Epigenetic Modifications at the Human Growth Hormone Locus Predict Distinct Roles for Histone Acetylation and Methylation in Placental Gene Activation (original) (raw)

Journal Article

,

1Departments of Genetics and Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104

Search for other works by this author on:

,

1Departments of Genetics and Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104

Search for other works by this author on:

1Departments of Genetics and Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104

*Address all correspondence and requests for reprints to: Nancy E. Cooke, 752b Clinical Research Building, University of Pennsylvania, 415 Curie Boulevard, Philadelphia, Pennsylvania 19104.

Search for other works by this author on:

Received:

03 December 2003

Accepted:

31 December 2003

• PDF
• Split View
• • Article contents
  • Figures & tables
  • Video
  • Audio
  • Supplementary Data
• Cite

Cite

Atsushi P. Kimura, Stephen A. Liebhaber, Nancy E. Cooke, Epigenetic Modifications at the Human Growth Hormone Locus Predict Distinct Roles for Histone Acetylation and Methylation in Placental Gene Activation, Molecular Endocrinology, Volume 18, Issue 4, 1 April 2004, Pages 1018–1032, https://doi.org/10.1210/me.2003-0468
Close

Navbar Search Filter Mobile Enter search term Search

Abstract

Developmental control of eukaryotic gene expression is tightly linked to alterations in chromatin structure. Studies of the hGH multigene cluster suggest that the four placental genes are activated by a pathway of histone modification distinct from the pathway leading to activation of the single pituitary hGH-N gene. The relationship between histone acetylation and hGH-N activation in the pituitary has been previously defined using a combination of epigenetic mapping and transgenic analyses. The repeated gene structures within the hGH cluster had been an impediment to comparable analysis of placental gene activation. In the present report we defined patterns of core histone acetylation and methylation within and flanking the hGH cluster in human placental chromatin. These data highlight differences between placental and pituitary pathways of transcriptional control at the hGH cluster and suggest that selective activation of the placental genes reflects distinct roles for histone acetyltransferase and histone methyltransferase coactivator complexes.

EUKARYOTIC GENE EXPRESSION is strictly controlled by mechanisms operating at levels of nuclear organization, chromosomal structure, and DNA modification. When organized in a highly compact chromatin configuration, DNA is inaccessible to most transcriptional activators, resulting in gene repression. Gene activation entails controlled decondensation of the encompassing chromatin environment. Modifications in chromatin involved in the activation process can be limited to promoter regions or can be expansive, creating activated chromatin domains. Histone modifications that accompany and regulate alterations in gene expression include acetylation, methylation, phosphorylation, and ubiquitinylation (1). These modifications may be established in concerted reactions or added sequentially (2, 3). When sequential, the order of modifications may be gene specific (4, 5). In either case, site-specific combinations of covalent modifications define a histone code that can alter higher-order chromatin structure and/or mediate subsequent targeting of transcriptional complexes (6–8).

The hGH multigene cluster represents a tightly controlled and robustly expressed model system for analysis of developmental controls. This cluster, located on chromosome 17q22–24, is composed of five structurally conserved genes: hGH-N, human chorionic somatomammatropin (hCS)-L, hCS-A, hGH-V, and hCS-B (Fig. 1A). These genes evolved subsequent to the rodent/primate divergence via successive local duplications of a precursor GH gene (9, 10). The five genes display distinct, mutually exclusive tissue specificities despite their close juxtaposition and strong structural conservation. hGH-N is expressed in somatotrope cells of the anterior pituitary; the remaining four genes are expressed in the syncytiotrophoblast (STB) cellular layer that lines the fetal-maternal interface of the placental villi. The four placental genes further differ in levels of expression. hCS-A and hCS-B are robustly transcribed in the mid- to late-gestation placenta. hGH-V mRNA is expressed at 100- to 1000-fold lower levels (9, 11), but its expression is sufficient to encode the major GH in maternal serum (12). hCS-L activity, although quite low, is difficult to accurately quantify because most of its transcripts are shunted along nonproductive splicing pathways (13). The B lymphocyte-specific CD79b gene immediately flanks the hGH cluster at its 5′-end and encodes _Ig_β, one of the two B cell receptor subunits. The striated muscle-specific SCN4A gene, encoding a sodium channel component, is located further 5′. The testis-specific TCAM gene flanks the hGH cluster at its 3′-end (Fig. 1A). This close packing of eight genes expressed in five different tissues presents a complex and potentially informative model of gene regulation.

Fig. 1.

Structure of the hGH Multigene Locus and the Amplimer Array Used in the ChIP Assays A, The hGH locus. A diagram of the 100-kb region encompassing the hGH gene cluster and its LCR is shown. Genes from 5′ to 3′ are: skeletal muscle-specific SCN4A, B lymphocyte-specific CD79b/_Ig_β; pituitary-specific hGH-N, placenta-specific hCS-L, hCS-A, hGH-V, and hCS-B; and testis-specific TCAM. _Ci_s determinants relevant to hGH cluster activation are also shown; five HS of the hGH LCR (vertical arrows), P elements (shaded ovals labeled P), and the putative hCS enhancers (white ovals labeled E). Tissue specificities of LCR HS (pituitary or placenta) are indicated. Because the hGH cluster is formed by repeated duplications of a single ancestral gene and its flanking sequences (9, 10 ), the structural genes and intergenic regions contain extensive conserved segments. These are indicated under the genes (shaded and patterned rectangles). Based upon these conserved segments, the hGH cluster can be structurally divided into five repeated domains (shaded horizontal ovals; light ovals include GH genes; dark ovals include CS genes). Positions of amplimers used in ChIP assays are indicated below the diagram. Due to the repeated nature of the hGH cluster, the amplimers corresponding to 3′GH, 3′CS1, 3′CS2, En, and hGH/CS each detect three to four regions within the locus. B, The five hGH cluster repeated domains. The region containing the five structural genes is expanded, and the positions of the repeated domains (shaded ovals) and the 5′ subdomains of each repeat are indicated (white rectangles). A detailed diagram of each 5′ subdomain is shown in expanded format; conserved regions are indicated by shading or hatching. Each 5′ subdomain contains three Alu repetitive elements. A single _Kpn_I repetitive sequence is present 5′ to hCS-B. The amplimers used to scan the 5′ repeat subdomains are shown (Pe0-Pe9 and hGH/CS). The 233GH/CS and the 930P amplimers (double-headed arrows) were used to distinguish each of the five genes and four P elements, respectively.

Activation of the hGH multigene cluster is under the control of a locus control region (LCR). The hGH LCR is marked by a set of DNase I-hypersensitive site (HSI-HSV) located between 15 kb and 32 kb 5′ to the cluster (14) (Fig. 1A). The sets of HS formed in the chromatin of pituitary somatotrope cells and placental STB cells are partially overlapping. HSV and HSIII are found in both cell types, HSIV is placenta specific, and HSI and -II are pituitary-specific. When linked to the full LCR, the hGH multigene cluster is expressed in transgenic mice in a copy number-dependent and site-of-insertion-independent manner; hGH-N is expressed specifically in the pituitary, and the hCS/hGH-V genes are selectively expressed in the placenta (14, 15).

