| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TECHNOLOGY DEVELOPMENT |
Institute of Basic Medical Sciences, Department of Biochemistry, Faculty of Medicine, University of Oslo, Oslo, Norway
Key Words. Chromatin immunoprecipitation • Differentiation • Embryonal carcinoma cell • Histone acetylation • Histone methylation
Correspondence: Philippe Collas, Ph.D., Institute of Basic Medical Sciences, Department of Biochemistry, University of Oslo, PO Box 1112 Blindern, 0317 Oslo, Norway. Telephone: +4722851066; Fax: +4722851058; e-mail: philc{at}medisin.uio.no
Received July 13, 2006;
accepted for publication December 22, 2006.
First published online in STEM CELLS EXPRESS February 1, 2007.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
Disclosure of potential conflicts of interest is found at the end of this article.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
107 per immunoprecipitation), exhaustive sample handling, antibody cost, time, and significant loss of material. This makes current ChIP protocols unsuitable for analysis of small or precious samples and tedious for analysis of many samples in parallel. Maintenance of a pluripotent state in embryos and embryonic stem (ES) cells is conferred by at least two developmentally regulated homeodomain transcription factors, Oct4 and Nanog [10–13]. Regulatory elements in the proximal promoter (PP) (including a hormone-responsive element resembling a retinoic acid [RA]-responsive element), the proximal enhancer (PE), and the distal enhancer (DE) of the Oct4 (Pou5f1) gene drive spatial and temporal Oct4 expression in mouse embryos and in ES cells [14]. The mouse Oct4 PE also contains an RA-repressive site involved in downregulation of Oct4 upon RA-mediated mouse embryonal carcinoma (EC) cell differentiation [15]. Oct4 and Nanog transcriptionally regulate one another in mouse ES cells through interactions on their own and each other's promoters [16, 17], and recent genome-wide analyses in murine ES cells have identified hundreds of target genes for Oct4 and Nanog [18, 19]. Oct4 is also expressed in mouse and human EC cells, and differentiation of mouse ES or EC cells with RA triggers Oct4 downregulation [10, 20]. This is accompanied by H3K9 deacetylation and increased H3K9 methylation on the promoter [8, 9]. In addition, while preparing this manuscript, a novel carrier ChIP protocol was reported to assay histone modifications on the Nanog promoter in murine blastocyst to ES cell transitions [21]. No report exists, however, on histone modifications on human OCT4 and NANOG regulatory elements.
Using a quick and quantitative (Q2)ChIP assay suitable for chromatin from cell numbers up to 105-fold lower than those required for conventional ChIP, we monitored changes in six histone modifications on the promoters of developmentally regulated genes in differentiating human EC cells. Q2ChIP unravels an unexpected biphasic heterochromatinization of OCT4 and NANOG in response to RA stimulation, whereas PAX6, a gene repressed in undifferentiated cells, acquires histone modifications compatible with transcriptional activation. Q2ChIP also shows a dynamic association of the Oct4 transcription factor with the NANOG promoter during differentiation. This study is, to our knowledge, the first to reveal histone modification changes on human OCT4 and NANOG regulatory sequences.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
Cells
Undifferentiated human teratocarcinoma NCCIT cells (American Type Culture Collection, Manassas, VA, http://www.atcc.org) were established from a mediastinal germ cell tumor and can differentiate into derivatives of the three germ layers [22]. Cells were cultured in RPMI 1640 supplemented with 10% fetal calf serum as described [23] and passaged 1:8 every 3–4 days. Cells were induced to differentiate in RPMI 1640 with 10 µM RA for 7 days in bacterial culture plates to allow sphere formation, trypsinized, and cultured for another 4 days on poly-L-lysine in RPMI 1640 without RA [24].
