Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M606803200 on July 29, 2006

J. Biol. Chem., Vol. 281, Issue 41, 30573-30580, October 13, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/41/30573    most recent
M606803200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhao, H.
Right arrow Articles by Dean, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhao, H.
Right arrow Articles by Dean, A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Enhancer Blocking by Chicken beta-Globin 5'-HS4

ROLE OF ENHANCER STRENGTH AND INSULATOR NUCLEOSOME DEPLETION*

Hui Zhao{ddagger}1, AeRi Kim§, Sang-hyun Song{ddagger}, and Ann Dean{ddagger}2

From the {ddagger}Laboratory of Cellular and Developmental Biology, NIDDK, National Institutes of Health, Bethesda, Maryland 20892 and the §Department of Molecular Biology, College of Natural Sciences, Pusan National University, Pusan 609-735 Korea

Received for publication, July 17, 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The 5'-HS4 chicken beta-globin insulator functions as a positional enhancer blocker on chromatinized episomes in human cells, blocking the HS2 enhancer of the human beta-globin locus control region from activating a downstream {epsilon}-globin gene. 5'-HS4 interrupted formation of a domain of histone H3 and H4 acetylation encompassing the 6-kb minilocus and inhibited transfer of RNA polymerase from the enhancer to the gene promoter. We found that the enhancer blocking phenotype was amplified when the insulated locus contained a weakened HS2 enhancer in which clustered point mutations eliminated interaction of the transcription factor GATA-1. The GATA-1 mutation compromised recruitment of histone acetyltransferases and RNA polymerase II to HS2. Enhancer blocking correlated with a significant depletion of nucleosomes in the core region of the insulator as revealed by micrococcal nuclease and DNase I digestion studies. Nucleosome depletion at 5'-HS4 was dependent on interaction of the insulator protein CCCTC-binding factor (CTCF) and was required for enhancer blocking. These findings provide evidence that a domain of active chromatin is formed by spreading from an enhancer to a target gene and can be blocked by a nucleosome-free gap in an insulator.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Domains of open chromatin and condensed heterochromatin must be appropriately circumscribed in the genome to assure correctly regulated gene expression. Chromatin insulators provide a barrier to the spread of repressive heterochromatin, and insulators can also function as enhancer blockers when positioned between an enhancer and a nontarget gene (1, 2). The most well characterized vertebrate insulator is 5'-HS4, which forms the boundary between the chicken beta-globin locus and an upstream 16-kb heterochromatic region (3–5). The barrier function of 5'-HS4 depends on interaction of the protein USF3 and the maintenance of a high level of acetylation within the insulator (6, 7). Interestingly, barrier function at the silenced mating type loci, HMR and HML, involves both histone acetyltransferases and alteration of chromatin structure resulting in a nucleosome gap at the barrier (8, 9). Enhancer blocking by 5'-HS4 depends on interaction of the zinc finger protein CTCF and is separable from barrier activity in ectopic assays (10, 11). It remains unclear how an enhancer blocker exerts its function at the level of chromatin structure.

Studies of the Drosophila gypsy insulator show that enhancer blocking does not affect the activity of an enhancer directly since an enhancer insulated from one gene may still activate another gene from which it is not insulated (12, 13). Instead, the positional nature of enhancer blocking by insulators suggests that they function by interrupting a processive signal from an enhancer to a target gene. Furthermore, the strength of the enhancer to be blocked is an important determinant of insulator function. For example, the ability of the gypsy insulator to block the major fat body enhancer FBE1 was reduced when the enhancer was multimerized and activated higher levels of transcription (14). These studies illustrate the intertwined nature of enhancer and insulator function (15). One common feature of enhancer blocking by gypsy and by 5'-HS4 is the ability of proteins bound to these insulators to tether them to elements of the nuclear architecture, with the presumptive isolation of an enhancer and unrelated genes in different chromatin domains (16, 17).

Chromatin structure and gene transcription are intimately linked. A "histone code" catalogues epigenetic modifications associated with either transcribed, "open" regions of chromatin or silent, "closed" regions (18). Methylation of histone H3 at lysine 9 is associated with silent or condensed chromatin. This histone mark spreads to create heterochromatic domains by a mechanism involving recognition of the methyl mark by the heterochromatin protein HP1, recruitment of SUV39 histone methyltransferase, and propagation of methylation to adjacent Lys-9 residues (19). Methylation of H3 on lysine 4 and acetylation of H3 and H4 are epigenetic modifications associated with active chromatin. Extensive domains marked by these modifications encompass transcribed genes and intergenic regions in loci of developmentally regulated gene families in mammals (20). Increasing evidence suggests that formation of the domains may be a property of remote enhancers and locus control regions (LCRs) to which histone acetyltransferases are recruited.

