Contents lists available at ScienceDirect

Plant Physiology and Biochemistry

journal homepage: www.elsevier.com/locate/plaphy

Research article

Functional analyses of PtRDM1 gene overexpression in poplars andevaluation of its effect on DNA methylation and response to salt stress

Ali Movahedia,b,1, Jiaxin Zhanga,b,1, Weibo Suna, Kourosh Mohammadia,Amir Almasi Zadeh Yaghutia, Hui Weia, Xiaolong Wua, Tongming Yinb, Qiang Zhugea,∗

a Co-Innovation Center for Sustainable Forestry in Southern China, Key Laboratory of Forest Genetics and Biotechnology, Ministry of Education, Nanjing ForestryUniversity, Nanjing, 210037, Chinab Co-Innovation Center for Sustainable Forestry in Southern China, College of Forestry, Nanjing Forestry University, Nanjing 210037, China

A R T I C L E I N F O

Keywords:PtRDM1PtROS1DNA methylationSalt stressPoplar

A B S T R A C T

Epigenetic modification by DNA methylation is necessary for all cellular processes, including genetic expressionevents, DNA repair, genomic imprinting and regulation of tissue development. It occurs almost exclusively at theC5 position of symmetric CpG and asymmetric CpHpG and CpHpH sites in genomic DNA. The RNA-directed DNAmethylation (RDM1) gene is crucial for heterochromatin and DNA methylation. We overexpressed PtRDM1 genefrom Populus trichocarpa to amplify transcripts of orthologous RDM1 in ‘Nanlin895’ (P. deltoides× P. eur-americana ‘Nanlin895’). This overexpression resulted in increasing RDM1 transcript levels: by ∼150% at 0mMNaCl treatment and by ∼300% at 60mM NaCl treatment compared to WT (control) poplars. Genomic cytosinemethylation was monitored within 5.8S rDNA and histone H3 loci by bisulfite sequencing. In total, transgenicpoplars revealed more DNA methylation than WT plants. In our results, roots revealed more methylated CGcontexts than stems and leaves whereas, histone H3 presented more DNA methylation than 5.8S rDNA in bothWT and transgenic poplars. The NaCl stresses enhanced more DNA methylation in transgenic poplars than WTplants through histone H3 and 5.8 rDNA loci. Also, the overexpression of PtRDM1 resulted in hyper-methylation,which affected plant phenotype. Transgenic poplars revealed significantly more regeneration of roots than WTpoplars via NaCl treatments. Our results proved that RDM1 protein enhanced the DNA methylation by chromatinremodeling (e.g. histone H3) more than repetitive DNA sequences (e.g. 5.8S rDNA).

1. Introduction

Studies have shown that plants must adjust to environmental con-ditions to survive (Santner and Estelle, 2009). Numerous environ-mental-stress-responsive genes have been used to improve plant gen-omes and proteomes using genomics, proteomics and transcriptomicsapproaches (Gu et al., 2017). Introduction of environmental stressescan cause damage to plants by reducing gene expression, remodelingchromatin and altering chromosomes, resulting in suppression ofgrowth and productivity (Hirakawa et al., 2017). Heritable epigeneticresistance exists in plants. One mechanism is chromatin remodeling,which regulates gene expression in the presence of stresses (Feng et al.,2010). Epigenetic marks such as histone modifications and DNA me-thylation are distributed along four states of chromatins: intergenicregions, repressed genes, active genes and silent repeat elements(Roudier et al., 2011). Histone H3 methylation accompanied by DNA

methylation will be associated with gene expression to promote or/andmaintain the differentiation status of plant cells (Ikeuchi et al., 2015).Already, lots of researches have been done on methylation of histoneH3 at lysine 9 (H3K9me), but to date there is no research on DNAmethylation of histone H3 gene body. This study is the first research oninvestigation of DNA methylation of histone H3 gene body in com-paring with 5.8S rDNA, impacted by abiotic stress.

The RNA-directed DNA methylation (RdDM) pathway involvesseveral important genes that regulate DNA methylation, which is amechanism of epigenetic regulation in eukaryotes (Movahedi et al.,2015a,b,c). One such gene is RNA-directed DNA methylation 1 (RDM1)(Gao et al., 2010; Movahedi et al., 2015a,b,c), some of mutations inRDM1 cause loss-of-function in the rdm1 gene, which suppresses ac-cumulation of 24 nt siRNAs, resulting in decreased DNA methylation(Gao et al., 2010). RDM1 is a small nuclear protein that associates withARGONAUTE (AGO 4/6/9), KOW domain-containing transcription

https://doi.org/10.1016/j.plaphy.2018.03.011Received 12 December 2017; Received in revised form 20 February 2018; Accepted 9 March 2018

∗ Corresponding author.

1 These authors contributed equally in this article.E-mail address: qzhuge@njfu.edu.cn (Q. Zhuge).

Abbreviations: RDM1, RNA-directed DNA methylation 1; ROS1, Repressor of Silencing; qPCR, quantitative Polymerase Chain Reaction; RdDM, RNA-directed DNA methylation

Plant Physiology and Biochemistry 127 (2018) 64–73

Available online 10 March 20180981-9428/ © 2018 Elsevier Masson SAS. All rights reserved.

T

factor 1 (KTF1) and DNA-dependent RNA polymerase V (PolV) to lo-calize domain rearranged methyltransferase 2 (DRM2), leading to DNAmethylation (Gao et al., 2010; Movahedi et al., 2015a,b,c). The re-pressor of silencing 1 (ROS1) gene family, which mediates demethyla-tion activities, and the RdDM pathway are maintained in equilibrium(Gao et al., 2010). Biologically, ROS1 is a DNA glycosylase/lyase thatactivates DNA demethylation, which is critical for demethylation ofpromoter sequences (Wang et al., 2017). This dynamic epigenetic reg-ulation may be necessary for efficient responses to environmental anddevelopmental stresses. Questa et al. (2013) showed that during theinitial stages of development, germination or post-germination inplants, ROS1 binds with pol IV to remodel chromatin in 5.8S rDNA.

Gao et al. (2010) reported that in the presence of ros1, heavy me-thylation occurs in all sequences, including CpG, CpHpG and CpHpH,where H represents A, T or C. The plant RDM1 protein has a highlyconserved sequence, homologs of which are present in both dicots andmonocots (Allard et al., 2005; Ma et al., 2015). According to Gao et al.(2010), suppression of RDM1 or loss-of-function caused by mutationresults in defective DNA methylation. These authors also showed that amutant RDM1 that contained alanine in place of methionine exhibiteddecreased DNA methylation due to the suppression of binding of RDM1to mCpHpH sequences. Gao et al. (2010) reported that the mutation ofrdm1-1 blocked asymmetric cytosine methylation CpHpH sites, but hadno impact on symmetric cytosine methylation CpG sites. That is whythe rdm1-1 genes cause a decrease in de novo methylation; thus, RDM1is an essential factor for DNA methylation. Also, it has been proved thatRDM1 is involved in a complex associated with DRM2 and AGO4 thatlocally catalyzes CG, CHG, and CHH methylation in plants (Gao et al.,2010). In total, the role of RDM1 gene in development of DNA me-thylation and also, the role of DNA methylation in response to en-vironmental stresses, led the authors to assess the role of PtRDM1 geneoverexpressing in improving resistant poplars against salt stress.

2. Materials and methods

2.1. PtRDM1 identification, RNA isolation, plasmid construction andassembly

A BLASTp search was carried out for detecting homologs of RDM1domain from Arabidopsis thaliana using Uniprot database (http://www.uniprot.org/blast/). Species containing characterized RDM1 were iso-lated to identify consensus sequences. The putative PtRDM1 was am-plified from P. trichocarpa using degenerate primers (Supplementary 1;DGF and DGR) (Geneious version 10.3 created by Biomatters devel-opment team. Available from https://www.geneious.com) and in-troduced into the pEASY-T3 cloning vector (pEASY®-T3 cloning kit) viathe TA cloning technique. The linked putative PtRDM1 was then se-quenced (GeneScript Company). Finally, the analyzed sequence wassubmitted to the National Center for Biotechnology Information (NCBI)under accession number KT633998.

Total RNA was extracted from P. trichocarpa using TRIzol (TiangenBiotech, Beijing, China) reagent according to the manufacturer's in-structions. RNA was treated with DNaseI (NEB, USA) and the RNAconcentration was determined using a BioDrop spectrophotometer(UK). Total RNA (2 μg) and oligo-dT primers were used to synthesizecDNA using a Prime Script One Step RT-PCR ver. 2 kit (TakaraBiotechnology, Dalian, China) according to the manufacturer's in-structions. Continuously, the full-length cDNA of PtRDM1 was ampli-fied from cDNA of P. trichocarpa, corresponding to the positions 48–575in XM-006379786 with a length of 528 bp (SnapGene™ 1.1.3 software,Chicago, USA). The amplified fragment was then introduced into thepGWB9 expression vector (accession number AB289772) via the infu-sion gateway technique in the sense orientation. The expression ofPtRDM1 was driven by the 35S promoter and terminated by the NOSterminator.

2.2. Plant preparation, transformation and regeneration

According to Movahedi et al. (2015a,b,c), Murashige and Skoog(MS) media containing 0.5 mg/L N-6-benzyladenine (6-BA) and0.004mg L−1 thidiazuron (TDZ) were used to sub-culture 20 plants ofhybrid clone ‘Nanlin895’ (P. deltoides× P. euramericana ‘Nanlin895’).Then, two cm grown shoots were transferred to shoot elongation mediasupplemented with 0.25mg L−1 6-BA and 0.002mg L−1 TDZ. Four cmgrown shoots were subjected on half MS media to regenerate roots(Movahedi et al., 2015a,b,c). Two weeks old transferred plants (firstexpanded leaves) and four weeks old grown poplars (immature leavesand roots) were used to investigate the expression of PtRDM1, whiletwo and three month old grown plants were assumed to be young andmature poplars, respectively.

Transformation of P. deltoides × P. euramericana ‘Nanlin895’ wascarried out using Agrobacterium tumefaciens strain EHA105 according to(Movahedi et al., 2014). Briefly, cut leaves were immersed in Agro-bacterium inducer liquid (MS) medium containing 5% sucrose and200 μM acetosyringone (AS) and incubated for 120min at 28 °C withshaking at 100 rpm. Explants were then transferred to semi-solid MSmedia supplemented with 200 μM AS and incubated for 2 days at 28 °Cin a dark condition. The putative transformants were then transferred toselective MS media supplemented with 0.5 mg l−1 6-BA, 0.004mg l−1

TDZ, 200mg l−1 cefotaxime and 50mg l−1 kanamycin at 23 °C and aperiod of 16/8 h light/dark. One microgram of genomic DNA, as de-termined using a BioDrop spectrophotometer (UK), was extracted fromyoung putative transformant leaves using the CTAB method. PCR am-plification was performed using the specific primers of the neomycinphosphotransferase II (NPT II) gene (Supplementary 1; NPTII-F andNPTII-R). The 3546 bp amplified fragments were then resolved in a1.5% agarose gel. Kanamycin-resistant and PCR positive plants wereconsidered transgenic. In total, 9 lines of transgenic poplars (including∼135 individuals) were obtained.

2.3. Analysis of gene expression by RT-PCR

Total RNA was extracted from putative transformant plants using aPlant RNA Kit (Omega Biotech No: R6827-01, China) according to themanufacturer's instructions. Total RNAs from 9 independent putativetransformant plants and 3 WT poplars were then treated with DNase I(Takara Biotechnology, China) and their concentrations were de-termined to be 1000 ng/μl using a BioDrop spectrophotometer (UK).The PrimeScript One Step RT-PCR ver. 2 kit (Takara Biotechnology,Dalian, China) was then used to synthesize cDNA with oligo-dT primers;the synthesized cDNA was stored in TE buffer. The β-actin gene (ac-cession number: XM-006370951.1) was used to standardize the cDNAconcentration from PCR using specific primers (Supplementary 1; Actin-F and Actin-R) with 2min denaturation at 95 °C, 30 cycles of (5 mindenaturation at 94 °C, 1min annealing at 55 °C and 1min extension at72 °C) and 5min extension at 72 °C. In addition, specific primers weredesigned to amplify the PtRDM1 from synthesized cDNA(Supplementary 1; PtRDM1-F and PtRDM1-R). Equal amounts of am-plified PCR products were resolved in a 1.5% TAE agarose gel and bandintensity was quantified using the ImageJ 1.5b image-analysis software(USA).

2.4. qPCR analysis

Synthesized cDNA was used to quantify the expression of exogenousPtRDM1 by quantitative real-time PCR (qPCR). Each NaCl treatment (0,20, 40 and 60mM) was carried out using three biological replicatesfrom independent transgenic and one non-transgenic (wild type) po-plars supplied with three repeats, and β-actin was used as an en-dogenous control (Movahedi et al., 2015a,b,c). An Applied Biosystemsreal-time PCR instrument (USA) and Fast Start Universal SYBR GreenMaster Mix (Rox; No. 04913914001: Roche, USA) were used to

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

65

compare the expression levels of genes of interest using the comparativequantitative analysis (ΔΔCt), and the β-actin housekeeping gene servingas an internal reference. The PtRDM1 CDS and the PtROS1 CDS (ac-cession number, KU587630) were primed to amplify 152 and 154 bp,respectively (Supplementary 1; Real-PtRDM1-F, Real-PtRDM1-R, Real-PtROS1-F and Real-PtROS1-R). Also, a 100 bp fragment of the β-actingene was amplified as a reference (Supplementary 1; Real-Actin-F andReal-Actin-F). The expressions of the genes of interest were calculatedaccording to Zhu et al. (2013) using independent cDNA samples.

2.5. DNA methylation assay

Southern blotting was carried out using genomic DNA extracted bythe CTAB method (5 μg) digested with the HpaII restriction en-donuclease (methylation sensitive) and a 150 bp probe with no HpaIIsite (Dimitrova et al., 2016; Mehrotra and Goyal, 2014). A 330 bpfragment with no HpaII site of ycf4 gene, Photosystem I gene isolatedfrom chloroplast with the lowest methylation in cytosines, has beenselected as a control to probe (Supplementary 1; Ycf4-F-probe and Ycf4-R-probe). Ngernprasirtsiri et al. (1988) reported that there was no de-tected DNA methylation in genes family encode proteins in chloroplastregion, which are related to photosynthesis in plants. Also, Omidvarand Fellner (2015) reported that the expression of ycf4 gene is not in-fluenced by changes of DNA methylation in plants affected by biotic orabiotic stresses. Transformants were grown on MS media containing20mM NaCl for 3 weeks.

A methylated DNA endonuclease that cleaves methylcytosine(McrBC with restriction site 5' … pumc(N40-3000)pumc … 3′) in eitherone or both strands of methylated DNA was used to digest genomicDNA (800 ng) for ≤5 h. 10% digested DNA was treated with 65 °C for20min to inactivate enzyme and subjected to PCR. Digested DNA wasdenatured at 95 °C for 5min, followed by 35 cycles to denature DNA at94 °C for 1min, anneal primers at 55 °C for 1min and extend fragmentsat 72 °C for 1min. PCR products were resolved in a 1.5% agarose gel.

The CTAB method was used to extract genomic DNA (2 μg) for bi-sulfite sequencing using the EZ DNA Methylation kit (Zymo Research,USA) (Li and Tollefsbol, 2011). Treated genomic DNA was then used asa template to perform PCR. PCR was performed for 14 h at 55 °C, fol-lowing pre-denaturing at 95 °C for 20min. Flanking specific primers forhomologs in P. trichocarpa (5.8S rDNA gene, accession no. AJ006440and histone H3 gene, accession no. XM_002299206) were designed andused to isolate genomic DNA (Supplementary 1; 5.8S-F, 5.8S-R, H3-Fand H3-R). In addition, specific primers (Supplementary 1; Ycf4-F andYcf4-R) have been used for isolating 555 bp of ycf4 gene to use ascontrol. The Wizard DNA Cleanup kit (Promega) was used to purify andremove PCR residue to prevent sequencing errors. A QIAquick Gel Ex-traction kit (Qiagen) was used to separate specific PCR products fromnonspecific fragments, followed by cloning sequencing using the pGEM-T Easy vector system II (Promega). The ligated pGEM-T Easy vector wastransformed into competent JM109 cells and cultured on LB agar. Pu-tative white colonies of transformants were selected and grown in LBmedium, followed by plasmid extraction and sequencing. In addition,purified PCR fragments were used to evaluate the DNA methylationstatus of histone H3 and 5.8S rDNA by cleavage with the methylation-sensitive restriction endonucleases HpaII and MspI (isoschizomerscleaving CCGG).

2.6. Phenotypic analyses

To analyze the phenotypic changes due to the overexpression ofPtRDM1, we transplanted 2 cm of the top of selected young WT (111plants) and transgenic poplars (114 plants) into the fresh half MSmedium, added with 0, 20, 40 and 60mM NaCl. To exhibit phenotypicdifferences between WT and transformants, we carried out 5 analyseson stems, roots and leaves at two steps (10 and 20 days). These analysesincluded the stem length (cm), main root regeneration, degraded stem

growth, number of leaves and number of wavy leaves.

2.7. Statistical analysis

The SPSS software ver. 16 and Microsoft Office Excel ver. 2013 wereused to analyze data by one-way analysis of variance (ANOVA) andDuncan's test. Significant differences were considered at p < 0.05 withno overlap of mean values.

3. Results

3.1. Identification of PtRDM1 in P. trichocarpa

We carried out a BLAST homology search of the RDM1 in A. thaliana(accession number: NP_188907, Locus: AT3G22680.1) and 10 plantspecies containing domains with similarity to the conserved RDM1domain of P. trichocarpa to identify homologs of the PtRDM1 gene(Fig. 1). The domains of RDM1 protein then were aligned to designdegenerate primers (Supplementary 2). In addition, the protein chargewas analyzed to demonstrate the similarity between PtRDM1 domainand the RDM1 homolog of the A. thaliana (Supplementary 3). The si-milarities between mentioned domains also were proved by analyzingthe isoelectric points (Supplementary 4) and the Hydrophobicity(Supplementary 5), using the Geneious software (version 10.3). Fur-thermore, a structural analysis of the PtRDM1 protein showed six alphahelixes, thirteen beta strands and twenty turns (supplementary 6). ThepGWB9 expression vector was used to overexpress the RDM1 transcripts(Supplementary 7A) and colony PCR was performed to detect trans-formant plants by amplifying 3546 bp of the T-DNA of interest (Sup-plementary 7B, C). The transgenic and WT plants were assayed for theexpression of PtRDM1 gene using RT-PCR. The putative transformant(No. #27) showed greater the expression of PtRDM1 than WT poplar

Fig. 1. Phylogenetic tree of RDM1 proteins from Arabidopsis lyrata (EFH59634),Arabidopsis thalian (NP_188907), Brassica rapa (XP_009135822), Capsella rubella(XP_006298700), Citrus clementina (XP_006419996), Eutrema salsugineum(XP_006406137), Glycine max (NP_001237231), Lotus japonicus (AFK36829), Medicagotruncatula (XP_003610752), Phaseolus vulgaris (XP_007157108), Populus trichocarpa(XP_002311634). Bootstrap analysis was performed using 1000 replicates to evaluate thereliability of the various phylogenetic groups.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

66

(No. #8) in leaves, immature roots and stems (Fig. 2A). Also, the in-tensity measurements of the gel bands revealed different expressionlevels in different organs of poplar plants. Our results showed thatmature roots exhibited a higher expression of PtRDM1 than leaves andstems in WT poplars (Fig. 2B). Also, transgenic plants revealed a higherexpression pf ptRDM1 in leaves, stems and immature roots, than WTpoplars (Fig. 2B). Transformant poplars exhibited the expression ofPtRDM1 by ∼120%, ∼250% and ∼250% in leaves, immature rootsand stems respectively, compared to the corresponding tissues of WTplants, while mature roots showed no significant differences betweenWT and transformant poplars (Fig. 2B).

3.2. Functional analysis of PtRDM1 expression and Gene Ontology analysis

To investigate the expression of PtRDM1 based on molecular func-tion, cellular component and biological process, we carried out GeneOntology (GO) analysis of RDM1 orthologs in plant species included inthe phylogenetic tree using EggNOG database v 4.5 (Supplementary 8).GO analysis revealed that the highest proportion of orthologous RDM1genes was involved in the biological processes of DNA modification,methylation and DNA methylation percentage (∼13%). Referring tothe results of RT-PCR, qPCR was performed to quantify the over-expression of endogenous PtRDM1 in leaves, stems and roots of trans-formants and in roots of WT poplars treated with various NaCl con-centrations (Fig. 3A). Results proved that roots presented the highestexpression of RDM1 in transgenic poplars. Furthermore, 60 mM NaCltreatment (severe stress) promoted the expression of RDM1 in trans-genic poplars especially in roots, but demoted it in WT poplars(Fig. 3A). On the other hand, qPCR was carried out to investigate theinteraction of the expression PtRDM1 on the expression of PtROS1(Fig. 3B). NaCl treatments enhanced the expression of PtROS1 gene inboth stems and leaves from transgenic poplars, whereas it was slightly

impacted in roots from both WT and transgenic poplars by the ex-pression of PtRDM1. Our results in 40 and 60mM NaCl treatmentssuggested that the increase of expression of PtRDM1 may demote theexpression of PtROS1.

3.3. Upregulated RDM1 causes DNA methylation in histone H3 andribosomal DNA

Modification of DNA methylation in Arabidopsis is associated withpleiotropic phenotypic changes (Resentini et al., 2017). Therefore, weinvestigated the DNA methylation status in PtRDM1-overexpressingtransformant poplars.

3.3.1. Endonuclease analysesThe repetitive 548 bp 5.8S rDNA and 719 bp histone H3 methyla-

tion statuses were analyzed to assess the effect of the overexpressionPtRDM1 on regulation of DNA methylation in constitutive hetero-chromatin. Therefore, methylation at 5′-CCGG-3′ sites was determinedusing McrBC and the methylation-sensitive MspI and HpaII iso-schizomer endonucleases. Both cytosines in CpNpG and CpG sequencescan be methylated. According to New England Biolabs®Inc, cleavage byMspI is blocked completely by methylation in asymmetric CpHpG se-quences, but occurs in CpG sequences (e.g., CCGG sites) only when theouter cytosine is methylated. In contrast, HpaII is blocked by methy-lation in both CpHpG and CpG sites. PCR was performed to assess DNAmethylation using the McrBC enzyme. Methylation at McrBC sites in thehistone H3 and 5.8S rDNA genes was significantly increased in trans-formant poplars compared with the WT (Fig. 4A).

3.3.2. Southern blot and qPCR analysesSouthern blot analysis of genomic DNA digested with HpaII revealed

that methylation in CpG sites was increased in PtRDM1 overexpressing

Fig. 2. The overexpression of PtRDM1 in transgenic plants as determined by RT-PCR. (A) Expression of genes encoding RDM1 was increased via transformation. (B) Transformant poplarsexhibited the greater expression of PtRDM1 in leaves, roots and stems respectively, compared to the corresponding tissues of WT plants. All the RT-PCR levels were similar and β-actin wasused as an endogenous PCR control.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

67

transformant poplars compared to WT plants (Fig. 4B). Southern blotanalysis also showed a significant increase in methylation in both WTand transgenic poplars treated with 40mM NaCl compared with un-treated plants. Furthermore, methylation of 5.8S rDNA and histone H3was analyzed by digestion with ScrFI (CCNGG) to detect methylation atsymmetric CpG sites. Digested genomic DNA was then used as a tem-plate to perform quantitative PCR to determine the rate of digestion(Fig. 5). qPCR was carried out to compare the DNA methylation fromdigested 5.8S and histone H3 loci by ScrFI in WT and transgenic po-plars; the three transformant plants exhibited increased methylationcompared with WT (Fig. 5A and B).

3.4. DNA methylation status at 5.8S rDNA and histone H3 loci

The expression of the RDM1 gene causes methylation of DNA in theRdDM pathway (Feng et al., 2010; Gao et al., 2010; Mehrotra andGoyal, 2014). The effect of PtRDM1 overexpression on methylation of5.8S rDNA and histone H3 located outside of the centrosome in P.

trichocarpa was assayed by genomic bisulfite sequencing following ex-position to various NaCl doses. The DNA methylation percentage atsymmetric CpG sites in 5.8S and histone H3 was increased significantlyfollowing treatment with increasing NaCl treatments (Fig. 6). In 0mMNaCl treatment, roots from transformant poplars revealed maximummethylated CpG sites in histone H3 by ∼50%, while the minimummethylated CpG sites has been shown in leaves from WT poplarsthroughout 5.8S rDNA by ∼10% (Fig. 6A, B and C). In 20mM NaCltreatment, roots from transformant poplars exhibited the most methy-lated CpG sites in histone H3 by ∼65%, while the less methylated CpGsites has been shown in leaves from WT poplars throughout 5.8S rDNAby ∼25% (Fig. 6D, E and F). In 40mM NaCl treatment, histone H3 and5.8S rDNA of roots from transgenic poplars revealed maximum me-thylated CpG sites by ∼75%–∼80%, while the minimum methylatedCpG percentage has been shown in leaves from WT poplars throughout5.8S rDNA by ∼40% (Fig. 6G, H and I). In 60mM NaCl treatment(severe conditions), histone H3 from both stems and roots in transfor-mants revealed the utmost percent of DNA methylation by

Fig. 3. (A) Roots showed more expression of PtRDM1 gene in transformant poplars (Gray bars) versus WT (non-transgenic, black line) plants via severe conditions (40 and 60mM NaCltreatments). (B) Stems and leaves in transformants exhibited more expression of PtROS1 gene compared to WT poplars via increasing NaCl treatments. Each NaCl treatment was carriedout using three independent transgenic and one WT poplars with three repeats; error bars represent SE and asterisks indicate significant differences.

Fig. 4. (A) DNA methylation assayed by PCR. Non-digested 5.8S rDNA and histone H3 genes, which lack CpG and CpHpG sites, served as the PCR control with the same amount ofdigested DNA in each sample. (B) Southern blot analysis of digested genomic DNA by HpaII (CCGG) to probe 5.8S rDNA. In addition, ycf4 gene was used as a control.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

68

∼80%–∼85%, respectively. In contrast, 5.8S rDNA from stems in WTpoplars revealed the less percent of DNA methylation by ∼40%(Fig. 6J, K and L).

Our results of percent of methylation on asymmetric CpHpG andCpHpH sites exhibited that NaCl treatments enhanced the DNA me-thylation in both WT and transformant poplars throughout 5.8S rDNAand histone H3 loci (Fig. 7). In 0mM NaCl treatment (common condi-tions), histone H3 and 5.8S rDNA loci exhibited more methylatedCpHpG sites in transformant and WT poplars respectively by∼45% and

∼15% (Fig. 7A and B). In total, histone H3 from transformant poplarsrevealed the maximum methylated CpHpG sites by ∼90% (Hyper-methylation) in severe condition, whereas the minimum methylatedCpHpG sites by ∼10% (Hypo-methylation) has been observed in 5.8SrDNA from WT poplars in common condition (Fig. 7A and B). Fur-thermore, in common condition, histone H3 and 5.8S rDNA loci re-vealed more methylated CpHpH sites in transformant poplars (respec-tively by ∼40% and ∼35%); while WT poplars with the samepercentage by ∼10% (Fig. 7C and D). In total, histone H3 from

Fig. 5. Investigation of DNA methylation using ScrFI (CCNGG). (A) Analyses of DNA methylation relative to WT poplars, resulted by ScrFI digestion of 5.8S. (B) Analysis of DNAmethylation relative to WT poplars, resulted by ScrFI digestion of histone H3. Average of three independent amplifications; error bars represent SD; Ycf4 gene was analyzed and used as acontrol.

Fig. 6. Investigation of symmetric methylated CpG sites using bisulfite sequencing. Salt stress enhanced DNA methylation throughout 5.8S rDNA and histone H3 in both WT andtransformant poplars. Ycf4 gene (Green lines) was analyzed to use as a control. Transformants (Orange lines) represented more DNA methylation than WT poplars (Blue lines). (A, B andC) In 0mM NaCl treatment, the maximum DNA methylation has been observed in histone H3 of roots from transformant poplars. (D, E and F) In 20mM NaCl treatment, the maximumDNA methylation was occurred in histone H3 of roots from transformant poplars. (G, H and I) In moderate NaCl treatment, the maximum DNA methylation was happened in histone H3 ofroots from transformant poplars. (J, K and L) the maximum DNA methylation was occurred in histone H3 of roots from transformant poplars. (For interpretation of the references tocolour in this figure legend, the reader is referred to the Web version of this article.)

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

69

transformant poplars presented the maximum methylated CpHpH sitesby ∼63% in severe condition, whereas 5.8S rDNA and histone H3 fromWT poplars presented the minimum methylated CpHpH sites incommon condition by ∼10% (Fig. 7C and D).

In addition, purified PCR fragments were digested by methylation-sensitive restriction enzymes HpaII and MspI to verify results obtainedby bisulfite sequencing of 5.8S rDNA and histone H3 loci. In this assaywe used HpaII and MspI enzymes, which recognize only unmethylatedcytosines in CpG contexts. In this assay, methylated DNA is not con-verted by bisulfite and can be cleaved by these enzymes. In contrast,

unmethylated DNA is converted and becomes resistant to digestion. Asshown in Fig. 8A and B, the cleaved PCR fragment 416 and 103 bpbands of 5.8s rDNA locus from PtRDM1 transformants indicate methy-lated cytosines, while the uncleaved PCR fragment 548 bp bands fromWT poplars represent unmethylated cytosines. Also, the cleaved PCRfragment 574 and 82 bp (plus an additional 63 bp band not visible on a2.5% agarose gel) bands of histone H3 locus from PtRDM1 transformantpoplars indicate methylated cytosines, while the uncleaved PCR frag-ment 719 bp bands from WT poplars indicate methylated cytosines.Genomic DNA from PtRDM1-overexpressing transgenic poplars (plants

Fig. 7. Genomic bisulfite sequencing of the 5.8S rDNA and histone H3 loci following treatment with the indicated NaCl concentrations. (A, C) Methylation percentage of asymmetricCpHpG and CpHpH sites in 5.8S. (B, D) Methylation percentage of asymmetric CpHpG and CpHpH sites in histone H3. Bars represent SE and asterisks indicate significant differences. Linesrepresent methylation percentage of ycf4 gene as a control.

Fig. 8. (A) Verification of bisulfite sequencing in 5.8S rDNA and histone H3 loci by HpaII restriction endonuclease. The ycf4 gene was digested with HpaII in WT and transformants poplarsas a control. (B) Sequences of 5.8S rDNA and histone H3 with relevant HpaII and MspI restriction sites are indicated. (C) The percent of DNA methylation of 5.8S and histone H3 loci intransformant and WT poplars resulted by HpaII and MspI digestion.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

70

11 and 17) exhibited hyper-methylation in both 5.8S and histone H3loci compared to WT (plants 5 and 7) (Fig. 8C).

3.5. Phenotypic analyses of PtRDM1 transgenic and WT poplars

Visible results showed that transformants exhibited more resistancethan WT poplars against NaCl treatments (0, 20, 40 and 60mM)(Supplementary 9). In addition, more experiments have been carriedout to show the role of increased methylated DNA in phenotypicchanges in comparing with WT poplars.

3.5.1. Stem lengthTransformants revealed significantly more stem length than WT

poplars via 0, 20, 40 and 60mM NaCl treatment. Severe condition(60mM NaCl treatment) stopped the growth of WT poplars and de-moted stem length by ∼50% after 10 and 20 days, while it exposed noimpact on the growth in transformant poplars in the same conditions(Fig. 9). In addition, transformants exhibited ∼2.5 fold stem length incomparing with WT poplars after 10 and 20 days in 60mM NaCltreatments.

3.5.2. Main root regenerationFurthermore, transgenic poplars revealed more regeneration of

roots than WT poplars via NaCl treatments. The regeneration of rootsexhibited significant differences between WT and transgenic poplars in20, 40 and 60mM NaCl treatments. Our results exhibited that severecondition (60mM NaCl treatment) promoted the regeneration of rootsin transformants by ∼2 fold after 10 and 20 days, but there was nomeaning differences in WT poplars (Fig. 10).

3.5.3. Degraded stem growthThe evaluation of degraded stem growth showed that 40 and 60mM

NaCl treatments enhanced significantly the interruption of stem growthin WT poplars in comparing with transformants. In 40 and 60mM NaCltreatments, WT poplars revealed ∼5 fold and ∼3.5 fold degraded stemgrowth in comparing with transformants after 10 and 20 days, re-spectively. There was no degraded stem growth in transformants viaNaCl treatments after 10 days (Fig. 11).

3.5.4. Number of leavesIn addition, transgenic poplars significantly presented more number

of leaves than WT poplars. Severe condition (60mM NaCl treatment)reduced the generation of leaves in WT poplars by ∼40% after 10 and20 days, whereas it was increased in transformants by ∼1.25 fold(Fig. 12).

3.5.5. Wavy leavesWavy leaves then was evaluated via NaCl treatments in WT and

transformant poplars. Sever condition (60mM NaCl treatment) in-creased wavy leaves in both WT and transformant poplars by ∼1.5 and3 fold, respectively (Fig. 13).

4. Discussion

The overexpression of PtRDM1 was carried out in poplar clones toanalyze changes of DNA methylation in repetitive DNA sequences suchas 5.8S rDNA and histone H3 under 0, 20, 40 and 60mM NaCl treat-ments. Quantitative PCR analysis revealed that the expression ofPtRDM1 was statistically increased in transgenic poplars compared toWT in roots, stems and leaves. It has been shown for the first time that

Fig. 9. In vitro phenotypic stem length changes of the PtRDM1 transgenic and WT poplars. Transformants presented more stem length (cm) than WT poplars via NaCl treatments after 10and 20 days. Five independent plants from WT and transformant poplars have been used for each NaCl treatment; asterisks represent significant differences; bars represents SE.

Fig. 10. In vitro phenotypic main root generation changes of the PtRDM1 transgenic and WT poplars. Transformants revealed more numbers of main root regeneration than WT poplarsafter 10 and 20 days. Five independent plants from WT and transformant poplars have been used for each NaCl treatment; asterisks represent significant differences; bars represents SE.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

71

NaCl treatments enhanced the expression of PtRDM1 in transgenic po-plars, while it reduced the expression of PtRDM1 in WT poplars, ingeneral. Also, it has been shown for the first time that the expression ofPtRDM1 in roots was more than leaves and stems in transformant andWT poplars. Lei et al. (2015) revealed that there is a balance betweenRdDM pathway and ROS1 gene in DNA methylation. For that is why,we investigated the expression of PtROS1 by qPCR and our results re-vealed that 40 and 60mM NaCl treatments boosted the expression ofPtROS1 in WT and transgenic poplars in comparison with the expres-sion of PtRDM1 in the same conditions.

The maximum DNA methylation will be occurred in repetitive DNAsequences such as transposons or retro-transposons, centromerictandem repeats and ribosomal DNA arrays (Fransz and de Jong, 2002;Lafon-Placette et al., 2013; Zheng et al., 2016). In this study, DNAmethylation was investigated by PCR on 5.8S rDNA and H3 histone loci.In regard with Ikeuchi et al. (2015), our results showed that DNA me-thylation was increased through loci in transformant poplars comparedto WT via fading gel bands resulted by digestion of McrBC restrictionenzyme. Also, southern blot analysis proved that HpaII digestion siteswas lacked in transformant poplars resulted by increasing DNA me-thylation compared to WT. Southern blot analysis also showed thatNaCl treatment can be a stimulant for DNA methylation in comparingboth NaCl treated transformants and WT poplars with untreated plants.Quantitative PCR was carried out using ScrFI restriction enzyme toprove the enhanced DNA methylation in transformant poplars by∼90–150% in 5.8S rDNA and by ∼115–180% in histone H3 comparedto WT poplars. Furthermore, bisulfite sequencing was carried out toshow the effect of the overexpression of PtRDM1 in plants on DNAmethylation via different stimulative NaCl treatments. It has beenproved that hyper-methylation mostly occurs in symmetric CpG sites(Gao et al., 2010). Our results revealed that transformant poplars pre-sented that methylated CpG sites were occurred in histone H3 more

than 5.8S rDNA in both WT and transformant poplars. Also, NaCltreatments enhanced the symmetric CpG sited in roots more than stemsand leaves in WT and transformant poplars.

Like to Cao and Jacobsen (2002) our results also revealed that thepoplar genome is also likely hyper-methylated in asymmetric CpHpGand CpHpH sites by methyltransferase proteins, such as CMT3, DRM1and DRM2 associated with RDM1. In this study we investigated theeffect of salt stress on DNA methylation in 5.8S rDNA repetitive se-quences for the first time. Methylation percentage of asymmetricCpHpG and CpHpH sites in 5.8S was significantly increased in trans-formant poplars compared to WT plants following NaCl treatments. Theresults of PCR also revealed that the overexpression of PtRDM1 has asignificant effect on 5.8S gene body to hyper-methylate. In addition, inthis study we investigated the effect of salt stress on DNA methylationin histone H3 locus for the first time. Methylation percentage of CpHpGand CpHpH sites in histone H3 was significantly increased in transfor-mant poplars compared to WT plants following NaCl treatments. Fur-thermore, the overexpression of PtRDM1 methylated histone H3 morethan 5.8S rDNA. Our results proved that RDM1 protein enhanced theDNA methylation by chromatin remodeling (e.g. histone H3) more thanrepetitive DNA sequences (e.g. 5.8S rDNA).

Furthermore, we investigated the overexpression of PtRDM1 onpleiotropic phenotypic changes in poplars. Transformant poplarsshowed more stem length than WT poplars in all NaCl treatments (0,20, 40 and 60mM). NaCl treatments reduced stem length in WT po-plars, while it had no impact on transgenic poplars. It has been shownthat salt stress in different concentrations affects on roots, then trans-formant plants respond to the stress by increasing number of roots(clawed roots) (Movahedi et al., 2015a,b,c). In addition, Demirkiranet al. (2013) explained that the lower concentration of NaCl treatmentincreased root growth. Furthermore, Demirkiran et al. (2013) provedthat higher concentration of salinity caused nucleotide variations and

Fig. 11. Investigation of degraded stem growth of the PtRDM1 transgenic and WT poplars. WT poplars exhibited more degraded stem growth than transformant poplars via increasing ofsalt stresses. Five independent plants from WT and transformant poplars have been used for each NaCl treatment; asterisks represent significant differences; bars represents SE.

Fig. 12. The number of leaves have been evaluated in thePtRDM1 transgenic and WT poplars. Transgenic poplars showedmore number of leaves than WT poplars at all salt stress condi-tions after 10 and 20 days. Five independent plants from WT andtransformant poplars have been used for each NaCl treatment;asterisks represent significant differences; bars represents SE.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

72

led to hypermethylation. Our results, showed that NaCl treatmentscaused to stop regeneration of roots in WT poplars, but it had no impacton transgenic poplars. In addition, our results proved that upregulationof PtRDM1 caused a hypermethylation of 5.8S rDNA and histone H3 inroots and revealed more regeneration of roots (number of main roots)than WT poplars via different NaCl treatments. Results revealed thatsalt stresses degraded the growth of stems in WT poplars much morethan transgenic poplars. Also, transformants showed more number ofleaves than WT poplars. NaCl treatments demoted number of leaves inWT poplars but it had no effect on transgenic poplars.

Therefore, DNA methylation in poplars in which RDM1 was upre-gulated demonstrated that the overexpression of PtRDM1 led to hyper-methylation in various cytosine contexts, which resulted in increasedresistance to salinity stresses via an alteration of phenotypes.

Author's contributions

In this research, Dr. Ali Movahedi directed totally the researchgroup and prepared the manuscript. All the experiments were carriedout by AM, WS, KM, AAY, HW and XW. The bioinformatics were de-signed and analyzed by AM. The ZQ and TY supervised this research.All the authors reviewed this manuscript and have no conflicts of in-terest to declare.

Acknowledgments

This work was supported by the National Science Foundation ofChina (No. 31570650)，the International Science & TechnologyCooperation Program of China (2014DFG32440) and the PriorityAcademic Program Development of Jiangsu Higher EducationInstitutions.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.plaphy.2018.03.011.

References

Allard, S.T., Bingman, C.A., Johnson, K.A., Wesenberg, G.E., Bitto, E., Jeon, W.B., PhillipsJr., G.N., 2005. Structure at 1.6 A resolution of the protein from gene locusAt3g22680 from Arabidopsis thaliana. Acta Crystallographica. F, Struct. Biol. cryst.commun. 61, 647–650.

Cao, X.F., Jacobsen, S.E., 2002. Locus-specific control of asymmetric and CpNpG me-thylation by the DRM and CMT3 methyltransferase genes. Proc. Natl. Acad. Sci.U.S.A. 99, 16491–16498.

Demirkiran, A., Marakli, S., Temel, A., Gozukirmizi, N., 2013. Genetic and epigeneticeffects of salinity on in vitro growth of barley. Genet. Mol. Biol. 36, 566–570.

Dimitrova, A.D., Georgiev, O., Mishev, K., Tzvetkov, S., Ananiev, E.D., Karagyozov, L.,2016. Mapping of unmethylated sites in rDNA repeats in barley NOR deletion line. J.Plant Physiol. 205, 97–104.

Feng, S., Jacobsen, S.E., Reik, W., 2010. Epigenetic reprogramming in plant and animaldevelopment. Science 330, 622–627.

Fransz, P.F., de Jong, J.H., 2002. Chromatin dynamics in plants. Curr. Opin. Plant Biol. 5,560–567.

Gao, Z., Liu, H.L., Daxinger, L., Pontes, O., He, X., Qian, W., Lin, H., Xie, M., Lorkovic,Z.J., Zhang, S., Miki, D., Zhan, X., Pontier, D., Lagrange, T., Jin, H., Matzke, A.J.,Matzke, M., Pikaard, C.S., Zhu, J.K., 2010. An RNA polymerase II- and AGO4-asso-ciated protein acts in RNA-directed DNA methylation. Nature 465, 106–109.

Gu, X., Gao, Z., Yan, Y., Wang, X., Qiao, Y., Chen, Y., 2017. RdreB1BI enhances droughttolerance by activating AQP-related genes in transgenic strawberry. Plant Physiol.Biochem. : PPB (Plant Physiol. Biochem.) 119, 33–42.

Hirakawa, T., Hasegawa, J., White, C.I., Matsunaga, S., 2017. RAD54 forms DNA repairfoci in response to DNA damage in living plant cells. Plant J. 90, 372–382.

Ikeuchi, M., Iwase, A., Sugimoto, K., 2015. Control of plant cell differentiation by histonemodification and DNA methylation. Curr. Opin. Plant Biol. 28, 60–67.

Lafon-Placette, C., Faivre-Rampant, P., Delaunay, A., Street, N., Brignolas, F., Maury, S.,2013. Methylome of DNase I sensitive chromatin in Populus trichocarpa shoot apicalmeristematic cells: a simplified approach revealing characteristics of gene-body DNAmethylation in open chromatin state. New Phytol. 197, 416–430.

Lei, M., Zhang, H., Julian, R., Tang, K., Xie, S., Zhu, J.K., 2015. Regulatory link betweenDNA methylation and active demethylation in Arabidopsis. Proc. Natl. Acad. Sci.U.S.A. 112, 3553–3557.

Li, Y., Tollefsbol, T.O., 2011. DNA methylation detection: bisulfite genomic sequencinganalysis. Meth. Mol. Biol. 791, 11–21.

Ma, L., Hatlen, A., Kelly, L.J., Becher, H., Wang, W., Kovarik, A., Leitch, I.J., Leitch, A.R.,2015. Angiosperms are unique among land plant lineages in the occurrence of keygenes in the RNA-directed DNA methylation (RdDM) pathway. Genome Biology andEvolution 7, 2648–2662.

Mehrotra, S., Goyal, V., 2014. Repetitive sequences in plant nuclear DNA: types, dis-tribution, evolution and function. Dev. Reprod. Biol. 12, 164–171.

Movahedi, A., Sun, W., Zhang, J., Wu, X., Mousavi, M., Mohammadi, K., Yin, T., Zhuge,Q., 2015a. RNA-directed DNA methylation in plants. Plant Cell Rep. 34, 1857–1862.

Movahedi, A., Zhang, J., Amirian, R., Zhuge, Q., 2014. An efficient Agrobacterium-mediated transformation system for poplar. Int. J. Mol. Sci. 15, 10780–10793.

Movahedi, A., Zhang, J.X., Gao, P.H., Yang, Y., Wang, L.K., Yin, T.M., Kadkhodaei, S.,Ebrahimi, M., Qiang, Z.G., 2015b. Expression of the chickpea CarNAC3 gene en-hances salinity and drought tolerance in transgenic poplars. Plant Cell Tissue OrganCult. 120, 141–154.

Movahedi, A., Zhang, J.X., Yin, T.M., Qiang, Z.G., 2015c. Functional analysis of two or-thologous NAC genes, CarNAC3, and CarNAC6 from cicer arietinum, involved inabiotic stresses in poplar. Plant Mol. Biol. Rep. 33, 1539–1551.

Ngernprasirtsiri, J., Kobayashi, H., Akazawa, T., 1988. DNA methylation occurred aroundlowly expressed genes of plastid DNA during tomato fruit development. Plant Physiol.88, 16–20.

Omidvar, V., Fellner, M., 2015. DNA methylation and transcriptomic changes in responseto different lights and stresses in 7B-1 male-sterile tomato. PLoS One 10 e0121864.

Questa, J.I., Fina, J.P., Casati, P., 2013. DDM1 and ROS1 have a role in UV-B induced- andoxidative DNA damage in A. thaliana. Front. Plant Sci. 4, 420.

Resentini, F., Cyprys, P., Steffen, J.G., Alter, S., Morandini, P., Mizzotti, C., Lloyd, A.,Drews, G.N., Dresselhaus, T., Colombo, L., Sprunck, S., Masiero, S., 2017.SUPPRESSOR of FRIGIDA (SUF4) supports gamete fusion via regulating ArabidopsisEC1 gene expression. Plant Physiol. 173, 155–166.

Roudier, F., Ahmed, I., Berard, C., Sarazin, A., Mary-Huard, T., Cortijo, S., Bouyer, D.,Caillieux, E., Duvernois-Berthet, E., Al-Shikhley, L., Giraut, L., Despres, B.,Drevensek, S., Barneche, F., Derozier, S., Brunaud, V., Aubourg, S., Schnittger, A.,Bowler, C., Martin-Magniette, M.L., Robin, S., Caboche, M., Colot, V., 2011.Integrative epigenomic mapping defines four main chromatin states in Arabidopsis.EMBO J. 30, 1928–1938.

Santner, A., Estelle, M., 2009. Recent advances and emerging trends in plant hormonesignalling. Nature 459, 1071–1078.

Wang, J., Blevins, T., Podicheti, R., Haag, J.R., Tan, E.H., Wang, F., Pikaard, C.S., 2017.Mutation of Arabidopsis SMC4 identifies condensin as a corepressor of pericen-tromeric transposons and conditionally expressed genes. Genes Dev. 31, 1601–1614.

Zheng, D., Ye, W., Song, Q., Han, F., Zhang, T., Chen, Z.J., 2016. Histone modificationsdefine expression bias of homoeologous genomes in allotetraploid cotton. PlantPhysiol. 172, 1760–1771.

Zhu, R.Q., Shevchenko, O., Ma, C., Maury, S., Freitag, M., Strauss, S.H., 2013. Poplarswith a PtDDM1-RNAi transgene have reduced DNA methylation and show aberrantpost-dormancy morphology. Planta 237, 1483–1493.

Fig. 13. In vitro phenotypic wavy leaves of the PtRDM1 trans-genic and WT poplars. Transformant poplars exhibited morewavy leaves affected by the over-expression of PtRDM1gene thanWT poplars at all salt stress conditions after 10 and 20 days. Fiveindependent plants from WT and transformant poplars have beenused for each NaCl treatment; asterisks represent significantdifferences; bars represents SE.

A. Movahedi et al. Plant Physiology and Biochemistry 127 (2018) 64–73

73