Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. 2006 Jan 17; 103(3): 574–579.
Published online 2006 Jan 5. doi:  10.1073/pnas.0509974103
PMCID: PMC1334668

Distribution of histone H3.3 in hematopoietic cell lineages


We have introduced the histone variant H3.3 into chicken erythroid cell lines and examined its distribution in the neighborhood of the folate receptor (FR) and β-globin genes by using high-resolution chromatin immunoprecipitation (ChIP). Marked incorporation of tagged H3.3 into the FR gene is confined to its upstream regulatory region and is observed whether or not the gene is transcriptionally active. Incorporation is also observed over locus control regulatory elements in the absence of transcription of genes regulated by these elements, suggesting that gene activity per se is not necessarily required to replace H3 with H3.3. Other active genes display various behaviors, either incorporating H3.3 over both the coding region and upstream regulatory region or over upstream sequences only. There is, however, no straightforward correlation between sites of H3.3 incorporation and regions of enrichment in H3 acetylation and lysine-4 methylation. In the case of FR and VEGF-D, in which incorporation is confined to upstream regions, the presence of exogenous H3 results in reduced expression, whereas H3.3 stimulates expression. This finding suggests that these histone variants can be active rather than passive participants in regulation of expression.

Keywords: folate receptor, vascular endothelial growth factor D, globin, chromatin

Histone variants are incorporated into chromatin at sites that suggest that the variants play an important role as epigenetic marks. Among these, the variants of histone H3 have been the focus of considerable attention. A family of H3-like proteins (CenH3) is localized at the centromeres of eukaryotes, where it replaces histone H3 and appears to play a critical role in defining centromeric identity (1, 2). Recently, it has been shown that another H3 variant, H3.3, is incorporated into chromatin throughout the cell cycle, in contrast to H3, which is incorporated only during S phase (3). This constraint ensures that if nucleosomes that are disrupted or lost during transcription must be replaced, the new nucleosomes will contain H3.3. Presumably, as a result, H3.3-containing nucleosomes are enriched in euchromatin (3). Consistent with this view, H3.3 isolated from a Drosophila cell line is enriched in covalent modifications associated with transcriptionally active chromatin such as acetylation at multiple lysines and methylation at Lys-4 (K4) (4). The mechanism of H3.3 deposition involves a specialized chaperone complex containing the protein HIRA. In contrast, the replication-dependent incorporation of H3 is effected by a replication-specific complex that includes CAF-1 (5).

Recent microarray studies in Drosophila provide strong evidence for a correlation between transcriptional activity and H3.3 incorporation (6). Because of the association of H3.3 with transcriptionally active chromatin domains, it has been suggested that there is a causal relationship between the presence of H3.3 and the accessibility of the chromatin template for transcription. It has been noted that if H3.3-containing chromatin is more easily transcribed, increased transcription may lead to further replacement of H3 by H3.3. During replication, the H3.3 nucleosomes would be diluted by newly deposited H3 nucleosomes, but the genomic domains with higher H3.3 content would preserve their active status postreplication, and the H3 nucleosomes would once again gradually be replaced. In this model, H3.3 would be an epigenetic mark for transcriptionally active chromatin. Alternatively, the incorporation of H3.3 into transcriptionally active genes might simply reflect the fact that H3.3 is the only available histone variant at that stage of the cell cycle. In addition to the studies in Drosophila, other recent papers have carried out more limited surveys of the distribution of H3.3 in cell lines derived from vertebrates. In one case (7), it was reported that for most genes surveyed, incorporation was confined to promoter regions, whereas in another (8), H3.3 was distributed rather uniformly over the transcribed regions of the genes surveyed. The more comprehensive results in Drosophila suggest that H3.3 can be found within coding regions but with particular concentration over the promoter and proximal regions 3′ of the promoter, as well as in other regions that appear to be free of genes (6).

For some years, we have made use of chicken erythroid cells, particularly the genes around the folate receptor/β-globin locus, as a test system for studying the interaction between chromatin structure and gene expression. We have mapped a variety of histone modifications across the locus as a function of developmental stage and attempted to correlate these maps with the state of local transcriptional activity. This system seemed to be an important system in which to examine the distribution of histone H3.3 to extend our knowledge of the state of chromatin within the locus. We, therefore, created transformed cells lines carrying tagged versions of both H3 and H3.3 for that purpose, and we measured H3.3 abundance at a resolution of one or two nucleosomes at a number of sites across the region. We detected H3.3 enrichment over regulatory elements at the folate receptor even at developmental stages when the gene itself was not expressed. We then explored a number of other genes active in these erythroid cells and found that, depending upon the gene, H3.3 was concentrated over the upstream regulatory region only, over both the upstream and coding regions, or, in one case, over neither. There was no obvious correlation with levels of transcription. We conclude that H3.3 incorporation over coding regions is not universally associated with transcription, that sites of incorporation probably reflect gene-specific mechanisms for histone displacement of nucleosomes, and that regulatory sites including distant enhancers or locus control elements may be targets for H3.3 incorporation independent of subsequent transcriptional activation. Furthermore, we found no simple correlation between sites of H3.3 incorporation and of histone H3 modifications (acetylation and methylation) associated with gene activation.

For two genes in which H3.3 incorporation was confined to the upstream regulatory elements, expression of exogenous, untagged H3 resulted in inhibition of expression, whereas expression of H3.3 stimulated expression. We suggest that in these cases, incorporation of H3.3 may play a direct role in activation of expression.


Incorporation of H3 and H3.3 over the β-Globin Locus. We studied the distribution of H3 and H3.3 over the β-globin locus and its surrounding genes (Fig. 1A) in the chicken erythroleukemia line, 6C2, which is arrested at the erythroid colony-forming unit stage of hematopoietic development. At that stage, the globin genes are not yet expressed, but an upstream, erythroid-specific folate receptor (FR) gene is active. Because antibodies that can distinguish the two H3 variants do not behave well in chromatin immunoprecipitation (ChIP) assays, we made use of expression vectors in which H3 or H3.3 had been fused to a sequence coding for a Flag element. Evidence from other laboratories suggested that these tagged versions of the histones would be incorporated equivalently to the untagged histones (5).

Fig. 1.
H3 variants in the chicken folate receptor (FR) and β-globin locus. (A) Locations of the upstream regulatory elements of the FR gene (HSA, large blue arrow) and HS′ (small blue arrow), the insulator element (HS4), and globin locus control ...

6C2 cell lines were stably transformed with these plasmids, and ChIP assays were carried out with antibodies to Flag. In these studies, we used a method (which we have described in ref. 9) in which formaldehyde cross-linking is not introduced. Nuclei are digested with micrococcal nuclease and fractionated on a sucrose gradient. The mono- and dinucleosome fractions are retained for immunoprecipitation followed by measurement by real-time PCR, permitting analysis at high resolution. Furthermore, there is evidently no need to correct for varying nucleosome density in the regions being studied: Results presented below reflect the abundance of all nucleosomes from that genomic site that carry the modification or variant being probed. Where necessary, as shown below, it is also possible to calculate the local nucleosome density from such data.

The distribution of DNA sequences in the immunoprecipitate was measured across the entire β-globin locus. Incorporation of H3-Flag was, as expected, fairly uniform across the entire FR β-globin domain (Fig. 1B): There are no sites with statistically significant preferred incorporation. In contrast, the map of H3.3-Flag distribution (Fig. 1C) reveals clearly preferred sites of incorporation. The largest of these are localized at positions 5.6/6.2 and 7.3/7.8 kb on the globin locus map. These correspond to HSA, a DNase I-hypersensitive site containing upstream regulatory elements of the FR gene, and HS′, a second and weaker hypersensitive site that appears to be connected with FR expression (10). Examination of the distribution of H3.3-Flag closer to the FR gene reveals that incorporation is confined to these upstream regulatory elements; as measured at five individual sites spanning the gene, there is no incorporation over the transcribed region (Fig. 1D). There are also smaller peaks at 25.7 and 34.7 kb, the positions, respectively, of the globin locus control region (LCR) HS3 element and of a CR1 chicken repeat sequence element (Fig. 1C).

We also introduced H3.3-Flag into HD24 cells, a chicken erythroleukemia line arrested at the erythroid burst-forming unit stage, earlier than 6C2. These cells express neither FR nor globin genes (10, 11). The pattern of incorporation of H3.3-Flag is fairly similar to that seen with 6C2 cells, especially at HSA (Fig. 2B). The smaller peak of incorporation at 35 kb is preserved, but in HD24 cells, the other small peak (at 25 kb in 6C2) is now shifted to 27.6 kb so that it lies over the LCR element HS2 rather than HS3 (Fig. 2B). These results show that H3.3 incorporation over the FR upstream regulatory region and the LCR element of β-globin are not coupled to transcription of the genes. We note that a small peak of preferential incorporation of H3 is also seen at 27.6 kb (Fig. 2A), although its deviation from the mean incorporation level is not as great.

Fig. 2.
ChIP analysis of H3-Flag and H3.3-Flag incorporation. (A) HD24 cells. Axes are same as those shown in Fig. 1. (B) Comparison of H3.3-Flag distribution in HD24 cells, in which FR is not expressed, and 6C2 cells, in which FR is expressed. y axis shows normalized ...

Incorporation of H3 Variants at Other Loci. Given the unusual distribution of H3.3-Flag over the FR gene (i.e., confinement to the upstream region), we also extended our measurements of H3.3-Flag deposition to a number of other genes expressed in our cell lines. These genes were selected as described in Materials and Methods, and they were all expressed in 6C2 cells as confirmed by RT-PCR measurements (see below). In particular, we measured incorporation over both the upstream regulatory regions and transcribed regions of these genes. Primers and probes to transcribed regions were in all but one case directed to exons rather than introns (Fig. 5, which is published as supporting information on the PNAS web site). As shown in Fig. 3, various combinations of patterns are observed: Genes can preferentially incorporate H3.3-Flag over regulatory regions, over the regulatory and coding regions, or over neither. Among the genes we surveyed, the only gene other than FR that showed incorporation of H3.3-Flag over an upstream regulatory region but not over the gene body was VEGF-D (Fig. 3); coding region incorporation was also quite weak in the case of β-spectrin. It is important to note that in the case of VEGF-D, incorporation of H3.3-Flag is found at a considerable distance (–397 bp) from the transcription start site, but not closer (–79 bp) to the start site. This distribution is similar to that shown in Fig. 1 for the FR gene, in which H3.3-Flag is concentrated at HSA and HS′, both some distance from the transcription start site. In principle, the absence of incorporation over the transcribed regions of FR and VEGF-D could be due simply to the absence of nucleosomes. However, direct measurement of the concentration of those sequences in the pentanucleosome fraction from a micrococcal nuclease digest shows that the representation of intronic sequences is similar to the genomic abundance for both FR and VEGF-D (respectively, Fig. 6 A and B and Supporting Materials and Methods, which are published as supporting information on the PNAS web site). Other genes for which we measured incorporation far enough upstream showed the presence of H3.3 at distal promoter elements and coding regions. These genes include IFNAR2 (IFN-α/β receptor 2), GAPDH, β-actin, PAI (plasminogen activator inhibitor), FOG (friend of GATA), and histone H5, where an upstream enhancer at base pair –2159 (12) contains H3.3-Flag (Fig. 3). One gene, DAP (death-associated protein), showed no detectable incorporation of H3.3-Flag at any of the three loci we examined, although transcripts from this gene could be detected (see Fig. 7, which is published as supporting information on the PNAS web site). It may be that the locus we examined is not far enough (base pair –239) from the transcription start site.

Fig. 3.
Distribution of H3.3-Flag over distal and proximal promoter elements and transcribed regions of a variety of genes. Anti-Flag immunoprecipitation was followed by PCR of the DNA by using TaqMan probes and primers shown in Table 3. Regions surveyed were ...

Histone Modification and H3 Variant Incorporation. There has been considerable interest in the relationship between the presence of H3.3 in the histone octamer and the extent of histone modifications typically associated with transcriptional activity, such as H3 acetylation and K4 methylation. Earlier data (13) suggested that there are only low levels of these modifications over the transcribed region of the FR gene. Higher-resolution studies of FR (Fig. 8A, which is published as supporting information on the PNAS web site) show that there is a pronounced gradient of decreasing modification extending 3′ of the promoter, so that the histones within the final exon are completely unmodified. It is interesting that acetylation levels are already close to zero at the beginning of the first intron, whereas K4 methylation extends somewhat farther into the gene. Similar behavior is observed for VEGF-D (Fig. 8B). Gradients of histone acetylation and K4 methylation have been reported by others (8, 14) and appear to be a common phenomenon of eukaryotic transcription. However, it is surprising that the coding regions of these active genes have no detectable incorporated H3.3.

We examined the distribution of histone modifications on the other genes in our survey. Here we made use of the same probes that had been used to detect H3 variant distributions. Some of these probes (for FOG, β-actin, β-spectrin) are located near the 5′ ends of the genes, whereas others are located farther downstream, where lower levels of modification could be expected. However, there was no simple correlation between modification and H3.3 incorporation (Fig. 4). In the case of PAI, for example, there is a high level of H3.3-Flag incorporation in the transcribed region near the 3′ end of the gene, which shows no H3 acetylation or K4 methylation. The same pattern is seen at the coding region of IFNAR2, where H3.3-Flag is highly incorporated toward its 3′ end, but there are no apparent histone modifications at the same site. Conversely, the DAP gene, as noted above, shows no H3.3-Flag incorporated at the upstream sites we examined, but these sites have high levels of histone modification. Similarly, at the proximal promoter of VEGF-D, the histones are strongly modified but H3.3-FLAG is not detected. There is, thus, no invariable relationship between H3 modification and the presence of H3.3-Flag.

Fig. 4.
Histone H3 K9/K14 diacetylation and H3 K4 methylation at the same sites shown in Fig. 3. ChIP was carried out with antibodies to diacetylated H3 or dimethyl K4 H3. Open bars, no-antibody control; filled bars, dimethyl K4 H3; gray bars, diacetylated H3. ...

Effects of H3 Variants on Gene Expression. We next examined the effect of overexpression of these variants on expression of the genes which we surveyed for variant incorporation. We had observed that overexpression of Flag-tagged (as well as GFP-tagged) H3.3 could, in some cases, affect gene expression levels (data not shown). To eliminate possible contributions from the tags to gene expression, we turned to experiments using untagged H3 and H3.3. As shown in Fig. 7 and summarized in Fig. 9A, which is published as supporting information on the PNAS web site, untagged H3.3 expression results in up-regulation by at least 2-fold of VEGF-D as well as FR, but reduces expression of both DAP and PAI. Of particular interest is the behavior of VEGF-D and FR, which show decreased expression in the presence of untagged H3 and increased expression with untagged H3.3. We note that these two genes share similar patterns of H3.3 incorporation (completely absent from transcribed regions). All of these responses to exogenous expression of the untagged H3 variants suggest that H3.3 is at least in some cases an active participant in the regulation of expression (see Discussion).


H3.3 Incorporation Occurs at Upstream Regulatory Elements in the Absence of Transcription. There is strong evidence for the incorporation of histone variant H3.3 into nucleosomes during the resting stages of the cell cycle. It has been suggested that this incorporation arises as a consequence of the replacement necessary after the displacement of nucleosomes during transcription; recent data obtained in Drosophila support this conclusion (6, 15). Because histone modifications have been well documented at the chicken FR/globin locus, it seemed important to examine the incorporation of histone H3 variants at these sites. For this purpose, we chose 6C2 cells, which are arrested at a developmental stage in which the FR gene is active and the embryonic globin genes are about to become active, and HD24 cells, representing an earlier stage at which neither FR nor globin genes are active.

Our first observation was that in both kinds of cells, the region upstream of the FR gene (corresponding to hypersensitive sites HSA and HS′) is a site of intense H3.3-Flag incorporation (Figs. (Figs.11 and and2).2). This is true despite the fact that FR is not expressed in HD24 cells. However, the entire transcribed region of the FR gene shows no sign whatsoever of H3.3-Flag incorporation (Fig. 1C). We also observed H3.3-Flag incorporation over elements of the β-globin LCR in both kinds of cells. These data indicated that there was no necessary relationship between transcriptional activity per se and H3.3 incorporation. The data further suggested that regulatory elements such as HSA and the LCR might be preferred sites of incorporation. These results are partially consistent with a quite recent report from Chow et al. (7) that promoters are the principal sites of H3.3 incorporation in a mouse pre-B-cell line. However, our data suggest a more complex relationship between H3.3 exchange and gene activity, inasmuch as some of the sites of H3.3 incorporation are associated with genes that will be active only at some later developmental stage. Furthermore, many of these sites are not at the proximal promoter, but at distal promoter elements, enhancers, or LCR elements farther removed from the gene. That is true not only for the β-globin locus but also for a number of other genes (Fig. 3). For example, the actin gene distal promoter, located 516 bp upstream of the transcription start site, has an even higher level of H3.3-Flag incorporation than the proximal promoter and coding region.

The majority of genes in our survey did not follow a pattern of preferential incorporation into upstream elements. Instead, they showed H3.3-Flag incorporation over the coding region of the gene as well as upstream regulatory regions. Genes with these patterns of incorporation were also observed in B cells by Chow et al. (7), but they concluded that only a small number of genes behaved in this way. We find, in contrast, that at least 6 of the 10 genes we examined exchanged H3.3 into nucleosomes over their coding regions. This exchange may reflect different patterns of activity in the two kinds of cells used in these studies. Our results are also consistent with those obtained in Drosophila (6, 15). Schwartz and Ahmad (15) found preferential incorporation of H3.3-GFP at heat shock genes and throughout large transcription units, notably at large ecdysone-responsive genes. Genome-wide H3.3 profiling for Drosophila genes (6) showed that H3.3 was prominent in the coding region in active gene and also enriched far upstream of transcribed regions. However, neither experiment addressed the question of whether incorporation occurred at regulatory elements not associated with high levels of transcriptional activity. As we have shown, these elements also may be sites of H3.3 incorporation.

It is striking that the highest levels of H3.3 incorporation at distal promoter elements, upstream enhancers, and other regulatory elements occur relatively distant from the genes. This finding may reflect intergenic transcriptional activity at these sites as found, for example, in the human β-globin intergenic regions (16) or may simply reflect the activity of histone remodeling complexes that are localized at such sites. The replacement of H3 by H3.3 at these sites may reflect the activity of histone remodeling complexes capable of mobilizing nucleosomes. We suggest that such activity may occur well before the genes themselves are turned on. It does not seem that high levels of transcription are necessary for replacement of H3 by H3.3, although low levels of transcription through these regulatory regions may well occur and be part of their maintenance mechanism. It is more difficult to explain why genes that are fully active, such as the FR and VEGF-D genes in 6C2 cells, concentrate H3.3 only over the HSA and HS′ and not over the coding region, despite the fact that the coding region is occupied by nucleosomes. We note the obvious fact that in those cases where there is no histone H3.3 deposition over the coding region, there is unlikely to have been appreciable histone displacement during transcription.

Lack of Correlation Between H3.3 Distribution and H3 Acetylation or K4 Methylation. We find that at sites enriched in H3.3, particularly toward the 3′ ends of genes (PAI, IFNAR2), there is no detectable acetylation or K4 methylation (Figs. (Figs.33 and and4),4), whereas some sites of marked histone modification (DAP, VEGF-D) have no H3.3 (a summary of all these data is provided in Fig. 9B). There is, therefore, no straightforward correlation between the distribution of these histone modifications and that of H3.3. Our results are partially consistent with the recently published observations of Wirbelauer et al. (8), who showed that, although in the genes they studied there was a decreasing gradient of histone acetylation and H3 K4 methylation extending 3′ from the promoter region (14), the distribution of H3.3 was uniform over the entire gene.

It had been suggested (17) that H3.3 might be modified before it was incorporated, making modification a consequence of H3.3 incorporation. The results presented here, however, would seem to indicate that acetylation or K4 methylation of H3 does not require the presence of H3.3. Conversely, the presence of H3.3 does not guarantee that the site of incorporation will carry modified H3, although in the latter case it could be argued that acetylation has been lost subsequent to incorporation.

Selective Sensitivity of Some Genes to Overexpression of H3 Variants. As part of this study, we were interested in the effects of overexpression of the H3 variants on transcriptional activity. Exogenous expression of untagged H3 and H3.3 had varied effects on expression of β-spectrin, DAP, and PAI. However, among the individual genes we examined for effects of H3 variant expression, the behavior of FR and VEGF-D was particularly interesting. In both cases, expression of untagged histone H3 caused decreased expression, and H3.3 expression resulted in increased expression. At these two loci, incorporation is confined to upstream regulatory elements. How could overexpression of H3 affect FR expression? The available evidence suggests that H3 can be deposited on chromatin only during the S phase of the cell cycle. This reaction requires both the presence of a nucleosome assembly complex containing CAF1 and DNA that is either undergoing replication or has acquired single-strand breaks (18). Endogenous H3 is made only during S phase, but H3 transcribed from the introduced transgene is probably not under this control. Overexpression of tagged H3 in Drosophila leads to a slight increase in H3 signal with increased RNA polymerase II density, and it was suggested (6) that a small amount of replication-independent H3 incorporation could be occurring, although its incorporation into chromatin is normally S phase-specific. H3.3, in contrast, must be delivered to DNA by a complex containing HIRA (5). This process can occur at all stages of the cell cycle, including S phase, but if there were a direct competition between H3 and H3.3, it could presumably occur only during S phase. We suggest that during S phase, the “excess” H3 is incorporated into nucleosomes in place of H3.3, and that in genes where H3.3 incorporation is confined to the promoter, this is sufficient to inhibit gene expression. In contrast, expression of H3.3 stimulates expression of FR and VEGF-D. Implicit in this model is the idea that H3.3, at least in the case of genes of this kind, has a specific regulatory function and does not merely substitute for H3 because it is available throughout the cell cycle. Perhaps in this case, H3.3 recruits, or is susceptible to, higher levels of activating histone modifications than H3. Co-deposition of H3.3 and other activated histones might also occur. An alternative model would involve an indirect effect in which H3 incorporation elsewhere in the genome affected expression at the subset of genes where H3.3 incorporation is confined to the upstream regulatory region. Although this is a more elaborate model, it also implies a difference in function for H3 and H3.3. Regardless of the model, the fact that overexpression of H3 can have an effect opposite to overexpression of H3.3 shows that the two are not in all cases biologically functionally interchangeable.

We have observed different combinations of incorporation patterns of H3.3 over regulatory and coding regions of individual genes. This observation suggests that H3.3 may be introduced into chromatin at various stages in the cell cycle with different functional consequences. It is possible that in some cases, H3.3 deposition is a passive phenomenon responsive to other reactions, whereas in other situations, such as that discussed above, it may play a more specific regulatory role. In some cases, the sites of “positive” histone modifications are also sites of H3.3 incorporation, but this is not always true. Further study of the effects of this and other histone variants on gene expression may allow us ultimately to distinguish various mechanisms of promoter activation.

Materials and Methods

Plasmid Constructs. Plasmids containing H3-Flag, H3.3-Flag, wild-type H3, or H3.3 were generated by inserting the fragment of H3.1 or H3.3 cDNA with or without C-terminal Flag and hemagglutinin epitope (HA) tags from pOZ constructs (kindly provided by Yoshihiro Nakatani, Dana–Farber Cancer Institute) into XhoI and XbaI sites, respectively, of the pcDNA3.1/Hygro vector (Invitrogen).