In mammalian beta-globin loci, domains of acH3, acH4, and H3 K4me encompass the LCR and the active genes, and the domains are extensively transcribed (21–24). In addition, RNA polymerase II (pol II) and histone acetyltransferases (HATs) CBP and p300 are recruited to the remote LCR DNase I hypersensitive sites (HSs) in vivo, and this recruitment is inhibited in cells null for the important LCR binding activator GATA-1 (25–28). We have shown that 5'-HS4 functions as a CTCF-dependent enhancer blocker when interposed between the human beta-globin LCR HS2 enhancer and a linked globin gene on replicating episomes, providing a convenient system to study enhancer-insulator antagonism (28). Here we show that 5'-HS4 is a more effective insulator in the presence of a weakened HS2 enhancer in which the GATA-1 site was mutated. CBP and p300 HATs and RNA pol II are not detected at the mutant HS2, indicating a direct role for GATA-1 in their recruitment. To investigate how the insulator antagonized the enhancer, we analyzed the chromatin structure of 5'-HS4. We found that the core region of 5'-HS4 is significantly depleted of nucleosomes. Formation of the nucleosome gap, similar to enhancer blocking, is CTCF-dependent. These findings support the idea that a domain of active chromatin spreads from an enhancer to a target gene and that enhancer blocking by an insulator involves formation of a nucleosome-free gap that interrupts the spread.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Minichromosome Construction, Cell Culture Conditions, and Transfection—Minichromosomes carrying the {epsilon}-globin gene (3.7-kb EcoRI fragment) and HS2 (374-bp HindIII to XbaI fragment or 1.46-kb KpnI to BglII fragment) have been described (29, 30). The 1.2-kb 5'-HS4 insulator was contained on an XbaI fragment of plasmid pJC5-4 (a kind gift of G. Felsenfeld) and was inserted into pGEMHS2{epsilon} between HS2 and {epsilon}-globin. Clone PCI ({Delta}II) containing the cHS4 core 250-bp element with a 23-bp deletion eliminating the CTCF site (a kind gift of A. West) was used to create cHS4{Delta}CTCF, which contains this deletion in the context of the 1.2-kb insulator fragment. The constructs were cloned into Epstein-Barr virus-based minichromosomes, which are stably maintained in K562 cells at 10–15 copies/cell. Transfection conditions, growth of K562 cells, and isolation of individual clones carrying minichromosomes have been described (29, 31). Between 12 and 15 clones of each type were tested for {epsilon}-globin transcription. Chromatin structure was studied for two or three representative clones, and one or two clones of each type were selected for chromatin immunoprecipitation (ChIP).

RNase Protection Assay—RNA was prepared from 3 x 106 cells of K562 clones carrying minichromosomes, and RNase digestion and gel analysis were performed as recommended by the manufacturer of the reagents (Ambion). The RNA probe used was generated with T7 polymerase from NcoI linearized pBS458 (29), which contains an EcoRV-SspI fragment from the {epsilon}-globin gene promoter and 5'-flanking region. The assay can distinguish endogenous and episomal transcripts since the episomal gene includes a 2-bp mismatch in the 5'-untranslated region, resulting in protection of a smaller fragment in the assay. RNase protection results were normalized to the actin control signal. Endogenous {epsilon}-globin signals varied among K562 cell clones but were unrelated to episomal transcription. The clones used in these studies stably maintained 12–15 copies of the episome.

Preparation of Nuclei and Nuclease Treatment—Nuclei of K562 cell clones (1–1.5 x 108 cells) were suspended in 0.3–0.5 ml of wash buffer and digested with 0, 6, or 12 µg/ml DNase I for 10 min at room temperature in the presence of 3 mM CaCl2 (29). Purified DNA was cut to completion with restriction enzymes and subjected to gel electrophoresis and Southern blotting. Blots were hybridized with probes labeled by random priming to a specific activity of 1–2 x 109 cpm/µg of DNA, and the intensity of bands was quantified with a PhosphorImager (GE Healthcare) using the ImageQuant software.

Chromatin Immunoprecipitation—Histone modifications were studied by ChIP as described (32). Nuclei from 5 x 107 cells were divided into three aliquots, which were digested with increasing concentrations of MNase at 37 °C for 10 min. The digests were combined, and mono- and dinucleosomes were purified on a 5–30% sucrose gradient. Chromatin was precleared by incubation with protein A-agarose. DNA was prepared from a sample of precleared chromatin and used as the "input" sample. Antibodies to acetylated H3 were incubated with chromatin for 2 h at 4 °C with gentle rocking. 100 µl of protein A-agarose was added, and incubation continued overnight. The protein A-agarose was washed, and immune complexes were eluted. DNA from input and bound samples was purified by phenol/chloroform extraction and ethanol precipitation. DNA samples were quantified using PicoGreen fluorescence (Molecular Probes, Eugene, OR).

For analysis of pol II interaction, 2.5 x 107 cells were cross-linked with 0.4% formaldehyde for 10 min at room temperature (33). The reaction was terminated by the addition of glycine to 0.125 M for 5 min at room temperature. Nuclei were prepared, and sonication was performed on ice to produce 200–500-bp fragments. Soluble chromatin was then diluted to 5 ml and incubated with 500 µl of protein A-agarose overnight at 4 °C. DNA was prepared from a sample of precleared chromatin and used as the input sample. pol II antibodies were incubated with chromatin for 3 h at 4 °C. 50 µl of protein A-agarose was added, and incubation continued for 2 h. The protein A-agarose was washed, and the immune complexes were eluted. Cross-links were then reversed, and samples were digested with proteinase K and DNA was purified by phenol/chloroform extraction and ethanol precipitation. ChIPs using antibodies to CBP, p300, H3, USF1, and USF2 were carried out using similar methods except that cells were cross-linked with 1% formaldehyde and sonication was followed by MNase digestion to produce primarily mononucleosome-sized fragments (28, 34).

Antibodies—Anti-acH3 (catalog number 06-599) was obtained from Upstate, Lake Placid, NY. pol II (catalog number sc-899), CBP (catalog number sc-369), p300 (catalog number sc-584), USF1 (catalog number sc-229), and USF2 (catalog number sc-862) antibodies were obtained from Santa Cruz Biotechnology, Santa Cruz, CA. The pol II antibody is against the N-terminal and recognizes both phosphorylated and unphosphorylated forms of the protein. Anti-H3 (ab1791) was obtained from AbCam Ltd., Cambridge, UK.

Real-time PCR and Data Analysis—Differences in DNA enrichment for histone ChIP samples were determined by real-time PCR on 1-ng samples of DNA using the ABI Prism 7900 (PE Applied Biosystems) (32). The threshold was set to cross a point at which PCR amplification was linear, and the number of cycles (Ct) required to reach the threshold was collected and analyzed using Microsoft Excel. The -fold difference of a given target sequence precipitated by the anti-histone antibodies was determined by comparing the amount of target sequence in the immunoprecipitated fraction with the amount of target sequence in input DNA. The relative enrichment of a given target sequence was then obtained by normalizing the -fold difference of the sequence by the -fold difference obtained for a primer in the endogenous inactive beta-globin gene. A similar analysis was carried out for the formaldehyde cross-linked chromatin samples, except that 2.5% of the precipitated sample DNA and 0.02% of the input DNA were used for the real-time PCR.

