Arabidopsis ORC1 is a PHD-containing H3K4me3 effector that regulates transcription

Abstract

Control of gene expression depends on a complex and delicate balance of various posttranslational modifications of histones. However, the relevance of specific combinations of histone modifications is not fully defined. Downstream effector proteins recognize particular histone modifications and transduce this information into gene expression patterns. Methylation of histone H3 at lysine 4 (H3K4me) is a landmark of gene expression control in eukaryotes. Its recognition depends on the presence in the effector protein of a motif termed plant homeodomain (PHD) that specifically binds to H3K4me3. Here, we establish that Arabidopsis ORC1, the large subunit of the origin recognition complex involved in defining origins of DNA replication, functions as a transcriptional activator of a subset of genes, the promoters of which are preferentially bound by ORC1. Arabidopsis ORC1 contains a PHD and binds to H3K4me3. In addition to H4 acetylation, ORC1 binding correlates with increased H4K20me3 in the proximal promoter region of ORC1 targets. This suggests that H4K20me3, unlike in animal cells, is associated with transcriptional activation in Arabidopsis. Thus, our data provide a molecular basis for the opposite role of ORC1 in transcriptional activation in plants and repression in animals. Since only ORC1 proteins of plant species contain a PHD, we propose that plant ORC1 constitutes a novel class of H3K4me3 effector proteins characteristic of the plant kingdom.

Regulation of gene expression depends on the reading of complex combinations of posttranslational modifications of histones (1). Modified histone surfaces can facilitate the specific recruitment of a variety of effector proteins. Understanding the mechanisms used by downstream effectors to recognize and transduce histone modifications into specific gene expression patterns constitutes a major challenge in the field. The consequences of individual histone modifications are still far from being fully understood.

Methylation of histone H3 at lysine 4 (H3K4me) is crucial for gene activation and repression (2–6). Recognition of H3K4me3 by proteins containing a plant homeodomain (PHD) is an initial event that ultimately leads to changes in the acetylation and/or methylation status in the vicinity of H3K4me3 (2–4, 6). Thus, the identification of H3K4me3 effector proteins has become a subject of primary importance. Some histone marks, e.g., the association of H3K4me3 with active genes, are common to animal and plant cells, but others are plant specific (5, 7). For example, H3K9me3 and H4K20me3, which in animal cells are detected in heterochromatin, localize to euchromatin in plants (5, 7, 8). Furthermore, H3K9me3 is detected in transcriptionally active genes in Arabidopsis (9, 10).

ORC1, the large subunit of the origin recognition complex (ORC), originally identified as a component of DNA replication initiation complexes (11, 12), also plays a role in transcriptional regulation. Thus, ORC1 participates in repression of yeast HMR and HML silent mating type loci (13), and human aldolase B (14) and c-Myc genes (15). Mutations of several ORC subunits in multicellular organisms cause pleiotropic phenotypes including chromosomal abnormalities, cell cycle arrest or zygotic lethality (16–18).

The six ORC subunits are highly conserved during evolution. Arabidopsis contains two ORC1 genes (ORC1a and ORC1b) that encode proteins possessing ≈92% amino acid similarity over the entire protein (19, 20). Here, we have uncovered a novel role of Arabidopsis ORC1 in transcriptional regulation as a H3K4me3 effector protein. This role is mediated by a PHD motif and depends on ORC1 binding at target promoters. ORC1-dependent gene activation is associated with an increase in H4 acetylation and H4K20 trimethylation. We propose that this mechanism is a general feature of the plant kingdom, as ORC1 proteins of unicellular algae, mosses and higher plants, but not other organisms, contain a highly conserved PHD motif.

Results

Arabidopsis ORC1 Contains a Functional PHD Motif that Specifically Binds to H3K4me3.

Both Arabidopsis ORC1a and ORC1b proteins, but not yeast and animal ORC1, contain two tandem Zn²⁺ fingers (Fig. 1A), whose predicted secondary structure constitutes a PHD motif (19, 21). According to 3D modeling, this region fits in the PHD fold of human CHD4 (Fig. 1B), a gene-silencing protein (22). The Arabidopsis ORC1 PHD motif also contains residues conserved in the “cage” present in some PHD-containing proteins (6), such as human ING2 (Figs. 1A, 1B), a protein that recognizes H3K4me3 and represses target genes (2, 23). Thus, we hypothesized that Arabidopsis ORC1 could be a novel PHD-containing H3K4me3 effector.