Cell Culture and Stable Transfection. HD24 and 6C2 cells were maintained as described (10) and were stably transfected with the DNA constructs by using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's protocol. Individual hygromycin-resistant colonies were picked after 2–3 weeks of culture in the culture medium plus 2% Methocel and 2,000 (for 6C2 cells) or 1,500 units/ml (for HD24 cells) hygromycin (Calbiochem) and expanded in hygromycin-containing medium (800 units/ml for 6C2 cells and 400 units/ml for HD24 cells) for an additional 6 days. Expression of Flag-tagged proteins was monitored by Western blotting. Expression of untagged histones was detected by RT-PCR. Cells were maintained in logarithmic-phase growth for all experiments. Separate experiments (not shown) with Flag-H3.3-transformed cells suggest that under our conditions, exogenous H3.3 represents about 10% of total H3 species. If the ratio of H3 to H3.3 in these cells is similar to that observed in Drosophila (4), exogenous H3.3 would be about 30% of all H3.3 present in the cell.

ChIP Analysis. Di- and mononucleosomes were prepared by micrococcal nuclease digestion followed by sucrose gradient purification. These procedures and ChIP assays for distribution of Flag-tagged H3.3 or H3 and for the modifications of core histones were performed as described in ref. 9 by using 100 μl of anti-Flag M2 affinity gel (Sigma, catalog no. A 2220), 25 μg of antibodies specific for histone H3 acetyl K9/acetyl K14 (Upstate Biotechnology, Lake Placid, NY, catalog no. 06-599), 20 μl of antibodies specific for dimethyl K4 histone H3 (Upstate Biotechnology, catalog no. 07-030). A detailed description of the methods used for calculation can be found in ref. 9. It is important to note that this method does not involve protein cross-linking, and that the purification of mono- and dinucleosomes for the analysis means that the ChIP data represent the fraction of nucleosomes with that modification at the particular site being measured.

Choice of Genes for Study. Six of the genes shown in Figs. Figs.3,3, ,4,4, and 7 were chosen for measurement because they were part of the β-globin domain (FR), because they were erythroid specific (FR, FOG, H5, and β-spectrin), or because they were housekeeping genes (GAPDH and β-actin). An additional four genes (VEGF-D, DAP, IFNAR2, and PAI) were chosen because a preliminary microarray survey in cells overexpressing GFP-tagged H3.3 revealed significant changes in expression of these genes (data not shown).

Primers and TaqMan Probes. Primers and TaqMan probes for ChIP assays and quantitative real-time RT-PCR were designed by using Applied Biosystems primer express software and obtained from Applied Biosystems. The primers and TaqMan probes used are reported in Tables 1–3, which are published as supporting information on the PNAS web site; those associated with the FR/globin locus are described elsewhere (9). For each of the other genes subject to analysis for H3 variant incorporation and for histone modifications, we chose three sites. The first was somewhere in the transcribed region, the second (proximal promoter) was within one nucleosome on either side of the transcription start site, and the third (distal promoter) was usually within two nucleosomes upstream of the start site, except in the case of histone H5, for which we chose the distal enhancer ≈2.1 kb upstream of the gene (12).

Quantitative TaqMan RT-PCR Assay. Total RNA was isolated by using the RNeasy Mini Kit (Qiagen) with on-column DNase digestion and was then treated for 30 min at 37°C with RQ1 DNase (Promega) at a final concentration of 0.1 unit/μl. Equal amounts of RNA were reverse transcribed and were amplified with TaqMan One-Step RT-PCR Master Mix reagents (Applied Biosystems) using specific primers and TaqMan probes (see Table 2). A standard curve for each specific primers and probe set was first obtained by amplification of serially diluted samples, and the concentrations of unknown samples were calculated according to each standard curve.

Supplementary Material

Supporting Information:


We thank Michael Litt and Bonnie Burgess-Beusse for their assistance, Yoshihiro Nakatani and Kami Ahmad for kindly providing reagents, and Michael Krause's laboratory for DNA sequencing. We also acknowledge Miklos Gaszner, Suming Huang, and Vesco Mutscov for their critical reading of our manuscript and all other members of the Felsenfeld laboratory for all their help. This research was supported by the Intramural Research Program of the National Institutes of Health (National Institute of Diabetes and Digestive and Kidney Diseases).


Conflict of interest statement: No conflicts declared.

Abbreviations: FR, folate receptor; DAP, death-associated protein; PAI, plasminogen activator inhibitor; ChIP, chromatin immunoprecipitation; LCR, locus control region.


1. Henikoff, S., Ahmad, K., Platero, J. S. & van Steensel, B. (2000) Proc. Natl. Acad. Sci. USA 97, 716–721. [PMC free article] [PubMed]
2. Warburton, P. E., Cooke, C. A., Bourassa, S., Vafa, O., Sullivan, B. A., Stetten, G., Gimelli, G., Warburton, D., Tyler-Smith, C., Sullivan, K. F., et al. (1997) Curr. Biol. 7, 901–904. [PubMed]
3. Ahmad, K. & Henikoff, S. (2002) Mol. Cell 9, 1191–1200. [PubMed]
4. McKittrick, E., Gafken, P. R., Ahmad, K. & Henikoff, S. (2004) Proc. Natl. Acad. Sci. USA 101, 1525–1530. [PMC free article] [PubMed]
5. Tagami, H., Ray-Gallet, D., Almouzni, G. & Nakatani, Y. (2004) Cell 116, 51–61. [PubMed]
6. Mito, Y., Henikoff, J. G. & Henikoff, S. (2005) Nat. Genet. 37, 1090–1097. [PubMed]
7. Chow, C. M., Georgiou, A., Szutorisz, H., Maia e Silva, A., Pombo, A., Barahona, I., Dargelos, E., Canzonetta, C. & Dillon, N. (2005) EMBO Rep. 6, 354–360. [PMC free article] [PubMed]
8. Wirbelauer, C., Bell, O. & Schubeler, D. (2005) Genes Dev. 19, 1761–1766. [PMC free article] [PubMed]
9. Litt, M. D., Simpson, M., Recillas-Targa, F., Prioleau, M. N. & Felsenfeld, G. (2001) EMBO J. 20, 2224–2235. [PMC free article] [PubMed]
10. Prioleau, M. N., Nony, P., Simpson, M. & Felsenfeld, G. (1999) EMBO J. 18, 4035–4048. [PMC free article] [PubMed]
11. Felsenfeld, G. (1993) Gene 135, 119–124. [PubMed]
12. Rousseau, S., Asselin, M., Renaud, J. & Ruiz-Carrillo, A. (1993) Mol. Cell. Biol. 13, 4904–4917. [PMC free article] [PubMed]
13. Litt, M. D., Simpson, M., Gaszner, M., Allis, C. D. & Felsenfeld, G. (2001) Science 293, 2453–2455. [PubMed]
14. Kim, T. H., Barrera, L. O., Zheng, M., Qu, C., Singer, M. A., Richmond, T. A., Wu, Y., Green, R. D. & Ren, B. (2005) Nature 436, 876–880. [PMC free article] [PubMed]
15. Schwartz, B. E. & Ahmad, K. (2005) Genes Dev. 19, 804–814. [PMC free article] [PubMed]
16. Ashe, H. L., Monks, J., Wijgerde, M., Fraser, P. & Proudfoot, N. J. (1997) Genes Dev. 11, 2494–2509. [PMC free article] [PubMed]
17. Henikoff, S., Furuyama, T. & Ahmad, K. (2004) Trends Genet. 20, 320–326. [PubMed]
18. Gaillard, P. H., Martini, E. M., Kaufman, P. D., Stillman, B., Moustacchi, E. & Almouzni, G. (1996) Cell 86, 887–896. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...