Primers and Taqman Probes—Primers and Taqman probes, selected using PE Applied Biosystems Primer Express software, were obtained from Invitrogen and PE Applied Biosystems, respectively. Amplicons were designed to be less than 147 bp with an average size of 86 bp. The primers and Taqman probe for cHS4 FIV were as follows: forward primer, 5'-CGG GAT CGC TTT CCT CTG A-3'; reverse primer, 5'-GGC ATA GGG GGT CCA CAG A-3'; TaqMan probe, 5'-6FAM-CGC TTC TCG CTG CTC TTT GAG CCT G-TAMRA (7). Other primers used have been described (28, 30).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
To study enhancer-insulator antagonism in a chromatin environment, we have exploited an episomal system; insulator function is maintained on extra-chromosomal elements (35–37). The episomes link the HS2 enhancer of the human beta-globin locus control region to the embryonic {epsilon}-globin gene and recapitulate salient features of their chromatin structure in chromosomes. The 5'-HS4 insulator blocked activation of {epsilon}-globin on episomes in a positional manner, i.e. only when placed between the enhancer and the gene (28). Transcription of the {epsilon}-globin gene was reduced 2.5-fold, an amount comparable with that seen in other systems for this insulator fragment (see for example Ref. 38).

Enhancer Blocking by 5'-HS4 Is Amplified by a Weakened Enhancer—In Drosophila, the effectiveness of an insulator block is dependent upon the strength of the enhancer-target gene interaction (14). Since the enhancer blocking we had observed with 5'-HS4 was not complete, we decided to exploit this phenomenon by making use of a previously characterized mutant HS2 in which clustered point mutations had destroyed the GATA-1 binding site (31). The mutation does not notably affect the nuclease sensitive structure of HS2, but it compromises promoter remodeling and reduces transcription of the linked {epsilon}-globin gene to about 20% of the wild type level.

We compared {epsilon}-globin transcription in uninsulated loci with the levels of transcription in the presence of the chicken 5'-HS4 insulator interposed between the gene and either a wild type or a GATA-1 mutant HS2 enhancer. RNase protection experiments showed reduction of the high level of transcription stimulated by HS2 to 40% by cHS4 (Fig. 1A) (28). However, when the wild type enhancer was replaced by the GATA-1 mutant enhancer, transcription was almost completely blocked, indicating that no enhancer activity bypasses the insulator block. Thus, 5'-HS4 is more effective when the enhancer is weakened by a mutation that prevents GATA-1 interaction there.

Since transcription of {epsilon}-globin was strongly inhibited, it was of interest to determine the effect of the GATA-1 mutant HS2 on the distribution of pol II in the model locus. ChIP was carried out using formaldehyde cross-linked K562 cells carrying insulated and uninsulated loci. The structure of the loci and the location of the TaqMan probes used in the quantitative PCR analysis are shown in Fig. 1B. In accord with earlier results, enhancer blocking by 5'-HS4 resulted in diminution of pol II at the gene promoter, consistent with reduced transcription, and increased pol II detection in HS2 and at nearby sequences (Fig. 1C) (28). The results also show that accumulation of pol II at HS2 does not take place if GATA-1 cannot interact there. This result shows that GATA-1 is directly involved in recruitment or stabilization of pol II at HS2 and is consistent with studies in GATA-1 null cells (26).

GATA-1 Mutation in HS2 Diminishes H3 Acetylation in the Locus and HAT Recruitment to HS2—The HS2 enhancer is required for formation of a domain of H3 and H4 acetylation across the {epsilon}-globin gene and sequences between the gene and the enhancer, although HS2 sequences themselves are not enriched in acetylated histones due to nucleosome depletion there (30). Fig. 2A illustrates, as reported, the reduced acetylation in the locus caused by the cHS4 insulator (28). When the GATA-1 site in HS2 was mutated, sequences downstream of the cHS4 core showed further reduction of H3 acetylation. This is particularly of note at sequences in the 3' of the insulator, more than 600 bp distant from the insulator core, and at sequences within the {epsilon}-globin gene, although the low level of modification at the promoter is not affected. Interestingly, histone acetylation at the 5'-HS4 core itself is unaffected by the enhancer mutation consistent with the ability of the 5'-HS4 core to maintain a high level of local histone acetylation (6, 32).

We and others have observed recruitment of CBP and/or p300 to beta-globin LCR HSs (27, 28). Interposing the 5'-HS4 insulator between HS2 and the target {epsilon}-globin gene reduced detection of the HATs, but they were still present at HS2 (Fig. 2, B and C) (28). However, when the GATA-1 site was mutated, the HATs were not recruited to HS2, indicating the direct involvement of GATA-1 in their recruitment and/or stabilization at the remote HS2 enhancer. GATA-1 and CBP occupancy at HS3 also coincided in chromosomes (27). Taken together, the data presented in Fig. 2 suggest a model in which HATs are recruited to the remote enhancer in a GATA-1-dependent fashion and then form a domain of acetylation in the model locus by a processive mechanism that an insulator can block.

A Nucleoprotein Complex at cHS4 Is Insufficient for Enhancer Blocking—The interruption of a processive signal from an enhancer to a gene is consistent with the positional nature of enhancer blocking. In one view, such a signal could be interrupted by a nucleoprotein complex that lies between the enhancer and gene and physically blocks propagation of the signal (39). The 5'-HS4 insulator is known to bind multiple factors at the five footprinted regions of the insulator core, among which are CTCF and USF1/2 (7, 10). CTCF, which binds to footprint II, is critical for enhancer blocking, and our deletion studies showed that all of the effects of 5'-HS4 on transcription and chromatin modifications in the model HS2/globin gene locus were abrogated upon deletion of the CTCF site (28). USF1/2 interacts at footprint IV and is involved in the barrier activity of HS4 (7). Thus, we were interested in whether USF1/2 was present on 5'-HS4 chromatin after deletion of the CTCF site and loss of enhancer blocking. If so, it would indicate that a nucleoprotein complex per se was insufficient to block enhancer action.