Fig. 1.

The Arabidopsis ORC1 proteins contain a PHD motif that binds H3K4me3. (A) Schema of ORC1 showing the location of DNA replication motifs (I-VI) and alignment of the two Arabidopsis ORC1 PHD motifs with that of several PHD-containing proteins. The main PHD regions appear color-coded, as indicated. Asterisks indicate two key cysteine residues of one of the two Zn²⁺ fingers and a phenylalanine of the cage that were mutated in this study. (B) Three-dimensional modeling of Arabidopsis ORC1b PHD using the crystal structure information of CHD4 (PDB 1mm2) and ING2 (PDB 2g6q) PHDs. Colored residues refer to those in panel A. White spheres in CHD4 and ING2 indicate the position of Zn²⁺ ions. Red lines in ING2 indicate the position of a histone H3K4me3 peptide. (C) Pull-down assay of plant histone extracts with Arabidopsis His-ORC1b. Bound histones were detected by Western blot with anti-H3K4me3, anti-H3K9me3, and anti-H4K20me3 antibodies. (D) Binding assay of Arabidopsis His-ORC1b with biotinylated H3 peptides either unmodified, mono-, di-, or trymethylated at K4. Bound peptides we detected by Western blot with anti-biotin antibodies. (E) Pull-down assay of plant histone extracts with Arabidopsis His-ORC1b, His-ORC1b^PHD(C/A), and His-ORC1b^PHD(F/A). His-ORC1b^PHD(C/A) contains two point mutations that change C183 and C186 to A. His-ORC1b^PHD(F/A) contains a F190A mutation (asterisks in A).

To investigate the functional relevance of the PHD motif of Arabidopsis ORC1 we first assessed its interaction with histones. Interaction of ORC1b with either unmodified histones H3 or H4 was not detectable (not shown). However, ORC1b was able to pull-down H3K4me3 from Arabidopsis extracts whereas binding of ORC1b to H3K9me3 or H4K20me3 was not detectable (Fig. 1C). Binding assays of Arabidopsis ORC1b to histone H3 peptides modified at K4 demonstrated the specific recognition of H3K4me3 residues (Fig. 1D). Finally, we studied the binding properties of ORC1b containing an altered PHD. Mutations of two critical cysteine residues to alanines (ORC1b^PHD(C/A) bears C183A and C186A substitutions) reduced, although did not completely abolish, ORC1b-H3K4me3 interaction (Fig. 1E). Down-regulation of protein activity by mutations in homologous Zn²⁺ ligands in the PHD has been reported for other proteins (24). Mutational analysis of the human ING2 PHD motif has demonstrated that residue W238 is part of the cage that contributes to the interaction with H3K4me3 residues (2, 23). Arabidopsis ORC1 proteins have a G residue instead of a W residue in the equivalent position (Fig. 1A). However, next to it they possess an F residue (F189 and F190 in ORC1a and ORC1b, respectively). An F residue is present also in other PHD-containing proteins and, in some of them, it is important for protein activity (24). Mutation of the F190 residue of ORC1b (ORC1b^PHD(F/A)) revealed its importance for H3K4me3 binding, although residual binding was still detected (Fig. 1E).

The predicted spatial location of the F residue in ORC1, relative to that of ING2 W238 (Fig. 1B), tentatively supports the idea that ORC1 may form a cage that stabilizes its interaction with H3K4me3, although full demonstration requires further structural analysis. ORC1a and ORC1b proteins differ in only two residues throughout the entire PHD motif in noncritical positions, suggesting that our results apply to both Arabidopsis ORC1 proteins. We conclude that Arabidopsis ORC1 contains a functional PHD that mediates its binding to H3K4me3, revealing an unanticipated role of Arabidopsis ORC1 as a potential H3K4me3 effector.

Expression of ORC1 Activates Transcription in a PHD-Dependent Manner.

To define the role of Arabidopsis ORC1 as a transcriptional regulator through H3K4me3 binding, we first searched for plants possessing reduced ORC1 mRNAs. A search in the Arabidopsis collections of lines with insertions throughout the genome, which in many cases disrupt a gene of interest, led us to identify plants bearing T-DNA insertions in the ORC1 genes. For ORC1a, insertions were located upstream the ORF and did not decrease ORC1a mRNA levels [supporting information (SI) Fig. S1]. Likewise, for ORC1b, the line with the lowest ORC1b mRNA level still retained ≈30% of wild-type expression (Fig. S1). The absence of individual orc1 knock-out plants suggests that both ORC1 genes are essential, as it occurs in other multicellular organisms. This observation, together with the expression patterns of the individual T-DNA lines that we found, hampered the selection of a double orc1a, orc1b knock-out line.

In the absence of viable orc1 loss-of-function mutant plants we generated Arabidopsis plants that expressed constitutively Myc-tagged ORC1b and ORC1b^PHD(C/A) proteins or control plants transformed with an empty vector. We chose to focus on the mutations affecting the Zn²⁺ finger because of their slightly higher effect on H3K4me3 binding. To avoid undesired effects derived from high levels of ectopic ORC1 expression, we selected homozygous plants that expressed the transgene mRNAs and the ORC1 protein to levels detectable only after a short preincubation with proteasome inhibitors (Figs. S2A, S2B). We found that the cotyledon epidermis of ORC1b transgenic plants contained cells of smaller size and an increased cell density (Fig. S2C). This phenotype was observed in several independent lines, as well as in transgenic plants expressing ORC1a (not shown). However, plants expressing the mutant ORC1b^PHD(C/A) protein did not show this phenotype (Fig. S2).

One possibility to explain this phenotype is that ectopic expression of ORC1 accelerates cell proliferation. Expression of genes required for initiation of DNA replication (25, 26) increase in proliferating cells (Fig. S3). Interestingly, ORC1 transgenic plants showed elevated mRNA levels of only some of the cell proliferation marker genes analyzed (CDT1a, MCM3, and ORC3). Other marker genes (CDT1b and CDC6a), which are also up-regulated in proliferating cells (Fig. S3), did not change their expression in ORC1b transgenic plants (Fig. 2A). A similar response was observed in plants expressing ORC1a protein (Fig. 2A). Furthermore, plants expressing the mutated ORC1b^PHD(C/A) protein did not show any significant stimulation of the expression of these genes (Fig. 2A). Similar results were obtained in several independent transgenic lines analyzed (Fig. S4). Therefore, we conclude that Arabidopsis ORC1 can act as a transcriptional activator of specific target genes in a PHD-dependent manner.

Fig. 2.

Expression of Arabidopsis ORC1 activates transcription. (A) Determination of mRNA levels of the indicated DNA replication genes by real-time RT-PCR in control and transgenic plants expressing Myc-ORC1a, Myc-ORC1b, and Myc-ORC1b^PHD(C/A). Values represent mean ± SD (n = 3). (B) Determination of CDT1a (At2g31270) and APG9 (At2g31260) mRNA levels by real-time RT-PCR in control and transgenic plants (10-day-old) expressing Myc-ORC1b and Myc-ORC1b^PHD(C/A). Values represent mean ± SD (n = 3). (C) Effect of ORC1a protein on the spatial expression pattern of CDT1a detected in pCDT1a:GUS reporter plants (7–10-day-old). Bars, 200 μm.

The fact that only a subset of genes up-regulated in proliferating cells are stimulated in ORC1 transgenic plants makes unlikely the possibility of a passive consequence of having more cells undergoing cell division. To investigate the mechanism behind ORC1-dependent transcriptional activation, we addressed our attention to one of the stimulated genes, CDT1a, a DNA replication licensing gene (11); the regulation and spatial pattern of expression in developing Arabidopsis of this gene has been reported (27).

To determine whether the effect of ORC1 was extended to neighbor genes we analyzed the expression of the APG9 gene, a divergent transcriptional unit ≈3 kb apart from CDT1a. We found that ORC1 showed a gene-specific effect, as APG9 expression was not affected (Fig. 2B; genomic map in Fig. 3A). We also assessed transcriptional activation of CDT1a in whole plants. Constitutive expression of ORC1 in pCDT1a:GUS reporter plants demonstrated an increase in CDT1a expression without modifying its spatial expression pattern (Fig. 2C), suggesting that ORC1-dependent transcriptional activation of target genes occurs in the same locations where these targets are normally expressed.

Fig. 3.

Identification of ORC1 binding sites in vivo. (A) ChIP assays were carried out using anti-Myc antibodies with control plants (transformed with an empty vector) and transgenic plants (10-day-old) expressing Myc-ORC1b and Myc-ORC1b^PHD(C/A). The genomic location of the CDT1a and APG9 genes is shown together with their direction of transcription. Letters and small black bars refer to location in the map of the fragment amplified by PCR and its size. A representative example of a ChIP experiment is shown in Fig. S5. Enrichment was calculated as (ChIP/Input)/(ChIP control/Input control) using the band intensity values of the “Input” and “ChIP” lanes after substracting the corresponding value in the “no Ab” lanes. Data shown are representative of at least two independent assays. (B) ChIP assays carried out as in (A) show ORC1 binding to the promoters of other ORC1-responsive or nonresponsive genes used in this study. Black boxes indicate position and size of PCR-amplified fragment.

Stimulation of Gene Expression by ORC1 Depends on Binding to Target Promoters.

The highly specific effect of ORC1 on CDT1a gene expression suggested targeting of ORC1 to this locus. To find whether activation of gene expression correlated with ORC1 binding, we carried out chromatin immunoprecipitation (ChIP) assays scanning the APG-CDT1a genomic region. This analysis revealed an enrichment of ORC1-bound DNA fragments covering ≈500 bp in the 5′ region of CDT1a ORF (Fig. 3A; Fig. S5). Interaction was site specific, as ORC1 was not detected upstream of the APG9 gene or in other locations within the CDT1a ORF or its 3′UTR (Fig. 3A). Furthermore, mutations in the PHD domain of ORC1b led to undetectable ORC1b binding under our experimental conditions (Fig. 3A; Fig. S5). We extended the ChIP assays to genes the expression of which was assessed earlier (Fig. 2A). ORC1 was detected bound to ORC1-responsive promoters (ORC3 and MCM3) but not to promoters of CDT1b and CDC6a, the expression of which did not change in an ORC1-dependent manner (Fig. 3B). Thus, our experiments so far led us to conclude that (i) binding of ORC1 to target promoters occurs in a PHD-dependent manner, likely through H3K4me3 residues, and (ii) it is associated with transcriptional activation of these target genes.

Transcriptional Activation by ORC1 Is Associated with Histone H4 Hyperacetylation.

Transcriptional activation normally depends on hyperacetylation of target promoters. Consistent with this, we first found that treatment with trichostatin A (TSA), a histone deacetylase inhibitor, synergistically increased CDT1a transcription in an ORC1-dependent manner (Fig. 4A). Then we assessed the acetylation status of ORC1-bound promoters and found that H3 acetylation did not change significantly (Fig. S6). However, we detected a drastic enrichment of H4 acetylation at the CDT1a gene, whereas we observed that plants expressing the mutant ORC1b^PHD(C/A) retained only a partial ability to increase H4 acetylation (Fig. 4B; Fig. S5). Other ORC1 target genes analyzed showed a comparable enrichment of H4ac residues at their proximal promoters (Fig. 4C).

Fig. 4.

Histone acetylation status in response to ORC1 binding. (A) Determination of CDT1a mRNA levels by real-time RT-PCR in control and transgenic plants (10-day-old) expressing Myc-ORC1b with and without treatment with the histone deacetylase inhibitor trichostatin A (TSA; 1 μg/ml). Values represent mean of two independent experiments, made relative to the control without or with TSA. (B) Quantification of histone H4ac throughout the CDT1a and APG9 loci. Fragments amplified are indicated in the map. A representative example of a ChIP experiment is shown in Fig. S5. Values were calculated as described in Fig. 3A. Data shown are representative of at least two independent assays. (C) ChIP assays were carried out as in (B) to determine histone H4ac in the promoters of other ORC1-responsive genes used in this study. Black boxes indicate position and size of PCR-amplified fragment.

H4K20me3 Is Present at the Promoters of Genes Activated in an ORC1-Dependent Manner.

In addition to changes in acetylation, the histone methylation status determines gene activity. ChIP assays showed that H3K4me3 increased in the same region where ORC1 binds in a PHD-dependent manner (Fig. S7). This enrichment, which might be a consequence of increased transcription, was specific of the proximal promoter region of the target genes (Fig. S7).

In human cells, H4K20me3 is a histone mark associated with gene silencing (28). In Arabidopsis, the H4K20me3 mark is found in euchromatin but not in heterochromatin (5, 7). However, its relevance in transcription has not been defined. To address this question, we determined the H4K20me3 status within the CDT1a genomic region and found that it was enriched just upstream of the ORF in an ORC1- and PHD-dependent manner (Fig. 5A; Fig. S5). A similar H4K20me3 enrichment was obtained when other ORC1-bound promoters were analyzed (Fig. 5B). Finally, we determined the H4K20me3 status of the same set of DNA replication genes the expression of which was significantly increased in cultured cells during the transition from arrested to proliferating cells (Fig. 5C). ChIP experiments revealed an enrichment of H4K20me3 in the promoter of these marker genes (Fig. 5D), suggesting that an association of H4K20me3 with active gene expression could be a general characteristic of the Arabidopsis histone code.

Fig. 5.

ORC1 binding and transcriptional activation is linked to enrichment in H4K20me3. (A) Quantification of histone H4K20me3 throughout the CDT1a and APG9 loci by ChIP analysis. Fragments amplified are indicated in the map. Representative example of a ChIP experiment is shown in Fig. S5. Values were calculated as described in Fig. 3A. Data shown are representative of at least two independent assays. (B) ChIP assays were carried out as in (A) to determine histone H4K20me3 in the promoters of other ORC1-responsive genes. Black boxes indicate position and size of PCR-amplified fragment. (C) Determination of mRNA levels by real-time RT-PCR of DNA replication genes in Arabidopsis MM2d-cultured cells arrested by sucrose deprivation for 24 hours and proliferating cells (2 hours after sucrose addition), during which sodium butyrate (10 mM) was added. (D) Histone H4K20me3 status at the promoter region of CDT1a, ORC3, MCM3 (test genes), and ACT2 (control) in Arabidopsis-cultured cells. ChIP assays were carried out in arrested and proliferating cells as described in (C). Fragments amplified correspond to fragment B of CDT1a (see A) and to an equivalent fragment in the promoters of ORC3 and MCM3. (E) Simplified model of Arabidopsis ORC1 function in transcriptional activation. We propose that Arabidopsis ORC1 is a H3K4me3 effector protein that binds to target promoters through its PHD motif. ORC binding is associated with an enrichment of H4 acetylation and H4K20me3, which are histone marks that determine transcriptional activation in Arabidopsis and, possibly, in the entire plant kingdom.

The ORC1 PHD Motif Is Characteristic of the Plant Kingdom.

ORC1 is highly conserved from archaea to yeast, animals, and plants (11, 12, 19–21, 29). All ORC1 proteins conserve the typical AAA⁺ ATPase domain as well as other motifs required for its function in DNA replication. The identification of a functional PHD in Arabidopsis ORC1 supports a novel role for this protein in transcriptional control. Moreover, the general relevance of this motif is evidenced by the fact that a PHD is found in all plant species surveyed, ranging from higher plants to mosses and unicellular algae (Fig. S8). The absence of a PHD in ORC1 of fungi and animals reinforces the idea that it is a general feature of all organisms within the plant kingdom. Therefore, we propose that the acquisition of a functional PHD in plant ORC1 proteins and its ability to mediate binding to H3K4me3, with its associated modifications in H4 acetylation and H4K20 trimethylation, may represent a crucial difference in transcriptional control among yeasts, animals, and plants.

Discussion

Understanding the mechanisms used by effector proteins that recognize specific histone modifications and translate this information into a specific gene expression response represents a major challenge in the field. In this study we found that Arabidopsis ORC1 acts as a transcriptional regulator of a subset of target genes but is unique in that it functions as a transcriptional activator. Moreover, it is also unique in the mechanism used, which depends on the recognition of histone H3K4me3 residues at ORC1 target promoters by a PHD motif located in the ORC1 N terminus. Interestingly, ORC1, of all plants surveyed but not of yeast or animals, contains a PHD. Binding of ORC1 to its target sites is associated with an increase in H4 acetylation and H4K20 trimethylation. Thus, contrary to the situation in animals, H4K20me3 associates with transcriptional activation. Our data provide a molecular basis for the role of Arabidopsis ORC1 in transcriptional regulation and lead us to propose that ORC1 is a novel class of PHD-containing H3K4me3 effector protein characteristic of the plant kingdom.

In yeast and human cells, where ORC1 participates in gene silencing (13–15), changes in the histone modification status occur in the vicinity of the genomic region affected. We have found that Arabidopsis ORC1 also plays a role as a transcriptional regulator; but there are two novel features in its mechanism of action. The function of ORC1 depends on its PHD motif and, in contrast to yeast and animal cells, it acts as a transcriptional activator in Arabidopsis, a role that may be a general property within the plant kingdom.

Trimethylation of histone H3 at lysine 4 (H3K4me3) is one of the modifications with a role in transcriptional regulation (2, 3, 30). Our data show that Arabidopsis ORC1 contains a functional PHD motif that mediates interaction with H3K4me3. Interestingly, the residue F190 (in ORC1b) may play a function analogous to W238 in the H3K4me3 cage of human ING2.

ChIP experiments revealed that binding of ORC1 occurs preferentially at certain promoters, leading to specific transcriptional activation of the downstream gene. Thus, for example, CDT1a gene expression was activated in a ORC1- and PHD-dependent manner, but not that of APG9, a divergent transcription unit located only ≈3 kb away from CDT1a. ORC1-mediated transcriptional activation occurred without altering its spatial expression pattern, suggesting that the observations made in plants ectopically expressing ORC1 may be an amplification of an event normally occurring in wild-type plants.

Transcriptional activation by ORC1 apparently occurs concomitantly with changes in the H4 acetylation and H4K20me3 status of a relatively small region close to the ORC1 binding site. H4K20me3 is a histone mark associated with gene silencing in human cells (28). However, in Arabidopsis, H4K20me3 and H3K9me3 are detected in euchromatin but not in heterochromatin (5, 7, 8). Whether these marks correlate with active or inactive genes has been so far a matter of debate. Our previous studies (9) and the present work strongly support the idea that H3K9me3 and H4K20me3 are associated with a subset of transcriptionally active euchromatic genes. However, our study still leaves open the question of whether H4K20me3 acts as a histone mark that activates transcription or whether it appears as a consequence of active transcription. The molecular basis for the opposite read-out of these histone marks in plant and animal cells is not presently known.

Collectively, our data support a proposal by which Arabidopsis ORC1 is a novel H3K4me3 effector protein that can activate transcription of target genes through its cage-containing PHD (Fig. 5E). Other Arabidopsis transcriptional regulators have been reported to contain a PHD motif. However these regulatory factors lack the cage motif and their ability to interact with H3K4me3 residues has not been shown (31–33). Whether H3K4me3 facilitates or it determines recruitment of ORC1 to target sites is not presently known.

Our study opens new roads to understand transcriptional control in eukaryotes. First, the transcriptional activation pathway uncovered in our study is in contrast to the silencing role of ORC1 in yeast and animals. This may represent a fundamental difference in the transcriptional regulatory strategies evolved in plants and other organisms. Second, defining whether H4K20me3 acts as a determinant of transcriptional activation or it appears in active genes located in accessible chromatin regions is another question that remains for the future. A major challenge is the genome-wide identification of ORC1 target sites. This issue is of special relevance given the function of ORC1 at DNA replication origins. Whether the ORC1-binding sites identified here are part of functional origins is not known yet but is an attractive possibility that will be addressed when appropriate tools are developed. ORC1 may bind in a sequence-specific manner, as it occurs in budding yeast; or it can recognize special DNA or chromatin environments, for example specific histone modifications. This may occur in association with promoters or other regulatory regions. An association of transcriptional regulatory elements and active origins of replication has been recently identified in different eukaryotic systems (34–36). Our approach should provide an adequate framework to address this question in the future based on the coordination between replication and transcription (37) and the association between H4 acetylation and origin activity (38, 39).

Materials and Methods

Plant Material.

Arabidopsis seedlings (Col-0 ecotype) were grown in MS salts medium supplemented with 1% sucrose (MSS) and 1% agar in a 16-hour light, 8-hour dark regimen at 22 °C. To generate Myc-ORC1a and Myc-ORC1b transgenic plants, the coding region of ORC1a (At4g14700) was cloned into the pROKII vector with a Myc tag and ORC1b (At4g12620) into the pGW18 vector (Gateway System). To generate mutant Myc-ORC1b^PHD(C/A), residues C183 and C186 of ORC1b were mutated to alanine using the QuikChange Site-Directed Mutagenesis Kit (Strategene) and then transferred to the pGW18 vector. To generate the mutant Myc-ORC1b^PHD(F/A), residue F190 was changed to alanine using a similar procedure. Constructs were introduced into Agrobacterium tumefaciens C58CRif^R to transform A. thaliana plants (Col-0 ecotype; T0 generation). Selection was carried on MSS agar plates containing kanamycin (50 μg/ml) and transformed plants were transferred to soil. Myc-ORC1a/pCDT1a:GUS plants were obtained by crossing. T3 and T4 homozygous lines were used in this work. In all cases, plants transformed with an empty vector were used as controls.

Real-Time Reverse Transcription-Polymerase Chain Reaction Analysis.

Total RNA from 10 day-old seedlings was extracted using the TRIzol reagent (Invitrogen) and reverse transcription-polymerase chain reaction (RT-PCR) was carried out with the ThermoScript RT system (Invitrogen) using 1 μg of total RNA as template and oligo(dT) as primer. The LightCycler system with the FastStart DNA Master Green I (Roche) was used. To normalize the differences in RNA amount we used the ubiquitin 10 gene (At4g05320). Primer sequences are described in Table S1. The data were generated from duplicates of three independent experiments.

Pull-Down Assays.

The coding region of ORC1b, ORC1b^PHD(C/A) and ORC1b^PHD(F/A) were cloned into pDEST17 Gateway vector (Invitrogen) to express the fusion proteins in bacteria. Nuclear extracts enriched in histones were prepared from MM2d Arabidopsis suspension cultured cells harvested 4 days after subculturing. After filtration, cells were resuspended in nuclei isolation buffer (10 mM Tris-HCl pH 9.5, 10 mM ethylenediaminetetraacetic acid (EDTA), 100 mM KCl, 0.5 M sucrose, 4 mM spermidine, 1 mM spermine, and 0.1% β-mercaptoethanol) and incubated for 20 minutes at 4 °C. Nuclei were then filtered through a 30-μm nylon mesh and resuspended in histone extraction buffer (320 mM (NH₄)₂SO₄, 200 mM Tris-HCl pH 8.0, 20 mM EDTA, 10 mM ethylene glycol tetraacetic acid, 5 mM MgCl₂, 1 mM DTT, 10% glycerol, 1:1000 dilution of Sigma plant protease inhibitor mixture). For pull-down assays, His-ORC1b, His-ORC1b^PHD(C/A) and His-ORC1b^PHD(F/A) proteins bound to Ni-NTA agarose beads were incubated with histone extracts in binding buffer (50 mM NaCl, 20 mM Tris-HCl, pH 7.5, 25% glycerol, 1.5 mM MgCl_2, 1 mM phenylmethylsulphonyl fluoride (PMSF), 0.02% Triton X-100, and 30 mM imidazole) for 2 hours at 4 °C. Beads were washed twice with 50 mM NaH₂P0₄, 300 mM NaCl, and 20 mM imidazole, and twice with 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% de Triton X-100. Samples were then fractionated in Tris-Tricine acrylamide gels and analyzed by Western blot.

Histone Peptide Binding Assays.

The biotinylated histone peptides, H3 peptide (12–357), H3K4me1 (12–563), H3K4me2 (12–460), and H3K4me3 (12–564) were from Upstate. The sequence corresponds to amino acids 1–21: ART(meK)QTARKSTGGKAPRKQLA. For peptide binding assay, each peptide (0.5 μg) was incubated with His-ORC1b bound to Ni-NTA agarose beads in 50 mM Tris-HCl, pH 7.5, 300 mM NaCl, 0.1% (vol/vol) Nonidet P-40, 1 mM PMSF during 4 hours at 4 °C. Beads were then washed five times, and the samples were separated in Tris-tricine polyacrylamide gel at 15% and subjected to Western blot analysis.

Chromatin Immunoprecipitation (ChIP).

ChIP assays were carried out using 10-day-old plants (9), except that they were preincubated with 50 μM MG132 before the fixation step. For immunoprecipitation, 10 μl of anti-Myc (05–724), anti-H3K4me3 (07–473), anti-H4K20me3 (07–463), anti-acetyl histone H4K5,8,12,16 (06–598), or anti-acetyl histone H3K9,14 (06–599) antibodies (Upstate Biotechnology), were incubated in phosphate-buffered saline solution with Protein A agarose, and then 1 mg of protein extract was added. One μl was used for each PCR assay. The input lanes contained a 1/100 dilution. PCR primers used in ChIP assays are described in Table S2. As negative controls we carried out the ChIP experiments using protein A-agarose without antibody. Enrichment was calculated as (ChIP/Input)/(ChIP control/Input control) using the band intensity values of the “Input” and “ChIP” lanes after substracting the corresponding value in the “no Ab” lanes.