HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 2007-0497v1 26/5/1174    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Atkinson, S. P. Articles by Armstrong, L. Search for Related Content PubMed PubMed Citation Articles by Atkinson, S. P. Articles by Armstrong, L.

STEM CELL GENOMICS AND PROTEOMICS

Epigenetic Marking Prepares the Human HOXA Cluster for Activation During Differentiation of Pluripotent Cells

^aNorth East Institute for Stem Cell Research and
^bInstitute of Human Genetics, University of Newcastle upon Tyne, International Centre for Life, Newcastle upon Tyne, United Kingdom;
^cThe Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, United Kingdom

Key Words. HOXA genes • Embryonic stem cells • Embryonal carcinoma cells

Correspondence: Correspondence: Lyle Armstrong, B.Sc., Ph.D., International Centre for Life, Central Parkway, Newcastle upon Tyne NE1 3BZ, United Kingdom. Telephone: 00-44-191-241-8695; Fax: 00-44-191-241-8666; e-mail: Lyle.Armstrong{at}ncl.ac.uk

Received on June 25, 2007; accepted for publication on February 14, 2008.

First published online in STEM CELLS EXPRESS  February 21, 2008.

	ABSTRACT

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
Activation of Hox gene clusters is an early event in embryonic development since individual members play important roles in patterning of the body axis. Their functions require precise control of spatiotemporal expression to provide positional information for the cells of the developing embryo, and the manner by which this control is achieved has generated considerable interest. The situation is different in pluripotent cells, where HOX genes are not expressed but are held in potentio as bivalent chromatin domains, which are resolved upon differentiation to permit HOX cluster activation. In this study we have used differentiation of the pluripotent embryonal carcinoma cell line NTera2SP12 and the human embryonic stem cell line H9 to examine epigenetic changes that accompany activation of the HOXA cluster and show that specific genomic loci are marked by lysine methylation of histone H3 (H3K4 tri- and dimethyl, H3K9 trimethyl) and acetylation of histone H4 even in the undifferentiated cells. The precise locations of such modified histones may be involved in controlling the colinear expression of genes from the cluster.

Disclosure of potential conflicts of interest is found at the end of this article.

	INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
The Homeobox, or Hox, genes were discovered from mutations occurring in the fruit fly Drosophila melanogaster that resulted in the appearance of body parts in inappropriate locations, such as the growth of legs where antennae should be (antennapedia). The Greek term for this interchange of body parts is "homeosis"; hence, the genes responsible for such mutations became known as homeotic selector genes and, more recently, Homeobox genes [1]. Homeobox genes show a high degree of evolutionary conservation across eukaryotic species [2], highlighting their primitive origin and their importance as master regulators of development. The Hox genes are expressed throughout embryonic development in a highly cocoordinated manner and continue to be expressed in the vast majority of adult tissues in organ-specific patterns.

In humans, the HOX genes are organized into four clusters (HOXA–HOXD) located on different chromosomes (7p15, 17q21.2, 12q13, and 2q31.). Each cluster contains 9–11 member genes encoding relatively small gene products containing a highly conserved 60-amino-acid region (the homeobox), with DNA-binding activity that contributes to their actions as transcription factors [3]. One of the major functions of Hox genes seems to be the formation of the body plan during embryonic development [4]; hence, HOX expression begins during early gastrulation, when the principal body axis is established [1].

A remarkable feature of all four HOX clusters is colinear gene activation [5], through which HOX genes are expressed in a precise spatiotemporal pattern along the anterior-posterior axis that corresponds to their position within the cluster. Expression of genes at the 3' end begins at an early stage of development in the anterior segments of the embryo, whereas those at the 5' end are upregulated later, in the posterior segments [6]. The mechanisms controlling this are poorly understood, although they are probably related to timing of activation, since Hox genes in ectopic locations maintain their spatial expression pattern, but their temporal expression is misregulated [7]. This also suggests that the clustered organization of HOX genes may be an essential part of the control mechanism, particularly since current evidence supports the theory that colinear expression requires coordinated changes in the higher-order chromatin structure across the cluster [8].

Hox gene transcription is regulated through enhancer elements at sites distal to the coding or promoter sequences of individual genes [9], with regulation by retinoic acid being of particular importance, and retinoic acid response elements (RAREs) have been identified in several HOX genes [10]. Mutation of the RAREs of murine HoxA1 and HoxB1 results in developmental abnormalities of the hindbrain and cranial nerves similar to knockouts of the genes themselves [11]. The retinoid-binding sequences are direct repeats of AGGTCA that are separated by two or five bases [12] and are found at multiple locations in all the HOX clusters [13]. Other factors are known to regulate Hox genes, typical examples being Pbx, Meis, Prep [14, 15], Krox20 [16], and the Sox-Oct family [17], but it has been hypothesized that the underlying mechanism controlling binding of many of these trans-acting factors to the Hox clusters is a sequential "opening" of the chromatin structure that allows access to the genes toward the 5' end according to a specific schedule [18, 19]. This is supported by data that demonstrate decondensation of the HoxB and HoxD clusters during the differentiation of mouse embryonic stem cells and their ability to "loop out" from their normal chromosome territories in a manner that seems to correlate with their temporal expression in the developing embryo [20–22].

Activation of the HOX clusters in this manner is probably an important mechanism for spatiotemporal expression control, but we understand very little about how this process is initiated. Some gene clusters, such as the β-globin loci, require a locus control region to which globin genes are recruited [23], but there is little evidence to suggest that similar elements control colinear HOX expression. The requisite molecular machinery must be able to have physical contact with the chromatin at a nucleosome level, which implies that such chromatin must already be in a relatively open conformation. However, with the exception of the studies on mouse embryonic stem cells, many of these data are derived from differentiated somatic cell types, in which many genes will be silenced by the formation of heterochromatin [24]. In pluripotent cells the chromatin architecture is rather different [25, 26]. Actively expressed genes (such as Oct4 or Nanog) in embryonic stem cells are characterized by acetylation and mono-, di-, and trimethylation of specific lysines in the N-terminal tails of histones H3 and H4 [26]. In addition, currently inactive genes whose expression will not be required until later stages of development also posses such "activating" histone modifications, although these are repressed by inactivating modifications, such as trimethylation of lysine 27 on histone H3, which is the basis of the "bivalent" chromatin status for many genes implicated in early development [27]. The members of all four HOX clusters are characterized by the presence of bivalent chromatin domains and as a consequence are mostly repressed in pluripotent cells. Embryonic stem cells have the capacity to either self-renew or differentiate into a wide range of cell types, making them an excellent model system to study the mechanisms that initiate HOX gene expression. Embryonal carcinoma (EC) cells may be regarded as the pluripotent stem cells of the teratocarcinoma from which they are derived [28] and are capable of differentiation into many different cell types in vitro. In this respect they may be considered the malignant counterparts of embryonic stem cells [29, 30]. Differentiation of EC cells to the neuroectodermal lineage may be induced by the addition of all-trans-retinoic acid (10^–8 M) to the growth medium, and this is also known to activate HOX gene expression [31–33]. EC cells can be grown in culture with less stringent requirements for growth factors to prevent differentiation; therefore, they are a useful counterpart to embryonic stem cells for studying HOX gene activation, albeit with some caution because of the known aneuploidy of EC cell lines, such as NTera2 SP12, used in this work.

It has been hypothesized that positions marked by activating histone modifications in developmentally specific genes function as recruitment sites for transcription factors during activation of those genes [34]. We show here that the human HOXA cluster is marked by several regions that are highly enriched in dimethylated H3K4, acetylated H3, and acetylated H4 in pluripotent EC cells and that these exist in pluripotent human embryonic stem cells (hESC), albeit at lower levels. These regions expand considerably when temporal HOXA gene expression is induced by retinoic acid in EC cells and by spontaneous hESC differentiation; however, there is no evidence to suggest that these regions serve to recruit transcription factors to the genomic regions responsible for controlling HOXA cluster expression. Examination of the epigenetic mechanisms that control not only the repression of the HOXA cluster in pluripotent cells but also the collinear activation of these genes in differentiating cells is therefore of considerable interest.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
Cell Culture
Human EC cells (NTera2-SP12) were differentiated for more than 4 weeks with 10^–8 M all-trans-Retinoic Acid (Sigma-Aldrich, Gillingham, U.K., http://www.sigmaaldrich.com) in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Paisley, U.K., http://www.invitrogen.com) with 10% fetal bovine serum (FBS) (Invitrogen). hESC (H9; WiCell Research Institute, Madison, WI, http://www.wicell.org) were grown on Matrigel-coated plates in Knockout DMEM (Invitrogen) with 20% Knockout serum replacement and 20 ng/ml basic fibroblast growth factor (Invitrogen) conditioned by mitotically inactivated mouse embryonic fibroblast monolayers. hESC differentiation was induced by removal of colonies from the Matrigel plates and transfer to six-well tissue culture plates (Iwaki, Camlab, Cambridge, U.K., http://www.camlab.co.uk) containing differentiation medium (Knockout DMEM plus 20% FBS). The colony fragments adhered to the culture surface and grew to form confluent monolayers of differentiating cells.

Chromatin Immunoprecipitation with Microarray Hybridization
Human EC cells (NTera2-SP12), human embryonic stem cells (WiCell line H9), and their differentiated progeny (1 x 10⁸ cells) (from three biological replicates) were suspended in serum-free medium (DMEM, Invitrogen) containing 0.37% formaldehyde (VWR, Lutterworth, U.K., http://uk.vwr.com) followed by incubation at ambient temperature (10 minutes). Glycine was added to a final concentration of 0.125 M, and then the cells were collected by centrifugation, washed with phosphate-buffered saline (Gibco, Grand Island, NY, http://www.invitrogen.com), and then resuspended in 1.5 volumes of cell lysis buffer (10 mM Tris-HCl, pH 8.0, 10 mM NaCl, 0.2% Nonidet P40, 10 mM sodium butyrate, 50 µg/ml phenylmethylsulfonyl fluoride [PMSF], 1 µg/ml leupeptin). Cells were incubated on ice (10 minutes) and then centrifuged (2,500 rpm, 5 minutes, 4°C) to collect nuclei, which were resuspended in nuclear lysis buffer (50 mM Tri-HCl, pH 8.1, 10 mM EDTA, 1% SDS, 10 mM sodium butyrate, 50 µg/ml PMSF, 1 µg/ml leupeptin). After incubation on ice (10 minutes), 0.72 ml of immunoprecipitation (IP) dilution buffer (20 mM Tris-HCl, pH 8.1, 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.01% SDS, 10 mM sodium butyrate, 50 µg/ml phenylmethanesulphonylfluoride, 1 µg/ml leupeptin) was added, and the chromatin content was sheared by sonication to fragments of approximately 300–600 base pairs (bp) in length. Chromatin was precleared using rabbit IgG (Upstate Biotechnology, Dundee, U.K., http://www.upstate.com) and protein G agarose (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) followed by centrifugation (3,000 rpm, 2 minutes, 4°C). Aliquots of 1.35 ml were combined with antibodies directed against various modified histone proteins and incubated overnight at 4°C with rotation. The samples were treated further with protein G agarose suspension (3 hours, 4°C) followed by washing (2x IP wash buffer 1 [20 mM Tris-HCl, pH 8.1, 250 mM LiCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS] and then 1x IP wash buffer 2 [same composition as IP wash 1 except that Triton X-100 and SDS were replaced by 1% Nonidet P40 and 1% deoxycholic acid]). The immune complexes were eluted (100 mM NaHCO₃, 1% SDS), treated with RNase A followed by proteinase K, and then isolated by phenol-chloroform extraction. The air-dried DNA samples were resuspended in water (50 µl), labeled, and hybridized to Encode (Sanger Centre, Cambridge, U.K.) polymerase chain reaction (PCR) tiling arrays as described previously [35]. Fluorescent intensities of spots on the arrays were interpreted using ScanArray express software (PerkinElmer UK, Beaconsfield, Buckinghamshire, U.K., http://las.perkinelmer.com), which automatically performs global normalization by correcting the intensity of each spot for variations in the overall intensity of the images with respect to the control image. The values used to generate chromatin immunoprecipitation (ChIP)-on-chip maps were background-corrected globally normalized signal intensities.

Gene Expression Analysis
Total RNA was extracted using TRIzol Reagent (Invitrogen) according to the manufacturer's instructions, and RNA concentration was assessed using a spectrophotometer (ND-1000; NanoDrop, Wilmington, DE, http://www.nanodrop.com). Genomic DNA was degraded by incubation with RQ1 RNase-Free DNase (Promega, Madison, WI, http://www.promega.com) for 30 minutes at 37°C. One microgram of total RNA was reverse-transcribed using Moloney murine leukemia virus reverse transcriptase and random hexamers following the manufacturer's instructions (Promega).

Gene expression was quantified by quantitative polymerase chain reaction (qPCR) on an Opticon2 DNA Engine (MJ Research, Inc., Waltham, MA, http://www.mjr.com) using the SYBR Green qPCR Buffer (Finnzymes, Espoo, Finland, http://www.finnzymes.fi) and is expressed as a value relative to the expression of glyceraldehyde-3-phosphate dehydrogenase. Primers used are noted in supplemental online Table 1. Error values were established from triplication within the qPCR from a representative experiment.

Chromatin Immunoprecipitation Assays
Chromatin immunoprecipitation assays were carried out as previously described [36] to study polymerase II (Pol II) occupancy of the HOXA gene cluster. Briefly, cells were used when at 70%–80% confluence and harvested following the instructions recommended by the kit supplier (Upstate Biotechnology). Sonication was optimized to give chromatin fragments of approximately 500 bp to 1 kilobase in length (6–10-second pulses, with a 20-second rest between pulses, on ice using a Sonicator 3000 [Misonix, Farmingdale, NY, http://www.misonix.com]). The antibodies used were RNA Polymerase II (ab5408; Abcam, Cambridge, U.K., http://www.abcam.com). Resultant DNA from each immunoprecipitate was purified using the QIAquick PCR Purification Kit (Qiagen, Crawley, West Sussex, U.K., http://www1.qiagen.com). Products from the ChIP assay were quantified by qPCR. The primers used to establish Pol II occupancy for the HOX genes were the same as those used for the expression analysis, as noted in supplemental online Table 1. Additional ChIP primers used are also noted in Table 1. Error values were established from triplication within the qPCR from a representative experiment. Also included in each ChIP experiment was a no-antibody control immunoprecipitate to detect any background, which, if present (as detected by qPCR), was subtracted from the values found for each immunoprecipitate within that experiment.

DNase I Hypersensitivity Assay
DNase I hypersensitivity assays were carried out as previously described [37, 38] and analyzed using qPCR. Briefly, nuclei were extracted from cells (2 x 10⁶ cells for one reaction) and were treated with a range of DNase I concentrations (1,000, 100, 10, and 1 unit per assay) in triplicate for 4 minutes. After proteinase K-SDS digestion, RNase A digestion and phenol:chloroform extraction, the DNA was precipitated and diluted in 50 ml of TE. Each sample was purified using the QIAquick PCR Purification Kit. Each sample was then analyzed by qPCR using primers listed in supplemental online Table 2 against predicted DNase I hypersensitivity sites in the HOXA cluster and standards of undigested genomic DNA of known concentration. Also included was a predicted DNase I-insensitive site as a control.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
ChIP-on-CHIP histone mapping reveals epigenetic marking of the inactive HOXA cluster in undifferentiated embryonal carcinoma cells. Hybridization of DNA obtained from chromatin immunoprecipitation experiments using antibodies directed against various post-translationally modified histone N-terminal tail sequences (acetylated H3 and H4, H3K4 tri- and dimethyl, H3K9 trimethyl) were hybridized to a single-stranded DNA array representing the 44 ENCODE regions (http://www.sanger.ac.uk/PostGenomics/encode/data-access.shtml) [35]. ChIP-on-chip experiments were performed with three biological replicates. Analysis of these data indicates genomic locations that are enriched for particular modified histones, such as tri- or dimethylation of lysine 3 on histone H3, over a large section of the genome in both undifferentiated embryonal carcinoma cells and H9 embryonic stem cells and their differentiated progeny. Specific analysis of the HOXA cluster and associated sequence in EC cells shows the presence of regions of H3K4 trimethylation that span the region from HOXA4 to HOXA11 with enrichment at the 3' end of HOXA4 and HOXA9 and the 5' end of HOXA10 (Fig. 1A). There is minimal enrichment of trimethylated H3K4 in the intervening sequence, although this is still higher than the region spanning HOXA1 to HOXA3 and the intergenic regions immediately outside the HOXA cluster. This enriched H3K4 trimethyl region expands after 10 days of retinoic acid-induced differentiation and increases only slightly by day 19, suggesting activation of the HOXA cluster, although in relation to the level of H3K4 trimethylation of the region by day 28, such increases are relatively modest (note: the scaling on Fig. 1 for day 28 is reduced to an appropriate size). The H9 hESC shows a different response to differentiation. The HOXA cluster during differentiation of hESC does show regions of H3K4 trimethyl enrichment associated with HOXA4 similar to those of undifferentiated EC cells but of greatly reduced intensity (Fig. 1B). The H3K4 trimethyl enrichment associated with the region between HOXA10 and HOXA11 is absent.

View larger version (22K): [in this window] [in a new window]   Figure 1. Distribution of trimethyl histone H3 lysine 4 across the human HOXA cluster during EC cell differentiation (A). The upper panel shows levels of this modified histone in EC cells followed by days 10, 19, and 28 of differentiation induced by ATRA. Chromosomal location is indicated at the top of the diagram, and the relative positions of members of the HOXA cluster are depicted at the bottom. (B): Distribution of trimethyl histone H3 lysine 4 across the human HOXA cluster during embryonic stem cell differentiation. The upper panel shows levels of this modified histone in H9 embryonic stem cells followed by days 10, 19, and 28 of spontaneous differentiation. Chromosomal location is indicated at the top of the diagram, and the relative positions of members of the HOXA cluster are depicted at the bottom. Abbreviations: EC, embryonic carcinoma; hESC, human embryonic stem cells.

View larger version (26K): [in this window] [in a new window]   Figure 2. Distribution of dimethyl histone H3 lysine 4 across the human HOXA cluster during EC cell differentiation (A). The upper panel shows levels of this modified histone in EC cells followed by days 10, 19, and 28 of differentiation induced by ATRA. Chromosomal location is indicated at the top of the diagram, and the relative positions of members of the HOXA cluster are depicted at the bottom. (B): Distribution of dimethyl histone H3 lysine 4 across the human HOXA cluster during embryonic stem cell differentiation. The upper panel shows levels of this modified histone in H9 embryonic stem cells followed by days 10, 19, and 28 of spontaneous differentiation. Chromosomal location is indicated at the top of the diagram, and the relative positions of members of the HOXA cluster are depicted at the bottom. Abbreviations: EC, embryonic carcinoma; hESC, human embryonic stem cells.

View larger version (26K): [in this window] [in a new window]   Figure 3. Distribution of acetylated histone H3 across the human HOXA cluster during EC cell differentiation (A). The upper panel shows levels of this modified histone in EC cells followed by days 10, 19, and 28 of differentiation induced by ATRA. Chromosomal location is indicated at the top of the diagram, and the relative positions of members of the HOXA cluster are depicted at the bottom. (B): Distribution of acetylated histone H3 across the human HOXA cluster during embryonic stem cell differentiation. The upper panel shows levels of this modified histone in H9 embryonic stem cells followed by days 10, 19, and 28 of spontaneous differentiation. Chromosomal location is indicated at the top of the diagram, and the relative positions of members of the HOXA cluster are depicted at the bottom. Abbreviations: EC, embryonic carcinoma; hESC, human embryonic stem cells.

View larger version (27K): [in this window] [in a new window]   Figure 4. Distribution of acetylated histone H4 across the human HOXA cluster during EC cell differentiation (A). Upper panel corresponds to undifferentiated EC cells followed by days 10, 19, and 28 of differentiation induced by ATRA. The scale of the upper panel is expanded since the level of acetylated histone H4 is much lower in the differentiated cells. (B): Distribution of acetylated histone H4 across the human HOXA cluster during embryonic stem cell differentiation. The upper panel shows levels of this modified histone in H9 embryonic stem cells followed by days 10, 19, and 28 of spontaneous differentiation. Abbreviations: EC, embryonic carcinoma; hESC, human embryonic stem cells.

Trimethylation of lysine 9 on histone H3 has been linked to gene repression via its ability to recruit heterochromatin protein 1 [39], so it is interesting that a region of the 5' region of HOXA3 is enriched for H3K9 trimethylation in undifferentiated EC cells (Fig. 5A). A smaller H3K9 trimethyl enrichment is present in the intergenic region between HOXA6 and HOXA7. Interestingly, the chromosomal locations that are enriched for H3K9 trimethylation seem to show low levels of H3K4 dimethylation/trimethylation in the undifferentiated cells (Figs. 1A, 2A). This is particularly true for regions associated with the HOXA4, HOXA6, HOXA9, and HOXA10 loci and perhaps suggests some exclusivity of function between these types of histone modification. The overall level of H3K9 trimethyl modification does not increase significantly upon differentiation, and indeed, the area 3' of HOXA4 seems to diminish slightly, whereas that between HOXA6 and HOXA7 expands (Fig. 5A). The association of seemingly repressive histone modifications with actively transcribing genes has recently been described [40]; in that report, it was suggested that sites enriched in H3K9 trimethylation serve to recruit HP1 in association with transcriptional elongation by phosphorylated RNA polymerase II, although it is not clear why this interaction is required. It is possible that a similar mechanism operates in the activated HOXA cluster, but this cannot explain the presence of the H3K9 trimethyl-enriched area in undifferentiated EC cells. Perhaps this region may serve as a "focus" for the recruitment of HP1/RNA Pol II so that the cluster can be activated rapidly. It is also interesting that the H3K9 trimethyl-enriched area coincides with the H4 acetylation boundary described earlier, suggesting a possible role in temporal expression control of the HOXA cluster.

View larger version (33K): [in this window] [in a new window]   Figure 5. Distribution of trimethyl histone H3 lysine 9 across the human HOXA cluster during EC cell differentiation (A). The upper panel shows levels of this modified histone in EC cells followed by days 10, 19, and 28 of differentiation induced by ATRA. (B): Distribution of trimethyl histone H3 lysine 9 across the human HOXA cluster during embryonic stem cell differentiation. The upper panel shows levels of this modified histone in H9 embryonic stem cells followed by days 10, 19, and 28 of spontaneous differentiation. Abbreviations: EC, embryonic carcinoma; hESC, human embryonic stem cells.

Expression of the HOXA genes in EC cells seems only partly dependent upon the time for which the cells have differentiated (at least for the four time points measured), since this is restricted to HOXA1–HOXA5. There is some variation of the expression levels during differentiation (Fig. 6A), starting with complete absence of expression in the undifferentiated cells (maximum at day 19 and slight decrease by day 28). Interestingly, for HOXA1 and HOXA2, which are outside the histone modification-enriched regions, expression in the differentiated EC cells is low (Fig. 6A), but HOXA3 expression is higher than that of HOXA4. HOXA3 has at least a part of its 3' sequence within the region of H3 and H4 acetylation, which may be linked to its higher expression level, but the occupancy level of RNA polymerase II is higher for HOXA1, HOXA2, and HOXA3 than any of the other members of the cluster at differentiation day 28. Precise expression levels may therefore be controlled by other factors in addition to chromatin conformation in the HOXA cluster. None of the HOXA cluster members are expressed at appreciable levels in the undifferentiated embryonal carcinoma cells, and retinoic acid-induced differentiation is required for their expression. Cell differentiation was confirmed by downregulation of the well-known pluripotency marker OCT4 (Fig. 6B). RNA polymerase II occupancy levels (determined by chromatin immunoprecipitation using antibodies directed against RNA polymerase II) are correspondingly low in the undifferentiated cells and at day 10 of differentiation, despite the increase in HOXA1–HOXA5 expression (Fig. 6A) RNA polymerase occupancy increases substantially at the promoters of these genes. Interestingly, even the nonexpressed members of the cluster show considerable increases in polymerase occupancy, although this is not equivalent to that of HOXA1–HOXA5. Global expression levels of the HOXA cluster genes during H9 hESC differentiation are much lower than for EC cell differentiation (less than 5% when expressed as a percentage of GAPDH expression), although transcripts of all cluster members are detectable at all differentiation stages (Fig. 6A). HOXA gene promoter occupancy by RNA polymerase II also occurs at levels around 10% of those detectable at the same promoter regions throughout EC differentiation, which correlates with the lower expression observed for hESC. Lower gene expression also correlates with the poorer enrichment of histone modification marks generally associated with active gene transcription (H3K4 trimethylation, dimethylation, and H3/H4 acetylation).

View larger version (31K): [in this window] [in a new window]   Figure 6. Changes in EXP of HOXA genes during human EC cell or hESC differentiation. (A): EXP levels of members of the human HOXA cluster in undifferentiated EC/hESC and progeny resulting from 10, 19, and 28 D of differentiation. These were measured by quantitative real-time reverse transcription-polymerase chain reaction (PCR) relative to the EXP level of GAPDH. Values are the result of triplicate PCRs, and errors are plotted as the SD of the three experiments. EXP levels in EC cells are close to zero; hence, they are not observable on the scale required to show EXP levels in the differentiated progeny. Also shown are the levels of RNA Pol II occupancy at the same chromosomal locations amplified by the primer sets used for EXP analysis. (B): Downregulation of the pluripotency-associated gene OCT4 during EC cell/hESC differentiation. Errors are plotted as the SDs of triplicate PCR experiments. Also shown are the levels of RNA Pol II occupancy at the same chromosomal locations amplified by the primer sets used for EXP analysis. (C): DNase I hypersensitivity assays of predicted hypersensitive sites throughout the human HOXA cluster as measured by numbers of real-time PCR cycles required to achieve a threshold amplification level. Errors are plotted as the SDs of triplicate DNase I digests of genomic DNA from EC cells and their differentiated progeny. Abbreviations: D, day(s); EC, embryonic carcinoma; EXP, expression; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Pol II, polymerase II.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
EC cells and human embryonic stem cells are generally regarded as pluripotent cell types; thus, they are valuable systems to study the epigenetic mechanisms used to control gene expression in both pluripotent cells and their differentiated progeny. EC cells are, of course, karyotypically abnormal, so we cannot be certain that their epigenetic regulation mechanisms are identical to those of pluripotent hESC; however, comparison of the two cell types may generate useful data. The HOXA cluster was chosen for further analysis since its pattern of histone modifications seems to be unique in comparison with the other genomic regions included in the ENCODE microarrays. The enrichment of defined regions of the human HOXA cluster in EC cells and H9 embryonic stem cells for specifically modified histones that increases upon differentiation is an interesting observation, but it does not immediately indicate a function for these epigenetic markers. The density of such modified histones is much greater than in other areas of the genome covered by the 44 ENCODE regions (Fig. 7A, 7B), suggesting that they have a role in HOXA cluster regulation. They may maintain HOXA cluster chromatin in an open conformation that is transcriptionally permissive, and the expansion of histone methylation and acetylation during cluster activation may serve to open this conformation yet further to permit access of transcription factors or other proteins that control the activity of HOXA genes, although this would appear to be a greater possibility for EC cell differentiation than hESC differentiation, since the expansion of permissive histone modification is attenuated in the latter. These are in accord with the observations of other workers, although additional studies are required to demonstrate such physical changes to the human HOXA cluster.

View larger version (27K): [in this window] [in a new window]   Figure 7. Distribution of trimethyl histone H3 lysine 4 across the human genomic region separate from the HOXA cluster during EC cell (A) and hESC (B) differentiation. These panels demonstrate the general "punctate" enrichment of this histone modification throughout the genome of EC cells or hESC. Abbreviations: EC, embryonic carcinoma; hESC, human embryonic stem cells.

Cell type-specific HOXA function can be controlled by the presence of retinoic acid, and in this work, differentiation of EC cells was induced by the addition of all-trans-retinoic acid to the culture medium, whereas hESC were allowed to undergo spontaneous differentiation following removal from mouse embryonic fibroblast feeder cells. This protocol was adopted since hESC are capable of multilineage differentiation and HOXA gene expression in the absence of additional ATRA. The human HOXA cluster contains several RAREs [47], with 11 of these lying in the region 5' of HOXA5 [47], although the regions 5' of HOXA4 are not highly enriched for any of the histone modifications examined in undifferentiated EC of hESC. Access to the RAREs may not require histone modifications associated with gene transcription, so their apparent absence could be consistent with ready access to the DNA sequence that forms the response element. Binding of the retinoic acid heterodimer to DNA has been linked to hyperacetylation of histones around the RARE, but neither H3 or H4 acetylation is highly enriched at these positions in undifferentiated EC or hESC. H3 and H4 acetylation increase markedly during EC cell differentiation (but not in the region of the first RARE), whereas hESC differentiation produces little acetylation increase. Nevertheless, HOXA expression does occur in differentiated hESC; thus, histone acetylation is not likely to control accessibility of the two RAREs described.

It is possible that access to the RAREs is controlled by the binding of polycomb repressive complexes to the HOXA cluster in pluripotent cells; this binding has been shown to associate with H3K27 trimethylation [48]. Removal of H3K27 trimethylation in response to differentiation (e.g., loss of OCT4 expression) may either permit or enhance RARE accessibility; however, it is not clear how this can contribute to the regulation of HOX gene colinearity since there is no evidence to suggest progressive H3K27 trimethyl loss that is coordinated to the HOX gene expression profile or selective accessibility to the RAREs. Literature evidence suggests that OCT4 might be involved in recruiting SUZ12 to its target genes to establish the H3K27 trimethyl mark; however, if this is the case, one might expect the H3K27 trimethyl-mediated repression of the HOXA cluster to be removed early in differentiation since OCT4 expression diminishes rapidly (Fig. 6B) for both EC and hESC, and this does not support progressive H3K27 demethylation.

	CONCLUSION

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
Other workers have suggested that narrow regions of histone methylation/acetylation may serve as recruitment sites for specific transcription factors and assembly of the preinitiation complex at the transcription start sites of developmentally specific genes [34]. We could detect only low levels of RNA polymerase II binding in undifferentiated EC or hESC across the HOXA cluster, which is in accord with the mutually exclusive binding of SUZ12 and RNA polymerase II to developmentally specific genes [49]. However, upon differentiation, RNA polymerase II became enriched at the promoters of HOXA1–HOXA3, which were outside the H3K4 dimethylated area; thus, it is unlikely that these serve as preinitiation complex assembly sites. HOX genes need to be expressed at very specific times and locations during development; thus, the fact that the HOXA cluster adopts a complex pattern of H3K4 dimethylated regions compared with the generally punctuate localization of H3K4 dimethylation observed for other areas of the EC and hESC genomes examined in this study (Fig. 7A) may reflect the requirement for precise expression control. The expression of noncoding RNAs from the strand opposite that of the HOXA genes has been suggested as a possible control mechanism [47, 50], and it is interesting to note that the positions of several of the noncoding RNAs described [51] coincide closely with H3K4 dimethyl-enriched regions of the HOXA cluster. Whether or not these regions contribute to the expression of noncoding RNAs is not yet clear but is currently under investigation in our group.

	DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
The authors indicate no potential conflicts of interest.

	ACKNOWLEDGMENTS

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
We are grateful to Stefan Przyborski of Durham University for the kind gift of the embryonal carcinoma cell line Ntera2SP12 and to ONE North East Regional Developmental Agency and Newcastle University for funding this work. S.P.A. and C.M.K. contributed equally to this work.

	REFERENCES

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 

1. Lappin TR, Grier DG, Thompson A et al. HOX genes: Seductive science, mysterious mechanisms. Ulster Med J 2006;75:23–31.[Medline]

2. Santini S, Boore JL, Meyer A. Evolutionary conservation of regulatory elements in vertebrate Hox gene clusters. Genome Res 2003;13:1111–1122.[Abstract/Free Full Text]

3. Lemons D, McGinnis W. Genomic evolution of Hox gene clusters. Science 2006;313:1918–1922.[Abstract/Free Full Text]

4. Akam M. Hox genes and the evolution of diverse body plans. Philos Trans R Soc Lond B Biol Sci 1995;349:313–319.[CrossRef][Medline]

5. Duboule D. Temporal colinearity and the phylotypic progression: A basis for the stability of a vertebrate Bauplan and the evolution of morphologies through heterochrony. Dev Suppl 1994;135–142.

6. Lee AP, Koh EG, Tay A et al. Highly conserved syntenic blocks at the vertebrate Hox loci and conserved regulatory elements within and outside Hox gene clusters. Proc Natl Acad Sci U S A 2006;103:6994–6999.[Abstract/Free Full Text]

7. Kmita M, Duboule D. Organizing axes in time and space; 25 years of colinear tinkering. Science 2003;301:331–333.[Abstract/Free Full Text]

8. Chambeyron S, Da Silva NR, Lawson K et al. Nuclear re-organisation of the Hoxb complex during mouse embryonic development. Development 2005;132:2215–2223.[Abstract/Free Full Text]

9. Maconochie M, Nonchev S, Morrison A et al. Paralogous Hox genes: Function and regulation. Annu Rev Genet 1996;30:529–556.[CrossRef][Medline]

10. Deschamps J, van Nes J. Developmental regulation of the Hox genes during axial morphogenesis in the mouse. Development 2005;132:2931–2942.[Abstract/Free Full Text]

11. Barrow JR, Stadler HS, Capecchi MR. Roles of Hoxa1 and Hoxa2 in patterning the early hindbrain of the mouse. Development 2000;127:933–944.[Abstract]

12. Mangelsdorf DJ, Umesono K, Evans RM. In: Sporn MB, Roberts AB, Goodman DS, eds. The Retinoids: Biology, Chemistry and Medicine. Retinoids in photosensitive systems. New York: Raven Press, 1994:319-350.

13. Langston AW, Thompson JR, Gudas LJ. Retinoic acid-responsive enhancers located 3' of the Hox A and Hox B homeobox gene clusters. Functional analysis. J Biol Chem 1997;272:2167–2175.[Abstract/Free Full Text]

14. Moens CB, Selleri L. Hox cofactors in vertebrate development. Dev Biol 2006;291:193–206.[CrossRef][Medline]

15. Featherstone M. In Lufkin T, ed. Murine Homeobox Gene Control of Embryonic Patterning and Organogenesis. Hox proteins and their cofactors in transcriptional regulation. Amsterdam: Elsevier Science, 2003:1-42.

16. Nonchev S, Maconochie M, Vesque C et al. The conserved role of Krox-20 in directing Hox gene expression during vertebrate hindbrain segmentation. Proc Natl Acad Sci U S A 1996;93:9339–9345.[Abstract/Free Full Text]

17. Di Rocco G, Gavalas A, Popperl H et al. The recruitment of SOX/OCT complexes and the differential activity of HOXA1 and HOXB1 modulate the Hoxb1 auto-regulatory enhancer function. J Biol Chem 2001;276:20506–20515.[Abstract/Free Full Text]

18. Duboule D, Deschamps J. Colinearity loops out. Dev Cell 2004;6:738–740.[CrossRef][Medline]

19. Duboule D. Vertebrate hox gene regulation: Clustering and/or colinearity? Curr Opin Genet Dev 1998;8:514–518.[CrossRef][Medline]

20. Morey C, Da Silva NR, Perry P et al. Nuclear reorganisation and chromatin decondensation are conserved, but distinct, mechanisms linked to Hox gene activation. Development 2007;134:909–919.[Abstract/Free Full Text]

21. Sproul D, Gilbert N, Bickmore WA. The role of chromatin structure in regulating the expression of clustered genes. Nat Rev Genet 2005;6:775–781.[CrossRef][Medline]

22. Chambeyron S, Bickmore WA. Chromatin decondensation and nuclear reorganization of the HoxB locus upon induction of transcription. Genes Dev 2004;18:1119–1130.[Abstract/Free Full Text]

23. Harju S, Navas PA, Stamatoyannopoulos G et al. Genome architecture of the human beta-globin locus affects developmental regulation of gene expression. Mol Cell Biol 2005;25:8765–8778.[Abstract/Free Full Text]

24. Feldman N, Gerson A, Fang J et al. G9a-mediated irreversible epigenetic inactivation of Oct-3/4 during early embryogenesis. Nat Cell Biol 2006;8:188–194.[CrossRef][Medline]

25. Meshorer E, Misteli T. Chromatin in pluripotent embryonic stem cells and differentiation. Nat Rev Mol Cell Biol 2006;7:540–546.[CrossRef][Medline]

26. Freberg CT, Dahl JA, Timoskainen S et al. Epigenetic reprogramming of OCT4 and NANOG regulatory regions by embryonal carcinoma cell extract. Mol Biol Cell 2007;18:1543–1553.[Abstract/Free Full Text]

27. Bernstein BE, Mikkelsen TS, Xie X et al. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 2006;125:315–326.[CrossRef][Medline]

28. Andrews PW. From teratocarcinomas to embryonic stem cells. Philos Trans R Soc Lond B Biol Sci 2002;357:405–417.[Abstract/Free Full Text]

29. Andrews PW, Matin MM, Bahrami AR et al. Embryonic stem (ES) cells and embryonal carcinoma (EC) cells: Opposite sides of the same coin. Biochem Soc Trans 2005;33:1526–1530.[CrossRef][Medline]

30. Andrews PW, Damjanov I, Simon D et al. Pluripotent embryonal carcinoma clones derived from the human teratocarcinoma cell line Tera-2. Differentiation in vivo and in vitro. Lab Invest 1984;50:147–162.[Medline]

31. Przyborski SA, Morton IE, Wood A et al. Developmental regulation of neurogenesis in the pluripotent human embryonal carcinoma cell line NTERA-2. Eur J Neurosci 2000;12:3521–3528.[CrossRef][Medline]

32. Mavilio F, Simeone A, Boncinelli E et al. Activation of four homeobox gene clusters in human embryonal carcinoma cells induced to differentiate by retinoic acid. Differentiation 1988;37:73–79.[CrossRef][Medline]

33. Simeone A, Acampora D, Nigro V et al. Differential regulation by retinoic acid of the homeobox genes of the four HOX loci in human embryonal carcinoma cells. Mech Dev 1991;33:215–228.[CrossRef][Medline]

34. Szutorisz H, Canzonetta C, Georgiou A et al. Formation of an active tissue-specific chromatin domain initiated by epigenetic marking at the embryonic stem cell stage. Mol Cell Biol 2005;25:1804–1820.[Abstract/Free Full Text]

35. Koch CM, Andrews RM, Flicek P et al. The landscape of histone modifications across 1% of the human genome in five human cell lines. Genome Res 2007;17:691–707.[Abstract/Free Full Text]

36. Atkinson SP, Hoare SF, Glasspool R et al. Lack of telomerase gene expression in alternative lengthening of telomere cells is associated with chromatin remodeling of the hTR and hTERT gene promoters. Cancer Res 2005;65:7585–7590.[Abstract/Free Full Text]

37. Schick BP, Petrushina I, Brodbeck KC et al. Promoter regulatory elements and DNase I-hypersensitive sites involved in serglycin proteoglycan gene expression in human erythroleukemia, CHRF 288-11, and HL-60 cells. J Biol Chem 2001;276:24726–24735.[Abstract/Free Full Text]

38. McArthur M, Gerum S, Stamatoyannopoulos G. Quantification of DNaseI-sensitivity by real-time PCR: Quantitative analysis of DNaseI-hypersensitivity of the mouse beta-globin LCR. J Mol Biol 2001;313:27–34.[CrossRef][Medline]

39. Snowden AW, Gregory PD, Case CC, Pabo CO. Gene-specific targeting of H3K9 methylation is sufficient for initiating repression in vivo. Curr Biol 2002;12:2159–2166.[CrossRef][Medline]

40. Vakoc CR, Mandat SA, Olenchock BA et al. Histone H3 lysine 9 methylation and HP1gamma are associated with transcription elongation through mammalian chromatin. Mol Cell 2005;19:381–391.[CrossRef][Medline]

41. Cheung P, Lau P. Epigenetic regulation by histone methylation and histone variants. Mol Endocrinol 2005;19:563–573.[Abstract/Free Full Text]

42. Jørgensen HF, Giadrossi S, Casanova M et al. Stem cells primed for action: Polycomb repressive complexes restrain the expression of lineage-specific regulators in embryonic stem cells. Cell Cycle 2006;5:1411–1414.[Medline]

43. Squazzo SL, O'Geen H, Komashko et al. Suz12 binds to silenced regions of the genome in a cell-type-specific manner. Genome Res 2006;16–900, 890.

44. Strathdee G, Holyoake TL, Sim A et al. Inactivation of HOXA genes by hypermethylation in myeloid and lymphoid malignancy is frequent and associated with poor prognosis. Clin Cancer Res 2007;13:5048–5055.[Abstract/Free Full Text]

45. Hassan NM, Hamada J, Murai T et al. Aberrant expression of HOX genes in oral dysplasia and squamous cell carcinoma tissues. Oncol Res 2006;16:217–224.[Medline]

46. Ohm JE, McGarvey KM, Yu X et al. A stem cell-like chromatin pattern may predispose tumor suppressor genes to DNA hypermethylation and heritable silencing. Nat Genet 2007;39:237–242.[CrossRef][Medline]

47. Sessa L, Breiling A, Lavorgna G et al. Noncoding RNA synthesis and loss of Polycomb group repression accompanies the colinear activation of the human HOXA cluster. RNA 2007;13:223–239.[Abstract/Free Full Text]

48. Kirmizis A, Bartley SM, Kuzmichev A et al. Silencing of human polycomb target genes is associated with methylation of histone H3 Lys 27. Genes Dev 2004;18:1592–1605.[Abstract/Free Full Text]

49. Lee TI, Jenner RG, Boyer LA et al. Control of developmental regulators by Polycomb in human embryonic stem cells. Cell 2006;125:301–313.[CrossRef][Medline]

50. Boyer LA, Lee TI, Cole MF et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 2005;122:947–956.[CrossRef][Medline]

51. Rinn JL, Kertesz M, Wang JK et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 2007;129:1311–1323.[CrossRef][Medline]

This article has been cited by other articles:

				J. M. Palmer, R. M. Perrin, T. R. T. Dagenais, and N. P. Keller H3K9 Methylation Regulates Growth and Development in Aspergillus fumigatus Eukaryot. Cell, December 1, 2008; 7(12): 2052 - 2060. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 2007-0497v1 26/5/1174    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Atkinson, S. P. Articles by Armstrong, L. Search for Related Content PubMed PubMed Citation Articles by Atkinson, S. P. Articles by Armstrong, L.