Figure 1
View larger version (48K):
[in this window]
[in a new window]
  		FIGURE 1.
Accumulation of pol II at HS2 in insulated loci is dependent on GATA-1. A, Shown are RNase protection results for several K562 clones carrying GATA-1-mutated HS2 and insulated from {epsilon}-globin by 5'-HS4 (HS2{Delta}G[HS4]{epsilon}). Controls include non-insulated loci with intact (HS2{epsilon}) or mutated (HS2{Delta}G{epsilon}) enhancers and HS2[HS4]{epsilon} insulated locus with an intact enhancer. {epsilon}M, minichromosomal {epsilon}-globin mRNA; {epsilon}, endogenous {epsilon}-globin mRNA; ac, actin mRNA. The graph illustrates the average (±S.E.) for between 6 and 15 K562 cell clones containing different minichromosomes when compared with the wild type HS2{epsilon} construct each normalized to the actin loading control. Clones carried similar numbers of minichromosomes. B, illustration of the structure of the wild type (WT) uninsulated locus and the chicken 5'-HS4-containing insulated locus (cHS4). Beneath the picture, the positions of TaqMan probes used for real-time PCR are shown. C, ChIP experiments were performed using K562 cells carrying uninsulated (WT) and insulated loci with or without the HS2 GATA-1 mutation (cHS4, cHS4{Delta}GATA). Chromatin was prepared from 0.4% formaldehyde cross-linked K562 cells carrying the different constructs, and sonicated chromatin was subjected to ChIP using an antibody to pol II. The relative enrichment for each primer pair was determined (see "Experimental Procedures"). The results for three chromatin preparations each assayed in duplicate or triplicate (±S.E.) are presented. Mid, cHS4 middle; IVR, intervening region, {epsilon} pro, {epsilon} promoter; {epsilon} int 2, {epsilon} intron 2; beta ex 2, beta exon 2.

 
We performed ChIP studies on insulated loci using antibodies to USF1 and USF2 and analyzed the results by quantitative real-time PCR amplification of the bound and input material using a probe centered on footprint IV. Fig. 3 shows that USF1/2 can be detected at similar levels at cHS4 both before and after deletion of the CTCF site. Interestingly, USF1/2 is detected at the promoter of the {epsilon}-globin gene when enhancer blocking is lost (cHS4{Delta}CTCF). USF is also a functional component of the HS2 enhancer (40–42), and its detection at the promoter of this gene could result from the resumption of enhancer-promoter communication after deletion of the CTCF site in the insulator. In any case, the results indicate that a nucleoprotein complex formed on the insulator, at least one involving USF factors, is insufficient to block enhancer action.

Chromatin Structure of 5'-HS4 Is Largely Determined by Interaction of CTCF—The above results prompted us to consider whether perturbation of chromatin structure at 5'-HS4 might be critical for enhancer blocking. Recent data indicate that in yeast, the propagation of repressive histone modifications is blocked, at least in part, by a nucleosome gap (9). To analyze chromatin structure at 5'-HS4 in the presence or absence of the CTCF site, we used DNase I hypersensitivity assays. Nuclei were prepared and digested with varying amounts of DNase I. After purification of the DNA, digestion with EcoRV and BglII restriction enzymes was used to produce a parent fragment. Fig. 4 shows that strong DNase I HSs exist at both HS2 and 5'-HS4 in the insulated locus. Strikingly, when the CTCF site was deleted, the DNase HS at 5'-HS4 was essentially lost, whereas HS2 was unaffected. Thus, CTCF is required to disrupt the chromatin structure of 5'-HS4, and to a large extent, CTCF determines the chromatin structure of 5'-HS4.


Figure 2
View larger version (42K):
[in this window]
[in a new window]
  		FIGURE 2.
GATA-1 mutation in HS2 diminishes H3 acetylation within the locus and HAT recruitment to HS2. A, chromatin was prepared from MNase-digested nuclei of K562 cells carrying uninsulated (WT) and insulated loci with or without the HS2 GATA-1 mutation (cHS4, cHS4{Delta}GATA), and ChIP was carried out using antibodies to acH3. The -fold difference and relative enrichment for each primer pair were determined (see "Experimental Procedures"). B, a ChIP experiment was performed as described in the legend for Fig. 1, except that chromatin was fixed by 1% formaldehyde and fragmented by MNase digestion plus sonication. Antibodies to CBP were used. C, ChIP was performed as above with antibodies to p300. The results for three chromatin preparations each assayed in duplicate or triplicate (±S.E.) are presented in each panel. FIV, footprint IV. Other abbreviations are the same as in the legend for Fig. 1.

 


Figure 3
View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 3.
Deletion of the CTCF sites in 5'-HS4 does not affect USF1/2 interaction. ChIP experiments were performed the same as described in the legend for Fig. 2, using antibodies to USF1 and USF2. TaqMan primers used for real-time PCR are as indicated in the legend for Fig. 1 with the addition of a primer and probe set for 5'-HS4 FIV (see "Experimental Procedures"). The results for three chromatin preparations each assayed in duplicate or triplicate (±S.E.) are presented. cHS4 FIV, cHS4 footprint IV; {epsilon} pro, {epsilon} promoter.