Distinct patterns of chromatin modification are associated with activation of the hGH cluster in the pituitary and the placenta. Pituitary somatotropes expressing hGH-N establish a 32-kb domain of acetylated chromatin. This domain encompasses the entire LCR and contiguous hGH-N promoter but excludes the four placental genes (16). The formation of this acetylated domain is dependent upon HSI, which is pituitary specific and located in the center of the acetylated domain. Selective deletion of HSI results in loss of acetylation throughout the region and loss of hGH-N expression in transgenic pituitaries (17). In placental chromatin, acetylation 5′ of the cluster is limited to the region encompassing HSIII–HSV (18). Regions internal to the multigene cluster are selectively acetylated in placental chromatin. One strongly acetylated region within the cluster encompasses the placental P element. The P elements are a set of conserved sequences situated 2 kb upstream from each of the four placental genes that have been implicated in the control of placental gene expression (Fig. 1). Whether the P element enhances gene expression in the placenta and/or represses gene expression in pituitary remain major unresolved questions (18, 19).

Establishing the profile of histone modifications at a locus can yield insights into epigenetic pathways leading to mechanisms of gene activation. This has been the case for the hGH-N gene in the pituitary (16, 17). It may be equally informative to define the patterns of epigenetic modification in the placenta and contrast them to those in the pituitary. However, prior attempts to establish a comprehensive picture of chromatin modification within the GH cluster were impeded by structural similarities among its five gene units and were limited in the repertoire of histone modifications studied (18, 20). In the present report we have substantially increased the density and resolution of the epigenetic map at the hGH multigene locus in human placental STB chromatin. The patterns of histone acetylation and methylation suggested that these modifications play distinct roles in selective activation of the placental genes from the hGH cluster.

RESULTS

The hGH LCR and Gene Cluster Are Selectively Acetylated in Placental Chromatin

Chromatin immunoprecipitation (ChIP) analysis of core histone H3 and H4 acetylation at the hGH cluster was carried out on chromatin preparations isolated from STB nuclei of normal human term placentas (STB chromatin is referred to throughout the text as “placental chromatin”). A set of 25 PCR primer sets was used to scan the 98-kb region encompassing the hGH cluster and flanking regions (Fig. 1 and Table 1). PCR amplimers located 5′ and 3′ to the cluster corresponded to unique genomic elements. The cluster itself is comprised of five repeated domains reflecting duplications of an initial precursor gene in primate species. The initial duplication resulted in GH and CS precursor genes. The 3′-flanking regions of these two genes diverged significantly, and the GH and CS units then underwent duplications to result in two GH repeated domains encompassing GH-N and GH-V and three CS repeated domains encompassing CS-A, CS-B, and CS-L (Fig. 1A). Thus amplimers located within the multigene cluster detect sites represented in three to five of these repeated domains (Fig. 1, A and B). The level of histone modification at each site was calculated as the ratio of DNA in the antibody-bound chromatin to that in the input sample. This ratio was normalized to the corresponding ratio for a constitutively positive promoter−−glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter, defined as 100. Parallel analysis of normal human fibroblast chromatin provided an assessment of tissue specificity. Each PCR signal was documented to be in the linear range by analysis of serial dilutions of each template. A representative set of studies scanning the entire hGH locus for acetylated histone H3 is shown in Fig. 2A. Compilations of multiple studies of histone H3 and H4 acetylation across the locus are represented in Fig. 2, panels B and C, respectively.

Fig. 2.

Acetylation of Histones H3 and H4 in Placental Chromatin Is Most Pronounced at the Placental LCR and in the Repeated Domains A, ChIP analysis of human placenta and fibroblast chromatin. An autoradiogram of a representative ChIP analysis is shown. The ChIP analyses used antiacetyl-H3 to immunoprecipitate chromatin from human term placental STB cells and from human skin fibroblasts. DNA purified from starting (Input) and immunoprecipitated (AcH3) samples were subjected to PCR amplification using primers summarized in Table 1 that correspond to the amplimers shown in Fig. 1. Quantification of each reaction was determined at the linear range of amplification as confirmed by serial dilutions of template DNA. B, HSIII, HSV, and the repeated domains of the hGH cluster in human placental STB chromatin are specifically associated with acetylated histone H3. Ratios of signal intensities of bound and input fractions (as in panel A) were calculated and displayed on the histogram. Each ratio was normalized to the experimental ratio of the constitutively active GAPDH promoter (defined as 100). The ratios for the ChIP data from human term placentas (black bars) and skin fibroblasts (white bars) are plotted. The data are shown as the average ± se from at least three independent experiments with two term placentas. C, The placental LCR and the repeated domains of the hGH cluster in human placenta chromatin are specifically associated with acetylated H4 histones. The ChIP assay was conducted using antisera to acetyl-H4. The data were analyzed and plotted as in panel B.

Table 1.

Oligonucleotide Primers for ChIP Assay

The primers for HSV, HSIV, P6, P3, P2, and P1 were described by Elefant et al. (16 ). Because the primer set for GAPDHPr did not work in human skin fibroblast cell CCDsk25, we used another primer set GAPDH3 for the cell line. The acetylation levels at GAPDHPr and GAPDH3 were not significantly different in human placental chromatin (data not shown).

Table 1.

Oligonucleotide Primers for ChIP Assay

The primers for HSV, HSIV, P6, P3, P2, and P1 were described by Elefant et al. (16 ). Because the primer set for GAPDHPr did not work in human skin fibroblast cell CCDsk25, we used another primer set GAPDH3 for the cell line. The acetylation levels at GAPDHPr and GAPDH3 were not significantly different in human placental chromatin (data not shown).

Histone H3 acetylation at the hGH locus.

ChIP analysis of the 35-kb region 5′ to the hGH cluster with a set of 10 amplimers revealed that acetylation of placental chromatin was limited to LCR elements HSV and HSIII. Surprisingly, the placenta-specific HSIV was only minimally acetylated. The 28-kb region between the placental LCR (HSV-HSIII) and the hGH cluster was modified in placental chromatin at levels that were low but significantly above modification at the flanking SCN4A and TCAM genes. The hGH cluster itself was robustly acetylated at levels equivalent to, or significantly above, the active GAPDH promoter. Analysis of primary human fibroblast chromatin failed to reveal significant levels of acetylation at any of the sites studied (Fig. 2, A and B, light bars). Thus, histone H3 acetylation in placental chromatin was prominently detected at HSV and HSIII of the LCR and throughout the hGH cluster. The acetylated LCR and gene cluster were separated by a 28-kb gap of minimally modified chromatin.

Histone H4 acetylation at the hGH locus.

The levels of H4 acetylation at the hGH locus were less robust than H3 acetylation, but the pattern was generally conserved; there were maximal acetylation at the placental LCR and at the hGH cluster with marginal levels of acetylation in the intervening 28 kb (compare panels B and C of Fig. 2, black bars). In contrast to the H3 acetylation study, HSIV was strongly H4-acetylated when compared with the GAPDH promoter control. Fibroblast chromatin was H4 acetylated at low and variable levels throughout the 98-kb region with highest levels at HSI, II and Pe8 (Fig. 2C, light bars). Thus, as was the case for H3 histones, the acetylation of H4 histones in placental chromatin was most prominent at the placental LCR and within the hGH cluster.

The hGH Cluster Is Encompassed in a Continuous Domain of Acetylated Chromatin

ChIP analyses within the hGH gene cluster revealed significant levels of H3 and H4 acetylation at all sites (Fig. 2). However, each of the amplimers within the cluster is represented in three to five repeated domains. Therefore, the data in Fig. 2 do not establish whether the modification of the cluster is continuous or is limited to specific repeated domains. To address this, each of the four P elements and each of the five gene units were individually assessed for H3 and H4 acetylation (Figs. 3 and 4, respectively).

Fig. 3.

All Four P Elements Are Associated with Acetylated Chromatin A, Hybridization probes specific to each of the P elements. The labeled arrows (PL, PA, PV, PB) indicate the oligoprobe predicted to be specific for each P element located 5′ to hCS-L, hCS-A, hGH-V, and hCS-B, respectively. Each of these probes is aligned with the corresponding sequences at the other three P elements to demonstrate the sequence divergence; only bases identical at all four P elements are indicated (*). Each probe has a specificity of three to eight nucleotides compared with the other aligned P elements. B, Control hybridizations confirm P element probe specificity. Each of the four P elements was amplified, cloned, and analyzed by Southern blot using the indicated 32P-endolabeled oligoprobes. Each oligoprobe demonstrates hybridization specificity for its corresponding P element. C, Representative ChIP analyses of individual P elements. Chromatin isolated in the ChIP analyses was amplified with the 930P primer set (Fig. 1B), which encompasses each of the P elements. The amplimers were quantified by Southern blots using each of the four specific P element oligoprobes. The GAPDH promoter was analyzed by ChIP in parallel as a control. D, All four P elements are associated with acetylated histone H3. Each ratio plotted on the histogram represents the average ± se from three independent experiments with two placentas. Values are normalized to the active GAPDH promoter control. E, All four P elements are associated with acetylated histone H4. Each value on the histogram represents the average ± se from three independent experiments using two placentas.

Fig. 4.

Acetylation Levels at the Individual hGH Cluster Gene Units A, Selective detection of each of the five hGH cluster genes by PCR and restriction digestions. The 32P-labeled 233GH/CS amplimer (Fig. 1B) was digested with _Hin_fI (H), _Dra_III (D), _Pst_I (P), and _Msc_I (M) to generate fragments specific to hGH-N (233 bp), hGH-V (214 bp), and hCS-L (159 bp). The amplimer was digested with _Hin_fI, _Dra_III, _Pst_I, and _Sml_I (S) to generate fragments specific to hCS-B (214 bp), hCS-A (199 bp), and hCS-L (159 bp). The labeled and digested products were separated by denaturing PAGE. Each signal was detected by autoradiography. A representative analysis is shown to the right; the analysis is carried out on three serial dilutions of chromatin to assure that the assay is in the linear range. The _hCS-L_-specific 159-bp fragment was common to both sets of enzymes (linked arrows) and was used to normalize the results between the two digestions. B, The four placental genes are more highly acetylated at histone H3 than pituitary hGH-N. Values on the histogram represent the average ± se from three independent experiments with two placentas. Values are normalized to the active GAPDH promoter control. C, All five hGH/hCS genes are acetylated to an equivalent degree at histone H4. Details are as in panel B.

Histones H3 and H4 are acetylated at all four P elements.

The P elements, located 2 kb 5′ to each of the four placentally expressed genes (hCS-L, hCS-A, hGH-V, and hCS-B), were coamplified from ChIP samples with a single primer pair to generate the 930P amplimer (Fig. 1B). Signals corresponding to each of the four individual P element determinants were then identified and quantified by sequential hybridization of the amplified fragment with oligoprobes corresponding to each of the four P elements (PL, PA, PV, and PB) (Fig. 3). This targeted detection revealed that each of the P elements was acetylated on histones H3 and H4 at levels equal to or greater than the GAPDH control. Levels of histone H4 acetylation were roughly equivalent at each of the P elements, whereas histone H3 acetylation varied by as much as 2-fold in pairwise comparisons (i.e. PV_vs_. PA).

The placental genes are selectively acetylated at histone H3.

Modifications at each of the five gene units were detected by amplification with a primer pair common to the 5′-terminus of all five genes (amplimer 233GH/CS; Fig. 1B). This was followed by digestion of the amplified DNA with two partially overlapping sets of restriction enzymes that recognize rare sequence differences specific to each of the five genes (Fig. 4A). The 159-bp hCS-L fragment, common to both sets of digestions, was used to normalize the two data sets (Fig. 4A, linked arrows). This assay revealed robust acetylation of histone H3 at each of the four placental genes and a significantly lower level of modification at hGH-N (Fig. 4B). Analysis of histone H4 acetylation demonstrated acetylation at all five genes at levels equivalent to the active gene (GAPDH) promoter (Fig. 4C).

The combined ChIP studies (Figs. 2, 3, and 4) showed that the four placental genes in the hGH cluster are encompassed within a continuous domain of acetylated chromatin. This domain does not form in fibroblast (Fig. 2) or pituitary chromatin (16). The acetylated H4 domain extends further 5′ in the cluster than the acetylated H3 domain, as indicated by the higher level of modification at the hGH-N gene (compare panels B and C of Fig. 4). LCR determinants HSV and HSIII, common to pituitary and placenta, are acetylated on histones H3 and H4, whereas the placenta-specific HSIV is more strongly modified at histone H4 than H3. The placental LCR is separated from the hGH cluster by a 28-kb intervening region that is only minimally acetylated.

Analysis of Histone H3K4 Methylation at the hGH Locus

Histone acetylation and methylation can play distinct roles in epigenetic pathways (8, 21). Di- and trimethylation of histone H3 at lysine 4 (H3K4-me2 and H3K4-me3) appear to be of particular importance in gene activation and/or prevention of gene silencing; di- and trimethylation at H3K4 may mediate distinct functions and/or reflect recruitment of distinct coactivator complexes (22–24). Therefore, ChIP studies were carried out on placental chromatin to specifically assess the levels and pattern of di- and trimethylation at H3K4. Remarkably, H3K4-me2 modification was essentially confined to the repeated domains of the gene cluster; there was minimal evidence of tissue-specific modification 5′ to the hGH cluster (Fig. 5, A and B). In this region only HSIII was dimethylated at levels that exceeded the flanking SCN4A and TCAM genes. However, HSIII was also modified in fibroblast chromatin. Thus the placental LCR lacked significant or specific H3K4-me2 modification.

Fig. 5.

Histone H3K4 Di- and Trimethylation Define Chromatin Subdomains within the hGH Cluster A, ChIP assay for H3K4-me2 in human placental and fibroblast chromatin samples. This study was performed as described in Fig. 2 using antibodies specific to H3K4-me2. A set of representative Southern blots of PCR products for each region is shown. B, Placenta-specific histone H3K4 dimethylation is restricted to a subregion of the repeated domain. The ratios of signal intensities from bound to input fractions were normalized to the ratio obtained at GAPDH promoter. The ratios for placenta (black) and fibroblasts (white) are plotted. The data represent the average ± se from at least three independent experiments using two placentas. C, Histone H3K4 trimethylation is detected at the placental LCR and within the repeated domain. The ChIP assay was performed as described above using anti-H3K4-me3 antibody. The Southern blot images are not shown. The ratios are plotted as in panel B and represent the average ± se from at least three independent experiments with chromatin isolated from two placentas.

Within the repeated domain, a restricted region of H3K4 dimethylation was observed that extended from 1 kb 5′ to the structural genes (amplimer Pe6) to a site 3.8 kb into the 3′-flanking regions (amplimer 3′CS1) (Fig. 5, A and B). Of note, sites immediately 3′ of the hGH-N and hGH-V genes (amplimer 3′GH) were unmodified. There was no significant modification in the fibroblast chromatin. As defined by histone H3K4-me2 modifications, these data indicated that the hGH gene cluster in placental chromatin is organized as a series of discontinuous domains encompassing individual gene units and their immediate flanking sequences.

ChIP was next carried out with antibodies to the H3K4-me3. HSV was robustly modified at levels comparable to the GAPDH promoter (Fig. 5C, black bars). HSIV and HSIII showed much lower, but still significant, levels of H3K4-me3 modification. Levels in the region extending from HSIII through the hGH-N promoter (hGHNPr amplimer) were quite low, but appeared to be significant because they remained above that at the flanking SCN4A locus (SCNex13) and above the matched fibroblast controls.

H3K4-me3 modification was detected at most sites within the hGH cluster and extended further 5′ and 3′ than the dimethylated region. The levels of H3K4-me3 modification within this region appeared to increase gradually from the 5′-end of the repeated domains toward the center. The most prominent modification was at the structural genes (hGH/CS amplimer) and immediately 3′ to the hCS genes (amplimer 3′CS1), and the least modified of any site in the cluster corresponded to the P element core (Pe2). There was no appreciable H3K4-me3 modification at any site within the hGH locus in fibroblast chromatin (Fig. 5C, white bars).

H3K4 Methylation in the hGH Cluster in STB Chromatin Is Restricted to the Placental Genes

The di- and trimethylation at H3K4 within the hGH cluster might encompass each of the five repeated domains or be more limited in distribution. To further define this epigenetic profile, the gene units were individually analyzed (Fig. 6). H3K4-me2 modification levels varied dramatically among the five structural genes. When compared with the controls, the modification at hGH-N was limited to background levels (6% of GAPDH). In contrast, hCS-B was modified at 20% of the GAPDH control, hCS-A was modified at 38%, and hGH-V and hCS-L regions were the most strongly modified at 65–72%. The analysis of trimethylation at H3K4 gave a more uniform result (Fig. 6C); all four placental genes in the cluster were robustly modified, ranging from 106–162% of the positive control. In contrast, the hGH-N gene was modified to a far lower extent (39% GAPDH).

Fig. 6.

Histone H3K4 Methylation Is Restricted to the Placentally Expressed Genes of the hGH Cluster A, Representative ChIP analyzed with the gene-specific labeled probes. The ChIP assay was carried out with human placental chromatin using anti-H3K4-me2 antibody. The chromatin DNA (input and immunoprecipitated) was amplified with the 233GH/CS primer set and the amplified DNA was then digested with two sets of restriction enzymes as in Fig. 4A. Digested products were fractionated by gel electrophoresis, and signals were quantified also as in Fig. 4, A and B. The four placental genes are selectively dimethylated at histone H3K4 in placental chromatin. Values on the histogram represent the average ± se from three independent experiments with two placentas. The values for hCS-L and hGH-V are significantly higher than those in hCS-A and hCS-B; all four of these values are significantly higher than hGH-N. C, The four placental genes are selectively enriched for trimethylated histone H3K4 in placental chromatin. Details are as in panel B. The levels of modification of the four placental genes are all significantly greater than hGH-N.

An overview of the ChIP studies of H3K4 methylation at the hGH cluster (Figs. 5 and 6) showed a modification pattern that was distinct from the acetylation pattern. Dimethylation of H3K4 defined four distinct subdomains in the hGH cluster in human placental chromatin corresponding to the four placentally expressed genes. Trimethylation at H3K4 encompassed the entire placental LCR (HSIII-HSV) with robust modification at HSV. It also defined the same four subdomains in the hGH cluster encompassing each placental gene unit. However, the trimethylated subdomains in the cluster extended more 5′ and 3′ than the dimethylated subdomains. The P element core was remarkable in that analyses of both di- and trimethylation at H3K4 demonstrated only minimal modifications at this site. A summary of the histone acetylation and methylation patterns at the hGH cluster in placental chromatin is presented in diagrammatic format in Fig. 7A.

Fig. 7.

Distinct Patterns of Epigenetic Modifications and Gene Activation at the hGH Cluster in Placenta and Pituitary A, Histone acetylation and methylation patterns in placental chromatin predict a looping mechanism. The hGH locus is shown as in Fig. 1. The epigenetic profile in placental chromatin is summarized below the gene diagram: histone H3 acetylation (AcH3), histone H4 acetylation (AcH4), histone H3K4 dimethylation (H3K4-me2), and histone H3K4 trimethylation (H3K4-me3). Robust acetylation of H3 and H4 is limited to the placental LCR (HSIV is selectively modified at histone H4), and the hGH multigene cluster is encompassed in a continuous acetylated domain. This domain excludes the pituitary hGH-N gene (gray rectangle) in the case of AcH3. In contrast, at histone H4 hGH-N is acetylated at an equivalent level with the placental genes. The H3K4 methylation pattern defines discrete subdomains within the cluster. The subdomains defined by H3K4 dimethylation are limited to the immediate vicinity of the four placental genes and exclude the P elements. The distribution of H3K4 trimethylation defines broader subdomains that extend into the 3′-flanking regions of the four genes. Of note, the 28-kb region between the placental LCR and the gene cluster contains a low level of acetylation and trimethylation (dashed lines). The model of placental gene activation proposes a long-distance looping between the modified LCR and the target placental genes over the intervening 28 kb. The establishment of a continuous acetylated domain over the placental genes by multisite HAT targeting and coalescence of the domains is discussed in the text. B, Model for activation of the hGH-N gene in the pituitary. In pituitary, HAT activity is recruited to pituitary-specific HSI and spreads bidirectionally to modify the broad region of chromatin extending from HSV to the hGH-N promoter (16, 17 ). This domain facilitates selective _trans_-factor access to the hGH-N promoter, activating its transcription. The pattern of histone methylation in pituitary chromatin is presently undefined.

DISCUSSION

The five genes of the hGH cluster are highly conserved in structure, yet are expressed selectively in the placenta or pituitary. hGH cluster expression in both tissues is dependent on long-range control by distinct subsets of LCR determinants (14, 15). In transgenic pituitary somatotropes, a single LCR determinant, HSI, recruits histone acetyltransferase (HAT) coactivators to the hGH locus (16, 17) (Fig. 7B). This HAT activity spreads to create a continuous 32-kb domain of modified chromatin that encompasses the hGH-N promoter, facilitating local _trans_-factor binding, and triggering gene transcription (17). In the present study, a distinct pattern of chromatin modifications is defined at the hGH locus in placental chromatin. This epigenetic profile is based on ChIP analyses of 73 sites spanning the locus. The resultant data support and extend a model in which activation of the hGH cluster in the pituitary and placenta are mediated by two distinct pathways of chromatin modification.

A Distinct Pattern of Histone Acetylation and Methylation at the hGH Locus in Placental Chromatin

A comparison of the epigenetic profiles in pituitary and placenta suggest that activation of the placental genes from the hGH cluster relies on a more complex pathway. There are several notable differences between these epigenetic profiles. The acetylated and methylated chromatin domains encompassing the LCR and the gene cluster in placental chromatin are separated by a 28-kb region of minimally modified chromatin (Fig. 7A). This contrasts significantly with the continuous, robustly acetylated domain extending from LCR to hGH-N in pituitary chromatin (Fig. 7B). The placental LCR is further removed from its target genes than the pituitary LCR. The hGH-N gene is maintained in a silent state in placenta despite its location between the LCR and activated placental genes. Finally, the placental genes themselves are encompassed in discrete chromatin subdomains, and they are activated to different magnitudes. Thus, clear differences distinguish the epigenetic pathways used by the partially shared LCR in placental and pituitary genes.

Aspects of the placental gene activation pathway may be inferred by considering the pattern of epigenetic modification. The gene cluster in placental chromatin is encompassed in a continuously acetylated domain, but is subdivided into four subdomains defined by H3K4 di- and trimethylation (Fig. 7A). The continuously acetylated domain suggests that HAT activity, once recruited to the cluster, may spread throughout the cluster. This model is reminiscent of HAT targeting to HSI in pituitary chromatin with its subsequent establishment of an expansive domain encompassing the hGH LCR (Fig. 7B). However, an alternative model based on multisite targeting and coalescence of subdomains may be more directly suited to the placental chromatin. Such multisite targeting could be mediated by determinants repeated in each of the four placental gene units (Fig. 1B). Although it should be possible to distinguish these two models by selective interruption of candidate targeting sites, this approach will be technically challenging due to the conserved nature of the hGH genes and their surrounding sequences.

The sites that recruit HAT coactivators to the hGH cluster in STB nuclei remain to be identified. A role for the P elements in this pathway is supported by its strong evolutionary conservation, its linkage to each of the four placental genes, its robustly acetylated state, and its ability to independently activate placental gene expression in transgenic models (14, 18). Prior work from our laboratory using a mixture of antibodies to acetylated forms of histones H3 and H4 indicated that the P element is robustly acetylated in placental chromatin (18). By using separate antibodies to acetyl-H3 and acetyl-H4, and monitoring at multiple sites throughout the cluster, the present study supported these findings and further revealed that the P element is embedded within an uninterrupted acetylated domain. A less extensive study of histone H4 acetylation by others is consistent with this observation (20). Studies from our laboratory and from others have failed to identify binding factors in human term placental nuclear extracts that might mediate P element activity (Ref.19 and our unpublished studies). More detailed studies in developmentally dynamic settings will be needed to determine whether the P element is a HAT recruitment site and to identify the underlying cis-trans interactions. Generation of a continuous acetylated domain by multisite HAT recruitment with subsequent fusion of acetylated subdomains would represent a novel pathway of epigenetic modification.

The domain of acetylated chromatin that encompasses the hGH gene cluster in STB nuclei is subdivided into smaller subdomains defined by H3K4 methylation (Figs. 5, 6, and 7A). The H3K4 dimethylated subdomains extend from approximately 1 kb 5′ to the transcription start sites of each of the placental genes (boundary between Pe4 and Pe6) to approximately 3.8 kb 3′ to the CS genes (boundary between 3′CS1 and 3′CS2). These dimethylated subdomains encompass the individual placental gene units with their respective promoters and are separated from each other by a segment of chromatin in which dimethylation is at marginal levels (Pe0 to Pe4). The H3K4 trimethylated subdomains overlap the dimethylated subdomains and extend further 5′ and, most markedly, further 3′. Remarkably, H3K4 di- and trimethylation domains both effectively exclude the P element core (Pe2) (Figs. 5 and 7A). These data suggest that H3K4 methylation is most likely mediated by multisite recruitment of histone methyltransferase (HMT) coactivator complexes and that any role(s) of the P element in HAT recruitment is divorced from recruitment of HMT complexes.

The relationship between chromatin acetylation and H3K4 methylation has been studied in a number of systems. The literature supports a model in which HMT recruitment and histone acetylation are linked processes and/or occur in parallel (25–30), although certain exceptions have been reported [_e.g._ the mouse β_-globin_ locus (31)]. Certain studies support a cooperative, one-step recruitment of HAT and HMT (32, 33), whereas others support sequential recruitment (34, 35). The pattern of di- and trimethylation at the hGH cluster is most consistent with four independent targeting events (Fig. 7A). The sites of HMT recruitment within the cluster might correspond to the robustly methylated peaks at the hGH/hCS gene units and the 3′CS1 regions (Fig. 5, B and C). The peak at the gene unit may reflect a linkage between PolII elongation and HMT activity, and the methylated chromatin domain may serve as a memory module or reinforcing structural alteration (36). The 3′CS1 site is located immediately 3′ to a putative hCS enhancer element known to be active after transient cell transfections (37); its in vivo epigenetic modification may relate to its enhancer activity. However, in vivo function of this putative enhancer element has not been demonstrated. Thus, the current data suggest that the domain of histone acetylation and the subdomains of histone H3K4 methylation at the hGH locus in the placenta are established by independent sets of determinants and reflect distinct facets of the placental gene activation pathway.

The low level of chromatin modification in the region between the placental LCR and the gene cluster may be of particular interest. In pituitary chromatin this region is encompassed in a continuous and robustly acetylated domain (Fig. 7B). The levels of histone H3 and H4 acetylation and H3K4 methylation in this region in placental chromatin are low but significant when compared with fibroblast chromatin or to modifications at the flanking SCN4A and TCAM. Studies in the literature suggest that certain modifications, such as H3K4-me3, may be generated during initial gene activation and persist for considerable time after the activation events have occurred (38, 39). It is possible that the low-level trimethylation and acetylation in the region between the LCR and the gene cluster may reflect residual modifications imparted by a prior chromatin activation event in placental cells. Establishing models in which the developmental process of hGH cluster activation can be followed temporally will allow this model to be tested.

Activation of the hGH Genes in Pituitary and Placenta

The hGH LCR is necessary for full activation of the multigene cluster in both pituitary and placenta (14 ,15). In pituitary, a tissue-specific transcription factor, Pit-1, binds to an array of sites at HSI and recruits HAT activity to establish the broad histone acetylation domain that is essential to hGH-N transcription (Fig. 7B and Refs.16 ,17 , and 40). This spreading of histone acetylation conforms to the tracking model for LCR action (41, 42). In contrast, the substantial separation of the placental LCR determinants (HSV-HSIII) from the target genes is more consistent with activation via a looping mechanism (Fig. 7A). Evidence that looping actually occurs in vivo has been recently provided in the mouse β_-globin_ locus where the β-globin LCR is separated from the β-globin target genes by 30 kb of unmodified intervening sequences (43–47). Surprisingly, deletion of the β_-globin_ LCR, although decreasing β-globin expression, does not diminish the levels of histone acetylation at the active β_-globin_ genes (44). These data have been interpreted to indicate that the direct contact of the β-globin LCR with the target β-globin promoters regulates the rate of transcription but does not play an active role in establishment of the acetylated domains (48). Thus, analysis of the β-globin cluster supports a looping interaction between its LCR and target promoters and suggests that the epigenetic imprint at the locus is established before this interaction.

The epigenetic structures of the hGH locus in pituitary and placental chromatin support a model in which looping and tracking pathways may be selectively used in a tissue-specific manner (Fig. 7). In contrast to the findings in the β-globin cluster, the continuous domain of acetylation in pituitary somatotropes is wholly dependent on LCR function; inactivation of HSI results in loss of this acetylated chromatin domain as well as loss of hGH-N expression (17). The general epigenetic structure of the cluster in the placenta is more closely parallel to that of the β_-globin_ locus with a spacer of minimally modified chromatin separating the LCR determinants from the target promoters (45). HSIII–HSV regions and the gene cluster regions may recruit histone modification complexes independently. A looping mechanism may then bring the two activated regions into physical proximity with resultant assembly of promoter initiation complexes and initiation of gene transcription (Fig. 7A). Looping of the LCR to the hGH cluster could be selectively directed to each of the placental gene promoters, yet effectively exclude the hGH-N gene. The question of whether the four subdomains of the hGH cluster are established and function as independent entities is now open to study. Furthermore, the model of independent targeting of modifiers to the LCR and cluster and selective activation of the four placental genes via a looping mechanism can also now be directly tested.

MATERIALS AND METHODS

Isolation of Nuclei from Human Term Placenta

Normal term placentas were obtained from the obstetrical service at the Hospital of the University of Pennsylvania in accordance with an approved Institutional Review Board (IRB) protocol. Intact nuclei were selectively released and isolated from the STB epithelial cells as previously described (14, 18). Briefly, villous fragments were excised from term placentas, rinsed with cold PBS containing 10 mm sodium butyrate, and finely minced. They were suspended in 150 mm NaCl and passed through a 10-gauge screen three times. After centrifugation at 1000 × g for 10 min, the fragments were resuspended in 150 mm NH4Cl containing 12.5 mg/ml ammonium bicarbonate, and incubated on ice for 45 min. This procedure causes selective osmotic lysis of the STB cells due to the expression of carbonic anhydrase at high levels specifically in these cells. The nuclei were collected by centrifugation at 1000 × g for 10 min and resuspended in cold RB [0.1 m NaCl; 50 mm Tris-HCl, pH 8.0; 3 mm MgCl2; proteinase inhibitor complex (Roche Molecular Biochemicals, Indianapolis, IN)] containing 20 mm sodium butyrate. The preparation was passed through a 40-μm nylon filter (Falcon) to remove tissue debris. The presence of intact nuclei was confirmed by microscopic observation. The isolated nuclei were stored at −80 C in glycerol storage buffer [40% glycerol; 50 mm Tris-HCl, pH 8.3; 5 mm MgCl2; 0.1 mm EDTA] containing 20 mm sodium butyrate before analysis.

Purification of Nuclei from Cultured Cells

The human skin fibroblast cell line, CCDsk-25 (American Type Culture Collection, Manassas, VA), was cultured at 37 C with 5% CO2 in MEM (Invitrogen Life Technologies, Carlsbad, CA) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Invitrogen Life Technologies). Cells were washed with PBS and lysed with Nonidet P-40 (NP-40) lysis buffer [10 mm Tris-HCl, pH 7.5; 10 mm NaCl; 3 mm MgCl2; 0.5% NP-40; and proteinase inhibitor cocktail (Roche Molecular Biochemicals)] containing 10 mm sodium butyrate on ice. The lysate was incubated on ice for 10 min, and then nuclei were collected by centrifugation. The nuclei were washed with NP-40 lysis buffer twice and stored in glycerol storage buffer containing 20 mm sodium butyrate before analysis.

Chromatin Immunoprecipitation Assay of Unfixed Chromatin

The ChIP assay was conducted as previously described with minor modifications (16). Digestions used 0.6 mg nuclei in 1 ml digestion buffer [50 mm NaCl; 20 mm Tris-HCl, pH 7.5; 3 mm MgCl2; 1 mm CaCl2; 10 mm sodium butyrate; 0.1 mm phenylmethylsulfonyl fluoride (PMSF)]. These were digested with 20 U of micrococcal nuclease (Amersham Biosciences, Piscataway, NJ) at 37 C for 6 min. Digestions were terminated by addition of EDTA to 0.5 mm, and samples was centrifuged at 12,000 × g for 10 min at 4 C to generate supernatant S1. The pellet was resuspended in 0.6 ml of low-salt lysis buffer [10 mm Tris-HCl, pH 7.5; 10 mm sodium butyrate; 0.25 mm EDTA; 0.1 mm PMSF], incubated on ice for 10 min, and then centrifuged as before. Repeating this step generated supernatants S2 and S3. The three supernatants were combined and concentrated with Amicon Microcon centrifugal filters (Millipore Corp., Bedford, MA). Fifteen percent of the resulting soluble chromatin was kept as input, and the rest was separated into two fractions, each adjusted to 0.5 ml with immunoprecipitation (IP) buffer (50 mm NaCl; 20 mm Tris-HCl, pH 7.5; 5 mm EDTA; 10 mm sodium butyrate). The immunoprecipitation reactions were performed on these fractions by adding 10 μl of antibody and incubating the samples overnight at 4 C with gentle rotation. The antibodies used were: antiacetyl histone H3 (catalog no. 06–599, Upstate Biotechnology, Inc., Lake Placid, NY), antiacetyl histone H4, ChIP Grade (catalog no. 06–866, Upstate Biotechnology, Inc.), antidimethyl histone H3 (Lys4) (catalog no. 07–030, Upstate Biotechnology, Inc.), and antitrimethyl histone H3 (Lys4) (catalog no. ab8580, Abcam, Cambridge, UK). The chromatin samples were added to 50 mg protein-A sepharose beads (Amersham Biosciences), prewashed with immunoprecipitation buffer, and incubated at 4 C for 3 h with gentle rotation. The beads were washed three times with 10 ml buffer A (50 mm Tris-HCl, pH 7.5; 10 mm EDTA; 10 mm sodium butyrate; 0.1 mm PMSF) containing increasing amounts of NaCl to 150 mm. Bound fractions were eluted twice by incubating the beads in 0.3 ml buffer A containing 1% sodium dodecyl sulfate (SDS) at room temperature for 20 min with constant inversion. DNA was isolated from the input and the bound fractions by phenol/chloroform extraction and ethanol precipitation. The precipitates were dissolved in water; aliquots containing 0.16%, 0.08%, and 0.04% of each sample were analyzed by PCR to confirm that each assay was within the linear range of amplification.

PCRs were performed as follows: 94 C for 2 min, 21–28 cycles of 94 C for 30 sec, 55–66 C for 30 sec, 72 C for 2 min, and 72 C for 7 min. The reaction mixtures (25 μl) contained 0.5 μl DNA samples, 1× PCR buffer, 0.2 mm deoxynucleotide triphosphate mixture, 0.5 μm paired primers, and 0.6 U of Ampli-Taq (PE Applied Biosystems, Norwalk, CT). The primers used in this study are listed in Table 1, and positions of their amplimers are indicated in Fig. 1. The amplified products were fractionated on 0.9% agarose gels and then transferred to Hybond-N+ nylon membrane (Amersham Biosciences) using 0.4 m NaOH as transfer buffer. The blots were hybridized at 65 C for 16 h in 0.5 m sodium phosphate, pH 7.2; 5% SDS; 1% BSA; 0.1 mm EDTA; and 0.1 mg/ml denatured salmon sperm DNA with 32P-labeled probes. The probes were amplified by PCR using the hGH/P1 clone (15) or human genomic DNA as template and were subsequently isolated from agarose gels after electrophoresis. The membranes were washed with 0.1× sodium chloride/sodium citrate (SSC), 0.1% SDS at 65 C. Signal intensities were quantified by PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA) using ImageQuant software. Ratios between bound and input fractions were generated using a linear PCR amplification range. Each ratio was normalized to the comparable signal detected at the ubiquitously expressed GAPDH promoter (defined as 100).

ChIP Assay to Distinguish the Four P Elements

The ChIP assay was performed as described above except that a lower amount of micrococcal nuclease (2.5 U) was used to prepare longer segments of chromatin (average length, 1.2 kb). The bound fraction together with the input fraction were subjected to PCR amplification using a common primer pair (5′-TGGAGAGGTGAGA GGAGACACTC-3′ and 5′-ATGGGGTGCCTTGTGGAAACCTC-3′) to generate the 930P amplimer (Fig. 1B) that anneals to identical sequences in all four P elements. The PCR was performed in 25 μl as follows: 94 C for 2 min followed by 23 cycles of 94 C for 30 sec, 66 C for 30 sec, 72 C for 2 min, and 72 C for 7 min. Portions (5 μl) of the products were applied to each lane of 0.9% agarose gels that were electrophoresed and Southern blotted. The DNA blots were hybridized with each specific oligoprobe at 42 C for 16 h in 5× SSC, 50 mm sodium acetate, pH 8.0; 0.1% SDS; 10× Denhardt’s solution; 0.01% sodium pyrophosphate; and 0.1 mg/ml denatured salmon sperm DNA. The sequences of the oligoprobes are shown in Fig. 3A. The oligoprobes were 5′-end labeled with T4 polynucleotide kinase (New England Biolabs, Beverly, MA) and [32P]γATP at 37 C for 60 min. The membranes were washed with 0.1× SSC, 0.1% SDS at 42 C, and the specific signals were detected as above.

PCR and Restriction Enzyme Digestions that Distinguish the Five hGH/hCS Genes

A common primer pair (5′-AGCATGTGTGGGAGGAGCTT-3′ and 5′-GATTTTAGGGGCGCTTACCT-3′) was identified that amplifies the region located between 146 bp upstream and 87 bp downstream of the transcription initiation site of each _GH_-related gene (amplimer 233GH/CS; Fig. 1B). After 5′-end labeling the sense primer with [32P]γATP as described previously, PCR was carried out as follows: 94 C for 2 min followed by 25 cycles of 94 C for 30 sec, 57 C for 30 sec, 72 C for 2 min, and 72 C for 7 min. The products were digested with restriction enzymes _Hin_fI, _Dra_III, and _Pst_I and divided into two portions. One portion was further digested with _Sml_I, the other with _Msc_I. The digestion products were separated on a 6% polyacrylamide, 7 m urea gels, and bands were detected by autoradiography. Based upon specific fragment sizes on the gel compared with the known DNA sequence of the region, individual bands were assigned to each of the five genes.

Acknowledgments

We thank Drs. Yugong Ho and Brian Shewchuk for critical reading of the manuscript.

This work was supported by NIH Grant HD25147 (to N.E.C. and S.A.L.).

Abbreviations:

• ChIP,
  Chromatin immunoprecipitation;
• CS,
  chorionic somatomammotropin;
• GAPDH,
  glyceraldehyde-3-phosphate dehydrogenase;
• HAT,
  histone acetyltransferase;
• HMT,
  histone methyltransferase;
• HS,
• LCR,
• PMSF,
  phenylmethylsulfonylfluoride;
• SDS,
• SSC,
  sodium chloride/sodium citrate;
• STB,

1

Turner

BM

2002

Cellular memory and the histone code.

Cell

111

:

285

–

291

2

Strahl

BD

,

Allis

CD

2000

The language of covalent histone modifications.

Nature

403

:

41

–

45

3

Jenuwein

T

,

Allis

CD

2001

Translating the histone code.

Science

293

:

1074

–

1080

4

Agalioti

T

,

Lomvardas

S

,

Parekh

B

,

Yie

J

,

Maniatis

T

,

Thanos

D

2000

Ordered recruitment of chromatin modifying and general transcription factors to the IFN-β promoter.

Cell

103

:

667

–

678

5

Cosma

MP

,

Tanaka

T

,

Nasmyth

K

1999

Ordered recruitment of transcription and chromatin remodeling factors to a cell cycle- and developmentally-regulated promoter.

Cell

97

:

299

–

311

6

Rice

JC

,

Allis

CD

2001

Histone methylation versus histone acetylation: new insights into epigenetic regulation.

Curr Opin Cell Biol

13

:

263

–

273

7

Berger

SL

2002

Histone modifications in transcriptional regulation.

Curr Opin Genet Dev

12

:

142

–

148

8

Lachner

M

,

O’Sullivan

RJ

,

Jenuwein

T

2003

An epigenetic road map for histone lysine methylation.

J Cell Sci

116

:

2117

–

2124

9

Chen

EY

,

Liao

Y-C

,

Smith

DH

,

Barrera-Saldaña

HA

,

Gelinas

RE

,

Seeburg

PH

1989

The human growth hormone locus: nucleotide sequence, biology, and evolution.

Genomics

4

:

479

–

497

10

Barrera-Saldaña

HA

1998

Growth hormone and placental lactogen: biology, medicine and biotechnology.

Gene

211

:

11

–

18

11

MacLeod

JN

,

Lee

AK

,

Liebhaber

SA

,

Cooke

NE

1992

Developmental control and alternative splicing of the placentally expressed transcripts from the human growth hormone gene cluster.

J Biol Chem

267

:

14219

–

14226

12

Lacroix

MC

,

Guibourdenche

J

,

Frendo

JL

,

Muller

F

,

Evain-Brion

D

2002

Human placental growth hormone—a review.

Placenta

23

(Suppl A):

S87

–

S94

13

Misra-Press

A

,

Cooke

NE

,

Liebhaber

SA

1994

Complex alternative splicing partially inactivates the human chorionic somatomammotropin-like (hCS-L) gene.

J Biol Chem

269

:

23220

–

23229

14

Jones

BM

,

Monks

BR

,

Liebhaber

SA

,

Cooke

NE

1995

The human growth hormone gene is regulated by a multicomponent locus control region.

Mol Cell Biol

15

:

7010

–

7021

15

Su

Y

,

Liebhaber

SA

,

Cooke

NE

2000

The human growth hormone gene cluster locus control region supports position-independent pituitary- and placenta-specific expression in the transgenic mouse.

J Biol Chem

275

:

7902

–

7909

16

Elefant

F

,

Cooke

NE

,

Liebhaber

SA

2000

Targeted recruitment of histone acetyltransferase activity to a locus control region.

J Biol Chem

275

:

13827

–

13834

17

Ho

Y

,

Elefant

F

,

Cooke

N

,

Liebhaber

S

2002

A defined locus control region determinant links chromatin domain acetylation with long-range gene activation.

Mol Cell

9

:

291

–

302

18

Elefant

F

,

Su

Y

,

Liebhaber

SA

,

Cooke

NE

2000

Patterns of histone acetylation suggest dual pathways for gene activation by a bifunctional locus control region.

EMBO J

19

:

6814

–

6822

19

Nachtigal

MW

,

Nickel

BE

,

Cattini

PA

1993

Pituitary-specific repression of placental members of the human growth hormone gene family: a possible mechanism for locus regulation.

J Biol Chem

268

:

8473

–

8479

20

Norquay

LD

,

Yang

X

,

Sheppard

P

,

Gregoire

S

,

Dodd

JG

,

Reith

W

,

Cattini

PA

2003

RFX1 and NF-1 associate with P sequences of the human growth hormone locus in pituitary chromatin.

Mol Endocrinol

17

:

1027

–

1038

21

Iizuka

M

,

Smith

MM

2003

Functional consequences of histone modifications.

Curr Opin Genet Dev

13

:

154

–

160

22

Kouzarides

T

2002

Histone methylation in transcriptional control.

Curr Opin Genet Dev

12

:

198

–

209

23

Santos-Rosa

H

,

Schneider

R

,

Bannister

AJ

,

Sherriff

J

,

Bernstein

BE

, Tolga

Emre

NC

,

Schreiber

SL

,

Mellor

J

,

Kouzarides

T

2002

Active genes are tri-methylated at K4 of histone H3.

Nature

419

:

407

–

411

24

Nishioka

K

,

Rice

JD

,

Sarma

K

,

Erdjumet-Broamge

H

,

Allis

CD

,

Tempst

P

,

Reinberg

D

2002

Set9, a novel histone H3 methytransferase that facilitates transcription by precluding histone tail modifications required for heterochromatin formation.

Genes Dev

16

:

479

–

489

25

Litt

MD

,

Simpson

M

,

Gaszner

M

,

Allis

CD

,

Felsenfeld

G

2001

Correlation between histone lysine methylation and developmental changes at the chicken β-globin locus.

Science

293

:

2453

–

2455

26

Boggs

BA

,

Cheung

P

,

Heard

E

,

Spector

DL

,

Chinault

AC

,

Allis

CD

2002

Differentially methylated forms of histone H3 show unique association patterns with inactive human X chromosomes.

Nat Genet

30

:

73

–

76

27

Perk

J

,

Makedonski

K

,

Lande

L

,

Cedar

H

,

Razin

A

,

Shemer

R

2002

The imprinting mechanism of the Prader-Willi/Angelman regional control center.

EMBO J

21

:

5807

–

5814

28

Fournier

C

,

Goto

Y

,

Ballestar

E

,

Delaval

K

,

Hever

AM

,

Esteller

M

,

Feil

R

2002

Allele-specific histone lysine methylation marks regulatory regions at imprinted mouse genes.

EMBO J

21

:

6560

–

6570

29

Nguyen

CT

,

Weisenberger

DJ

,

Velicescu

M

,

Gonzales

FA

,

Lin

JCY

,

Liang

G

,

Jones

PA

2002

Histone H3-lysine 9 methylation is associated with aberrant gene silencing in cancer cells and is rapidly reversed by 5-aza-2′-deoxycytidine.

Cancer Res

62

:

6456

–

6461

30

Fahrner

JA

,

Eguchi

S

,

Herman

J

,

Baylin

SB

2002

Dependence of histone modifications and gene expression on DNA hypermethylation in cancer.

Cancer Res

62

:

7213

–

7218

31

Kiekhaefer

CM

,

Grass

JA

,

Johnson

KD

,

Boyer

ME

,

Bresnick

EH

2002

Hematopoietic-specific activators establish an overlapping pattern of histone acetylation and methylation within a mammalian chromatin domain.

Proc Natl Acad Sci USA

99

:

14309

–

14314

32

Milne

TA

,

Briggs

SD

,

Brock

HW

,

Martin

ME

,

Gibbs

D

,

Allis

CD

,

Hess

JL

2002

MLL targets SET domain methyltransferase activity to Hox gene promoters.

Mol Cell

10

:

1107

–

1117

33

Nakamura

T

,

Mori

T

,

Tada

S

,

Krajewski

W

,

Rozovskaia

T

,

Wassell

R

,

Dubois

G

,

Mazo

A

,

Croce

CM

,

Canaani

E

2002

ALL-1 is a histone methyltransferase that assemble a supercomplex of proteins involved in transcriptional regulation.

Mol Cell

10

:

1119

–

1128

34

Wang

H

,

Cao

R

,

Xia

L

,

Erdjument-Bromage

H

,

Borchers

C

,

Tempst

P

,

Zhang

Y

2001

Purification and functional characterization of a histone H3-lysine 4-specific methyltransferase.

Mol Cell

8

:

1207

–

1217

35

Martens

JHA

,

Verlaan

M

,

Kalkhoven

E

,

Zantema

A

2003

Cascade of distinct histone modifications during collagenase gene activation.

Mol Cell Biol

23

:

1808

–

1816

36

Gerber

M

,

Shilatifard

A

2003

Transcriptional elongation by RNA polymerase II and histone methylation.

J Biol Chem

278

:

26303

–

26306

37

Walker

WH

,

Fitzpatrick

SL

,

Saunders

GF

1990

Human placental lactogen transcriptional enhancer: tissue specificity and binding with specific proteins.

J Biol Chem

265

:

12940

–

12948

38

Ng

HH

,

Robert

F

,

Young

RA

,

Struhl

K

2003

Targeted recruitment of Set1 histone methylase by elongating Pol II provides a localized mark and memory of recent transcriptional activity.

Mol Cell

11

:

709

–

719

39

Hampsey

M

,

Reinberg

D

2003

Tails of intrigue: phosphorylation of RNA polymerase II mediates histone methylation.

Cell

113

:

429

–

432

40

Shewchuk

BM

,

Asa

SL

,

Cooke

NE

,

Liebhaber

SA

1999

Pit-1 binding site at the somatotrope-specific DNase I hypersensitive sites I, II of the human growth hormone locus control region are essential for in vivo hGH-N gene activation.

J Biol Chem

274

:

35725

–

35733

41

Bulger

M

,

Groudine

M

1999

Looping versus linking: toward a model for long-distance gene activation.

Genes Dev

13

:

2465

–

2477

42

Li

Q

,

Peterson

KR

,

Fang

X

,

Stamatoyannopoulos

G

2002

Locus control regions.

Blood

100

:

3077

–

3086

43

Forsberg

EC

,

Downs

KM

,

Christensen

HM

,

Im

H

,

Nuzzi

PA

,

Bresnick

EH

2000

Developmentally dynamic histone acetylation pattern of a tissue-specific chromatin domain.

Proc Natl Acad Sci USA

97

:

14494

–

14499

44

Schübeler

D

,

Groudine

M

,

Bender

MA

2001

The murine β-globin locus control region regulates the rate of transcription but not the hyperacetylation of histones at the active genes.

Proc Natl Acad Sci USA

98

:

11432

–

11437

45

Bulger

M

,

Sawado

T

,

Schübeler

D

,

Groudine

M

2002

ChIPs of the β-globin locus: unraveling gene regulation within an active domain.

Curr Opin Genet Dev

12

:

170

–

177

46

Bulger

M

,

Schübeler

D

,

Bender

MA

,

Hamilton

J

,

Farrell

CM

,

Hardison

RC

,

Groudine

M

2003

A complex chromatin landscape revealed by patterns of nuclease sensitivity and histone modification within the mouse β-globin locus.

Mol Cell Biol

23

:

5234

–

5244

47

Tolhuis

B

,

Palstra

R-J

,

Splinter

E

,

Grosveld

F

, de

Laat

W

2002

Looping and interaction between hypersensitive sites in the active β-globin locus.

Mol Cell

10

:

1453

–

1465

48

Sawado

T

,

Halow

J

,

Bender

MA

,

Groudine

M

2003

The β-globin locus control region (LCR) functions primarily by enhancing the transition from transcription initiation to elongation.

Genes Dev

17

:

1009

–

1018

Copyright © 2004 by The Endocrine Society

Citations

Views

Altmetric

Metrics

Total Views 890

653 Pageviews

237 PDF Downloads

Since 2/1/2017

Citations

53 Web of Science

×

Email alerts

More on this topic

Related articles in PubMed

Citing articles via

• Latest

• Most Read

• Most Cited

More from Oxford Academic