Salt-induced transcription factor MYB74 is regulated by the RNA-directed DNA methylation pathway in Arabidopsis

1High salinity is a crucial problem affecting plant growth and crop production in many parts of the world. Recently, many salt-stress responsive genes and protein have been identified by both forward and reverse genetics approaches. The transcription factors, such as CCAAT-binding transcription factors (CBFs), NAM, ATAF1/2 CUC2 transcription factors (NACs), and WRKYs, act as key regulators in response to salt stress in plants. The MYB proteins also function as transcription factors that play regulatory roles in the defence responses of plants. R2R3-MYB is the largest subfamily of the MYB family, which includes 126 members and can be further categorized into 22 subgroups. To date, several R2R3-MYB proteins have been reported to be involved in the abiotic stress responses of Arabidopsis. Overexpression of AtMYB2 results in the enhanced expression of RD22 (a dehydration-responsive gene) and AtADH1 (alcohol dehydrogenase1), and the transgenic plants display higher sensitivity to abscisic acid. The mutants of AtMYB108 are hypersensitive to salt, drought, and oxidative stresses, which are possibly mediated by reactive oxygen intermediates. Subgroup 11 consists of three members, AtMYB41, AtMYB102, and AtMYB74. AtMYB41 is involved in the control of primary metabolism and negative regulation of short-term transcriptional responses to osmotic stress. AtMYB102 functions as a key factor in both osmotic stress and wounding signalling pathways in Arabidopsis. AtMYB74, as one of the stress-upregulated genes, has been reported in a general profile of the expression pattern of the MYB family. However, the molecular mechanism of the response of AtMYB74 to abiotic stress is largely unknown.

2Genome expression is mainly influenced by chromatin structure, which is governed by processes often associated with epigenetic regulation, including histone post-translational modification and DNA methylation. In histone modification, arginine and lysine methylation is also involved in transcriptional regulation. Recent evidence indicates that DNA methylation and siRNA participate in the regulation of gene expression in plants in response to environmental stresses. DNA methylation occurs in the contexts of CG, CHG, and CHH (where H is adenine, cytosine, or thymine) in plants. Methyltransferase 1 (MET1), chromomethylase 3(CMT3), and domains rearranged methyltransferase (DRM) 2 have been characterized to function as DNA methyltransferases that transfer a methyl group to the cytosine bases of DNA to form 5-methylcytosine. MET1 and CMT3 are mainly in charge of the maintenance of CG and CHG methylation. DRM2 is responsible for de novo DNA methylation and exhibits the most prominent role in CHH methylation.

3RNA-directed DNA methylation (RdDM), which was first discovered in viroid-infected tobacco, is an important regulatory phenomenon involved in repressive epigenetic modifications that can trigger transcriptional gene silencing (TGS). Many key components of the RdDM pathway, such as nuclear RNA polymerase (NRP) D1/NRPE1, RNA-dependent RNA polymerase 2 (RDR2), argonaute 4/6, and dicer-like 3 (DCL3), have been identified, and its molecular mechanism has been established. In plants, DNA demethylation depends on four bifunctional 5-methylcytosine glycosylases, repressor of silencing 1 (ROS1), DEMETER, DEMETER-like protein (DML) 2, and DML3. ROS1 has been found to counteract the robust RdDM pathway at hundreds of discrete regions across the plant genome together with DML2 and DML3. Recent evidence reveals that Arabidopsis zinc finger DNA 3′ phosphoesterase is a DNA phosphatase that interacts with ROS1 and functions downstream of ROS1 in one branch of the active DNA demethylation pathway.

4The gain or loss of DNA methylation is correlated with a considerable decrease or increase in the corresponding amount of mRNA abundance and with the presence or absence of 24-nt siRNAs at each silenced epiallele. Withdrawal of the inducing siRNA signal could result in active or passive demethylation, which in turn leads to the loss of TGS. Approximately one-third of methylated DNA loci in Arabidopsis is associated with siRNA clusters, implying a primary determinant role of siRNAs in DNA methylation. In plants, the most abundant class of siRNAs includes heterochromatic siRNAs that originate from different sources, such as inverted repeats, pseudogenes, and natural cis-antisense transcript pairs. Increasing evidence has shown that the activity of siRNAs could be triggered by various environmental stimuli to affect the targeting chromatin structure. In Craterostigma plantagineum, an endogenous siRNA is induced during dehydration, which may contribute to dehydration tolerance. Salt stress also results in a dramatic change in the accumulation of three siRNAs in wheat seedlings. In Arabidopsis, the 24-nt SRO5-P5CDH natural antisense transcript siRNA is involved in the cleavage of P5CDH mRNA, which in turn increases proline accumulation under salt stress. Exposure of Arabidopsis plants to salt, UVC, cold, heat, and flood stresses increases global genome methylation in the progeny, in which Dicer-like protein is involved.

5With the development of bisulphite sequencing, recent studies have revealed previously uncharted subsets of the epigenome and provided insights into the complex interplay between DNA methylation and transcription. Although the salt stress signal transduction pathway has been intensively studied, whether or not DNA methylation/demethylation is involved in this pathway remains unclear. Despite the fact that salinity stress alters the global DNA methylation level in plants, the mechanism of how siRNA mediates DNA methylation by RdDM to respond to salt stress also needs to be elucidated. The present study shows that the expression of AtMYB74 is regulated by RdDM, and the accumulation of 24-nt siRNAs is the major contributor to RdDM. The findings provide a new insight into salt stress regulation in plants and allow a better understanding of the role of siRNAs in controlling the RdDM pathway to regulate gene expression in response to abiotic stress.

6Arabidopsis thaliana (Col-0) was used as the wild type (WT) and the genetic background for transgenic plants in this study. Dry seeds were collected and stored in a dehumidifier cabinet for at least 2 months before the seed germination test was performed. Seeds were stratified and sown on agar plates containing 1% Suc as described previously at 4 °C for 2 days and then transferred to 23 °C. Arabidopsis seedlings were grown under continuous light (70 μmol·m–2·s–1) at 23±1 °C. Soil-grown Arabidopsis and Nicotiana benthamiana plants were grown under a 16h light/8h dark photoperiod at 23±1°C. For the germination assay, at least 100 seeds of each genotype were sterilized and sown on Murashige and Skoog (MS) medium supplemented with or without phytohormones or chemicals. Germination was defined as the first sign of radicle tip emergence and scored daily, and the germination results were calculated based on at least three independent experiments. At 3 days post-germination, the plants of the 5-azacytidine (5-azaC) group were transferred to 50mM 5-azaC. 14-day-old seedlings were subjected to NaCl treatments by transfer to MS liquid medium with 150mM NaCl for durations as indicated, respectively. All these treatments were carried out under a growth condition of 16h light/8h dark at 23 °C unless otherwise mentioned.

7The binary vector pBI121 used for overexpression of AtMYB74 and the binary vector pFGC5941 used for RNAi were introduced into Agrobacterium tumefaciens strain GV3101 and the Arabidopsis thaliana (Col-0) plants were transformed by floral dipping. All constructs were verified by sequencing. The transgenic plants were screened on MS medium containing 50 µg ml–1 kanamycin for pBI121 and 10 µg ml–1 glufosinate ammonium for pFGC5941. T1 transgenic Arabidopsis plants were identified by quantitative real-time reverse transcription PCR (qRT-PCR). The corresponding T2 transgenic seedlings that segregated at a ratio of 3:1 (resistant:sensitive) were selected to propagate T3 individuals. RNAi-3 and RNAi-6 were used for further analysis.

8Different constructs were transformed into A. tumefaciens strain GV3101. Overnight cultures were harvested and mixed at a 1:1 ratio with the different construct groups. After incubation for 3h at room temperature in 10mM MgCl2, 10mM 2-(N-morpholino) ethanesulfonic acid hydrate, pH 5.6, and 150mM acetosyringone, the Agrobacterium suspension was coinfiltrated into 3-week-old N. benthamiana leaves. Infected leaves were harvested 48h after the infiltration. β-Glucuronidase (GUS) activity and small RNA extraction were performed as described below. Each assay was obtained from at least five independent lines and repeated three times.

9The promoter sequence of AtMYB74 was acquired from the TAIR database (http://www.arabidopsis.org/). The promoter–GUS recombinant construct was transformed into A. tumefaciens strain GV3101 and then introduced into Arabidopsis by the floral dip method. Primers for amplifying the promoter sequence are shown in Supplementary Table S1 at JXB online. Histochemical localization of GUS activities in the transgenic seedlings or different tissues was performed after the transgenic plants had been incubated overnight at 37°C in 1mg ml–1 5-bromo-4-chloro-3-indolyl-glucuronic acid, 5mM potassium ferrocyanide, 0.03% Triton X-100 and 0.1M sodium phosphate buffer, pH 7.0. After incubation, the tissues were cleared with 70% ethanol. The cleaned tissues were then observed and photographs were taken by using a stereoscope. For examination of the detailed GUS staining, the tissues were observed with a bright-field microscope and photographed. These GUS staining data were representative of at least five independent transgenic lines for each construct.

10Tobacco transgenic plants (100mg) were ground with a mortar with 1000 μl GUS extraction buffer (50mM NaH2PO4, 10mm EDTA, 0.1% Triton X-100, 0.1% sarcosyl) and centrifuged at 5000rpm for 10min at 4 °C. The crude extract of total protein was obtained from the supernatant. The protein concentration of the extract was determined using a NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific). Fluorometric GUS assays were performed as previously described. GUS activity was measured with 4-methylumbelliferyl-β-D-glucuronide as substrate with a Hitachi F-4500 Fluorescence Spectrofluorometer. The standard curves were prepared with 4-methylumbelliferone. GUS activity was expressed as pmol methylumbelliferone min–1 mg–1 protein. Average GUS activity was obtained from at least five independent transformants and each assay was repeated three times.

11For RNA isolation, leaves and roots of seedlings were harvested separately, frozen in liquid nitrogen, and stored at –80°C until use. Total RNA was isolated from different Arabidopsis thaliana seedlings (100mg) with TRIZOL reagent (Invitrogen 15596-026). Contaminated DNA was removed with RNase-free DNase I. First-strand cDNA synthesis was performed with 1 μg RNA using oligo(dT) primer or gene-specific primers. A Prime script RT reagent kit with gDNA Easer (Takara RR047A) was used for all reactions according to the manufacturer’s protocol. cDNAs were diluted 1/50 for qRT-PCR, and qRT-PCR was performed using the FastStart Essential DNA Green Master (Roche 6402712001) and a CFX96 Real-Time System (Bio-Rad). The reaction volume was 15 μl and included 2×FastStart Essential DNA Green Master, 6 μl diluted cDNA, and 1.5 μl of 10 μM primers. Cycle conditions were: 95 °C for 10min; 45 cycles of 95 °C for 10 s, 60 °C for 15 s, and 72 °C for 20 s. Fluorescence was read following each annealing and extension phase. Melt curve analysis of real-time PCR products was performed to verify amplification of a single product. The qRT-PCR experiment was carried out at least three times under identical conditions using tubulin as an internal control. Primers for amplifying genes were designed according to the sequences from the TAIR database (http://www.arabidopsis.org/). Details of primers are listed in Supplementary Table S1. Gene expression was normalized by subtracting the CT value of the control gene from the CT value of the gene of interest. Average expression ratios were obtained from the equation 2–ΔΔCT, according to a previously described protocol.

12The RNA blot analysis was carried out as described previously. 50 μg of each RNA was subjected to electrophoresis on a 15% TBE–urea, Criterion gel (Bio-Rad 345-0091) and electroblotted onto Hybond-N+ filter paper (Amersham RPN303B) using a TransBolt-SD apparatus (Bio-Rad 170–3940). The filter then was hybridized at 37 °C in Hybexpression buffer with a 32P-labeled probe to detect the five 24-nt siRNAs targeting the AtMYB74 promoter. The 300bp probe was made by labelling a DNA template that was amplified by primers (listed in Supplementary Table S1) with the Prime-a-Gene Labeling System (Promega U1100) and 32P. The filters were washed twice at 37 °C in buffer containing 2×SSC (0.3M NaCl and 0.03M sodium citrate) and 0.5% SDS.

13Small RNA for siRNA qRT-PCR was isolated with the miRcute miRNA Isolation kit (TIANGEN DP501), and first-strand cDNA synthesis was performed with the miRcute miRNA first-strand cDNA synthesis kit (TIANGEN KR201-02). siRNA qRT-PCR analysis was performed with 0.5 μg small RNA and a miRcute miRNA qRCR detection kit (TIANGEN FP401). The experiments were performed at least three times under identical conditions using U6 RNA as an internal control. Details of primers are listed in Supplementary Table S1.

14Aliquots of 800ng DNA were treated with sodium bisulphite using the EZ DNA Methylation-Gold kit (ZYMO RESEARCH D5005) according to the manufacturer’s instructions. The chloroplast genome of every sample was used to calculate the conversion efficiency. Conversion efficiency was >98% for each bisulphite-treated sample. DNA was amplified by PCR with ExTaq (TaKaRa RR01CM). Primer sequences are shown in Supplementary Table S1. PCR products were cloned into the pMD18-T Simple Vector (TaKaRa D103B) and the clones were sequenced. For each region, more than 20 independent top-strand clones were sequenced from each sample. Sequenced results were calculated by using CyMATE (http://www.cymate.org/).

15Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: MYB74 (At4g05100), drm1-2 (SALK_031705), drm2-2 (SALK_150863), cmt3-11 (SALK_148381), rdr2-1 (SAIL_1277H08), dcl3-1 (SALK_005512).

16Transcriptome analysis revealed that AtMYB74 expression level in Arabidopsis was increased dramatically by NaCl treatments. To determine the biological function of AtMYB74 and to test the responses to NaCl, overexpression (OE) and RNAi transgenic Arabidopsis lines were generated. In the T3 generation, two independent OE lines and two independent RNAi lines were selected for further experiment (Fig. 1A). Overexpression of AtMYB74 also induced the expression of a set of known stress marker genes, including AtRD29B, AtRAB18, and AtRD20, all of which contain the conserved MYB recognition sites (TAACTG) in their promoter regions (Fig. 1B). These observations suggest that AtMYB74 acts as a transcription factor of the salt stress-induced marker genes involved in the salt signalling pathway. Under normal growth conditions, the germination rates of both the OE and RNAi lines were almost the same as those of WT plants, which all achieved rates of 100% at day 2 (Fig. 1C). However, after NaCl treatment, the germination of OE transgenic seeds was inhibited more severely than that of WT seeds, reaching 80% at day 7, when the germination of WT seeds reached almost 100%. The germination rates of RNAi transgenic seeds were similar to those of WT seeds (Fig. 1D). In addition, for 21-day-old seedlings, the percentage seedling survival of OE transgenic lines was much lower than that of WT and RNAi plants under salt stress conditions (Fig. 1E). The reduced germination rates and seedling survival of the OE lines further indicate that AtMYB74 is involved in the response to salt stress in plants.

17The full-length cDNA corresponding to the AtMYB74 mRNA is 975bp in length and encodes a putative protein of 324 amino acids. The R2 and R3 MYB domains (amino acids 13–65 and 66–117, respectively) of AtMYB74 are highly conserved with all other MYB proteins in Arabidopsis and in other plant species (Fig. 2A).

18To describe the temporal and spatial expression patterns of AtMYB74 in greater detail, qRT-PCR and promoter–GUS analysis were performed. The highest number of AtMYB74 transcripts was found in flowers, followed by rosette leaves and cauline leaves, with the roots, stems, and siliques exhibiting the fewest transcripts (Fig. 2B). The tissue pattern of GUS staining was consistent with the qRT-RCR analysis (Fig. 2C). These results suggest that AtMYB74 is constitutively expressed in various tissues at low abundance.

19To detect the subcellular localization of the AtMYB74 protein in plant cells, the AtMYB74 coding region was fused in the frame to the coding region for the C-terminal side of GFP under the control of the cauliflower mosaic virus 35S promoter. Onion epidermal cells transformed with an expression plasmid for the AtMYB74–GFP fusion protein exhibited GFP fluorescence in the nucleus (Fig. 2D). However, GFP fluorescence was observed in the entire region of the cell when intact GFP was expressed. These results illustrate that AtMYB74 is localized in the nucleus.

20Nucleotide sequence analysis revealed that substantial DNA methylation and siRNA target sites exist in the AtMYB74 promoter region (http://neomorph.salk.edu/epigenome/epigenome.html; http://bioinfo.uni-plovdiv.bg/starpro/). To investigate whether the RdDM pathway regulates AtMYB74 responses to salt stress, qRT-PCR was used to compare the expression of AtMYB74 in WT Col-0, 5-azaC (an inhibitor of DNA methylation)-treated WT, ddc (drm1/drm2/cmt3 triple mutant), dcl3 (dicer-like 3 mutant), rdr2 (RNA-dependent RNA polymerase 2 mutant), ros1-4 (ros1 mutant), and rdd (ros1/dml2/dml3 triple mutant) in response to 150mM NaCl treatment. As shown in Fig. 3A, the level of AtMYB74 transcripts increased significantly (~8-fold) in response to NaCl treatment in WT plants, indicating that AtMYB74 responds to salt stress signals at the transcriptional level. Additionally, the GUS staining assay confirmed that NaCl could enhance AtMYB74 promoter activity in all tested tissues (Fig. 2C). WT plants treated with 5-azaC, which were used as a control, exhibited a small increase in AtMYB74 expression under salt stress, indicating the effect of DNA methylation on AtMYB74 (Fig. 3A). Besides DNA methylation, this result may suggest the presence of other minor factors that may control the expression of AtMYB74. For RdDM mutants, the accumulation of AtMYB74 mRNA in methylation mutants (ddc, rdr2, and dcl3) showed much less change than that in WT plants under salt stress, whereas the deficiency of active DNA demethylation in ros1-4 and rdd mutants resulted in an obvious decrease in AtMBY74 expression after salt treatment (Fig. 3A). These results reveal that dynamic DNA methylation results in AtMYB74 activation in response to salt stress in Arabidopsis.

21To further understand the mechanism by which RdDM regulates AtMYB74 expression, the 200bp promoter region approximately 500bp upstream of the transcription initiation site of AtMYB74 was analysed by bisulphite sequencing. WT plants treated with NaCl exhibited a visible reduction in total 5-methylcytosine content compared with controls (Fig. 3B, Supplementary Fig. S1, Supplementary Fig. S2). Interestingly, the percentage of CHH methylation was nearly halved in the treated WT plants (Fig. 3C), whereas only ~10% reduction in CG contexts was detected (Fig. 3D). No CHG contexts were found in the 200bp region. Moreover, fewer methylated sites in CHH contexts were observed in the RdDM mutants and pharmaceutically treated plants, whether under salt stress or not. Additionally, the three DNA demethylases might not participate in the demethylation of CHH sites in the AtMYB74 promoter region. The reduction of AtMYB74 expression in ros1-4 and rdd mutants under salt stress might be regulated by its upstream regulators, which are affected by the demethylases (Fig. 3A, C). These findings suggest that the dynamic balance of DNA methylation and demethylation is crucial for the transcriptional regulation of AtMYB74.

22Time-course analysis revealed that the level of AtMYB74 mRNA under salt stress showed a peak from 0.5h, with the highest levels at 3h, accompanied by the lowest percentage of CHH methylation. After 6h, AtMYB74 expression under salt stress maintained a high level (~7-fold) in treated plants compared with that in the controls, and a lower percentage of CHH methylation was maintained in WT plants (Fig. 3E). These results demonstrate that the increase of AtMYB74 transcripts was correlated with the lower level of CHH methylation during the NaCl treatment, suggesting that the stress-increased expression of AtMYB74 resulted from a reduction in dynamic DNA methylation, mainly in the CHH context.

23To investigate why CHH hypomethylation occurs in WT plants under salt stress, a sequence analysis of AtMYB74 was performed using starPRO DB v1.0 (http://bioinfo.uni-plovdiv.bg/starpro/). Five 24-nt siRNAs (ASRP215119, ASRP41948, ASRP27256, ASRP13208, and ASRP2423) were identified to target a narrow region (–603 to –477bp) of the 2.9kb promoter of AtMYB74. These five 24-nt siRNAs were located in a cluster in the promoter region (Fig. 4A). Moreover, bisulphite sequencing analysis of individual clones showed that DNA methylation considerably changed in or near the siRNA target region (Fig. 4B). However, no obvious changes were detected in 5-azaC-treated WT or RdDM mutants (Supplementary Fig. S1), indicating that CHH hypomethylation is dependent on RdDM in or near the siRNA target region.

24To reveal the role of the siRNAs in RdDM, siRNA qRT-PCR and northern blot analysis were performed. The accumulation of five 24-nt siRNAs was substantially reduced under salt stress in WT plants compared with that in dcl3 plants (Fig. 4C, D), suggesting that decreases in DNA methylation caused by the reductions in 24-nt siRNA accumulation lead to the activation of AtMYB74 under salt stress.

25To demonstrate whether the ectopic expression of 24-nt siRNAs affects AtMYB74 promoter activity, A. tumefaciens-mediated transient cotransformation in tobacco (N. benthamiana) leaves was performed. Considering that the siRNAs diced from the hairpin RNA could induce transcriptional silencing of the target genes, two types of expression cassettes were constructed to generate siRNAs and express the reporter gene (Fig. 5A). For siRNA expression cassettes in pFGC5941, the 300bp AtMYB74 promoter containing the siRNA target region that harbors an inverted DNA repeat was used to generate a double-strand RNA as a silencer (R1). A 300bp inverted cDNA sequence of AtMYB74 was employed as a control (R2). In the reporter gene cassettes of pBI121, gusA was driven by the 35S promoter (P1), AtMYB74 promoter (P2), and mutated AtMYB74 promoter with the deletion of a 200bp region where these siRNAs are targeted (P3). As shown in Fig. 5B, GUS activities between groups P1 and P1-R1, P2 and P2-R2, and P2 and P3-R1 did not show obvious changes after transient cotransformation. The GUS activity of group P2-R1 was obviously lower than that of group P2, indicating that promoter activity is affected by the targeted siRNA. In addition, bisulphite sequencing analysis revealed that DNA methylation in the 200bp promoter region increased by 43% due to the siRNAs yielded by the hairpin RNA (Fig. 5C, D). To confirm the accumulation of 24-nt siRNAs in the infection zones of transformed leaves, siRNA northern blot analysis was carried out. The constructs R1, P1-R1, P2-R1, and P3-R1 could efficiently generate siRNAs in the transformed tobacco (Fig. 5E). Overall, these findings demonstrate that the ectopic expression of artificial siRNAs targeting the AtMYB74 promoter also can regulate GUS expression through RdDM.

26RdDM, which controls the expression of a number of genes in many developmental processes and abiotic stress responses, has been considered an important epigenetic pathway. However, only a few transcription factors regulated by RdDM, such as b1 and C-LEC1, have been identified. Although many members of the R2R3-MYB family, such as AtMYB2, AtMYB108, AtMYB41, and AtMYB102, are reported to be involved in abiotic stress responses in plants, little is known about whether and how RdDM participates in this response pathway. The present study characterized AtMYB74, a putative R2R3-MYB member. DNA methylation of the promoter of AtMYB74 negatively correlates with gene expression under salt stress. Moreover, when the RdDM region is deleted from the AtMYB74 promoter, transgenic plants display higher GUS activity, suggesting that RdDM as a repressive epigenetic modification affects transcription of AtMYB74 (Fig. 5). Using transgenic Arabidopsis overexpressing siRNAs (R1) further confirmed this conclusion (Fig. S3). Overall, this study demonstrates that RdDM directly controls the expression of AtMYB74, providing a novel regulatory pathway in the R2R3-MYB family. Using a gain-of-function approach, and according to the phenotypes of AtMYB74 overexpressing and RNAi transgenic plants, the findings suggest that there may be functional redundancy of the R2R3-MYB family, as reported by. These results may point to the existence of a network of inter-regulated MYB genes, including AtMYB2 and other members of the subgroup of AtMYB74, that respond to abiotic stresses in plants.

27Previous studies have indicated that siRNAs and long non-coding RNAs could be involved in de novo DNA methylation. Comparison of genome-wide methylation patterns and small RNAs in Arabidopsis reveals that ~37% of the methylated loci are related to siRNA clusters, suggesting a close relationship between DNA methylation and siRNAs. The present study identified a cluster of siRNAs that directly mediates DNA methylation in or near the target region of AtMYB74 in Arabidopsis. Transformation in tobacco using artificial siRNAs yielded by hairpin RNA further demonstrated the involvement of RdDM in the native siRNA-targeted region. Interestingly, artificial siRNAs can also direct DNA methylation in the 200bp region, separately from the native siRNA-targeted region (Fig. 5), implying that any artificial siRNA may regulate promoter activity via DNA methylation in its target region. The accumulation of 24-nt siRNAs and changes in CHH DNA methylation have similar tendencies in Arabidopsis under salt stress, suggesting that the accumulation of 24-nt siRNAs is the trigger of RdDM, in agreement with a previous study. In rice, it was reported that siRNAs generated by OsRDR6 regulate the post-transcriptional gene silencing of the isocitrate lyase (ICL) gene. Taken together, the present observations reveal that the RdDM component may affect the accumulation of 24-nt siRNAs and then trigger DNA methylation of AtMYB74 under salt stress, providing strong evidence of epigenetic modification in the plant response to abiotic stress. Further work is necessary to determine the mechanisms of RdDM components that affect the accumulation of 24-nt siRNAs and the biochemical function of AtMYB74 in response to salt stress.

28The salt stress signal transduction pathway is far more complex than has been suggested previously. As shown in Fig 6, previous studies have shown that some R2R3-MYB transcription factors, such as AtMYB2, AtMYB20, and AtMYB73, participate in the regulation of gene expression under salt stress. The present study demonstrated the involvement of AtMYB74, a R2R3-MYB family member, which is regulated by RdDM in controlling the positive regulation of transcriptional responses to salt stress. Salt stress induces the RdDM pathway via repressing the accumulation of siRNAs to activate AtMYB74 expression, which then transduces the related signalling into downstream processes that endow resistance to such stress.

29Supplementary data are available at JXB online.

30Supplementary Fig. S1. Bisulphite sequencing analysis of promoter methylation status of AtMYB74 promoter in 14-day-old 5-azaC treated WT, ddc, dcl3, rdr2, ros1-4, and rdd mutants after 150mM NaCl treatment.

31Supplementary Fig. S2. A schematic of the AtMYB74 promoter sequence with bisulphite sequencing region.

32Supplementary Fig. S3. RdDM regulation of AtMYB74 expression in transgenic Arabidopsis.

33Supplementary Table S1. Primers and probes used in this study.