 
Enhancer Blocking by cHS4 Is Coincident with a Nucleosome Gap—DNase I HS are often assumed to be nucleosome-free sites that are bound by activators and co-activators. However, we previously observed that the transcribed promoter in the model HS2/{epsilon}-globin locus retained a nucleosome, although it is a DNase I HS, whereas at HS2, a nucleosome is evicted (30). This conclusion rested on analysis of MNase sensitivity throughout the locus. Mono- and dinucleosome-sized MNase-digested chromatin, which served as the input for ChIP studies, was compared with total genomic DNA by real-time PCR. MNase normally cleaves in linker regions between nucleosomes. However, when a nucleosome is absent from the DNA, more extensive digestion of chromatin by MNase occurs. Depletion of MNase-treated material when compared with the genomic material was interpreted to indicate loss of a nucleosome at the sequence corresponding to the probe.

Fig. 5A shows, as expected, that the {epsilon}-globin promoter is not depleted from MNase-digested material, whereas HS2 sequences are 6-fold depleted, consistent with loss of a nucleosome. This depletion is unaffected by the presence of the 5'-HS4 insulator with or without an intact CTCF site. Fig. 5 also shows that the CTCF binding region in 5'-HS4 is very highly depleted (up to 15-fold), indicating a strong sensitivity to MNase digestion. When the CTCF site was deleted from 5'-HS4, this elevated sensitivity was completely lost. This result is consistent with loss of a nucleosome at the 5'-HS4 core dependent upon interaction of CTCF. We note that in Fig. 2, a signal is observed for acH3 at this same location. We believe that this signal results from the retention of a nucleosome on a minority of the templates. In support of this explanation, we observed a very large increase in the acetylation signal at this position when the CTCF site was deleted from 5'-HS4, consistent with our conclusion that this mutation restores a nucleosome (28).

To further assess the presence or absence of a nucleosome at the 5'-HS4 core, we performed ChIP analysis using antibodies to total H3 after cross-linking chromatin with formaldehyde (Fig. 5B). When compared with the silent, brain-specific necdin gene, H3 is somewhat reduced at many positions across the HS2/{epsilon}-globin locus. This phenomenon is observed at the endogenous globin locus in human erythroid cells4 and is consistent with a reduced nucleosome density across active globin loci. However, detection of H3 at HS2 and at 5'-HS4 was 5-fold lower than at the necdin gene control. When the CTCF site was deleted, H3 detection at the 5'-HS4 core returned to a level similar to sites surrounding it. Together, the MNase and H3 ChIP studies indicate that CTCF interaction in 5'-HS4 results in formation of a nucleosome-free gap at the core of cHS4. Thus, the activity of enhancer blocking by 5'-HS4 depends on the disruption of chromatin structure in 5'-HS4 caused by interaction of CTCF.


Figure 4
View larger version (53K):
[in this window]
[in a new window]
  		FIGURE 4.
The 5'-HS4 core DNase I-hypersensitive site is lost after CTCF deletion. DNase I hypersensitivity at the HS2 enhancer and 5'-HS4 insulator was determined by indirect end labeling. Nuclei from K562 cell clones carrying insulated loci with or without CTCF deletion (cHS4, {Delta}CTCF) were incubated with varying amounts of DNase I. DNA was then purified and cut to completion with EcoRV and BglII to yield a parent fragment. Expected subfragments are diagrammed at the top. The black rectangle represents the CTCF site. Representative results are depicted. M, molecular weight markers. RV, EcoRV; BgI, BgIII.

 


Figure 5
View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 5.
MNase sensitivity and exclusion of H3 from the 5'-HS4 core require CTCF. A, real-time PCR was performed on genomic DNA and MNase-digested chromatin input DNA used in an acetylated H3 ChIP experiment from K562 cells carrying uninsulated (WT) and insulated loci with or without CTCF mutation (cHS4, cHS4{Delta}CTCF). The -fold difference between genomic DNA and MNase-digested input DNA was determined. B, ChIP experiments were performed as described in the legend for Fig. 1, except that antibody to histone H3 was used and -fold difference was normalized to the necdin gene signal to obtain the relative enrichment.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Enhancer Blocking and Enhancer Strength—Experiments in Drosophila lead to the proposal that enhancer strength is an important modulator of enhancer blocking by insulators (14). In those experiments, enhancer strength was augmented by multimerization, effectively increasing the number of transcription activator sites, and enhancer blocking was weakened as measured by increased transcriptional output of the target gene (14). We further tested this proposal by, inversely, weakening the beta-globin HS2 enhancer by mutation of the GATA-1 motif, which is known to decrease enhancer function (31). Although the 5'-HS4 insulator reduced transcriptional output of the insulated {epsilon}-globin gene to about 40%, in the context of the weakened enhancer, 5'-HS4 was able to block transcription almost completely, consistent with the original proposal.

GATA-1 is a zinc finger transcription factor required for erythroid differentiation (43). Motifs for this activator are found in the beta-globin LCR HSs and the globin gene promoters, although only a subset appears to be occupied in vivo (26, 44). GATA-1 does not play a major role in chromatin structure formation at the HS2 enhancer (31). Instead, a very important role for GATA-1 appears to be in recruitment and/or stabilization of the HATs CBP/p300 and pol II at HS2. Assays in cell lines null for GATA-1 show that CBP and GATA-1 co-occupy LCR HS3 and that recruitment of pol II to HS1-HS3 depends on expression of GATA-1 (26, 27). Our studies strongly support the notion that GATA-1 directly recruits HATs and pol II to the LCR HSs and that this remote recruitment is important for target gene function.

Formation of a Domain of Histone Acetylation—The positional nature of enhancer blocking implies that an insulator interrupts a processive signal from the enhancer to the promoter. Our observation that 5'-HS4 reduced histone acetylation between the HS2 enhancer and the {epsilon}-globin gene is consistent with histone acetylation being an example of such a signal. That acetylation is further reduced when recruitment of HATs to HS2 is impaired also suggests that the domain of histone modifications spreads from the enhancer toward the gene rather than resulting from random acetylation (45). In support of this proposal, it has been demonstrated that histone acetylation spreads unidirectionally from the HNF4{alpha} enhancer to the gene as a function of time after induction (46), whereas at the growth hormone and major histocompatibility complex class II loci, domains of modified histones are proposed to spread bidirectionally from the LCRs (47, 48).