Q2ChIP
Antibody-Bead Complexes. Paramagnetic beads (Dynabeads Protein A; Dynal Biotech, Oslo, Norway, http://www.dynalbiotech.com) were washed twice in RIPA buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% SDS, 0.1% Na-deoxycholate, 140 mM NaCl) and resuspended in 1 volume of RIPA buffer. Beads (10 µl) were added to 90 µl of RIPA buffer and 2.4 µg of primary antibody (specified in the text) in a 0.2-ml polymerase chain reaction (PCR) tube and incubated on a rotator (Stuart SB3; Barloworld Scientific, Staffordshire, U.K., http://www.barloworld-scientific.com; 40 rpm) for 2 hours at 4°C. Note that, for this and subsequent procedures (except when noted), PCR tubes were in the form of 8-tube strips and handled in parallel in an aluminum rack fitted with a magnet and chilled on ice.
DNA-Protein Cross-Linking. Immediately before cell harvest, 20 mM the histone deacetylase inhibitor sodium butyrate was added to the culture medium (Fig. 1). Butyrate was also added to all solutions thereafter unless otherwise indicated. Cells were harvested by trypsinization, washed, and resuspended to 1–2 x 106 cells per milliliter in phosphate-buffered saline (PBS). Formaldehyde was added to 1% (vol/vol) and, after 8 minutes of fixation, was stopped with 125 mM glycine for 5 minutes. All following steps were performed on ice or at 4°C.
|
Lysis, Sonication.
Cross-linked cells were washed twice by sedimentation and suspension in 0.5–1 ml of PBS/20 mM butyrate and lysed for 5 minutes by a sixfold dilution in lysis buffer (50 mM Tris-HCl, pH 8, 10 mM EDTA, 1% SDS, protease inhibitor cocktail 1 mM PMSF) containing 20 mM butyrate. Aliquots of 200 µl in 1.5-ml tubes were sonicated, each for 10 x 30 seconds on ice (Labsonic-M, 3-mm probe; cycle 0.5, 30% power; Sartorius AG, Goettingen, Germany, http://www.sartorius.com) with 30-second pauses on ice to produce chromatin fragments of 500 base pairs. Note that, when preparing chromatin from 100,000 cells, samples (120 µl) were sonicated in 0.5-ml tubes. The lysate was sedimented at 10,000g for 10 minutes, and the supernatant, except the upper lipid layer, was collected and DNA (chromatin) concentration determined by A260 from an aliquot diluted 100-fold.
Immunoprecipitation. Chromatin was diluted to a concentration of 2 A260 units in RIPA buffer containing 20 mM butyrate. Chromatin (100 µl, 2 A260 units) was transferred to a 0.2-ml tube containing antibody-bead complexes in RIPA buffer held to the tube wall by a magnet. Beads were released into the chromatin suspension and rotated at 40 rpm for 2 hours at 4°C. In dilution experiments, chromatin was further diluted in RIPA buffer to produce chromatin from 104, 103, and 102 cell equivalents (eq.) as specified in the text.
Washes. Immune complexes (ChIP material) were washed three times by capturing and releasing the beads in fresh 100 µl of RIPA buffer and once in TE buffer (10 mM Tris-HCl, pH 8.0, 10 mM EDTA). Each wash lasted for 4 minutes on a rotator (40 rpm) at 4°C.
DNA Elution, Cross-Link Reversal, Proteinase K Digestion. ChIP material suspended in TE, pH 8.0, was transferred to a new 0.2-ml tube. Beads were captured, TE was removed, and 150 µl of elution buffer (20 mM Tris-HCl, pH 7.5, 5 mM EDTA, 20 mM sodium butyrate, 50 mM NaCl) containing 1% SDS and 50 µg/ml proteinase K were added. Samples were incubated for 2 hours at 68°C on a Thermomixer (Eppendorf AG, Hamburg, Germany, http://www.eppendorf.de; 1,300 rpm) with constant vortexing. After capturing the beads, the supernatant was recovered and the ChIP material was reincubated in 150 µl of elution buffer/SDS/proteinase K for 5 minutes on the Thermomixer. Both supernatants were pooled.
DNA Isolation. Elution buffer (200 µl) was added to the eluted ChIP material (300 µl), and DNA was extracted once with phenol-chloroform isoamylalcohol and once with chloroform isoamylalcohol and was ethanol-precipitated. The Q2ChIP procedure is summarized in Figure 2 with indications on which step is compatible with –20°C freezing of the preparations.
|
|
|
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
Histone lysine acetylation is a labile modification that is degraded unless histone deacetylase (HDAC) inhibitors are used in ChIP buffers. With the aim of preserving histone acetylation, we compared the amounts of GAPDH promoter DNA associated with immunoprecipitated H3K9ac or H3K9m2 from NCCIT cells collected and cross-linked without or with 20 mM the HDAC inhibitor sodium butyrate. Butyrate was also added to all buffers after either cross-linking treatment. Figure 1A shows that cross-linking in the presence of 20 mM butyrate enhanced H3K9ac ChIP efficiency by approximately threefold, leaving H3K9m2 ChIP efficiency unaffected.
Background DNA caused by nonspecific binding of chromatin to tube walls is expected to increase with reduced ChIP reaction volumes due to increased surface-to-volume ratio. To reduce background, while in TE buffer, H3K9ac and H3K9m2 ChIP complexes were transferred to a clean 0.2-ml tube before DNA elution ("tube shift") rather than being processed in the same tube. PCR analysis of precipitated GAPDH promoter DNA shows that tube shift resulted in a more than fivefold increase in ChIP specificity (Fig. 1B). We interpret enhanced specificity with tube shift as a result of elimination of contaminating chromatin bound to the tube wall.
Conventional ChIP protocols include SDS elution steps of the precipitated DNA followed by a 6-hour cross-link reversal at 68°C and a 2-hour proteinase K digestion at 55°C [27]. In order to shorten the procedure and minimize sample handling, we combined SDS elution, cross-link reversal, and proteinase K digestion into a single step. We first showed that shortening the 68°C cross-link reversal step from 6 to 2 hours, and combining cross-link reversal and proteinase K digestion, did not affect ChIP efficiency or specificity (data not shown). Secondly, SDS (1%) elution, cross-link reversal, and proteinase K (50 µg/ml) digestion were combined into a single 2-hour step at 68°C in 150 µl of elution buffer. We refer to this procedure as the "short protocol." Alternatively, ChIP samples were processed according to a standard "long protocol" as follows: 2 x 15-minute elution in 1% SDS, 6-hour cross-link reversal at 68°C, and 2-hour proteinase K digestion at 55°C (long protocol). Enrichment of GAPDH promoter DNA associated with H3K9ac or H3K9m2 achieved with the long protocol was similar with the short protocol (p > .1; t test; Fig. 1C). Furthermore, variation in the amount of recovered DNA was significantly reduced with the short protocol (p < .001; t test; Fig. 1C). We concluded, therefore, that an 9-hour procedure could be shortened to 2 hours without loss of ChIP efficiency or specificity.
As a first validation of the Q2ChIP assay, we directly compared the specificity of H3K9ac and H3K9m2 Q2ChIP of the GAPDH promoter with that achieved by conventional ChIP with in-suspension cross-linking [27]. When comparing Q2ChIP with conventional ChIP in a 500-µl reaction volume, both protocols yielded similar specificities (p > .05, t test; Fig. 1D). Moreover, a direct comparison between Q2ChIP and conventional ChIP carried out in 100 µl (i.e., with the same amount of input chromatin), Q2ChIP specificity was enhanced fivefold (p < .0001, t test; Fig. 1D).
Integration of all steps examined above followed by real-time PCR analysis of precipitated and purified DNA resulted in a quick and quantitative ChIP procedure we refer to as Q2ChIP. The Q2ChIP assay may be carried out over 1 day with PCR and data analysis taking place the day after, in contrast to conventional ChIP, which may take as long as 3–5 days. A flow diagram of Q2ChIP is provided in Figure 2, and the procedure is detailed under Materials and Methods. Additional validation of the Q2ChIP assay is provided in the following experiments.
Epigenetic Histone Modifications on OCT4, NANOG, and Differentiation Marker Genes in Undifferentiated EC Cells
To further validate Q2ChIP, we first monitored six histone H3 modifications on the promoters of the developmentally regulated genes OCT4 and NANOG expressed in undifferentiated NCCIT cells [23]. PAX6, in contrast, is upregulated during differentiation and is essentially not expressed in NCCIT cells (see below). Promoter regions examined are illustrated in Figure 3A. Q2ChIP analysis of five regions across the OCT4 PP, PE, and DE unraveled consistent histone marks associated with transcription (Fig. 3B). Acetylated H3K9 (H3K9ac) was enriched 8–22-fold over input chromatin levels, with a trend toward highest acetylation on the PP (detected with OCT4-E:PP primers; Fig. 3B). Di- and trimethylated forms of H3K4 (H3K4m2 and H3K4m3) were also strongly detected (p < .001; one-sample t tests) throughout the OCT4 regulatory region, albeit at the highest level in the PP (p < .001, ANOVA; Fig. 3B). A background level of H3K9m2, H3K9m3, and H3K27m3 was also detected on the OCT4 DE, PE, and PP (Fig. 3B). Furthermore, Q2ChIP analysis of the NANOG promoter revealed acetylated H3K9 together with di- and trimethylated H3K4, whereas methylation marks associated with inactive loci were only detected at background levels (p > .2; Fig. 3C). This is in agreement with NANOG expression in undifferentiated NCCIT cells. Therefore, in undifferentiated NCCIT cells, regulatory regions of OCT4 and NANOG are enriched in euchromatic marks.
We next examined histone modifications on the promoter of the LMNA gene, which encodes A-type lamins only in lineage committed or differentiated cells [28]. We previously showed that LMNA is repressed or transcribed at a basal level in undifferentiated NCCIT cells [23]. The LMNA promoter region examined (Fig. 3A) displayed nearly basal levels of H3K9 acetylation and significant di- and trimethylation of H3K4 (p < .001), in agreement with a potential for expression of the gene (Fig. 3D). The LMNA promoter was also dimethylated on H3K9 (p < .02) and trimethylated on H3K27 (p < .01), indicative of the positioning of a transcriptional brake on the LMNA promoter (Fig. 3D).
We also determined histone changes on the promoter of PAX6 (Fig. 3A), a developmentally controlled gene upregulated during pancreas, eye, and brain development [29, 30]. PAX6 was not expressed in undifferentiated NCCIT cells (data not shown). Similarly to LMNA, the PAX6 promoter harbored primarily di- and trimethylated H3K4, a low-level of H3K9ac, and little H3K9m3 (Fig. 3E). This, combined with trimethylation of H3K27 (Fig. 3E), was indicative of a repressed gene with a potential for activation [31, 32]. Lastly, we showed that the GAPDH housekeeping gene promoter (Fig. 3A) was marked by H3K9ac, H3K4m2, H3K4m3, and by a lack of heterochromatin histone modifications, as anticipated from a constitutively expressed gene (Fig. 3F).
Collectively, these results indicate that Q2ChIP can be applied to identify histone modifications on regulatory sequences of pluripotency and differentiation marker genes. Undifferentiated NCCIT cells are characterized by euchromatic packaging of OCT4 and NANOG consistent with expression of these genes and assembly of a facultative repressive mark (H3K27m3) on LMNA and PAX6 with nevertheless a potential for expression upon differentiation.
Q2ChIP Unravels Heterochromatinization of OCT4 Regulatory Elements upon RA-Mediated Differentiation
Seven days of stimulation with 10 µM RA induced differentiation of NCCIT cells. The cells formed spheres in bacterial culture plates and acquired an early neurogenic phenotype upon subsequent replating on poly-L-lysine (Fig. 4A). Induction of differentiation was confirmed by downregulation of OCT4 and NANOG expression and upregulation of PAX6 expression (p < .0001 for each; one-sample t tests; Fig. 4B). LMNA mRNA levels were surprisingly not altered (Fig. 4A).
|
|
20-fold enrichment in H3K9m3 in the DE (OCT4-A:DE), PE (OCT4-D:PE), and PP (OCT-E:PP) (p < .001); (e) increased H3K9m2 on the distal PE (OCT4-C:PE; p < .001) (note that H3K9m3 remained unchanged in the OCT4-C:PE region, reflecting a different equilibrium between di- and trimethylated H3K9); and (f) increased H3K27m3 on the DE and PE (OCT4-A:DE to OCT4-D:PE; p < .005) but, notably, not on the PP (OCT4-E:PP). We concluded that RA removal and replating cells on poly-L-lysine potentiated heterochromatinization on the OCT4 regulatory locus initiated by exposure to RA.
The NANOG Promoter Is More Resistant to Assembly of Repressive Histone Marks
Despite downregulation of expression, the dynamics of histone H3 modifications on NANOG differed from that of OCT4. RA stimulation maintained H3K9ac, H3K4m2, and H3K4m3 levels, enhanced H3K9m2 level (p < .01), and moderately (but not significantly) increased H3K9m3 and H3K27m3 levels (p > .05; Fig. 4D). RA removal and cell replating on poly-L-lysine, however, triggered 90% deacetylation of H3K9 (p < .001) and 40% reduction of di- and trimethylated H3K4 (p < .001; Fig. 4D, inset). Furthermore, H3K27 trimethylation was potentiated (p < .01), whereas the moderate increases in di- and trimethylated H3K9 remained unaltered (Fig. 4D). Collectively, these results suggest that NANOG is more refractory than OCT4 to histone modification changes upon RA-mediated differentiation. Alternatively, histone modification changes on NANOG may take place at later time points.
PAX6 and LMNA Promoters Undergo H3K9 Acetylation upon RA-Mediated Differentiation
LMNA mRNA levels were not significantly upregulated upon RA stimulation (Fig. 4B). Acetylation of H3K9 on the LMNA promoter was observed only after RA stimulation; however, this effect was not maintained after replating on poly-L-lysine (Fig. 4E). No significant changes in H3K9m2, H3K9m3, and H3K27m3 were observed.
In contrast to LMNA, PAX6 was transcriptionally activated after RA stimulation (Fig. 4B). Accordingly, the promoter was strongly hyperacetylated on H3K9 (p < .001) while remaining di- and trimethylated on H3K4 (Fig. 4F). Surprisingly, H3K27 remained trimethylated. H3K27m3 on PAX6 was apparently compatible with transcriptional activation of the gene, which may have been brought about by the marked H3K9 acetylation. Lastly, we found that, as expected, the GAPDH promoter retained modifications detected in undifferentiated cells (Fig. 4G), consistent with its ubiquitous expression. We concluded that epigenetic alterations of upregulated, downregulated, or unaffected genes examined by Q2ChIP in response to RA stimulation reflect transcriptional changes or the lack thereof.
Determination of Amount of Chromatin Compatible with Q2ChIP
We next determined the lower limit of input chromatin amount compatible with Q2ChIP. Sonicated chromatin was prepared from 105 NCCIT cells and diluted to 10–1,000-fold to produce chromatin amounts corresponding to that of 104, 103, or 102 cells (10,000, 1,000, or 100 cell eq.). Modified histones were immunoprecipitated by Q2ChIP and association with two regions of OCT4, NANOG, PAX6, and GAPDH was determined by real-time PCR as reported earlier (Fig. 5). Histone binding pattern to OCT4-D:PE detected in chromatin from 104 cell eq. was similar to that detected with 105 cell eq. (see Fig. 3B). Furthermore, the Q2ChIP pattern was preserved with dilutions to 103 and 102 cell eq. We also noted that enrichment of precipitated DNA was greater with increasing dilutions of input chromatin (Fig. 5A). This was most likely due to an increased antibody-bead complex/histone ratio in the assay as chromatin was diluted, because antibody concentration remained constant under all conditions. Similar observations were consistently made for Q2ChIP analysis of OCT4-E:PP (Fig. 5B) and NANOG (Fig. 5C) as well as for the inactive PAX6 promoter (Fig. 5D) and for the constitutively expressed GAPDH promoter (Fig. 5E). Moreover, in each example, standard deviations of the relative amounts of precipitated DNA tended to augment as chromatin dilution increased (in particular for 100 cell eq.; Fig. 5), reflecting more variable quantitative PCR efficiencies with lower amount of template. Enhanced variation was particularly evident in the analysis of PAX6 (Fig. 5D) due to intrinsic lower PCR efficiency with the PAX6 primers (not shown). Detection limit is probably reached with 100 cell eq. in our assay, as shown by enhanced variation and, in some instances, slightly enhanced background (no antibody control; Fig. 5A, 5C, 5D). Nevertheless, the results demonstrate that Q2ChIP can be used with up to 105-fold reduced amounts of input chromatin compared with standard ChIP protocols without affecting patterns of histone binding.
Q2ChIP Reveals Dynamic Oct4 Binding to the NANOG Promoter
We examined whether Q2ChIP was also applicable to assess transcription factor binding to DNA. The developmentally regulated transcription factor Oct4 would be expected to bind the NANOG promoter on the Sox/Oct binding site [17] in undifferentiated, but not in differentiated, cells. Indeed, Q2ChIP analysis of 105 cells showed that Oct4 bound to the NANOG promoter in undifferentiated NCCIT cells but was dissociated from the promoter in RA-differentiated cells (Fig. 6A). In contrast, Oct4 was not significantly associated with the control GAPDH promoter, as anticipated (Fig. 6A). In addition, we showed that Oct4 binding to NANOG was also clearly detected by Q2ChIP analysis of 10,000 and 1,000 cell eq. (Fig. 6B). We concluded that Q2ChIP is suitable for analysis of both histone modifications and transcription factor binding from greatly reduced amounts of chromatin relative to conventional ChIP.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
2-hour step, which substitutes for the 9-hour process of current protocols. ChIP efficiency is maintained, and we observe a marked reduction in the variation of the amount of H3K9ac recovered compared with conventional ChIP. Chromatin can easily be prepared from as low as 105 cells and diluted 1,000-fold to produce reliable templates for Q2ChIP. Indeed, histone binding patterns are maintained with chromatin from as low as 100 cell equivalents. Furthermore, efficiency of Q2ChIP is enhanced with diluted chromatin due to the increased ratio of precipitating antibody-bead complex to target histone (Fig. 5E). Lastly, we demonstrate that Q2ChIP can be used for analysis of transcription factor binding to DNA. While this paper was in preparation, a carrier ChIP (CChIP) procedure designed for small cell numbers was reported [21]. CChIP is based on the high efficiency NChIP and relies on the coprecipitation of carrier chromatin with the sample of interest. CChIP has allowed analysis of histone modifications on the Nanog promoter to monitor embryonic-to-ES cell epigenetic transitions in the mouse [21]. CChIP may be applicable to cross-linked chromatin, but this remains to be shown. The Q2ChIP procedure reported here is fast and includes a cross-linking step and thus, unlike NChIP, is applicable to the analysis of histone and nonhistone proteins. A combination of CChIP with Q2ChIP should create possibilities to unravel epigenetic modifications unexplored to date in embryos and ES cells.
Remodeling of the NANOG Promoter During EC Cell Differentiation
Cooperative interaction between Nanog, Oct4, and Sox2 on their respective promoters is essential for transcriptional activation of these genes [17, 18]. Thus, removal of one component from one locus is expected to result in downregulation of that locus. Consistent with this view, we found that RA-mediated differentiation of NCCIT cells promotes removal of Oct4 from the NANOG promoter. How dissociation of Oct4 from NANOG is elicited remains unclear but may involve chromatin remodeling on the Sox/Oct binding site in the NANOG promoter [17]. In undifferentiated human carcinoma cells (this paper) and in mouse ES cells [8, 21], NANOG harbors the euchromatic marks H3K9ac, H3K4m2, H3K4m3, and unmethylated H3K9, consistent with expression of the gene. In agreement with changes reported by CChIP during mouse ES cell differentiation [21], we observed deacetylation of H3K9 and a threefold increase in H3K9 di- and trimethylation of NANOG, adjacent to the Sox/Oct binding site. RA-mediated differentiation of NCCIT cells also causes hypermethylation of H3K27m3, potentially creating binding sites for polycomb-group transcriptional repressors [33].
Histone Modifications on OCT4 During EC Cell Differentiation
As with NANOG, the OCT4 regulatory locus contains epigenetic marks indicative of transcriptional activation in undifferentiated NCCIT cells. Likewise, in mouse ES and EC cells, Oct4 displays acetylated and unmethylated H3K9 and di- and trimethylated H3K4 [8, 9, 21]. Furthermore, we consistently detected more H3K9ac and di- and trimethylated H3K4 on the OCT4 PP than on the PE or DE in NCCIT cells, suggesting the existence of a gradient of euchromatin assembly on OCT4 in these cells. Enhanced histone acetylation in the OCT4 PP may facilitate the recruitment of transcriptional regulators in this region [34].
Our results unexpectedly unravel a biphasic epigenetic modulation of the OCT4 regulatory region upon RA stimulation. RA causes hypermethylation of H3K9 and H3K27 yet maintains H3K9 acetylation, and only subsequent removal of RA and replating on poly-L-lysine results in H3K9 deacetylation throughout the OCT4 PP, PE, and DE and enhances assembly of heterochromatin marks. We found consistent deacetylation of H3K9 and demethylation of H3K4m2 throughout the OCT4 promoter, reflecting heterochromatin assembly. Marked increases in H3K9m3 were also prominent in the PP (detected by OCT4-E:PP), the 3' region of the PE (detected by OCT4-D:PE), and the 5' region of the DE (detected by OCT4-A:DE). The OCT4-C:PE and OCT4:B:DE regions displayed a moderate increase in H3K9m3, whereas H3K9m2 was more prominent. These results suggest a highly localized assembly of heterochromatin marks. The OCT4 promoter harbors RA-responsive elements in the PP and PE, which in murine EC and ES cells are responsible for RA-mediated inactivation of Oct4 [14, 15] by repressors transiently induced at the onset of differentiation [35, 36]. No RA-responsive element has been identified to date in the DE of OCT4; however, we observed a clear methylation response on lysine 9 and 27 of H3 in this region after RA stimulation. Yet, the absence of clear H3K9m3 in the distal PE (OCT4-C:PE) and in the proximal DE (OCT4-B:DE) argues against a uniform spreading of heterochromatin on OCT4. Rather, H3K9m3 in the distal DE (OCT4A:DE) may be brought about by the deposition of histone methyltransferases or by the action of enzymes assembled in the PP/PE region through DNA looping.
In accordance with our findings, the early stage of RA differentiation of mouse ES cells is associated with H3K9 methylation to a greater extent than H3K9 deacetylation or H3K4 demethylation [9]. H3K9 methylation and acetylation are exclusive marks, and it is believed that inactivation of transcription is mediated through histone deacetylation, which favors the establishment of more long-term, repressive methylation marks [37]. Thus, the Feldman et al. [9] and our observations may reflect the analysis of a mixed cell population or differential marking of OCT4 alleles. Alternatively, the data suggest that specific lysine (K9) residues in a nonacetylated state on nucleosomes are "reserved" for immediate methylation upon transcriptional inactivation, whereas acetylated lysines must undergo deacetylation prior to methylation. Either putative mechanism would result in epigenetic mosaicism between cells in response to RA.
|
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. A. Dahl and P. Collas {micro}ChIP--a rapid micro chromatin immunoprecipitation assay for small cell samples and biopsies Nucleic Acids Res., February 11, 2008; 36(3): e15 - e15. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. R. Lee, F. E. Murdoch, and M. K. Fritsch High Histone Acetylation and Decreased Polycomb Repressive Complex 2 Member Levels Regulate Gene Specific Transcriptional Changes During Early Embryonic Stem Cell Differentiation Induced by Retinoic Acid Stem Cells, September 1, 2007; 25(9): 2191 - 2199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Attema, P. Papathanasiou, E. C. Forsberg, J. Xu, S. T. Smale, and I. L. Weissman Epigenetic characterization of hematopoietic stem cell differentiation using miniChIP and bisulfite sequencing analysis PNAS, July 24, 2007; 104(30): 12371 - 12376. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
withdrawal). Bars, 25 µm. (B): Quantitative reverse transcription-polymerase chain reaction analysis of OCT4, NANOG, PAX6, and LMNA expression after 7 days of RA stimulation (yellow bars) and after replating cells (red bars). Data are expressed as mean ± SD fold change mRNA level relative to untreated cells (triplicate experiment). (C–G): Q2ChIP analysis of indicated histone modifications on the (C) OCT4, (D) NANOG, (E) LMNA, (F) PAX6, and (G) GAPDH promoters in unstimulated cells (–RA) after RA stimulation for 7 days (+RA) and after replating (+RA
