CHR17 clade (Fig 1B), the latter belonging to the SWR1 complex that deposit H2A.Z in genes [28]. Altogether, these findings suggest that PpDDM1 is the only DDM1 gene in P. patens.
A. Phylogenic analysis of all P. patens SWI/SNF2 protein family members (including PpDDM1), and previously characterized DDM1 orthologues from flowering plants and mammals (LSH). The DDM1 clade is marked in green and PpDDM1 is highlighted in blue. B. Phylogenic analysis of A. thaliana SWI/SNF2 protein family members, PpDDM1, and additional DDM1 orthologs from flowering plants and mammals. DDM1 clade is marked in green and PpDDM1 is highlighted in blue. C. Coverage data from BS-seq and RNA-seq show the depletion of DDM1 sequences within the recombination region (Chr17:13053812-Chr:13059096). (D-H). Morphological phenotypic analysis of the wild type and Ppddm1 plants. Protoplast-recovered protonematal tissue (D), gametophore tissue (E), single gametophore (F), sporophyte (G), and germinated spores of T1 (H). Scales for D/E, F/H, and G, are 2 mm, 100 uM, and 50 uM, respectively.
To knockout DDM1 in P. patens, we performed gene replacement using homologous recombination through the transformation of P. patens with the Zeocin resistance gene flanked by PpDDM1 sequences (Fig 1C). Following antibiotic selection, we identified a PpDDM1 knockout plant. Despite the identification of 609 differentially expressed genes (S1 File), morphological investigation of Ppddm1 found it to be developed normally throughout the plant life cycle as well as in subsequent generations (Fig 1D–1H).
DNA methylation is severely disrupted in Ppddm1
To check the role of DDM1 on DNA methylation in P. patens, we profiled the methylome of Ppddm1 using whole genome bisulfite sequencing (WGBS). DNA methylation was substantially reduced in Ppddm1 (Fig 2A). In transposons, which is the main target for methylation in P. patens, CG, CHG, and CHH methylation levels in Ppddm1 were reduced by 76.7%, 77.8%, and 24.1%, respectively (Fig 2B). This result implies that the symmetric CG and CHG sites are both hypomethylated to a similar extent, which is substantially stronger than that of asymmetric CHH methylation. Meta analyses of methylation across TEs and chromosomes substantiated these findings (Fig 2C–2G). The similar hypomethylation effect in CG and CHG in Ppddm1 is an interesting result, as these methylation sites are independently targeted and maintained in P. patens by a distinct type of methylases; CG sites are methylated by PpMET, whereas CHG sites by PpCMT [25]. CG and CHG methylation levels can be influenced by site density. Thus, to check for the relative hypomethylation effect of CG versus CHG, we next tested it within 50 bp windows with a similar number of CG and CHG sites. Within those sequences, CG and CHG methylation levels were similar in wild type and Ppddm1 mutant (Fig 2H), with a median at zero for comparison between mCG and mCHG (CG minus CHG) in wild type and in Ppddm1 (Fig 2I). Together, these results suggest that CG and CHG methylation are similarly affected by the deletion of PpDDM1.
A. Box plots of DNA methylation (separated to CG, CHG, and CHH contexts) in wild type and Ppddm1 in TEs. B. Box plots of percent methylation change in TEs between Ppddm1 versus wild type calculated for 50-bp windows. C-D. Patterns of CG and CHG methylation (C) and CHH methylation (D) in wild type and Ppddm1 across TEs. TEs were aligned at the 5′ or 3′ ends, and methylation was averaged within 50-bp intervals. The dashed lines represent points of alignment. E-F. Average of CG and CHG methylation (E) and of CHH methylation (F) in wild type and Ppddm1 along P. patens chromosome 1 averaged in a 100-kb sliding window. G. A genomic representative snapshot of hypo-methylated TEs and a hyper-methylated TE (marked in dotted line) in Ppddm1. DNA methylation, methylation difference (Ppddm1 minus wild type), RNA level, GC content, and histone H3 are plotted in 50-bp windows. Genes and TEs oriented 5’ to 3’ and 3’ to 5’ are shown above and below the line, respectively. H. Box plots of CG and CHG methylation in wild type and Ppddm1 within methylated 50-bp windows (>0.5 methylation in either of the samples) and with an equal number of CG and CHG sites. I. Percent-methylation change between mCG and mCHG in wild type and Ppddm1 (i.e. WT-mCG minus WT-mCHG (red) and ddm1-mCG minus ddm1-mCHG (blue)), within windows with an equal number of CG and CHG sites. J. Violin plots of percent-methylation-change in 50-bp windows of CHH methylation in TEs between Ppddm1 and wild type, separated based on the level (quantiles) of indicated attributes. Histone H3 and H3K9me2 data were derived from Widiez et al., [29].
In comparison to symmetric sites, which were consistently hypomethylated in Ppddm1 (Fig 2B, 2C, 2E and 2G), asymmetric CHH sites showed some hyper-methylation effect in addition to its hypo-methylation (Fig 2B, 2D, 2F and 2G). By dissecting the methylation sequences to different chromatin features, we located the CHH hyper-methylation effect to euchromatic-TE sequences, i.e., regions depleted of GC content, nucleosomes, or the heterochromatic mark H3K9me2 (Fig 2J). CHH hyper-methylation was particularly negatively associated with CHH methylation in wild type, i.e. CHH sites with normal average methylation lower than 15% were mainly hyper-methylated, whereas CHH sites with normal average methylation between 23–47% were mostly hypo-methylated in Ppddm1 (Fig 2J, rightmost panel). These results align with previous findings on the redistribution of CHH methylation in ddm1 mutants [16,18].
PpDDM1 knockout upregulated a small number of TEs
Ddm1 mutants in flowering plants caused a substantial upregulation of hypomethylated TEs, i.e., more than 1000 upregulated TEs in Arabidopsis, tomato, and rice ([2,16,30]; Fig 3A). To check the effect of PpDDM1 on TE regulation, we profiled the transcriptome of Ppddm1 via RNA-seq. A differential expression analysis identified only 35 TEs that were upregulated in Ppddm1 (Fig 3A and S1 File). TE downregulation was negligible in either of the ddm1 mutants (S1A Fig). Next, we checked for a meta-transcriptional effect of Ppddm1, i.e., on entire TE families rather than on single TEs. To do so, we quantified the total number of reads mapped to entire TE families (including multi-mapped reads). This analysis did not discover a considerable TE upregulation in Ppddm1 (S1B Fig), thus substantiating the trivial transcriptional effect of Ppddm1 also among non-unique TE sequences, such as of recently duplicated elements. Ppddm1-upregulated TEs were not more hypomethylated than unaffected TEs (Fig 3B), suggesting that not the level of hypomethylation triggered the expression in this group of TEs. Alternatively, we found that Ppddm1-upregulated TEs were depleted of methylation within their long terminal repeat (LTR) sequences in wild type plants (Fig 3C and 3D). Furthermore, Ppddm1-upregulated TEs were slightly expressed within their LTRs already in wild type (Fig 3D). Accordingly, these results suggest the minute effect of TE regulation in Ppddm1 is focused on a small group of TEs that are predisposed to be expressed.
A. Number of upregulated TEs in indicated DDM1 and DNMT mutants. Osddm1 and Slddm1 are Oryza sativa and Solanum lycopersicum plants, respectively, that are each mutated in two DDM1 genes (ddm1a and ddm1b) [16,21]. Atddm1 is an Arabidopsis ddm1 mutant [2]. P. patens met, cmt, and dnmt3 mutations are notated with the Prefix Pp [26]. Osddm1, Slddm1, and Atddm1 data are from leaves or roots, and all P. patens samples were from protonema. B. Percent methylation change between Ppddm1 and wild type in all versus Ppddm1-upregulated TEs. C. Wild type DNA methylation in all TEs versus Ppddm1-upregulated TEs across the entire TE sequence (left panel), internal TE sequences (middle panel), or only within long terminal repeat (LTR) sequences (right panel). D. A genomic representative snapshot of hypo-methylated upregulated and unaffected TEs in Ppddm1. The dotted line marks an LTR region that is already hypomethylated and expressed in wild type. E. Total methylation (CG, CHG, and CHH methylation combined) across TEs in wild type, Ppddm1, and Ppcmt. F. Venn diagram of upregulated TEs in Ppddm1 and all Ppdnmts (left diagram) or Ppcmt (right diagram).
Such a result could have suggested that DNA methylation is not necessary to silence TEs in P. patens. However, mutations in the DNA methylases (DNMTs) in P. patens caused to upregulation of up to about 3600 TEs (Fig 3F) [26], that included most of upregulated TEs in ddm1 (Fig 3F). Accordingly, the strong TE-upregulation in Ppdnmt mutants implies the importance of DNA methylation in the silencing of P. patens TEs. DNA hypomethylation in Ppddm1 is different than any of that of the methylases mutants. For example, Ppcmt and Ppmet mutants were entirely, and almost exclusively, depleted in CHG and CG methylation, respectively [25]. In comparison, in Ppddm1 both CG and CHG sites were strongly but also partially hypomethylated (Fig 3B). At the CHH context, Ppcmt and Ppddm1 had a similar complex effect, that is hypo and hyper-methylation in hetero- and eu-chromatic TE sequences, respectively [25]. When averaging the combined methylation level of all three methylation contexts (referred to as total methylation), Ppddm1 had a lower total methylation level than Ppcmt (Fig 3E), implying that close to 100% reduction of methylation in CHG has a bigger effect on TE expression than a 75% reduction in both CG and CHG methylation. This result suggests a lack of a complete redundancy between CG and CHG methylation in the silencing of P. patens TEs and that the pattern of hypomethylation among the three methylation contexts, rather than the total methylation level, has a stronger influence on TE regulation.
Discussion
By investigating DDM1 in the moss P. patens, our data substantiate the role of DDM1 in mediating DNA methylation in plants. Similar to flowering plants, PpDDM1 is required for maintaining all three DNA methylation contexts in TEs. CHH methylation in flowering plants is mediated by DRMs and CMTs, where the activity of the latter is most dependent on DDM1. In comparison to flowering plants, CHH methylation in P. patens is mediated primarily by DNMT3, i.e., hardly by CMT or DRMs. Thus, the disruption of CG, CHG, and CHH methylation in Ppddm1, suggests that DDM1 is required by all types of plant methylases that function within heterochromatic TE sequences, that is MET1, CMTs, and DNMT3 enzymes. Nevertheless, the level of DDM1 dependency on particular methylases could be different. In the flowering plant A. thaliana, heterochromatic CHH methylation mediated by CMT2 is strongly dependent on DDM1, and to about a similar extent to MET1’s CG methylation and CMT3’s CHG methylation [2]. Here, we found CHH methylation in Ppddm1 to be reduced three times less than CG and CHG methylation, suggesting that PpDDM1 is more crucial for symmetric versus asymmetric methylation, as well as that PpDNMT3 heterochromatic CHH methylation activity is less dependent on DDM1 than that of the A. thaliana CMT2’s.
In flowering plants CG and CHG sites are usually methylated to a different level in transposons and are interdependent, CG methylation influences CHG methylation and vice versa [31–35]. Therefore, the relative role of DDM1 on CG methylation versus CHG methylation could have not been identified. In comparison to flowering plants, CG and CHG sites in P. patens TEs are methylated to a similar extent and are regulated independently, i.e. complete removal of CG methylation, as shown for Ppmet, hardly affects CHG methylation and vice versa, when Ppcmt is compromised [25,36]. Accordingly, profiling the methylome of ddm1 in P. patens can reveal the relative role of DDM1 on CG and CHG methylation. Our data show that both CG and CHG methylation are being depleted by about 75% in Ppddm1, suggesting their similar dependency on DDM1. CMT and MET1 are distinct in their protein structure and mode of action. CMTs contain a chromodomain within their catalytic domain, which recruits them to heterochromatin by binding to H3K9me2 [37], whereas MET1s are targeted to chromatin via UHRF1/VIM proteins [38]. Accordingly, the similar extent of CG and CHG hypomethylation in Ppddm1 illustrates the convergence of two distinct and independent methylation enzymatic pathways to be equally regulated by the chromatin remodeler DDM1.
Mutations of DDM1 in flowering plants caused massive upregulation of TEs associated with abnormal developmental phenotypes [2,16,18,30]. In comparison to published data in flowering plants, our data show that mutation of ddm1 in moss, does not cause a substantial activation of TEs (at least not in the protonema tissue) nor an obvious developmental phenotype. This trivial transcriptional and developmental effect of Ppddm1 suggests that the residual methylation is sufficient to silence TEs and maintain proper development. Furthermore, we show that total methylation of Ppddm1 is lower than Ppcmt, while the latter has a greater effect on TEs. In comparison to Ppddm1 which lost 75% of CG and CHG methylation, Ppcmt lost close to 100% of CHG methylation and 0% of CG methylation. These results support our previous finding that CHG methylation is a stronger TE-silencer than CG methylation [26], and imply that not the total amount of methylation is important to silence TEs but rather its distribution among the different methylation contexts. Accordingly, while the molecular effect of DDM1 on DNA methylation could be similar among different plants, its biological role can be different and is determined by the methylation landscape of the particular organism.
Conclusion
Our results extend our knowledge of the role of DDM1 in plants. We show that DDM1 is necessary for any heterochromatic-DNA methylases, including for CHH methylation by plant DNMT3. We further show that separate methylation pathways can be similarly regulated by chromatin and DDM1. Finally, the trivial effect of Ppddm1 on TE expression and plant development suggests that the biological effect of DDM1 is species dependent.
Materials and methods
Phylogenic reconstruction and motif finding
Protein sequences of DDM1/LSH were obtained using the Phytozome and Swissprot databases. Arabidopsis thaliana SWI/SNF2 members were found using the SMART and Eggnog databases. Physcomitrium patens SWI/SNF2 members were found using Pfam. Full protein sequences (S1 Table) were aligned by MAFFT using the BLOSUM62 algorithm [39] and trees were built using PhyML [40] (NNJ) with bootstrap value 100, and were illustrated using Figtree v1.4 [41]. Motifs were obtained by using MEME-suite under default options [42].
Biological materials
Plant material: Physcomitrium patens line “Grandsen2004” was grown on either BCD, BCDAT, or germination media [43] in 16/8 hours of light to the dark regime at a light intensity of 10–50 μmol/m2/s. For sporulation, the light regime was 8/16 hours light to dark at a light intensity of 10–50 μmol/m2/s.
Generation of Ppddm1 knockout mutant and growth conditions
Ppddm1 was obtained through the replacement of the gene coding region with Zeocin resistance antibiotic cassette through homologous recombination. Homologous flanking 5’ (1068bp upstream to ATG site) and 3’ (252 bp upstream to stop site and 536bp downstream to stop site) were amplified using KOD polymerase (Sigma-Aldrich) and cloned into a PMBL5-Zeo vector. Integration of flanking regions was validated using PCR. Linearized plasmid was introduced to moss via PEG-mediated transformation protocol [43]. After three days, the regenerated protoplasts were transferred to the BCDAT medium. After seven days the moss was subjected to selection using Zeocin (25ng\μl). After three weeks, surviving plants were subjected to PCR to validate the correct integration of the construct into the genome and gene replacement. This was done through primers targeting the endogenous gene sequence and primers targeting regions flanking the 5’ and 3’ homologous regions and Zeocin resistance cassette.
BS-seq library preparation
Genomic DNA of 7-day old protonemata was extracted using the Plant II midi kit (Machery-Nagel) according to the manufacturer’s instructions. About 1 μg of purified genomic DNA was fragmented by sonication, end-repaired, and ligated to custom synthesized methylated adaptors. Adaptor ligated libraries were subjected to sodium bisulfite treatment using the Methylation-Gold Bisulfite Kit (ZYMO) as outlined in the manufacturer’s instructions. The converted libraries were then amplified by PCR, gel-purified to select the 350- to 400-bp size range, quantified using real-time quantitative PCR with TaqMan probe, and validated with a Bioanalyzer (Agilent). The libraries were amplified on cBot to generate the cluster on the flow cell (TruSeq PE Cluster Kit V3–cBot–HS, Illumina) and sequenced as a paired end on the HiSEq. 2000 System (TruSeq SBS KIT-HS V3, Illumina).
BS-seq data analysis
The quality of BS-seq reads (S2 Table) was evaluated using FastQC v0.11.8 [44]. Trimming of low-quality reads was conducted using TrimGalore v0.6.0 [45]. Duplicated reads were removed using Prinseq-lite v0-20-4 [46]. Methylation data processing of qualified reads was performed as described previously [26]. Briefly, we converted all of the Cs in the “forward” reads (and in the scaffold) to Ts, and all of the Gs in the “reverse” reads and scaffold to As. The converted reads were aligned to the converted scaffold (P. patens v3.3) using Bowtie2 v2.3.4.1 [47], allowing two mismatches and multimapping reads with up to 10 hits. We then recovered the original sequence information and, for each C (on either strand), counted the number of times it was sequenced as a C or a T. Fractional methylation was calculated within a 50-bp sliding window for each sequence context separately (CG, CHG, or CHH) or all sequence contexts together (total methylation). The percent of methylation change was calculated by dividing the difference in methylation level between two samples by the level of methylation in the sample with the higher methylation level, e.g. (mCG in WT–mCG in ddm1)/mCG in WT if mCG in WT>mCG in ddm1 and (mCG in ddm1 –mCG in WT)/mCG ddm1 if mCG in ddm1>mCG in WT. Meta-analysis of DNA methylation relative to TE edges was performed by aligning TEs at either their 5′ or 3′ ends and averaging methylation in a sliding 100 bp windows upstream and downstream of the point of alignment. Elements were included in the meta-analysis either when they reached the end of their sequence, another annotated element, or the indicated selected length. Box and density plots were generated on STATA v14. Public BS-seq raw data of A. thaliana (SRR771524, SRR771518) [2], S. lycopersicum (ERS2066779, ERS2066938) [16], and O. sativa (SRR3503137 SRR3503134) [21], were downloaded from the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) and processed as described above.
RNA-seq library preparation
Three repetitions of total RNA were extracted from 7-day old protonema using the SV Total RNA Isolation System (Promega) and DNase treatment (Thermo Fisher Scientific) according to the manufacturers’ instructions. The polyA fraction (mRNA) was purified from 500 ng of total input RNA, followed by fragmentation and the generation of double-stranded cDNA. Next, Agencourt Ampure XP beads cleanup (Beckman Coulter), end repair, A base addition, adaptor ligation, and PCR amplification steps were performed. Libraries were quantified by Qubit (Thermo Fisher Scientific) and TapeStation (Agilent). The libraries were sequenced as single-end reads on HiSEq 2500.
RNA-seq data analysis
The quality of RNA-seq reads (S3 Table) was evaluated using FastQC v0.11.8 [44]. Trimming of low-quality reads, PolyA-tails, and adapter sequences was conducted using TrimGalore v0.6.0 [45]. The postprocessed reads were aligned to the genome using STAR v2.7.0.f [48]. Multimapped reads (<100) have been randomly assigned. RNA-seq stats are summarize in For differential gene and TE expression, we used the Limma R package [49]. Genes and TEs with a read count below 10 were omitted. Genes and TEs with 2-fold change and p-val<0.05 were considered as differentially expressed and used in downstream analyses. Public RNA-seq raw data of A. thaliana root tissue (SRR578947, SRR578948, SRR578941, and SRR578942) [2], S. lycopersicum leaf tissue (ERS2066781, ERS2067976, ERS2067977, ERS2066782, ERS2067978, ERS2067979) [16], O. sativa leaf tissue (SRR3503149, SRR3503151, SRR3503146, and SRR3503148) [21], and P. patens (GSM4218374- GSM4218393), were downloaded from NCBI SRA and processed as described above. To achieve comparable sequencing depths between genotypes, we used seven million random reads from each of the public RNA-seq libraries. The genome assemblies used for P. patens, A. thaliana, O. sativa and S. lycopersicum were v3.3, TAIR10, v7.0 and 3.0, respectively.
Annotations
P. patens gene and TE annotations (v3.3) were downloaded from CoGe https://genomevolution.org/coge/GenomeView.pl?gid=33928. A. thaliana TE annotation (TAIR10) was downloaded from https://www.arabidopsis.org. S. lycopersicum TE annotation (Slycopersicum_514_ITAG3.2) was downloaded from Phytozome. O. sativa TEs were annotated de-novo using REPET software. TEs shorter than 100 bp were excluded for expression analysis. Intact LTR retrotransposons annotation (used in Fig 3B and 3C) was kindly provided by Prof. Stefan A. Rensing (University of Marburg, Germany).
Supporting information
S1 Fig. TE expression in Atddm1, Slddm1, Osddm1, Ppddm1 and Ppdnmt mutants.
A. Number of downregulated TEs in O. sativa, S. lycopersicum, A. thaliana and P. patens ddm1 mutants, as well as in P. patens dnmt mutants. B. Total number of RNA-seq reads mapped to indicated TE family types. Reads per million (RPM) values represent the average of replicates per genotype, error bars represent standard error.
https://doi.org/10.1371/journal.pone.0279688.s001
(DOCX)
S1 Table. Gene ID and accessions numbers.
Gene ID and accessions numbers for sequences used in evolutionary tree (Fig 1).
https://doi.org/10.1371/journal.pone.0279688.s002
(DOCX)
S2 Table. Summary of BS-seq data generated in this study.
https://doi.org/10.1371/journal.pone.0279688.s003
(DOCX)
S3 Table. Summary of RNA-seq data generated in this study.
https://doi.org/10.1371/journal.pone.0279688.s004
(DOCX)
S1 File. Up and down regulated TEs and genes in Ppddm1.
https://doi.org/10.1371/journal.pone.0279688.s005
(XLSX)
References
- 1. Law JA, Jacobsen SE. Establishing, maintaining and modifying DNA methylation patterns in plants and animals. Nat Rev Genet. 2010;11: 204–220. pmid:20142834
- 2. Zemach A, Kim MY, Hsieh P-H, Coleman-Derr D, Eshed-Williams L, Thao K, et al. The Arabidopsis nucleosome remodeler DDM1 allows DNA methyltransferases to access H1-containing heterochromatin. Cell. 2013;153: 193–205. pmid:23540698
- 3. Stroud H, Do T, Du J, Zhong X, Feng S, Johnson L, et al. Non-CG methylation patterns shape the epigenetic landscape in Arabidopsis. Nat Struct Mol Biol. 2013;21: 64–72. pmid:24336224
- 4. Lindroth AM, Cao X, Jackson JP, Zilberman D, McCallum CM, Henikoff S, et al. Requirement of CHROMOMETHYLASE3 for Maintenance of CpXpG Methylation. Science (80-). 2001;292: 2077–2080. pmid:11349138
- 5. Ronemus MJ, Galbiati M, Ticknor C, Chen J, Dellaporta SL. Demethylation-Induced Developmental Pleiotropy in Arabidopsis. Science (80-). 1996;273: 654–657. pmid:8662558
- 6. Finnegan EJ, Peacock WJ, Dennis ES. Reduced DNA methylation in Arabidopsis thaliana results in abnormal plant development. Proc Natl Acad Sci. 1996;93: 8449–8454. pmid:8710891
- 7. Groth M, Stroud H, Feng S, Greenberg MVC, Vashisht AA, Wohlschlegel JA, et al. SNF2 chromatin remodeler-family proteins FRG1 and -2 are required for RNA-directed DNA methylation. Proc Natl Acad Sci. 2014;111: 17666–17671. pmid:25425661
- 8. Zhou M, Palanca AMS, Law JA. Locus-specific control of the de novo DNA methylation pathway in Arabidopsis by the CLASSY family. Nat Genet. 2018;50: 865–873. pmid:29736015
- 9. Kanno T, Mette MF, Kreil DP, Aufsatz W, Matzke M, Matzke AJM. Involvement of Putative SNF2 Chromatin Remodeling Protein DRD1 in RNA-Directed DNA Methylation. Curr Biol. 2004;14: 801–805. pmid:15120073
- 10. Smith LM, Pontes O, Searle I, Yelina N, Yousafzai FK, Herr AJ, et al. An SNF2 Protein Associated with Nuclear RNA Silencing and the Spread of a Silencing Signal between Cells in Arabidopsis. Plant Cell. 2007;19: 1507–1521. pmid:17526749
- 11. Jeddeloh JA, Stokes TL, Richards EJ. Maintenance of genomic methylation requires a SWI2 / SNF2-like protein. 1999;22: 94–97.
- 12. Stroud H, Greenberg MVC, Feng S, Bernatavichute Y V., Jacobsen SE. Comprehensive Analysis of Silencing Mutants Reveals Complex Regulation of the Arabidopsis Methylome. Cell. 2013;152: 352–364. pmid:23313553
- 13. Lippman Z, Gendrel AV, Black M, Vaughn MW, Dedhia N, McCombie WR, et al. Role of transposable elements in heterochromatin and epigenetic control. Nature. 2004;430: 471–476. pmid:15269773
- 14. Lyons DB, Zilberman D. DDM1 and Lsh remodelers allow methylation of DNA wrapped in nucleosomes. Elife. 2017;6. pmid:29140247
- 15. Osakabe A, Jamge B, Axelsson E, Montgomery SA, Akimcheva S, Kuehn AL, et al. The chromatin remodeler DDM1 prevents transposon mobility through deposition of histone variant H2A.W. Nat Cell Biol. 2021;23: 391–400. pmid:33833428
- 16. Corem S, Doron-Faigenboim A, Jouffroy O, Maumus F, Arazi T, Bouché N. Redistribution of CHH Methylation and Small Interfering RNAs across the Genome of Tomato ddm1 Mutants. Plant Cell. 2018;30: 1628–1644. pmid:29875274
- 17. Long JC, Xia AA, Liu JH, Jing JL, Wang YZ, Qi CY, et al. Decrease in DNA methylation 1 (DDM1) is required for the formation of m CHH islands in maize. J Integr Plant Biol. 2019;61: 749–764. pmid:30387549
- 18. Fu F-F, Dawe RK, Gent JI. Loss of RNA-Directed DNA Methylation in Maize Chromomethylase and DDM1-Type Nucleosome Remodeler Mutants. Plant Cell. 2018;30: 1617–1627. pmid:29884624
- 19. Tan F, Zhou C, Zhou Q, Zhou S, Yang W, Zhao Y, et al. Analysis of Chromatin Regulators Reveals Specific Features of Rice DNA Methylation Pathways. Plant Physiol. 2016/05/12. 2016;171: 2041–2054. pmid:27208249
- 20. Kakutani T, Jeddeloh JA, Flowers SK, Munakata K, Richards EJ. Developmental abnormalities and epimutations associated with DNA hypomethylation mutations. Proc Natl Acad Sci. 1996;93: 12406–12411. pmid:8901594
- 21. Tan F, Zhou C, Zhou Q, Zhou S, Yang W, Zhao Y, et al. Analysis of Chromatin Regulators Reveals Specific Features of Rice DNA Methylation Pathways. Plant Physiol. 2016;171: 2041–2054. pmid:27208249
- 22. Niederhuth CE, Bewick AJ, Ji L, Alabady MS, Kim K Do, Li Q, et al. Widespread natural variation of DNA methylation within angiosperms. Genome Biol. 2016;17: 194. pmid:27671052
- 23. Takuno S, Ran J-H, Gaut BS. Evolutionary patterns of genic DNA methylation vary across land plants. Nat Plants. 2016;2: 15222. pmid:27249194
- 24. Noy-Malka C, Yaari R, Itzhaki R, Mosquna A, Auerbach Gershovitz N, Katz A, et al. A single CMT methyltransferase homolog is involved in CHG DNA methylation and development of Physcomitrella patens. Plant Mol Biol. 2014;84: 719–35. pmid:24370935
- 25. Yaari R, Katz A, Domb K, Harris KD, Zemach A, Ohad N. RdDM-independent de novo and heterochromatin DNA methylation by plant CMT and DNMT3 orthologs. Nat Commun. 2019;10: 1613. pmid:30962443
- 26. Domb K, Katz A, Harris KD, Yaari R, Kaisler E, Nguyen VH, et al. DNA methylation mutants in Physcomitrella patens elucidate individual roles of CG and non-CG methylation in genome regulation. Proc Natl Acad Sci. 2020;117: 33700–33710. pmid:33376225
- 27. Parihar V, Dangwal M, Arya D, Kapoor S, Kapoor M. Decrease in DNA methylation 1 interacts with chromomethylase and like heterochromatin protein 1 in Physcomitrella patens. FEBS Lett. 2019;593: 2686–2697. pmid:31271647
- 28. Luo Y, Hou X, Zhang C, Tan L, Shao C, Lin R, et al. A plant-specific SWR1 chromatin-remodeling complex couples histone H2A.Z deposition with nucleosome sliding. EMBO J. 2020;39: e102008. pmid:32115743
- 29. Widiez T, Symeonidi A, Luo C, Lam E, Lawton M, Rensing SA. The chromatin landscape of the moss Physcomitrella patens and its dynamics during development and drought stress. Plant J. 2014;79: 67–81. pmid:24779858
- 30. Tan F, Lu Y, Jiang W, Wu T, Zhang R, Zhao Y, et al. DDM1 Represses Noncoding RNA Expression and RNA-Directed DNA Methylation in Heterochromatin. Plant Physiol. 2018;177: 1187–1197. pmid:29794169
- 31. Hu L, Li N, Xu C, Zhong S, Lin X, Yang J, et al. Mutation of a major CG methylase in rice causes genome-wide hypomethylation, dysregulated genome expression, and seedling lethality. Proc Natl Acad Sci U S A. 2014;111: 10642–10647. pmid:25002488
- 32. Anderson SN, Zynda GJ, Song J, Han Z, Vaughn MW, Li Q, et al. Subtle Perturbations of the Maize Methylome Reveal Genes and Transposons Silenced by Chromomethylase or RNA-Directed DNA Methylation Pathways. G3 (Bethesda). 2018;8: 1921–1932. pmid:29618467
- 33. Gouil Q, Baulcombe DC. DNA Methylation Signatures of the Plant Chromomethyltransferases. PLoS Genet. 2016;12: 1–17. pmid:27997534
- 34. Cheng C, Tarutani Y, Miyao A, Ito T, Yamazaki M, Sakai H, et al. Loss of function mutations in the rice chromomethylase OsCMT3a cause a burst of transposition. Plant J. 2015;83: 1069–1081. pmid:26243209
- 35. Yang Y, Tang K, Datsenka TU, Liu W, Lv S, Lang Z, et al. Critical function of DNA methyltransferase 1 in tomato development and regulation of the DNA methylome and transcriptome. J Integr Plant Biol. 2019;61: 1224–1242. pmid:30652405
- 36. Zemach A, McDaniel IE, Silva P, Zilberman D. Genome-Wide Evolutionary Analysis of Eukaryotic DNA Methylation. Science (80-). 2010;328: 916–919. pmid:20395474
- 37. Du J, Zhong X, Bernatavichute Y V, Stroud H, Feng S, Caro E, et al. Dual binding of chromomethylase domains to H3K9me2-containing nucleosomes directs DNA methylation in plants. Cell. 2012;151: 167–180. pmid:23021223
- 38. Woo HR, Dittmer TA, Richards EJ. Three SRA-Domain Methylcytosine-Binding Proteins Cooperate to Maintain Global CpG Methylation and Epigenetic Silencing in Arabidopsis. Kakutani T, editor. PLoS Genet. 2008;4: e1000156. pmid:18704160
- 39. Rozewicki J, Li S, Amada KM, Standley DM, Katoh K. MAFFT-DASH: integrated protein sequence and structural alignment. Nucleic Acids Res. 2019;47: W5–W10. pmid:31062021
- 40. Guindon S, Dufayard J-F, Lefort V, Anisimova M, Hordijk W, Gascuel O. New Algorithms and Methods to Estimate Maximum-Likelihood Phylogenies: Assessing the Performance of PhyML 3.0. Syst Biol. 2010;59: 307–321. pmid:20525638
- 41.
Rambaut A. FigTree 1.4.4 software. Institute of Evolutionary Biology, Univ. Edinburgh. 2018.
- 42. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 2009;37: W202–W208. pmid:19458158
- 43. Nishiyama T. Tagged Mutagenesis and Gene-trap in the Moss, Physcomitrella patens by Shuttle Mutagenesis. DNA Res. 2000;7: 9–17. pmid:10718194
- 44.
Andrews S, Krueger F, Seconds-Pichon A, Biggins F, Wingett S. FastQC. A quality control tool for high throughput sequence data. Babraham Bioinformatics. In: Babraham Institute. 2015.
- 45. Krueger F. Trim Galore!: A wrapper tool around Cutadapt and FastQC to consistently apply quality and adapter trimming to FastQ files. Babraham Inst. 2015.
- 46. Schmieder R, Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27: 863–864. pmid:21278185
- 47. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9: 357–359. pmid:22388286
- 48. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29: 15–21. pmid:23104886
- 49. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43: e47–e47. pmid:25605792