Although in globin loci at the adult stage, the acetylated LCR and distant adult beta-globin genes are separated by an unmodified domain encompassing the silent embryonic genes (49, 50), chromatin looping creates a pseudo-continuous relationship between the LCR and the adult genes (51). Thus, the current data accommodate models in which domains of active chromatin are propagated from an LCR (21) or, alternatively, form as sequences between an LCR and an active gene are "reeled" past LCR-associated chromatin-modifying complexes (52).

How could a domain of acetylated histone modifications be propagated? The formation of condensed heterochromatin marked by H3 Lys-9 methylation provides one possible model (19). H3 Lys-9 methylation is recognized by the N-terminal chromodomain of HP1 at the same time as its c-terminal chromo shadow domain interacts with the specific histone methyltransferase required for this modification. These interactions position the transferase to modify a subsequent nucleosome and spread the heterochromatin mark. Similarly, lysine acetylation of histones generates recognition sites for bromo-domain proteins, examples of which include the CBP and p300/CBP-associated factor (PCAF) HATs, but further details of these interactions remain to be elucidated (53). Alternatively, histone modifications could be established via HAT activity associated with elongating pol II (54). Extensive non-genic transcription is known to occur in the endogenous human beta-globin locus and on transgenes in mice (23, 24, 55), but the association of HATs with elongating pol II in intergenic regions has not been demonstrated. Quite recently, it has been shown that CTCF sites in the DM1 locus decrease the extension of antisense (but not sense) intergenic transcripts by an unknown mechanism (60). In this case, it is not known whether the CTCF sites coincide with DNase HS, but this seems likely. The mechanism underlying this block will be important to investigate further.

Role of CTCF in Enhancer Blocking by 5'-HS4—How does an insulator block a processive signal? In one view, an insulator might form a passive physical road block consisting of a nucleoprotein complex (39). We have shown that a nucleoprotein complex, at least one involving USF1/2, is insufficient to mediate enhancer blocking by cHS4. Alternatively, the insulator could serve as a point of recruitment for enzymatic activities that covalently modify histones or alter nucleosomes. We observed a correspondence between enhancer blocking and formation of a nucleosome-free gap in the 5'-HS4 insulator. Interestingly, recent data show that barrier activity in yeast is at least partially dependent on a nucleosome gap (9) and that the gap created by a DNA sequence known to disfavor nucleosome formation could function as an efficient boundary of silent chromatin (8). Most recently, these authors have shown that the spread of histone acetylation from a tethered HAT can be opposed by excluding nucleosomes (56), consistent with our observations. However, the CTCF insulator in the H19 imprinting control region appears to function by a mechanism that does not include formation of a gap since the CTCF sites map to linker regions that exist within an array of highly positioned nucleosomes (57).

We think that nucleosome disruption that results from CTCF interaction at 5'-HS4 contributes to enhancer blocking by interrupting the spread of acetylation. A region free of nucleosomes would disfavor propagation of histone modification and could halt the advance of the domain. Evidence exists supporting the proposal that a modified histone domain is important to transcription of LCR target genes, which is consistent with the inhibitory role of an enhancer blocker (28, 58). CTCF sites also coincide with DNase I HSs that exist at the 5' and 3' flanks of mouse and human globin loci, yet these regions have varying enhancer blocking capability (59). Although it is not known whether these sites represent nucleosome gaps, if so, this might indicate that a nucleosome gap is not sufficient for strong enhancer blocking. Tethering of an insulator to a subnuclear structure can be viewed as related to the roadblock mechanism or the altered chromatin mechanism of insulator function. Since nucleolar tethering by 5'-HS4 is lost when the CTCF site is mutated (17), we cannot rule out the possibility that the nucleosome gap we observe is secondary to a tethering event, which serves to isolate regulatory elements or genes either upstream or downstream of an insulator in separate domains.

Complex Activities of the Chicken 5'-HS4 Insulator—Barrier activity and enhancer blocking by 5'-HS4 are separable activities in ectopic assays, but it is reasonable to suppose that at the endogenous site in the chicken beta-globin locus, 5'-HS4 might function both to stop encroachment of heterochromatin on the 5' border of the locus and to restrict LCR activity to the globin genes therein (1). If, as we propose, enhancer blocking involves interrupting the spread of positive-acting histone modifications from an enhancer, then both blocking and barrier activity might involve opposing the spread of (different) histone modifications.

5'-HS4 is a focus of H3 and H4 histone acetylation, which was proposed to underlie the antagonism of the insulator to heterochromatic histone modifications (32). The HATs PCAF and CBP are recruited to 5'-HS4 by USF1/2, which binds to footprint IV of the insulator (7). However, the role of this recruitment in enhancer blocking is not clear because footprint IV can be deleted without affecting this activity (10). It is also unclear why there is a different outcome when HATs are recruited to an enhancer compared to an insulator. In yeast, a tethered HAT is sufficient to create a sizeable, bidirectional domain of acetylated histones (61). Although histone acetylation does not appear to spread from the 5'-HS4 insulator, this issue remains to be addressed directly.

In our experiments, loss of enhancer blocking by mutation of the 5'-HS4 CTCF site coincided with loss of a nucleosome gap. However, this gap is unlikely to play a role in barrier activity of 5'-HS4 since deletion of footprint II including the CTCF site does not affect barrier activity. This supposition is interesting in light of experiments showing that barrier activity at the silent HMR locus in yeast is dependent on both histone acetylation and formation of a nucleosome gap (9). Thus, some aspects of insulator function in different species may vary.


   FOOTNOTES
 
* This work was supported by a grant from the Intramural Program of the NIDDK, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 Present address: Division of Basic Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N, Seattle, WA 98109. Back

2 To whom correspondence should be addressed: NIDDK, National Institutes of Health, Bldg. 50, Rm. 3154, 50 South Dr., MSC 8028, Bethesda, MD 20892. Tel.: 301-496-6068; Fax: 301-496-5239; E-mail: anndean{at}helix.nih.gov.

3 The abbreviations used are: USF, upstream transcription factor; CTCF, CCCTC-binding factor; LCR, locus control region; HAT, histone acetyltransferase; CREB, cAMP-response element-binding protein; CBP, CREB-binding protein; ChIP, chromatin immunoprecipitation; HS, hypersensitive site; FAM, carboxyfluorescein; TAMRA, carboxy-tetramethylrhodamine. Back

4 A. Kim, C. M. Kiefer, and A. Dean, unpublished results. Back


   ACKNOWLEDGMENTS
 
We thank Adam West and Gary Felsenfeld for plasmids and Catherine Farrell for critical comments.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Oki, M., and Kamakaka, R. T. (2002) Curr. Opin. Cell Biol. 14, 299–304[CrossRef][Medline] [Order article via Infotrieve]
  2. West, A. G., Gaszner, M., and Felsenfeld, G. (2002) Genes Dev. 16, 271–288[Free Full Text]
  3. Hebbes, T. R., Clayton, A. L., Thorne, A. W., and Crane-Robinson, C. (1994) EMBO J. 13, 1823–1830[Medline] [Order article via Infotrieve]
  4. Prioleau, M. N., Nony, P., Simpson, M., and Felsenfeld, G. (1999) EMBO J. 18, 4035–4048[CrossRef][Medline] [Order article via Infotrieve]
  5. Chung, J. H., Whiteley, M., and Felsenfeld, G. (1993) Cell 74, 505–514[CrossRef][Medline] [Order article via Infotrieve]
  6. Mutskov, V. J., Farrell, C. M., Wade, P. A., Wolffe, A. P., and Felsenfeld, G. (2002) Genes Dev. 16, 1540–1554[Abstract/Free Full Text]
  7. West, A. G., Huang, S., Gaszner, M., Litt, M. D., and Felsenfeld, G. (2004) Mol. Cell 16, 453–463[CrossRef][Medline] [Order article via Infotrieve]
  8. Bi, X., Yu, Q., Sandmeier, J. J., and Zou, Y. (2004) Mol. Cell. Biol. 24, 2118–2131[Abstract/Free Full Text]
  9. Oki, M., and Kamakaka, R. T. (2005) Mol. Cell 19, 707–716[CrossRef][Medline] [Order article via Infotrieve]
  10. Bell, A. C., West, A. G., and Felsenfeld, G. (1999) Cell 98, 387–396[CrossRef][Medline] [Order article via Infotrieve]
  11. Recillas-Targa, F., Pikaart, M. J., Burgess-Beusse, B., Bell, A. C., Litt, M. D., West, A. G., Gaszner, M., and Felsenfeld, G. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 6883–6888[Abstract/Free Full Text]
  12. Cai, H., and Levine, M. (1995) Nature 376, 533–536[CrossRef][Medline] [Order article via Infotrieve]
  13. Scott, K. S., and Geyer, P. K. (1995) EMBO J. 14, 6258–6267[Medline] [Order article via Infotrieve]
  14. Scott, K. C., Taubman, A. D., and Geyer, P. K. (1999) Genetics 153, 787–798[Abstract/Free Full Text]
  15. Zhao, H., and Dean, A. (2005) Biochem. Cell Biol. 83, 516–524[CrossRef][Medline] [Order article via Infotrieve]
  16. Gerasimova, T. I., Byrd, K., and Corces, V. G. (2000) Mol. Cell 6, 1025–1035[CrossRef][Medline] [Order article via Infotrieve]
  17. Yusufzai, T. M., Tagami, H., Nakatani, Y., and Felsenfeld, G. (2004) Mol. Cell 13, 291–298[CrossRef][Medline] [Order article via Infotrieve]
  18. Jenuwein, T., and Allis, C. D. (2001) Science 293, 1074–1080[Abstract/Free Full Text]
  19. Elgin, S. C., and Grewal, S. I. (2003) Curr. Biol. 13, R895–R898[CrossRef][Medline] [Order article via Infotrieve]
  20. Dean, A. (2006) Trends Genet. 22, 38–45[CrossRef][Medline] [Order article via Infotrieve]
  21. Forsberg, E. C., and Bresnick, E. H. (2001) BioEssays 23, 820–830[CrossRef][Medline] [Order article via Infotrieve]
  22. Bulger, M., Sawado, T., Schubeler, D., and Groudine, M. (2002) Curr. Opin. Genet. Dev. 12, 170–177[CrossRef][Medline] [Order article via Infotrieve]
  23. Ashe, H. L., Monks, J., Wijgerde, M., Fraser, P., and Proudfoot, N. J. (1997) Genes Dev. 11, 2494–2509[Abstract/Free Full Text]
  24. Gribnau, J., Diderich, K., Pruzina, S., Calzolari, R., and Fraser, P. (2000) Mol. Cell 5, 377–386[CrossRef][Medline] [Order article via Infotrieve]
  25. Johnson, K. D., Christensen, H. M., Zhao, B., and Bresnick, E. H. (2001) Mol. Cell 8, 465–471[CrossRef][Medline] [Order article via Infotrieve]
  26. Johnson, K. D., Grass, J. A., Boyer, M. E., Kiekhaefer, C. M., Blobel, G. A., Weiss, M. J., and Bresnick, E. H. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 11760–11765[Abstract/Free Full Text]
  27. Letting, D. L., Rakowski, C., Weiss, M. J., and Blobel, G. A. (2003) Mol. Cell. Biol. 23, 1334–1340[Abstract/Free Full Text]
  28. Zhao, H., and Dean, A. (2004) Nucleic Acids Res. 32, 4903–4919[Abstract/Free Full Text]
  29. Gong, Q. H., McDowell, J. C., and Dean, A. (1996) Mol. Cell. Biol. 16, 6055–6064[Abstract]
  30. Kim, A., and Dean, A. (2003) Mol. Cell. Biol. 23, 8099–8109[Abstract/Free Full Text]
  31. McDowell, J. C., and Dean, A. (1999) Mol. Cell. Biol. 19, 7600–7609[Abstract/Free Full Text]
  32. Litt, M. D., Simpson, M., Recillas-Targa, F., Prioleau, M. N., and Felsenfeld, G. (2001) EMBO J. 20, 2224–2235[CrossRef][Medline] [Order article via Infotrieve]
  33. Johnson, K. D., and Bresnick, E. H. (2002) Methods (Orlando) 26, 27–36
  34. Morshead, K. B., Ciccone, D. N., Taverna, S. D., Allis, C. D., and Oettinger, M. A. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 11577–11582[Abstract/Free Full Text]
  35. Parnell, T. J., and Geyer, P. K. (2000) EMBO J. 19, 5864–5874[CrossRef][Medline] [Order article via Infotrieve]
  36. Kanduri, C., Holmgren, C., Pilartz, M., Franklin, G., Kanduri, M., Liu, L., Ginjala, V., Ulleras, E., Mattsson, R., and Ohlsson, R. (2000) Curr. Biol. 10, 449–457[CrossRef][Medline] [Order article via Infotrieve]
  37. Steinwaerder, D. S., and Lieber, A. (2000) Gene Ther. 7, 556–567[CrossRef][Medline] [Order article via Infotrieve]
  38. Filippova, G. N., Thienes, C. P., Penn, B. H., Cho, D. H., Hu, Y. J., Moore, J. M., Klesert, T. R., Lobanenkov, V. V., and Tapscott, S. J. (2001) Nat. Genet. 28, 335–343[CrossRef][Medline] [Order article via Infotrieve]
  39. Fourel, G., Magdinier, F., and Gilson, E. (2004) BioEssays 26, 523–532[CrossRef][Medline] [Order article via Infotrieve]
  40. Bresnick, E. H., and Felsenfeld, G. (1993) J. Biol. Chem. 268, 18824–18834[Abstract/Free Full Text]
  41. Lam, L. T., and Bresnick, E. H. (1996) J. Biol. Chem. 271, 32421–32429[Abstract/Free Full Text]
  42. Elnitski, L., Miller, W., and Hardison, R. (1997) J. Biol. Chem. 272, 369–378[Abstract/Free Full Text]
  43. Cantor, A. B., and Orkin, S. H. (2002) Oncogene 21, 3368–3376[CrossRef][Medline] [Order article via Infotrieve]
  44. Im, H., Grass, J. A., Johnson, K. D., Kim, S. I., Boyer, M. E., Imbalzano, A. N., Bieker, J. J., and Bresnick, E. H. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 17065–17070[Abstract/Free Full Text]
  45. Struhl, K. (1998) Genes Dev. 12, 599–606[Free Full Text]
  46. Hatzis, P., and Talianidis, I. (2002) Mol. Cell 10, 1467–1477[CrossRef][Medline] [Order article via Infotrieve]
  47. Elefant, F., Cooke, N. E., and Liebhaber, S. A. (2000) J. Biol. Chem. 275, 13827–13834[Abstract/Free Full Text]
  48. Krawczyk, M., Peyraud, N., Rybtsova, N., Masternak, K., Bucher, P., Barras, E., and Reith, W. (2004) J. Immunol. 173, 6200–6210[Abstract/Free Full Text]
  49. Forsberg, E. C., Downs, K. M., Christensen, H. M., Im, H., Nuzzi, P. A., and Bresnick, E. H. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 14494–14499[Abstract/Free Full Text]
  50. Bulger, M., Schubeler, D., Bender, M. A., Hamilton, J., Farrell, C. M., Hardison, R. C., and Groudine, M. (2003) Mol. Cell. Biol. 23, 5234–5244[Abstract/Free Full Text]
  51. Tolhuis, B., Palstra, R. J., Splinter, E., Grosveld, F., and de Laat, W. (2002) Mol. Cell 10, 1453–1465[CrossRef][Medline] [Order article via Infotrieve]
  52. West, A. G., and Fraser, P. (2005) Hum. Mol. Genet. 14, Spec. No. 1, R101–R111[Abstract/Free Full Text]
  53. Yang, X. J. (2004) BioEssays 26, 1076–1087[CrossRef][Medline] [Order article via Infotrieve]
  54. Orphanides, G., and Reinberg, D. (2000) Nature 407, 471–475[CrossRef][Medline] [Order article via Infotrieve]
  55. Kim, A., and Dean, A. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 7028–7033[Abstract/Free Full Text]
  56. Yu, Q., Sandmeier, J., Xu, H., Zou, Y., and Bi, X. (2006) J. Biol. Chem. 281, 3980–3988[Abstract/Free Full Text]
  57. Kanduri, M., Kanduri, C., Mariano, P., Vostrov, A. A., Quitschke, W., Lobanenkov, V., and Ohlsson, R. (2002) Mol. Cell. Biol. 22, 3339–3344[Abstract/Free Full Text]
  58. Masternak, K., Peyraud, N., Krawczyk, M., Barras, E., and Reith, W. (2003) Nat. Immunol. 4, 132–137[CrossRef][Medline] [Order article via Infotrieve]
  59. Farrell, C. M., West, A. G., and Felsenfeld, G. (2002) Mol. Cell. Biol. 22, 3820–3831[Abstract/Free Full Text]
  60. Cho, D. H., Thienes, C. P., Mahoney, S. E., Analau, E., Filippova, G. N., and Tapscott, S. J. (2005) Mol. Cell 20, 483–489[CrossRef][Medline] [Order article via Infotrieve]
  61. Chiu, Y. H., Yu, Q., Sandmeier, J. J., and Bi, X. (2003) Genetics 165, 115–125[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles: