Abstract Arabidopsis ASYMMETRIC LEAVES2 (AS2) plays a critical role in leaf adaxial–abaxial partitioning by repressing expression of the abaxial-determining gene ETTIN/AUXIN RESPONSE FACTOR3 (ETT/ARF3). We previously reported that six CpG dinucleotides in its exon 6 are thoroughly methylated by METHYLTRASFERASE1, that CpG methylation levels are inversely correlated with ETT/ARF3 transcript levels and that methylation levels at three out of the six CpG dinucleotides are decreased in as2-1. All these imply that AS2 is involved in epigenetic repression of ETT/ARF3 by gene body DNA methylation. The mechanism of the epigenetic repression by AS2, however, is unknown. Here, we tested mutations of NUCLEOLIN1 (NUC1) and RNA HELICASE10 (RH10) encoding nucleolus-localized proteins for the methylation in exon 6 as these mutations enhance the level of ETT/ARF3 transcripts in as2-1. Methylation levels at three specific CpGs were decreased in rh10-1, and two of those three overlapped with those in as2-1. Methylation levels at two specific CpGs were decreased in nuc1-1, and one of those three overlapped with that in as2-1. No site was affected by both rh10-1 and nuc1-1. One specific CpG was unaffected by these mutations. These results imply that the way in which RH10, NUC1 and AS2 are involved in maintaining methylation at five CpGs in exon 6 might be through at least several independent pathways, which might interact with each other. Furthermore, we found that AS2 binds specifically the sequence containing CpGs in exon 1 of ETT/ARF3, and that the binding requires the zinc-finger-like motif in AS2 that is structurally similar to the zinc finger-CxxC domain in vertebrate DNA methyltransferase1. Introduction Organ development in multicellular systems is controlled not only by the activation of a new genetic program but also by the repression of the previously functioning program that is mostly controlled by epigenetic systems. A developmental process of the Arabidopsis leaf having adaxial–abaxial (dorsal–ventral) polarity that is epigenetically regulated by the repressor complex ASYMMETRIC LEAVES1 (AS1)–AS2 (Guo et al. 2008, Yang et al. 2008, Iwasaki et al. 2013, Lodha et al. 2013) is a model to study such positive and negative controlling programs (Machida et al. 2015). AS1 and AS2 proteins are localized to the nucleoplasm, but some speckle-like AS2 bodies are found in regions adjacent to and sometimes inside the nucleolus (Ueno et al. 2007, L. Luo et al. 2012), suggesting a potential role for AS1 and AS2 in gene silencing through nucleolus-associated events. Leaves develop as lateral organs from the peripheral zone of a shoot apical meristem. Initially, a group of cells is patterned along the proximal–distal axis establishing the adaxial–abaxial axis, which is crucial for further leaf development. Subsequent cell proliferation along the medial–lateral axis results in flat and mediolaterally symmetric leaves. The AS1–AS2 complex regulates the proper confinement of the stem cell fate and adaxial development. At an initial step of leaf development, the abaxial identity appears to proceed by the expression of abaxial-determining genes such as ETTIN/AUXIN RESPONSE FACTOR3 (ETT/ARF3), ARF4 (a functionally redundant gene of ARF3) and FILAMENTOUS FLOWER/YABBY (FIL/YAB) in shoot apices (Eshed et al. 1999, Sawa et al. 1999, Eshed et al. 2001), then the adaxial (dorsal) domain is allowed to develop through the repression of these abaxial genes by the AS1–AS2 complex (Iwakawa et al. 2007, Yang et al. 2008, Iwasaki et al. 2013). The complex directly binds the promoter region of the ETT/ARF3 gene and represses its expression. In addition, AS1–AS2 also indirectly represses the expression of ETT/ARF3 and the functionally redundant gene ARF4 through activating the miR390–tasiR-ARF pathway, essential for adaxial development. Our recent study (Iwasaki et al. 2013) showed that all six CpG sites in exon 6 of the ETT/ARF3 coding region are highly methylated and that levels of methylation at three of the CpG sites are decreased in the as2-1 mutant, suggesting that AS2 is involved in maintenance of methylation at these sites. The CpG methylation at all six sites of exon 6 is abolished in the mutant of METHYLTRANSFERASE1 (met1-1) (Iwasaki et al. 2013) responsible for maintenance of CpG methylation (Ronemus et al. 1996), and ETT/ARF3 transcript levels in shoot apices of met1-1 are higher than those of Col-0 (wild type) (Iwasaki et al. 2013). These results suggest an inverse correlation between the levels of CpG methylation in the ETT/ARF3 gene body and levels of ETT/ARF3 transcripts. Many factors (so-called modifiers) are involved in leaf adaxialization in an AS1–AS2-dependent manner. In the as1 or as2 mutation background, various modifier mutations have been identified that markedly enhance the defects of adaxial development to generate abaxialized filamentous leaves with loss of the adaxial domain. Causative mutations occur in genes that are involved in the biogenesis of small RNAs, chromatin modification and cell cycle progression (Machida et al. 2015). Mutations or disruptions of some ribosomal protein genes (Pinon et al. 2008, Yao et al. 2008, Szakonyi et al 2010, Horiguchi et al. 2011, Szakonyi and Byrne 2011) and genes for ribosome biogenesis, such as nucleolar RNA HELICASE10 (RH10) and NUCLEOLIN1 (NUC1), an abundant nucleolar protein involved in various molecular processes in the nucleolus (Matsumura et al. 2016), also act as modifiers of the as1 and as2 phenotypes. These findings suggest that AS1–AS2 and modifier proteins act co-operatively on adaxial development through epigenetic repression of the ETT/ARF3 and ARF4 genes. How the co-operative repression of the ETT/ARF3 transcript levels and the levels of MET1-dependent CpG methylation in exon 6 of ETT/ARF3 are correlated, however, remains to be demonstrated. AS2 encodes a nuclear protein made up of 199 amino acid residues. It has a plant-specific AS2/LOB domain comprising 99 amino acid residues near the N-terminus, which is conserved in all of the other 41 members [designated AS2-LIKE/LOB-DOMAIN proteins (ASL/LBD proteins)] that belong to the AS2/LOB protein family (Iwakawa et al. 2002, Shuai et al. 2002, Matsumura et al. 2009). The AS2/LOB domain (so-called AS2D) of AS2 contains a zinc-finger-like motif including four cysteine residues [ZFL-motif: formerly designated the C-motif or C-block (Iwakawa et al. 2002, Shuai et al. 2002)] at the N-terminus. The ZFL-motif of AS2 is similar, in terms of the cysteine-repeat pattern, to some extent to the zinc-finger motifs of CxxC Finger Protein 1 (CFP1) and DNA methyltransferase1 (Dnmt1), which are members of a family of ZF-CxxC domain-containing proteins in vertebrates (Xu et al. 2011, Clouaire et al. 2012, Long et al. 2013) and have the capacity to bind non-methylated, hemi-methylated and/or methylated CpG (Long et al. 2013, Brown et al. 2017). Intervals between the cysteine residues in the ZFL-motif are perfectly conserved in all 42 members of the AS2/LOB protein family. AS2 domains of AS2 and some ASL/LBDs are capable of binding to a highly conserved specific DNA sequence (Husbands et al. 2007). Although AS2 is involved in maintenance of CpG methylation of the ETT/ARF3 gene body (Iwasaki et al. 2013), it is not known, however, whether the AS2 protein could bind CpG sequences in the ETT/ARF3 gene. In the present study, we found that RH10 and NUC1 are also involved in the maintenance of cytosine methylation at several different specific CpG sites in exon 6 of ETT/ARF3, implying the involvement of multiple pathways mediated by these factors in CpG methylation. Results of analyses of the as2 rh10 and as2 nuc1 double mutants suggest that pathways mediated by AS2 and these nucleolar modifiers interact with each other for maintenance of the CpG methylation status in exon 6. We also report that the AS2 protein binds the short sequence containing the CpG repeat in exon 1 of ETT/ARF3. The ZFL-motif of AS2 is essential for this binding. Since the AS2/LOB protein family seems to be found specifically in plants, the present study might represent the plant-specific CpG methylation system. There might be similar mechanisms, however, for the methylation systems in plants and animals because AS2 can bind to CpG sequences, and is involved in maintaining CpG methylation mediated by MET1 that is a homolog of Dnmt1 of vertebrates. Results Co-operative action of AS2 and NUC1 plays a role in leaf polarity establishment through a function of ETT/ARF3 The as2-1 nuc1-1 mutant efficiently formed abaxialized filamentous leaves, and transcript levels of the genes, ETT/ARF3, ARF4, KANADI1 (KAN1), KAN2, FIL and YAB5, that are involved in leaf abaxialization are increased by 2- to 4-fold in as2-1 nuc1-1 (Matsumura et al. 2016). In the present study, we confirmed the efficient formation of filamentous leaves in the as2-1 nuc1-1 double mutant: all the plants with the double mutation generated filamentous leaves (Fig. 1A, B). We then examined the effect on the leaf phenotype of introducing a loss-of-function mutation (ett-13) into the double mutant. Fig. 1B shows that ett-13 alleviated the abnormal leaf phenotype of as2-1 nuc1-1: eight out of 23 plants with the as2-1 nuc1-1 ett-13 triple mutation (35%) generated filamentous leaves. Filamentous leaves were not observed in any single mutants. These results suggest that a co-operative action of AS2 and NUC1 plays a role in repression of ETT/ARF3 gene expression and the formation of flat symmetric leaves through a function of ETT/ARF3 in the wild-type plant. Fig. 1 View largeDownload slide Suppression of leaf phenotypes of as2-1 nuc1-1 by ett-13. (A) Gross morphologies of the wild type (Col-0) and the indicated mutants. Two and three double and triple mutant plants, respectively, that exhibited representative phenotypes are shown. Plants were grown on soil at 22°C for 21 d. Arrowheads indicate filamentous leaves. Scale bars = 5 mm in Col-0 and single mutants; and 10 and 5 mm in as2-1 nuc1-1 and as2-1 nuc1-1 ett-13, respectively. (B) Quantitative analysis of filamentous leaf formation in Col-0 and the indicated mutants. Plants with the indicated mutations were grown on soil at 22°C for 20 d. The number of plants with more than one filamentous leaf and the total number of plants examined are indicated. Percentages of the plants with filamentous leaves were calculated. Plants were classified into three types (Types I, II and III) depending on the indicated severities of the formation of filamentous leaves. Fig. 1 View largeDownload slide Suppression of leaf phenotypes of as2-1 nuc1-1 by ett-13. (A) Gross morphologies of the wild type (Col-0) and the indicated mutants. Two and three double and triple mutant plants, respectively, that exhibited representative phenotypes are shown. Plants were grown on soil at 22°C for 21 d. Arrowheads indicate filamentous leaves. Scale bars = 5 mm in Col-0 and single mutants; and 10 and 5 mm in as2-1 nuc1-1 and as2-1 nuc1-1 ett-13, respectively. (B) Quantitative analysis of filamentous leaf formation in Col-0 and the indicated mutants. Plants with the indicated mutations were grown on soil at 22°C for 20 d. The number of plants with more than one filamentous leaf and the total number of plants examined are indicated. Percentages of the plants with filamentous leaves were calculated. Plants were classified into three types (Types I, II and III) depending on the indicated severities of the formation of filamentous leaves. AS2, RH10 and NUC1 maintain cytosine methylations at specific sets of CpG sites in exon 6 of ETT/ARF3 The AS1–AS2 complex is involved in the maintenance of DNA methylations in exon 6 of the ETT/ARF3 gene (Iwasaki et al. 2013). CpG sites in only exons 6, 9 and a part of 10 in the ETT/ARF3 locus are methylated in Col-0 (Zhang et al. 2006, Cokus et al. 2008). Since all six CpG sites of exon 6 were highly methylated (>82% at each site) (Fig. 2A, C) and CpG sites of exons 9 and 10 were partially methylated, we intensively examined the effects of as2-1, rh10-1 and nuc1-1 mutations on the level of methylation at each CpG site in exon 6 in the present study. We prepared genomic DNAs from above-ground parts of 14-day-old Col-0 and mutant plants and measured the proportion of the methylated CpG dinucleotide at each site of exon 6 by bisulfite next-generation sequencing (NGS; see the Materials and Methods; Supplementary Table S1). As shown in Fig. 2B, the average levels of CpG methylations in exon 6 in as2-1, rh10-1 and nuc1-1 were decreased to 57.8, 60.5 and 60.1%, respectively. Fig. 2 View largeDownload slide Levels of CpG methylation in exon 6 of ETT/ARF3 were decreased in rh10 and nuc1. (A) Schematic representation of the ETT/ARF3 locus is shown. Gray boxes indicate coding exons. The region outlined by a dashed line (exon 6) was examined for cytosine methylation in genomic DNAs from the above-ground parts of 14-day-old plants by bisulfite NGS, as described in the Materials and Methods. (B) Average levels of methylation in the region of exon 6 of ETT/ARF3 for all six (6a–6f) CpG sites in Col-0, as2-1, rh10-1, as2-1 rh10-1, nuc1-1 and as2-1 nuc1-1. (C) Levels of DNA methylation in Col-0 and the mutants are indicated. Vertical bars indicate the percentage of methylated cytosines in CG, CHG and CHH, whereas levels below 1% are not visible. 6a–6f indicate positions of CpG sequences relative to the first nucleotide of exon 6. (D) Percentages of CpG methylation levels. Severe and moderate reductions are marked by bold (<5%) and thin (5–75%) underlines, respectively. Fig. 2 View largeDownload slide Levels of CpG methylation in exon 6 of ETT/ARF3 were decreased in rh10 and nuc1. (A) Schematic representation of the ETT/ARF3 locus is shown. Gray boxes indicate coding exons. The region outlined by a dashed line (exon 6) was examined for cytosine methylation in genomic DNAs from the above-ground parts of 14-day-old plants by bisulfite NGS, as described in the Materials and Methods. (B) Average levels of methylation in the region of exon 6 of ETT/ARF3 for all six (6a–6f) CpG sites in Col-0, as2-1, rh10-1, as2-1 rh10-1, nuc1-1 and as2-1 nuc1-1. (C) Levels of DNA methylation in Col-0 and the mutants are indicated. Vertical bars indicate the percentage of methylated cytosines in CG, CHG and CHH, whereas levels below 1% are not visible. 6a–6f indicate positions of CpG sequences relative to the first nucleotide of exon 6. (D) Percentages of CpG methylation levels. Severe and moderate reductions are marked by bold (<5%) and thin (5–75%) underlines, respectively. As shown in Fig. 2C, D, some similarities in terms of patterns of the reduction of methylation levels were observed between as2-1 and rh10-1. On the other hand, the pattern of the reduction in the nuc1-1 mutant was different from those of as2-1 and rh10-1. In both as2-1 and rh10-1, methylation levels were reduced at three methylated CpG sites. Methylation levels at two CpG sites (6d and 6f) were decreased in both mutants: those at 6f were greatly decreased in both as2-1 (3.8%) and rh10-1 (1.3%); those at 6d were decreased moderately in both as2-1 (40.3%) and rh10-1 (41.0%). Levels of methylation were decreased moderately at 6c (48.1%) and 6e (53.2%) sites in as2-1 and rh10-1, respectively, showing the mutant-specific reductions in methylation at these two sites. Levels of CpG methylation in nuc1-1 were greatly decreased at 6a (0.7%) and moderately at 6c (15.7%), which was distinct from the reduction pattern of the methylated CpG in rh10-1. The reduction at 6c was also observed in as2-1. Note that the methylation at, at least, one out of six methylated CpG sites (6a or 6f) in exon 6 was almost completely abolished in each of the single mutants. The methylation at 6b was not affected by any mutation tested. We further analyzed levels of methylated CpGs in exon 6 in as2-1 rh10-1 and as2-1 nuc1-1 double mutants (Fig. 2C, D). Methylation levels at four sites (6a, 6c, 6d and 6f) in as2-1 rh10-1 were reduced: those at three sites (6c, 6d and 6f) in as2-1 and those at two sites (6d and 6f) in rh10-1 single mutants were reduced in the double mutant. It is worth noting that the methylation level at 6a was greatly reduced by 26.6% and that at 6d was further reduced (26.4%) in the double mutant. Thus, the number of affected CpG sites at which levels of cytosine methylation were reduced, was increased and levels of the methylation at 6a and 6d were markedly decreased in as2-1 rh10-1 compared with those in each single mutant. In contrast, the reduced methylation at 6e in the rh10-1 single mutant (53.2%) was rescued by as2-1 (86.2%) in the double mutant. Methylation at 6f was consistently abolished in the double mutant. Methylation levels in as2-1 nuc1-1 were reduced at three sites including 6d in addition to sites 6a and 6c that were affected by the nuc1-1 single mutant. In addition, methylation levels at sites 6e and 6f were significantly reduced (80.9% from 92.4% and 78.4% from 88.5%, respectively) in as2-1 nuc1-1 compared with those in Col-0. Thus, the overall number of the affected CpG sites at which methylation levels were reduced was increased in the as2-1 nuc1-1 double mutant compared with each single mutant. In addition, it is worth noting that the severe reduction (3.8%) in methylation at 6f by as2-1 was clearly rescued by nuc1-1 in the as2-1 nuc1-1 double mutant (78.4%). In summary, the present results described above have shown that AS2, RH10 and NUC1 are involved in maintenance of methylation at five out of six CpG sites in exon 6. Each of the pathways mediated by these genes might be differentially involved in maintenance of several specific combinations of CpG sites, but some pathways might overlap at least partially. For example, the pathway for 6f might be controlled by both AS2 and RH10. Since the level of methylation at 6a was severely reduced in the as2-1 rh10-1 double mutant, two independent and parallel pathways that might be mediated by AS2 and RH10, respectively, might be involved in the methylation at this site. The result of as2-1 nuc1-1 suggests that AS2 and NUC1 might interact negatively with each other in the context of the methylation at 6f. These results imply that the CpG methylation in exon 6 might be supported by differential actions of AS2, RH10 and NUC1. The AS2 domain specifically interacts with the short sequences containing three CpG repeats in exon 1 of ETT/ARF3 The ZFL-motif of AS2 is similar, with respect to the pattern of the cysteine repeat, to the zinc-finger (ZF) motif of a ZF-CxxC domain protein family in vertebrates (Xu et al. 2011, Clouaire et al. 2012, Long et al. 2013), which has the capacity to bind non-methylated CpG islands (Long et al. 2013, Brown et al. 2017). We tested the recombinant AS2 proteins synthesized in vitro by the wheat germ system for binding to the ETT/ARF3 coding region through the AlphaScreen (amplified luminescent proximity homogeneous assay screen) system. Model experiment. Since the AS2 domain in AS2 (designated AS2D) has been shown to bind double-stranded DNAs (dsDNAs) containing the sequence 5'-GCGGCG-3' as the core motif (Husbands et al. 2007), we first carried out model experiments using the wild-type AS2D protein, which was tagged with FLAG at the C-terminus (designated AS2D-FLAG) (Fig. 3A), and the biotinylated and non-biotinylated dsDNAs of 50 nucleotides as potential binding targets, shown in Fig. 3B. These DNAs included a polynucleotide containing a single core motif (designated Core); that containing its mutant (designated Mcore); those containing two core motifs in two types of palindromic organizations (designated Palindromes 1 and 2); that containing two core motifs in the direct orientation (designated Core repeat); and that containing two mutated core motifs in the direct orientation (designated Mcore repeat). We also used the mutant AS2D-FLAG protein, in which four conserved cysteine residues in the ZFL-motif were replaced with alanine residues [designated as2(4CA)D-FLAG] (Fig. 3A). We used these proteins and synthetic DNAs to examine whether interactions of these proteins with the synthesized DNAs could be detected by the AlphaScreen system, as described in the Materials and Methods. In our experiments, if AS2D-FLAG binds to the biotinylated target DNA that we synthesized, fluorescent signals from the acceptor beads should be highly amplified. When non-biotinylated DNAs are used, signals might not be amplified. Fig. 3 View largeDownload slide Detection of interactions between the AS2 domain and various arrays of the binding motif by the AlphaScreen system. (A) Motif organization in the AS2 domain (designated AS2D) of AS2 protein. Four cysteine residues in the zinc-finger-like (ZFL) motif were substituted with alanine residues in the as2(4CA) mutant. AS2D in this mutant was designated as2(4CA)D. AS2D and as2(4CA)D were fused to FLAG tags to generate the AS2D-FLAG and the as2(4CA)D-FLAG, respectively. These fusion proteins were synthesized by using the wheat germ protein synthesis system as described in the Materials and Methods. (B) Sequences of synthetic polynucleotides used for the AlphaScreen system. Sequences of only one strand are shown, although dsDNAs were synthesized. The sequences in bold in the polynucleotides represent the core binding motif of the AS2 domain (Husbands et al. 2007) and its mutated motifs. See details in the text. The synthetic polynucleotides shown here were biotinylated at the 5' ends. Non-biotinylated polynucleotides and complementary strands were also synthesized. (C) Interactions between AS2D-FLAG [or as2(4CA)D-FLAG] and biotinylated dsDNAs listed in (B) were tested by the AlphaScreen system (see the Materials and Methods). Emission signals generated by reactions of the indicated combinations of FLAG-tagged proteins and biotinylated DNAs were measured. Signals generated by reactions of the indicated combinations of FLAG-tagged proteins and corresponding non-biotinylated DNAs were also measured. Signals generated by mock reactions containing only the WGE, biotinylated DNAs and non-biotinylated DNAs were measured. Relative signal values were calculated as ratios of signals generated by biotin DNAs to those generated by non-biotin DNAs. Bars represent the mean ± SE of triplicate experiments. Fig. 3 View largeDownload slide Detection of interactions between the AS2 domain and various arrays of the binding motif by the AlphaScreen system. (A) Motif organization in the AS2 domain (designated AS2D) of AS2 protein. Four cysteine residues in the zinc-finger-like (ZFL) motif were substituted with alanine residues in the as2(4CA) mutant. AS2D in this mutant was designated as2(4CA)D. AS2D and as2(4CA)D were fused to FLAG tags to generate the AS2D-FLAG and the as2(4CA)D-FLAG, respectively. These fusion proteins were synthesized by using the wheat germ protein synthesis system as described in the Materials and Methods. (B) Sequences of synthetic polynucleotides used for the AlphaScreen system. Sequences of only one strand are shown, although dsDNAs were synthesized. The sequences in bold in the polynucleotides represent the core binding motif of the AS2 domain (Husbands et al. 2007) and its mutated motifs. See details in the text. The synthetic polynucleotides shown here were biotinylated at the 5' ends. Non-biotinylated polynucleotides and complementary strands were also synthesized. (C) Interactions between AS2D-FLAG [or as2(4CA)D-FLAG] and biotinylated dsDNAs listed in (B) were tested by the AlphaScreen system (see the Materials and Methods). Emission signals generated by reactions of the indicated combinations of FLAG-tagged proteins and biotinylated DNAs were measured. Signals generated by reactions of the indicated combinations of FLAG-tagged proteins and corresponding non-biotinylated DNAs were also measured. Signals generated by mock reactions containing only the WGE, biotinylated DNAs and non-biotinylated DNAs were measured. Relative signal values were calculated as ratios of signals generated by biotin DNAs to those generated by non-biotin DNAs. Bars represent the mean ± SE of triplicate experiments. As shown in Fig. 3C, only when the AS2D protein and the Core repeat were incubated was the high relative signal value (biotin DNA/non-biotin DNA) detected (at a ratio of 3.5). When AS2D and other DNAs including the Mcore repeat were incubated, the detected signals were lower than 1.5. When all combinations of the as2(4CA)D-FLAG protein and these DNAs were incubated, the relative signal values were constantly lower than 1.6. Thus, the direct repeat of the core motif and the wild-type ZFL-motif of the AS2D protein are both necessary for the generation of the detectable signal levels under our assay conditions. These results suggest a molecular interaction between the AS2D protein and the direct repeat of the core DNA. Incubation of the full-length AS2 protein that was fused with FLAG at the C-terminus and any of the DNAs in Fig. 3B did not generate detectable signal levels (our unpublished data). Binding of AS2D to the ETT/ARF3 genomic sequence. For the binding test with the ETT/ARF3 genomic DNA, we searched for DNA sequences that are similar to the core sequence (5'-GCGGCG-3' and 5'-CGCCGC-3') within the entire genomic region at positions –2,987 to +3,413 from the start codon. We found only one perfectly matched sequence, CGCCGC, at position +264 in exon 1 (Fig. 4A, B), and no GCGGCG sequence in this region. We searched for genomic sequences that included five nucleotides in the core motif, and found the CGCCGA sequence in exon 3 at position +1,096 and the GGCCGC sequence in exon 10 at position +2,727. We also searched for genomic sequences that include four nucleotides in the core motif, and found the TCGGCT sequence in exon 6 at position +1,662 and two sequences, TCGGCA and ACGGCT, in exon 10 at positions +2,822 and +3,045, respectively. Finally, we found the CGCCGTsequence, which includes five nucleotides in the core motif, at position –1,444 in the untranslated region. As depicted in (Fig. 4B), we synthesized biotinylated and non-biotinylated DNA fragments of 50 nucleotides including these core-related sequences (designated Noncod_-1444 for DNA in the non-coding region; Ex1_264 for DNA in exon 1; Ex3_1096 for DNA in exon 3; Ex6_1662 for DNA in exon 6; Ex10_2727 for DNA in exon 10; Ex10_2822 for DNA in exon 10; and Ex10_3045 for DNA in exon 10, respectively) (Fig. 4B) and mutant Ex1_264 DNA fragments (designated Ex1_264 m), in which the CGCCGC sequences were replaced with ATAATA sequences. We examined whether AS2D-FLAG and as2(4CA)D-FLAG proteins could bind to these genomic DNA fragments by using the AlphaScreen system. Fig. 4C shows that incubation of AS2D-FLAG and the Ex1_264 DNA generated a high signal level. Incubations of AS2D-FLAG and the Ex1_264 m DNA and other genomic DNAs did not generate significant signal levels (Fig. 4C), Incubation of the mutant protein as2(4CA)D-FLAG and any DNAs, including Ex1_264, did not generate detectable signal levels. We examined whether AS2D-FLAG could physically bind to the biotinylated Ex1_264 DNA by pull-down assay with streptavidin-conjugated donor beads (see the Materials and Methods). The result showed the physical interaction between AS2D-FLAG and the Ex1_264 DNA (Fig. 4D). These results suggest that the AS2 domain binds to the Ex1_264 DNA, which is mediated by the interaction between the ZFL-motif of AS2 and the core motif. The binding of AS2D-FLAG to the Mcore repeat could not be detected with the pull-down assay; this might be due to the weaker binding affinity. We used the WRKY18 transcription factors as a positive control, since its target oligomer had been reported (Lebel et al. 1998, Xu et al. 2006, Kesarwani et al. 2007, Pape et al. 2010). As shown in Fig. 4D in lane 11, physical binding of N-terminal-FLAG-WRKY18 protein to its target oligomer was observed. Note that the Ex1_264 fragment contains the G residue next to the 3' end of the Core sequence to generate CGCCGCG that contains three CpG repeats. Fig. 4 View largeDownload slide A search for AS2D-binding sites in the coding region of the ETT/ARF3 gene. (A) Schematic representation of the exon–intron (box-thin line) organization of ETT/ARF3 and candidate binding sites of AS2D. Numbers below the gene organization correspond to the distance from the start codon of the first nucleotides in the candidate binding sites. (B) Synthesized DNAs containing genomic sequences used for the AlphaScreen system. Numbers in the symbols of DNAs indicate the genomic distance of the first nucleotides in the candidate binding sites from the start codon. Numbers above the DNA sequences correspond to distances of the 5' ends of genomic sequences from the start codon. Sequences that exhibit the highest similarities to the core motif in the coding region are shown in bold. Note that exon 1 contains the sequence of the core motif (designated Ex1_264). We replaced the core motif with the ATAATAT sequence to generate the mutant DNA (designated Ex_264 m). These synthesized oligonucleotides were biotinylated at the 5' ends. Non-biotinylated DNAs and complementary strands were also synthesized. (C) Interactions between AS2D-FLAG [or as2(4CA)D-FLAG] and biotinylated DNAs listed in (B) were tested as described in (C). Relative signal values were calculated as described in (C). Bars represent the mean ± SE of triplicate experiments. (D) Examination for physical interaction of the AS2D-FLAG protein with Ex1_264 DNA. Core repeat DNA, and the WARKY target as a control. Input lanes show each protein solution before pull-down. Streptavidin-conjugated beads were incubated with biotinylated or non-biotinylated DNAs as indicated below the gel photograph, before incubation with AS2D-FLAG proteins. After recovering proteins from streptavidin-conjugated beads, a Western blot with anti-FLAG antibodies was performed to detect the AS2D-FLAG protein. When the Ex1_264 DNA was used, AS2D-FLAG was detected. The combination of the WRKY18 target oligonucleotide and FLAG-WRKY18 protein was also used for the positive control. Fig. 4 View largeDownload slide A search for AS2D-binding sites in the coding region of the ETT/ARF3 gene. (A) Schematic representation of the exon–intron (box-thin line) organization of ETT/ARF3 and candidate binding sites of AS2D. Numbers below the gene organization correspond to the distance from the start codon of the first nucleotides in the candidate binding sites. (B) Synthesized DNAs containing genomic sequences used for the AlphaScreen system. Numbers in the symbols of DNAs indicate the genomic distance of the first nucleotides in the candidate binding sites from the start codon. Numbers above the DNA sequences correspond to distances of the 5' ends of genomic sequences from the start codon. Sequences that exhibit the highest similarities to the core motif in the coding region are shown in bold. Note that exon 1 contains the sequence of the core motif (designated Ex1_264). We replaced the core motif with the ATAATAT sequence to generate the mutant DNA (designated Ex_264 m). These synthesized oligonucleotides were biotinylated at the 5' ends. Non-biotinylated DNAs and complementary strands were also synthesized. (C) Interactions between AS2D-FLAG [or as2(4CA)D-FLAG] and biotinylated DNAs listed in (B) were tested as described in (C). Relative signal values were calculated as described in (C). Bars represent the mean ± SE of triplicate experiments. (D) Examination for physical interaction of the AS2D-FLAG protein with Ex1_264 DNA. Core repeat DNA, and the WARKY target as a control. Input lanes show each protein solution before pull-down. Streptavidin-conjugated beads were incubated with biotinylated or non-biotinylated DNAs as indicated below the gel photograph, before incubation with AS2D-FLAG proteins. After recovering proteins from streptavidin-conjugated beads, a Western blot with anti-FLAG antibodies was performed to detect the AS2D-FLAG protein. When the Ex1_264 DNA was used, AS2D-FLAG was detected. The combination of the WRKY18 target oligonucleotide and FLAG-WRKY18 protein was also used for the positive control. The AlphaScreen system that we adopted in the present binding test may have limitations for assaying the interaction between DNA and a protein. For example, the positions of target sequences in DNA fragments should be critical in general: the sequences to be examined should be placed between the middle and the 3' end of DNA fragments, which allows detection of sufficient signals generated by the interaction of the donor beads with the acceptor beads. Therefore, we cannot exclude the possibility that other regions in the 50 nucleotide sequences we used here could have binding affinity for the AS2 domain. Discussion Multiple pathways mediated by AS2, RH10 and NUC1, including positive and negative interactions, are implicated in the maintenance of methylation Genetic and expression analyses carried out with double and triple mutants including as2-1, rh10-1, nuc1-1, rid2-1 and ett-13 imply that certain molecular interactions between AS2 and the nucleolar proteins RH10, NUC1 and RID2 might function in the partitioning of leaf adaxial–abaxial domains and repression of expression of the abaxial-determining gene ETT/ARF3 (Iwasaki et al. 2013, Matsumura et al. 2016; Fig. 1). In the present study, we examined roles of AS2 and RH10 and NUC1 in maintenance of six highly methylated CpG dinucleotides in exon 6 with the165 nucleotide sequence of the ETT/ARF3 coding region. The present results have shown that AS2, RH10 and NUC1 are positively involved in the maintenance of methylations in five out of six methylated CpG sites. Each of these proteins seems to be differentially involved in the maintenance of several specific combinations of CpG sites. These results imply that the molecular events of CpG methylation in exon 6 are achieved by differential actions of AS2, RH10 and NUC1. Since genomic DNAs used in the present analysis were prepared from whole aerial parts of plants 14 d after sowing, the genomic DNAs should be derived from developmentally distinct cells, implying that the present data of methylation analyses with as2-1, rh10-1 and nuc1-1 mutants might reflect average values obtained from DNAs of mixed cells with different developmental states. The results of the analyses with as2-1 rh10-1 and as2-1 nuc1-1 double mutants, however, suggest strong genetic interactions between these mutations. For example, a typical result to support this suggestion should be the observation that the level of methylation at the CpG site at 6a in exon 6 was greatly reduced (26.6%) in as2-1 rh10-1 as compared with methylation levels in single mutants (79.1% in as2-1; 90.1% in rh10-1); a marked decrease in methylation at 6f in as2-1 (3.8%) was rescued by nuc1-1 (83.2%) in the as2-1 nuc1-1 double mutant (78.4%). We have also reported that these double mutations of as2-1 rh10-1 and as2-1 nuc1-1 synergistically enhance the level of ETT/ARF3 transcript, which causes much more severe leaf phenotypes than those of single mutants, suggesting genetic interactions among these factors (Matsumura et al. 2016). Therefore, it might be difficult to explain the present results by hypothesizing heterogeneity of genomic DNAs prepared from developmentally distinct cells in these mutants, although the cells from which DNAs were prepared should be heterogeneous, as described above. It would be intriguing to elucidate the mechanism that operates in differential methylation at CpG sites in exon 6 by actions of AS2 and these nucleolar proteins. In both as2-1 and rh10-1 single mutants, methylation levels at two CpG sites (6d and 6f in Fig. 2C, D) were decreased. These results suggested that AS2 and RH10 might share, at least in part, a common pathway to maintain the methylations at these specific CpG sites. Since the methylation level at site 6c was also decreased in both as2-1 and nuc1-1, AS2 and NUC1 might also share a common pathway to maintain the CpG methylation at 6c. The methylation level at 6c was not decreased in rh10-1. These observations predict that there might be at least two pathways which are independently involved in the maintenance of methylations at 6d–6f (the 6d–6f pathway involving AS2 and RH10) and in that of methylation at 6c (the 6c pathway involving AS2 and NUC1). Since the strong reduction in methylation at 6a was observed in the as2-1 rh10-1 double mutant, the methylation at 6a might be controlled by two independent and parallel pathways mediated by AS2 and RH10, respectively, suggesting an interaction of the pathway mediated by these genes. In addition, the methylation at 6a might also be controlled strongly by NUC1 alone. Interestingly, the methylation at 6f is sufficiently controlled by AS2, and this process should be affected by NUC1 because the abolishment of methylation by as2-1 was almost completely rescued by the nuc1-1 mutation. This suggests the involvement of a tight negative interaction between AS2 and NUC1, whereby AS2 might repress NUC1 from upstream, which might in turn repress a positive reaction for the methylation at a position further downstream. The methylation at 6b was independent from these factors. AS2 belongs to the AS2/LOB protein family that is plant specific and includes at least 42 members. The ZFL-motif is highly conserved in all of the members. It would also be interesting to investigate roles of other members in the context of gene body DNA methylation and the nature of DNA binding. Correlation between the ETT/ARF3 gene body methylation and its expression We have previously reported that the transcript levels of ETT/ARF3 increase significantly in the background of met1, in which methylation at all six sites is also abolished, which implies the involvement of MET1 in cytosine methylation at all sites and an inverse correlation between expression and methylation (Iwasaki et al. 2013). In addition, the reduced levels of methylation at two CpG sites (6a, and 6d) were further decreased in the as2-1 rh10-1 double mutant, and the number of CpG sites at which methylation levels were significantly decreased (6a, 6c, 6d and 6f) was increased (Fig. 2). These results also suggest parallel relationships among the reduction in the CpG methylation level, the increase in ETT/ARF3 transcripts and the enhancement of leaf abnormalities (Matsumura et al. 2016). These results are also in line with the hypothesis that the methylation at CpG sites in exon 6 might contribute at least in part to ETT/ARF3 repression. Although effects by as2-1 nuc1 double mutation on the overall methylation levels were not obvious, NUC1 participates in the methylations at two CpG sites (6a and 6c in Fig. 2C, D). It has been reported that NUC1 is also required for DNA methylation in the regulatory region of rRNA transcription and is involved in the silencing of rDNAs to be inactive in Arabidopsis (Pontvianne et al. 2010). Recently, it has been reported that a significant proportion of expressed genes are negatively controlled by gene body DNA methylation at CpG dinucleotides through functions of histone H3.3 (Wollmann et al. 2017). Regulatory mechanisms for more individual genes that are developmentally and environmentally controlled should be elucidated in the context of gene body methylation. In our laboratory, analyses of ETT/ARF3 variants with synonymous mutations at methylated CpG sites in exon 6 are being carried out, which will provide useful information on clear relationships between these molecular and developmental events. Possible mechanisms of ETT/ARF3 DNA methylation mediated by AS2, RH10 and NUC1 The AS2, RH10 and NUC1 proteins do not have DNA methyltransferase activity, although they are positively involved in maintenance of cytosine methylation at six CpG dinucleotides in exon 6 of ETT/ARF3. How, then, can these factors be involved in the maintenance of DNA methylation in exon 6? As described in the previous section, MET1 could be responsible for the methylation in exon 6. MET1 is an Arabidopsis homolog of vertebrate Dnmt1, which has activity to methylate hemi-methylated CpG, converting it to methylated CpG during DNA replication, and is part of a protein complex that is involved in the maintenance of DNA methylation (Long et al. 2013, Nishiyama et al. 2013, Song et al. 2015, Zhang et al. 2015, Du et al. 2016, Ferry et al. 2017). MET1 has sequences homologous to those of methyltransferase (Finnegan and Dennis 1993) and two other domains present in Dnmt1 (Ryazanova et al. 2012). Nevertheless, MET1 has no sequence that could encode the domain corresponding to the ZF-CxxC of Dnmt1, which directly binds to unmethylated CpG dinucleotides (Song et al. 2011, Long et al. 2013). If a protein complex including AS2 and MET1 would be hypothesized, AS2 might provide the ZFL-motif, which is similar to the ZF-CxxC of Dnmt1, to the MET1-containing protein complex. It has been reported that the ZF-CxxC domain of Dnmt1 and sequences adjacent to the ZF-CxxC play roles in inhibition of the methyltransferase activity of Dnmt1 (Song et al. 2011, Long et al. 2013, Zhang et al. 2015), preventing de novo methylation of unmethylated CpG dinucleotides (Song et al. 2011). Since AS2 binds to the CpG sequences in exon 1 of ETT/ARF3, which are not methylated (Zhang et al. 2006, Cokus et al. 2008), AS2 might play a similar inhibitory role in CpG methylation in exon 1. In the present study, our results show, however, that AS2 acts as the positive factor for maintaining CpG methylation in exon 6. As described below, VIM proteins, which have the ability to bind to hemi-methylated CpG, might be candidate factors that are involved in CpG methylation in exon 6 (Yao et al. 2012, Kim et al. 2014). In Arabidopsis, it is worth noting that AS2, AS1, NUC1, RH10 and MET1 are localized in the nucleolus and/or its peripheries, and physical and functional interactions between these proteins have been reported (Kojima et al. 2007, Petricka and Nelson 2007, Pontvianne et al. 2007, Matsumura et al. 2016). The regulatory regions of inactive rDNAs are highly methylated by MET1 and organized in heterochromatic states, and MET1 requires NUC1 and nucleolar histone deacetylase HDA6 for the methylation (Pontvianne et al. 2010, Pontvianne et al. 2013). MET1 directly interacts with HDA6 (To et al. 2011, Liu et al. 2012), which is also associated with AS1 and AS2 (M. Luo et al. 2012). AS2 forms the nucleolus-associated speckles designated AS2 bodies, which include AS1 protein, and are involved in proper leaf development (Ueno et al. 2007, L. Luo et al. 2012). These physical and functional interactions might provide a spatial basis for the methylation in exon 6 by MET1. In contrast, the maintenance of cytosine methylation might also be achieved by the inhibition of a de-methylation process for methylated CpG. If AS2 has a capacity to bind methylated CpG sites, they might be protected from this de-methylation. This possibility could be tested using methylated CpG dinucleotides as substrates. Is the AS2 binding to exon 1 related to CpG methylation in exon 6? Although Dnmt1 of vertebrates binds hemi-methylated CpG dinucleotides, it also binds to unmethylated CpG dinucleotides (Gruenbaum et al. 1982, Yoder et al. 1997, Song et al. 2011, Schrader et al. 2015, Zhang et al. 2015). Therefore, we examined whether AS2, a positive regulator for maintaining DNA methylation, could interact with unmethylated CpG dinucleotides in vitro. Although the binding of AS1, an AS2-interacting molecule, and that of AS2 to exon 1 of ETT/ARF3 have been reported (Iwasaki et al. 2013, O’Malley et al. 2016), interacting sites between exon 1 and AS2 protein have yet to be reported. The present results have shown that AS2 strongly binds in vitro the short sequence containing a three-CpG repeat in exon 1 of ETT/ARF3, which requires the ZFL-motif of AS2. This observation in the present study is consistent with the finding that the ZF-CxxC domains of Cfp1, Dnmt1 and Kdm bind to the CpG dinucleotide (Long et al. 2013). It could also be hypothesized that AS2, which binds to exon 1 in the ETT/ARF3 locus, might play an inhibitory role in CpG methylation in exon 1: AS2 binding to these CpG dinucleotides might prevent MET1 from acting on the potential substrates in exon 1 and/or nearby exons. If AS2 forms a complex with MET1, AS2 might recruit MET1 to exon 1 and, subsequently, MET1 in the complex could be transferred by an unknown mechanism to exon 6 that might contain hemi-methylated CpG dinucleotides created after DNA replication of the ETT/ARF3 locus. This hypothesis might explain AS2-dependent methylation of CpG sites (6c, 6d and 6f) in exon 6, because other sites are methylated in the as2-1 mutant (Fig. 2). As regards methyltransferase activity of Dnmt1, it requires binding to the SRA (SET and RING finger-associated) domain of Uhrf1, which has the ability to bind to hemi-methylated CpG dinucleotides (Song et al. 2011, Nishiyama et al. 2013, Berkyurek et al. 2014). In order to understand a role for AS2 in the maintenance of methylation in exon 6, it should be important to examine how efficiently AS2 could bind to hemi-methylated CpG dinucleotides in addition to unmethylated CpG dinucleotides. In addition, it should be informative that factors, which might interact with AS2 protein, are isolated and characterized at the molecular level. The VIM1 protein, which is an Arabidopsis homolog of Uhrf1 and includes the SRA domain, binds to hemi-methylated CpG in a MET1-dependent manner and is also a prerequisite for maintenance methylation mediated by MET1 (Yao et al. 2012, Kim et al. 2014). The molecular mechanism whereby these proteins interact with each other, however, is unknown in plants. Future challenges will include investigations of molecular relationships among AS2, MET1, VIMs and replication factors. Alternatively, the AS2 binding we observed might not be related to CpG methylation; rather, AS2 might be involved in the direct repression of ETT/ARF3 expression through its binding to the CpG repeat in exon 1 at transcriptional and/or translational steps. This possibility could be tested by using the wheat germ in vitro translation system. Materials and Methods Plant materials and growth condition Arabidopsis thaliana ecotype Col-0 (CS1092) and the mutant as2-1 (CS3117) were obtained from the Arabidopsis Biological Resource Center (ABRC). We outcrossed as2-1 with Col-0 three times and used the progeny for our experiments (Kojima et al. 2011). Details of ett-13 and arf4-1 (Pekker et al. 2005), rh10-1, as2-1 rh10-1 (Matsumura et al. 2016), nuc1-1 (Kojima et al. 2007, Durut et al. 2014) and as2-1 nuc1-1 (Matsumura et al. 2016) were described previously. rh10-1 and nuc1-1 were on the Col-0 background. For phenotypic analyses, seeds were sown on soil. After 2 d at 4°C in darkness, plants were transferred to a regimen of white light at 50 μmol m–2 s–1 for 16 h and darkness for 8 h daily at 22°C, as described previously (Semiarti et al. 2001). Ages of plants are given in terms of number of days after sowing. Genomic DNA extraction DNA was extracted from about 100 mg of the whole aerial part of 14-day-old plant seedlings. The plant samples were frozen in liquid nitrogen and then crushed into powder in a mortar. Total DNA was isolated with a DNeasy Plant Mini Kit (QIAGEN) according to the manufacturer’s instructions. Bisulfite treatment and sample preparation for sequencing Approximately 300 ng of genomic DNA was used for bisulfite conversion with the EZ DNA Methylation-Gold kit (Zymo Research). Immediately after the conversion, we amplified the fragments of interest by using the Epitaq bisulfite kit (TAKARA BIO INC.). The PCRs were carried out with different sets of primers for each DNA sample. Primer sets are listed in Supplementary Table S2. A sequence of four different nucleotides, hereinafter referred to as the ‘index,’ was added to the 5' end of the primers in order to identify the original DNA sample of the sequences during sequencing (Supplementary Table S2). We targeted ETT exon 6, ETT exon 9, ETT exon 10 and the AP1 promoter region. The primers were designed specifically to target the coding strand (Iwasaki et al. 2013; Supplementary Table S2). The conversion of each DNA sample was tested for completeness by using the APETALA1 promoter, a region previously shown to be non-methylated (APETALA1, AT1G69120). A solution containing approximately the same amount of each fragment from each DNA sample was prepared and purified with a WIZARD SV Gel and a PCR Clean-up System kit (Promega) to eliminate small DNA fragments, according to the manufacturer’s instructions. The purified solution was then adjusted to a concentration of 200 ng µl–1 of DNA and a volume of 20 µl. NGS The KAPA HyperPlus Library Preparation Kit (Kapa Biosystems) was used to prepare a sequencing library. A 500 ng aliquot of DNA was subjected to the End Repair and A-Tailing reaction, according to the Kapa HyperPlus Library Preparation Kit specification. The DNA ligase mix including annealed adaptor was added to the A-tailed library. The library products were purified with AMPure XP beads (La Jolla Institute Next Generation Sequencing Facility) to remove adaptor dimers. After four-cycle PCR amplification, the library products were purified with AMPure XP beads. Libraries were sequenced on an Illumina MiSeq system with a MiSeq Reagent Kit v3 (Illumina Inc.) generating 2 × 300 bp paired-end sequences. Data processing We used Fast Length Adjustment of Short reads (FLASH) proposed by Magoč and Salzberg (2011) to combine paired reads. FLASH has two important parameters: m, the minimum overlap length; and M, the maximum overlap length. In this study, parameter m is set to 70 and M is set to 240, according to the expected sizes of PCR products. A total of 19,481,435 sequenced read pairs were first combined by FLASH, and then 16,024,086 combined reads were constructed. The combined reads were filtered on the basis of the expected sizes for each PCR product. Furthermore, we determined the direction of each filtered sequence on the basis of pairs of PCR primers and of the classified sequences on the basis of each pair of the 4 bp indexes that we designed. A total of 7,942,143 sequences remained following these processes. We selected 7,001,947 sequences with a sequence error rate <1%. All programs, except for FLASH, were written in R (www.r-project.org). We also used R libraries, sangerseqR (Hill et al. 2014), CrispRVariants (Lindsay et al. 2016), ShortRead (Morgan et al. 2009), Biostrings and seqinr (Charif and Lobry 2007). Among the 7,001,947 selected sequences, 3,880,195 sequences corresponding to ETT/ARF3 exon 6, 9, 10 and the AP1 promoter in Col-0, as2-1, rh10-1, nuc1-1, as2-1 rh10-1, as2-1 nuc1-1 plants were used for determination of methylation. In vitro protein–dsDNA interaction assay The amplified luminescence proximity homogeneous assay was performed using an AlphaScreen® FLAG® (M2) Detection Kit provided by Perkin Elmer (Eglen et al. 2008, Hornung et al. 2009, Tokizawa et al. 2015) to show the interaction of the AS2/LOB domain of AtAS2 with dsDNAs designed from the literature and the ETT/ARF3 gene locus. The C-terminal FLAG (DYKDDDDK)-tagged proteins were expressed in wheat germ extract (WGE) from in vitro transcribed mRNA obtained from PCR-generated cDNA (Nomoto and Tada 2018). The peptide corresponds to amino acid residues 1–119 of AS2 protein fused with FLAG and hereinafter referred to as AS2D-FLAG. We also expressed a mutated protein in which the cysteine residues C10, C13, C20 and C24 were turned into alanine residues, referred to as as2(4CA)-FLAG. The protein quality (i.e. efficient synthesis with the expected molecular mass) was confirmed by Western blotting analysis with an anti-FLAG antibody. The proteins were diluted 2.8-fold before being used in the assay at a 5.6 dilution in a final volume of 25 µl. Biotinylated 50-mer cis-elements and the non-biotinylated complementary fragments were obtained from Eurofins (Supplementary Table S2). A 20 µl aliquot of 50 µM biotinylated oligonucleotide (sense strand) or non-biotinylated sense strand and the same volume of 50 µM antisense oligonucleotide were mixed and incubated at 60°C for 20 min, followed by an overnight incubation at room temperature to obtain the biotinylated double-stranded nucleotides. The FLAG-tagged proteins and dsDNA were mixed with 10× control buffer provided in the kit, 0.1% (v/v) Tween-20 (Sigma-Aldrich) and 10% (w/v) bovine albumin serum (BSA) in MilliQ-water and then incubated for 1 h at room temperature. Acceptor beads coated with anti-FLAG antibody provided in the kit were then added to the reaction mix and incubated for 1 h at room temperature. Under subdued laboratory lighting (dark room), the streptavidin donor beads were added to the reaction mix and incubated for 1 h in the dark at room temperature. The final volume of 25 µl comprised 1× phosphate-buffered saline (PBS), 0.005% Proclin-300, 0.01% (v/v) Tween-20, 0.1% (w/v) BSA, 5.8-fold diluted proteins, 50 nM dsDNA, 500 ng of acceptor beads and 500 ng of donor beads. The AlphaScreen signals (chemiluminescence between the donor and the acceptor beads conjugated by the binding of labeled AS2/LOB domain and dsDNA oligos) were determined with the Spark 10 M plate reader (TECAN). The AlphaScreen signals for the control (non-biotinylated) dsDNA oligos in the labeling step were used to estimate the background luminescence. WGE without expressed proteins was used to estimate the luminescence caused by endogenous wheat germ protein interactions with the assay. Relative AlphaScreen signals were defined as the ratio of luminescence of the biotinylated dsDNA oligos to the background. In vitro pull-down analysis Invitrogen Dynabeads® M-280 are streptavidin-coated magnetic beads that can be used to trap biotinylated molecules for multiple purposes (Kobayashi et al. 2009, Kalb et al. 2015, So et al. 2016). We used them to capture biotinylated DNA and checked whether AS2D-FLAG could bind to this DNA. We used non-biotinylated beads to make sure that the DNA does not bind directly to the beads in detectable amounts. A 25 µl slurry of Dynabeads® M-280 Streptavidin (Invitrogen) and 0.6 µl of 25 µM biotinylated cis-element were suspended in 200 µl of Buffer A [1× PBS and 0.01% Tween-20 (w/v) (Sigma-Aldrich)], and rotated at room temperature for 1 h. The beads were isolated from the supernatant on a magnetic bench and washed three times with 1 ml of Buffer A, and then further rotated in 200 µl of Buffer B [1× PBS, 0.01% (w/v) Tween-20 and 0.1% (w/v) BSA] at room temperature for 1 h to saturate the beads and prevent non-specific binding. A solution comprising 775 µl of Buffer B and 5 µl of a FLAG-tagged protein expressed in WGE (2.8-fold dilution) is then added to the beads and the mixture is incubated under rotation at 4°C for 2 h. The beads were then isolated from the supernatant on a magnetic bench, washed three times with 1 ml of Buffer A and incubated in 30 µl of sample buffer [75 mM Tris–HCl, pH 6.8, 15% (w/v); glycerol, 0.03% (w/v); bromophenol blue, 3% (w/v); SDS; and 200 mM dithiothreitol] at 70°C for 20 min. Samples were then subjected to SDS–PAGE together with the input, the FLAG-tagged proteins in Buffer B before the addition to beads. The immunoblotting for FLAG-tagged proteins was performed using an anti-FLAG antibody. Supplementary Data Supplementary data are available at PCP online. Funding This work was supported by Chubu University [a Grant-in Aid D, 2016–2017 (Nos. 28IM03D and 29IM05D) to S.V-P]; Japan Society for the Promotion of Science (JSPS) KAKENHI [grant Nos. JP15K07116, JP26291056, JP16K14574 and JP13J10800]; the Ministry of Education, Culture, Sports, Science and Technology (MEXT) KAKENHI [grant Nos. JP19060015, JP16H01246, JP17H05659, JP15H05956 and JP15H01223]; and the Research Foundation for the Electrotechnology of Chubu. Acknowledgments The authors are grateful to Ms. Yamakawa and Mr. Harayama for their helpful technical support. Disclosures The authors have no conflicts of interest to declare. References Berkyurek A.C. , Suetake I. , Arita K. , Takeshita K. , Nakagawa A. , Shirakawa M. , et al. ( 2014 ) The DNA methyltransferase Dnmt1 directly interacts with the SET and RING finger-associated (SRA) domain of the multifunctional protein Uhrf1 to facilitate accession of the catalytic center to hemi-methylated DNA . J. Biol. Chem. 289 : 379 – 386 . Brown D.A. , Di Cerbo V. , Feldmann A. , Ahn J. , Ito S. , Blackledge N.P. , et al. ( 2017 ) The SET1 complex selects actively transcribed target genes via multivalent interaction with CpG island chromatin . Cell Rep . 20 : 2313 – 2327 . Charif D. , Lobry J.R. ( 2007 ) SeqinR 1.0-2: a contributed package to the R project for statistical computing devoted to biological sequences retrieval and analysis. In Structural Approaches to Sequence Evolution: Molecules, Networks, Populations . Edited by Bastolla U. , Porto M. , Roman H.E. , Vendruscolo M. pp. 207 – 232 . Springer Verlag , New York . Clouaire T. , Webb S. , Skene P. , Illingworth R. , Kerr A. , Andrews R. , et al. ( 2012 ) Cfp1 integrates both CpG content and gene activity for accurate H3K4me3 deposition in embryonic stem cells . Genes Dev . 26 : 1714 – 1728 . Cokus S.J. , Feng S. , Zhang X. , Chen Z. , Merriman B. , Haudenschild C.D. , et al. ( 2008 ) Shotgun bisulphite sequencing of the Arabidopsis genome reveals DNA methylation patterning . Nature 452 : 215 – 219 . Du Q. , Wang Z. , Schramm V.L. ( 2016 ) Human DNMT1 transition state structure . Proc. Natl. Acad. Sci. USA 113 : 2916 – 2921 . Durut N. , Abou-Ellail M. , Pontvianne F. , Das S. , Kojima H. , Ukai S. , et al. ( 2014 ) A duplicated NUCLEOLIN gene with antagonistic activity is required for chromatin organization of silent 45S rDNA in Arabidopsis . Plant Cell 26 : 1330 – 1344 . Eglen R.M. , Reisine T. , Roby P. , Rouleau N. , Illy C. , Bossé R. , et al. ( 2008 ) The use of AlphaScreen technology in HTS: current status . Curr. Chem. Genomics 1 : 2 – 10 . Eshed Y. , Baum S.F. , Bowman J.L. ( 1999 ) Distinct mechanisms promote polarity establishment in carpels of Arabidopsis . Cell 99 : 199 – 209 . Eshed Y. , Baum S.F. , Perea J.V. , Bowman J.L. ( 2001 ) Establishment of polarity in lateral organs of plants . Curr. Biol. 11 : 1251 – 1260 . Ferry L. , Fournier A. , Tsusaka T. , Adelmant G. , Shimazu T. , Matano S. , et al. ( 2017 ) Methylation of DNA ligase 1 by G9a/GLP recruits UHRF1 to replicating DNA and regulates DNA methylation . Mol. Cell. 67 : 550 – 565 . Finnegan E.J. , Dennis E.S. ( 1993 ) Isolation and identification by sequence homology of a putative cytosine methyltransferase from Arabidopsis thaliana . Nucleic Acids Res. 21 : 2383 – 2388 . Gruenbaum Y. , Cedar H. , Razin A. ( 1982 ) Substrate and sequence specificity of a eukaryotic DNA methylase . Nature 295 : 620 – 622 . Guo M. , Thomas J. , Collins G. , Timmermans M.C. ( 2008 ) Direct repression of KNOX loci by the ASYMMETRIC LEAVES1 complex of Arabidopsis . Plant Cell 20 : 48 – 58 . Hill J.T. , Demarest B.L. , Bisgrove B.W. , Su Y.C. , Smith M. , Yost H.J. ( 2014 ) Poly peak parser: method and software for identification of unknown indels using Sanger sequencing of polymerase chain reaction products . Dev. Dyn. 243 : 1632 – 1636 . Horiguchi G. , Mollá-Morales A. , Pérez-Pérez J.M. , Kojima K. , Robles P. , Ponce M.R. , et al. ( 2011 ) Differential contributions of ribosomal protein genes to Arabidopsis thaliana leaf development . Plant J . 65 : 724 – 736 . Hornung V. , Ablasser A. , Charrel-Dennis M. , Bauernfeind F. , Horvath G. , Caffrey D.R. , et al. ( 2009 ) AIM2 recognizes cytosolic dsDNA and forms a caspase-1-activating inflammasome with ASC . Nature 458 : 514 – 518 . Husbands A. , Bell E.M. , Shuai B. , Smith H.M. , Springer P.S. ( 2007 ) LATERAL ORGAN BOUNDARIES defines a new family of DNA-binding transcription factors and can interact with specific bHLH proteins . Nucleic Acids Res . 35 : 6663 – 6671 . Iwakawa H. , Iwasaki M. , Kojima S. , Ueno Y. , Soma T. , Tanaka H. , et al. ( 2007 ) Expression of the ASYMMETRIC LEAVES2 gene in the adaxial domain of Arabidopsis leaves represses cell proliferation in this domain and is critical for the development of properly expanded leaves . Plant J . 51 : 173 – 184 . Iwakawa H. , Ueno Y. , Semiarti E. , Onouchi H. , Kojima S. , Tsukaya H. , et al. ( 2002 ) The ASYMMETRIC LEAVES2 gene of Arabidopsis thaliana, required for formation of a symmetric flat leaf lamina, encodes a member of a novel family of proteins characterized by cysteine repeats and a leucine zipper . Plant Cell Physiol . 43 : 467 – 478 . Iwasaki M. , Takahashi H. , Iwakawa H. , Nakagawa A. , Ishikawa T. , Tanaka H. , et al. ( 2013 ) Dual regulation of ETTIN (ARF3) gene expression by AS1–AS2, which maintains the DNA methylation level, is involved in stabilization of leaf adaxial–abaxial partitioning in Arabidopsis . Development 140 : 1958 – 1969 . Kalb S.R. , Krilich J.C. , Dykes J.K. , Lúquez C. , Maslanka S.E. , Barr J.R. ( 2015 ) Detection of botulinum toxins A, B, E, and F in foods by endopep-MS . J. Agric. Food Chem. 63 : 1133 – 1141 . Kesarwani M. , Yoo J. , Dong X. ( 2007 ) Genetic interactions of TGA transcription factors in the regulation of pathogenesis-related genes and disease resistance in Arabidopsis . Plant Physiol . 144 : 336 – 346 . Kim J. , Kim J.H. , Richards E.J. , Chung K.M. , Woo H.R. ( 2014 ) Arabidopsis VIM proteins regulate epigenetic silencing by modulating DNA methylation and histone modification in cooperation with MET1 . Mol. Plant 7 : 1470 – 1485 . Kobayashi M. , Aida M. , Nagaoka H. , Begum N.A. , Kitawaki Y. , Nakata M. , et al. ( 2009 ) AID-induced decrease in topoisomerase 1 induces DNA structural alteration and DNA cleavage for class switch recombination . Proc. Natl. Acad. Sci. USA 106 : 22375 – 22380 . Kojima S. , Iwasaki M. , Takahashi H. , Imai T. , Matsumura Y. , Fleury D. , et al. ( 2011 ) ASYMMETRIC LEAVES2 and Elongator, a histone acetyltransferase complex, mediate the establishment of polarity in leaves of Arabidopsis thaliana . Plant Cell Physiol . 52 : 1259 – 1273 . Kojima H. , Suzuki T. , Kato T. , Enomoto K.-I. , Sato S. , Kato T. , et al. ( 2007 ) Sugar-inducible expression of the nucleolin-1 gene of Arabidopsis thaliana and its role in ribosome synthesis, growth and development . Plant J . 49 : 1053 – 1063 . Lebel E. , Heifetz P. , Thorne L. , Uknes S. , Ryals J. , Ward E. ( 1998 ) Functional analysis of regulatory sequences controlling PR-1 gene expression in Arabidopsis . Plant J. 16 : 223 – 233 . Lindsay H. , Burger A. , Biyong B. , Felker A. , Hess C. , Zaugg J. , et al. ( 2016 ) CrispRVariants charts the mutation spectrum of genome engineering experiments . Nat. Biotechnol. 34 : 701 – 702 . Liu X. , Yu C.W. , Duan J. , Luo M. , Wang K. , Tian G. , et al. ( 2012 ) HDA6 directly interacts with DNA methyltransferase MET1 and maintains transposable element silencing in Arabidopsis . Plant Physiol . 158 : 119 – 129 . Lodha M. , Marco C.F. , Timmermans M.C. ( 2013 ) The ASYMMETRIC LEAVES complex maintains repression of KNOX homeobox genes via direct recruitment of Polycomb-repressive complex2 . Genes Dev . 27 : 596 – 601 . Long H.K. , Blackledge N.P. , Klose R.J. ( 2013 ) ZF-CxxC domain-containing proteins, CpG islands and the chromatin connection . Biochem. Soc. Trans . 41 : 727 – 740 . Luo L. , Ando S. , Sasabe M. , Machida C. , Kurihara D. , Higashiyama T. , et al. ( 2012 ) Arabidopsis ASYMMETRIC LEAVES2 protein required for leaf morphogenesis consistently forms speckles during mitosis of tobacco BY-2 cells via signals in its specific sequence . J. Plant Res. 125 : 661 – 668 . Luo M. , Yu C.W. , Chen F.F. , Zhao L. , Tian G. , Liu X. , et al. ( 2012 ) Histone deacetylase HDA6 is functionally associated with AS1 in repression of KNOX genes in Arabidopsis . PLoS Genet. 8 : e1003114 . Machida C. , Nakagawa A. , Kojima S. , Takahashi H. , Machida Y. ( 2015 ) The complex of ASYMMETRIC LEAVES (AS) proteins plays a central role in antagonistic interactions of genes for leaf polarity specification in Arabidopsis . Wiley Interdiscip. Rev. Dev. Biol. 4 : 655 – 671 . Magoč T. , Salzberg S.L. ( 2011 ) FLASH: fast length adjustment of short reads to improve genome assemblies . Bioinformatics 27 : 2957 – 2963 . Matsumura Y. , Iwakawa H. , Machida Y. , Machida C. ( 2009 ) Characterization of genes in the ASYMMETRIC LEAVES2/LATERAL ORGAN BOUNDARIES (AS2/LOB) family in Arabidopsis thaliana, and functional and molecular comparisons between AS2 and other family members . Plant J . 58 : 525 – 537 . Matsumura Y. , Ohbayashi I. , Takahashi H. , Kojima S. , Ishibashi N. , Keta S. , et al. ( 2016 ) A genetic link between epigenetic repressor AS1–AS2 and a putative small subunit processome in leaf polarity establishment of Arabidopsis . Biol. Open 5 : 942 – 954 . Morgan M. , Anders S. , Lawrence M. , Aboyoun P. , Pagès H. , Gentleman R. ( 2009 ) ShortRead: a bioconductor package for input, quality assessment and exploration of high-throughput sequence data . Bioinformatics 25 : 2607 – 2608 . Nishiyama A. , Yamaguchi L. , Sharif J. , Johmura Y. , Kawamura T. , Nakanishi K. , et al. ( 2013 ) Uhrf1-dependent H3K23 ubiquitylation couples maintenance DNA methylation and replication . Nature 502 : 249 – 253 . Nomoto M. , Tada Y. ( 2018 ) Cloning-free template DNA preparation for cell-free protein synthesis via two-step PCR using versatile primer designs with short 3'-UTR . Genes Cells 3 : 46 – 53 . O’Malley R.C. , Huang S-sC. , Song L. , Lewsey M.G. , Bartlett A. , Nery J.R. , et al. ( 2016 ) Cistrome and epicistrome features shape the regulatory DNA landscape . Cell 165 : 1280 – 1292 . Pape S. , Thurow C. , Gatz C. ( 2010 ) The Arabidopsis PR-1 promoter contains multiple integration sites for the coactivator NPR1 and the repressor SNI1 . Plant Physiol . 154 : 1805 – 1818 . Pekker I. , Alvarez J.P. , Eshed Y. ( 2005 ) Auxin response factors mediate Arabidopsis organ asymmetry via modulation of KANADI activity . Plant Cell 17 : 2899 – 2910 . Petricka J.J. , Nelson T.M. ( 2007 ) Arabidopsis nucleolin affects plant development and patterning . Plant Physiol . 144 : 173 – 186 . Pinon V. , Etchells J.P. , Rossignol P. , Collier S.A. , Arroyo J.M. , Martienssen R.A. , et al. ( 2008 ) Three PIGGYBACK genes that specifically influence leaf patterning encode ribosomal proteins . Development 135 : 1315 – 1324 . Pontvianne F. , Abou-Ellail M. , Douet J. , Comella P. , Matia I. , Chandrasekhara C. , et al. ( 2010 ) Nucleolin is required for DNA methylation state and the expression of rRNA gene variants in Arabidopsis thaliana . PLoS Genet. 6 : e1001225 . Pontvianne F. , Blevins T. , Chandrasekhara C. , Mozgová I. , Hassel C. , Pontes O.M. , et al. ( 2013 ) Subnuclear partitioning of rRNA genes between the nucleolus and nucleoplasm reflects alternative epiallelic states . Genes Dev . 27 : 1545 – 1550 . Pontvianne F. , Matı´a I. , Douet J. , Tourmente S. , Medina F.J. , Echeverria M. , et al. ( 2007 ) Characterization of AtNUC-L1 reveals a central role of nucleolin in nucleolus organization and silencing of AtNUC-L2 gene in Arabidopsis . Mol. Biol. Cell 18 : 369 – 379 . Ronemus M.J. , Galbiati M. , Ticknor C. , Chen J. , Dellaporta S.L. ( 1996 ) Demethylation-induced developmental pleiotropy in Arabidopsis . Science 273 : 654 – 657 . Ryazanova A.Y. , Abrosimova L.A. , Oretskaya T.S. , Kubareva E.A. ( 2012 ) Diverse domains of (cytosine-5)-DNA methyltransferases: structural and functional characterization. In Methylation—From DNA, RNA and Histones to Diseases and Treatment . Edited by Dricu A. InTech . doi: 10.5772/52046. Sawa S. , Watanabe K. , Goto K. , Kanaya E. , Morita E.H. , Okada K. ( 1999 ) FILAMENTOUS FLOWER, a meristem and organ identity gene of Arabidopsis, encodes a protein with a zinc finger and HMG-related domains . Genes Dev . 13 : 1079 – 1088 . Schrader A. , Gross T. , Thalhammer V. , Längst G. ( 2015 ) Characterization of Dnmt1 binding and DNA methylation on nucleosomes and nucleosomal arrays . PLoS One 10 : e0140076 . Semiarti E. , Ueno Y. , Tsukaya H. , Iwakawa H. , Machida C. , Machida Y. ( 2001 ) The ASYMMETRIC LEAVES2 gene of Arabidopsis thaliana regulates formation of a symmetric lamina, establishment of venation and repression of meristem-related homeobox genes in leaves . Development 128 : 1771 – 1783 . Shuai B. , Reynaga-Peña C.G. , Springer P.S. ( 2002 ) The lateral organ boundaries gene defines a novel, plant-specific gene family . Plant Physiol . 129 : 747 – 761 . So B.R. , Wan L. , Zhang Z. , Li P. , Babiash E. , Duan J. , et al. ( 2016 ) A U1 snRNP-specific assembly pathway reveals the SMN complex as a versatile hub for RNP exchange . Nat. Struct. Mol. Biol. 23 : 225 – 230 . Song J. , Du Z. , Ravasz M. , Dong B. , Wang Z. , Ewing R.M. ( 2015 ) A protein interaction between β-catenin and Dnmt1 regulates Wnt signaling and DNA methylation in colorectal cancer cells . Mol. Cancer Res . 13 : 969 – 981 . Song J. , Rechkoblit O. , Bestor T.H. , Patel D.J. ( 2011 ) Structure of DNMT1–DNA complex reveals a role for autoinhibition in maintenance DNA methylation . Science 331 : 1036 – 1040 . Szakonyi D. , Byrne M.E. ( 2011 ) Ribosomal protein L27a is required for growth and patterning in Arabidopsis thaliana . Plant J. 65 : 269 – 281 . Szakonyi D. , Moschopoulos A. , Byrne M.E. ( 2010 ) Perspectives on leaf dorsoventral polarity . J. Plant Res. 123 : 281 – 290 . To T.K. , Kim J.M. , Matsui A. , Kurihara Y. , Morosawa T. , Ishida J. , et al. ( 2011 ) Arabidopsis HDA6 regulates locus-directed heterochromatin silencing in cooperation with MET1 . PLoS Genet. 7 : e1002055 . Tokizawa M. , Kobayashi Y. , Saito T. , Kobayashi M. , Iuchi S. , Nomoto M. , et al. ( 2015 ) SENSITIVE TO PROTON RHIZOTOXICITY1, CALMODULIN BINDING TRANSCRIPTION ACTIVATOR2, and other transcription factors are involved in ALUMINUM-ACTIVATED MALATE TRANSPORTER1 expression . Plant Physiol. 167 : 991 – 1003 . Ueno Y. , Ishikawa T. , Watanabe K. , Terakura S. , Iwakawa H. , Okada K. , et al. ( 2007 ) Histone deacetylases and ASYMMETRIC LEAVES2 are involved in the establishment of polarity in leaves of Arabidopsis . Plant Cell 19 : 445 – 457 . Wollmann H. , Stroud H. , Yelagandula R. , Tarutani Y. , Jiang D. , Jing L. , et al. ( 2017 ) The histone H3 variant H3.3 regulates gene body DNA methylation in Arabidopsis thaliana . Genome Biol. 18 : 94 . Xu C. , Bian C. , Lam R. , Dong A. , Min J. ( 2011 ) The structural basis for selective binding of non-methylated CpG islands by the CFP1 CXXC domain . Nat. Commun. 2 : 227. Xu X. , Chen C. , Fan B. , Chena Z. ( 2006 ) Physical and functional interactions between pathogen-induced Arabidopsis WRKY18, WRKY40, and WRKY60 transcription factors . Plant Cell 18 : 1310 – 1326 . Yang J.Y. , Iwasaki M. , Machida C. , Machida Y. , Zhou X. , Chua N.H. ( 2008 ) betaC1, the pathogenicity factor of TYLCCNV, interacts with AS1 to alter leaf development and suppress selective jasmonic acid responses . Genes Dev . 22 : 2564 – 2577 . Yao Y. , Ling Q. , Wang H. , Huang H. ( 2008 ) Ribosomal proteins promote leaf adaxial identity . Development 135 : 1325 – 1334 . Yao Q. , Song C.-X. , He C. , Kumaran D. , Dunn J.J. ( 2012 ) Heterologous expression and purification of Arabidopsis thaliana VIM1 protein: in vitro evidence for its inability to recognize hydroxymethylcytosine, a rare base in Arabidopsis DNA . Protein Expr. Purif . 83 : 104 – 111 . Yoder J.A. , Soman N.S. , Verdine G.L. , Bestor T.H. ( 1997 ) DNA (cytosine-5)-methyltransferases in mouse cells and tissues. Studies with a mechanism-based probe . J. Mol. Biol . 270 : 385 – 395 . Zhang Z.M. , Liu S. , Lin K. , Luo Y. , Perry J.J. , Wang Y. , et al. ( 2015 ) Crystal structure of human DNA Methyltransferase 1 . J. Mol. Biol. 427 : 2520 – 2531 . Zhang X. , Yazaki J. , Sundaresan A. , Cokus S. , Chan S.W. , Chen H. , et al. ( 2006 ) Genome-wide high-resolution mapping and functional analysis of DNA methylation in Arabidopsis . Cell 126 : 1189 – 1201 . Abbreviations Abbreviations AlphaScreen amplified luminescent proximity homogeneous assay screen AS2 ASYMMETRIC LEAVES2 BSA bovine serum albumin Dnmt1 DNA methyltransferase1 dsDNA double-stranded DNA ETT/ARF3 ETTIN/AUXIN RESPONSE FACTOR3 MET1 METHYLTRASFERASE1 NGS next-generation sequencing NUC1 NUCLEOLIN1 PBS phosphate-buffered saline RH10 RNA HELICASE10 WGE wheat germ extract Footnote Footnote The output data generated by next-generation sequencing in this paper have been submitted to the DDBJ Sequence Read Archive (DRA) under the accession number DRA006505 © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact email@example.com
Plant and Cell Physiology – Oxford University Press
Published: Feb 5, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera