Transcription Factor Interactions and Chromatin Modifications ...

MOLECULAR AND CELLULAR BIOLOGY, Mar. 2005, p. 2147–2157 0270-7306/05/$08.00⫹0 doi:10.1128/MCB.25.6.2147–2157.2005 Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Vol. 25, No. 6

Transcription Factor Interactions and Chromatin Modifications Associated with p53-Mediated, Developmental Repression of the Alpha-Fetoprotein Gene Thi T. Nguyen, Kyucheol Cho, Sabrina A. Stratton, and Michelle Craig Barton* Department of Biochemistry and Molecular Biology, Program in Genes and Development, Graduate School of Biological Sciences, University of Texas M. D. Anderson Cancer Center, Houston, Texas Received 10 September 2004/Returned for modification 18 October 2004/Accepted 10 December 2004

We performed chromatin immunoprecipitation (ChIP) analyses of developmentally staged solid tissues isolated from wild-type and p53-null mice to determine specific histone N-terminal modifications, histonemodifying proteins, and transcription factor interactions at the developmental repressor region (ⴚ850) and core promoter of the hepatic tumor marker alpha-fetoprotein (AFP) gene. Both repression of AFP during liver development and silencing in the brain, where AFP is never expressed, are associated with dimethylation of histone H3 lysine 9 (DiMetH3K9) and the presence of heterochromatin protein 1 (HP1). These heterochromatic markers remain localized to AFP during developmental repression but spread to the upstream albumin gene during silencing. Developmentally regulated decreases in levels of acetylated H3 (AcH3K9) and H4 (AcH4) and of di- and trimethylated H3K4 (DiMetH3K4 and TriMetH3K4) occur at both the core promoter and distal repressor regions of AFP. Hepatic expression of AFP correlates with FoxA interaction at the repressor region and the binding of RNA polymerase II and TATA-binding protein to the core promoter. p53 acts as a developmental repressor of AFP in the liver by binding to chromatin, excluding FoxA interaction and targeting mSin3A/HDAC1 to the distal repressor region. p53-null mice exhibit developmentally delayed AFP repression, concomitant with acetylation of H3K9, methylation of H3K4, and loss of DiMetH3K9, mSin3A/ HDAC1, and HP1 interactions. methylation, and increased H3K9 methylation, concomitant with gene-specific silencing during differentiation (65). Active alpha-fetoprotein (AFP) gene expression occurs during the rapid growth and replication periods of fetal hepatic development. After concomitant tissue-specific activation of the highly homologous AFP and albumin (ALB) genes in fetal liver, the expression levels of each are regulated autonomously. As the postnatal liver differentiates and gains full metabolic function, AFP gene activity is repressed to nearly undetectable levels and ALB expression is maintained (reviewed in references 8, 66, and 73). Renewed AFP expression occurs only when the differentiated liver exits G0 and enters a program of resumed cellular proliferation, as a consequence of hepatic tumorigenesis, liver regeneration, or tissue damage due to chronic disease (reviewed in references 4 and 63). Reactivation of AFP gene expression in hepatocellular carcinomas is widespread, with 70 to 85% of all hepatocellular carcinoma patients expressing the gene at varying levels (48), underscoring one role of AFP as a diagnostic tumor marker. Previously, we showed that the tumor suppressor p53 is a sequence-specific repressor of AFP transcription which binds at ⫺850 within the identified developmental repressor and represses transcription of chromatin-assembled AFP in vitro (39, 52). Chromatin immunoprecipitation (ChIP) and expression analyses of developmentally staged wild-type (WT) and p53-null liver tissue, presented here, support the function of p53 as a direct developmental repressor of AFP expression in vivo. These studies suggest a mechanism whereby repressionassociated histone modifications, histone-modifying complexes, and heterochromatin protein 1 (HP1) are targeted specifically to AFP by p53 during hepatic development.

Regulated alterations of chromatin structure are integral to activation and repression of gene expression. Temporal studies of gene expression and chromatin structure during the cell cycle, in response to hormones, tissue-specific activators, and cellular differentiation, have offered insights into the orchestration of transcription factors, chromatin remodeling complexes, and histone modifiers, which combine to alter chromatin and regulate transcription (1, 13, 20, 30, 55, 58, 62, 65). The patterns of progressive chromatin alterations and transcription factor interactions can be both gene and signal specific. These parameters, including specific histone modifications, establish a chromatin structure amenable to active transcription or an exclusion of gene expression (59, 64, 67). To date, there is considerably less knowledge of changes in chromatin structure that promote repression, versus activation, of ongoing gene expression, especially during mammalian development. Early studies of rat skeletal muscle differentiation in vivo documented a decrease in global levels of acetylated histones from 3 days postpartum (pp) to the age of 1 month (54). More recently, the essential roles that class II histone deacetylases (HDACs) play in muscle differentiation and the signaling pathways that control their activities have been defined (reviewed in reference 44). Detailed studies of thymocyte maturation have shown that an ordered process of histone modification occurs: deacetylation of H3K9, loss of H3K4

* Corresponding author. Mailing address: Department of Biochemistry and Molecular Biology, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030. Phone: (713) 794-1161. Fax: (713) 563-2969. E-mail: [email protected] .tmc.edu. 2147

2148

NGUYEN ET AL. MATERIALS AND METHODS

RT-PCR analysis. Total RNA from C57BL/6J WT and p53⫺/⫺ mouse liver tissue and Hepa 1-6 cells was extracted with TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. cDNA was obtained by reverse transcription of 2 ␮g of RNA using the SuperScript first-strand synthesis system (Invitrogen). Reverse transciption-PCR (RT-PCR) was carried out using primers for AFP, ALB, or Brn-3b simultaneously with GAPDH primers. RT-PCR primers were as follows: AFP (forward primer, 5⬘-CCCACTTCCAGCACTGCCTGCG G-3⬘; reverse primer, 5⬘-GGCTGCAGCAGCCTGAGAGTC-3⬘), Albumin (forward primer, 5⬘-TGTCCGTCAGAGAATGAAGTGC-3⬘; reverse primer, 5⬘-A AGACATCCTTGGCCTCAGCA-3⬘), Brn-3b (forward primer, 5⬘-AGTCTCTC ACGCTGTCACACAA-3⬘; reverse primer, 5⬘-AGAGGCAGAAGAGACAAG AAGG-3⬘), and GAPDH (forward primer, 5⬘-TTCACCACCATGGAGAAGG C-3⬘; reverse primer, 5⬘-GGCATGGACTGTGGTCATGA-3⬘). PCR products were separated on 6% polyacrylamide gels and stained with Sybr Green (Sigma). Multiple dilutions of input cDNA and respective PCR products were analyzed, and results were quantified, by using ImageQuant 5.2 and NIH Image 1.63 software, with comparison to levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). ChIP of solid tissues. Liver and brain tissues from C57BL/6J WT and p53⫺/⫺ mice were isolated, cross-linked, and processed for ChIP analyses as described previously, with some modifications (71). Modifications to this protocol included sonication of the cross-linked chromatin with glass beads (Sigma G-1277), followed by digestion with micrococcal nuclease (4 U/␮l; Worthington Biochemicals) to generate fragments with an average length of less than 400 to 500 bp, as determined empirically by agarose gel electrophoresis of fragmented chromatin samples. The fragmented chromatin lysate was incubated with nonspecific immunoglobulin G (IgG) for 1 h to preclear the lysate. This precleared lysate was incubated overnight with specific antibodies, and protein A-Sepharose beads (preblocked with salmon sperm DNA) were added for a 1-h incubation, followed by recovery of antibody precipitates (6, 7). Antibodies against the following proteins and residues were added to 20 ␮l of lysate for ChIP: p53 (Ab1 OP03, 15 ␮l; Oncogene), normal sheep IgG (12-369, 5 ␮l; Upstate), acetylated histone H4 (AcH4) (06-860, ChIP grade, 5 ␮l; Upstate), acetylated histone H3 lysine 9 (AcH3K9) (06-942, 5 ␮l; Upstate), dimethylated H3 K4 (DiMetH3K4) (07-030, 5 ␮l; Upstate), DiMetH3K9 (07-212, 5 ␮l; Upstate), trimethylated H3 K4 (TriMetH3K4) (ab8580-50, 5 ␮l; Abcam), H3 (ab1791-100, 10 ␮l; Abcam), FoxA (ab5089-100, 6 ␮l; Abcam), RNA polymerase II phosphoserine 5 (MMS-134R, 2 ␮l; Covance), mSin3A (06-913, 3 ␮l; Upstate), HDAC1 (06-720, 5 ␮l; Upstate), HP1␣ (05-689, 5 ␮l; Upstate), and TFIID/TATA-binding protein (TBP) (sc-273, 30 ␮l; Santa Cruz). To analyze specific antibody-bound DNA fractions, quantitative PCRs were performed using Taq polymerase (Continental Labs) and primers were generated to detect the AFP SBE/p53RE region of ⫺887 to ⫺762 (forward primer, TAAAAAATAAACTCAACTACATATG; reverse primer, GAAAACTTTTAA AACTTCCC), the AFP start site region of ⫺82 to ⫹94 (forward primer, CAT ATGTTTGCTCACTGAAGGTTAC; reverse prime, CGCAGCGAAATGTAG CAGGAGGA), the ALB enhancer region of 151 bp, 11 kb upstream of the transcription start site (7, 40, 45) (forward primer, GGGACGAGATGTACTT TGTG; reverse primer, GATCAGTCCAAACTTCTTTCTG), and the Brn-3b region from 151 bp 5⬘ of the stop codon to 186 bp downstream (46) (forward primer, TCTGGAAGCCTACTTCGCCA; reverse primer, CCGGTTCACAAT CTCTCTGA). To ensure that the PCRs were in linear range, several serial dilutions of the input DNA and two dilutions of each of the bound DNA fractions were used to quantify products. In addition, 24 to 27 PCR cycles were performed for each bound DNA fraction and input. PCR products were separated on 6% polyacrylamide gels and stained with Sybr Green (Sigma). DNA bands were visualized and quantified by using ImageQuant 5.2 and NIH Image 1.63 software. The percentage of the input that was bound was calculated by dividing the value of the bound fraction by the average of values for the input dilutions in the linear range and multiplying by the respective dilution factor. Additionally, we used an antibody that recognizes the C-terminal tail of histone H3, which is not modification specific, as an internal control and a measure of chromatin recovery for each lysate. Quantifications from multiple individual experiments were taken, and the average values were plotted graphically using Microsoft Excel software. Error bars represent standard deviations.

RESULTS Chromatin modification during developmental repression of AFP. The time course of developmental regulation of AFP

MOL. CELL. BIOL.

expression in WT and transgenic mouse livers was previously determined by Northern blotting (3, 5, 37) and RNase protection analyses (35). We chose several time points for RNA analyses by RT-PCR to establish relative AFP expression levels, from 2 days pp to 4 months, for both WT and p53-null mouse liver. The amounts of stable AFP RNA were compared to those of GAPDH RNA, a gene constitutively expressed during liver development (Fig. 1), at varying numbers of cycles of PCR amplification (data not shown). Our previous studies showed that the p53 tumor suppressor protein acted as a direct repressor of AFP expression by sequence-specific binding to an overlapping FoxA/p53/Smad binding element (SBE/p53RE), centered at ⫺850 of the AFP upstream region (39, 52). Clearly, p53 does not act alone in developmental repression of AFP, but its absence delays the time course of repression, with 1.7-fold-higher expression in p53-null liver than in WT liver at 12 days pp (Fig. 1B). At 8 days pp, AFP expression is robust and equivalent in both WT and null livers, as well as in cultured mouse Hepa 1-6 hepatoma cells, where AFP is transcribed as a tumor marker gene (16). Albumin RNA serves as a positive control for a liverspecific gene with nearly constitutive expression during development, following its initial activation shortly after AFP in fetal liver (74, 75). The tissue-specificity of AFP and albumin expression was affirmed by analysis of RNA from mouse brain tissue and was compared with that of a gene expressed only in developing mid-brain and retinal cells, Brn-3b (51), as a positive control (Fig. 1A). In order to assess chromatin structural alterations, which occur during development and in specific tissues, and the transcription factors that regulate and target these changes in chromatin and expression, we performed ChIP analyses of mouse liver and brain tissue at specific times of development. Formaldehyde-cross-linked chromatin fragments were isolated from mouse tissues at 8 days and 2 months after birth and were sheared by sonication and micrococcal nuclease digestion to less than 500 bp. We chose 8-day pp liver as our source of AFP-expressing tissue, because the amount of tissue needed for ChIP analyses is significant (70), and AFP expression levels at this stage are comparable to those in tumor-derived Hepa 1-6 cells. At the age of 2 months, AFP expression is fully repressed in WT mouse liver. DNA-associated histones and transcription factors were compared at two specific regions of the AFP gene: the SBE/ p53RE region (⫺850) within the previously identified developmental repressor of AFP, which we found is a regulatory center for cooperative p53- and transforming growth factor ␤-mediated repression of AFP (data not shown), and the core promoter surrounding the transcription start site at position ⫹1 of AFP. Antibodies that recognize specific modifications of residues within histone H3 amino tails—AcH3K9, DiMetH3K9, DiMetH3K4, and tetra-acetylated histone H4 (K5, K8, K12, and K16)—were used to precipitate cross-linked chromatin fragments. Multiple analyses were performed for each test antibody, and a titration of the input lysate was used to establish the linear range of PCR detection for each lysate. An antibody that recognizes the C-terminal tail of histone H3, which is not modification specific, was also used to standardize chromatin lysates with regard to protein-chromatin recovery.

VOL. 25, 2005

p53 TARGETS REPRESSION-ASSOCIATED HISTONE MODIFICATION

2149

FIG. 1. Tissue- and developmental-stage-specific regulation of AFP expression in WT and p53-null liver. (A) RT-PCR analysis was performed with RNAs isolated from liver tissues excised from 2-, 8-, and 12-day-old and 2- and 4-month-old WT and p53-null mice. Brain tissue RNAs from 8-day- and 2-month-old mice served as tissue-specific negative controls. Tumor-derived Hepa 1-6 cellular RNA served as a positive control for AFP expression. Primers specific for AFP, ALB, Brn-3b, and GAPDH were used to determine the relative levels of stable RNA expressed from each gene. (B) Multiple dilutions of input cDNA were analyzed as for panel A, over the linear range of PCR amplification. Levels of AFP expression were determined relative to GAPDH levels in the appropriate samples.

Representative data are shown in each figure, along with quantified results of multiple lysates and analyses. The general paradigms of the “histone code,” which holds that histone acetylation is indicative of transcription activation and that H3K9 methylation, concomitant with histone deacetylation, is associated with repression, are illustrated by analysis of the AFP developmental repressor (Fig. 2A, SBE/p53RE) and core promoter (Fig. 2B, start site) regions. At 8 days of development, histones H3 and H4 are highly acetylated, but they are considerably less so at 2 months, when AFP is fully repressed. Methylation of H3K4 is correlated with active transcription, as well, and decreases with repression. Histone H3K9 cannot be simultaneously acetylated and methylated within an individual histone amino tail, and reciprocal changes in acetylation and methylation of H3K9 occur during develop-

mental repression at both the AFP distal repressor and the core promoter. DiMetH3K9 levels are very low in both chromatin regions at 8 days but are readily detectable at the age of 2 months. The distribution of histone modifications, both acetylation and methylation, is roughly equivalent between the core promoter and the distal repressor at both times. However, DiMetH3K4 levels showed a greater decrease at the transcription start site region (⫹1) than at SBE/p53RE (⫺850) when AFP was repressed. This modification of histone H3 is frequently, but not always, associated with transcription activation. Modification of H3K4 to create trimethylated H3K4 has been more closely associated with active chromatin within the chick globin locus (58). We assessed di- and trimethylation of H3K4 during AFP repression, focusing on the core promoter

2150

NGUYEN ET AL.

MOL. CELL. BIOL.

FIG. 2. Changes in histone modifications during developmental repression of AFP conform to the paradigms of active and repressed chromatin. (A) ChIP analysis of the AFP developmental repressor region. (Top) Graph summarizing the results of multiple ChIP determinations. Each bar represents the average result from three to eight ChIP experiments; error bars, standard deviations. Cross-linked lysates were prepared from 8-day pp (gray bars) and 2-month (white bars) WT mouse liver tissue. Primers specific for the ⫺850 SBE/p53RE were used in PCR amplification. Percent bound (% bound) represents a quantified value for antibody-precipitated, specific DNA divided by the amount of specific DNA present in the input lysate. (Bottom) Representative PCR analysis of the DNA present in chromatin fragments precipitated by each specific antibody, as indicated above each lane. Negative controls include nonspecific IgG, no antibody addition to the immunoprecipitation (no Ab), and no precipitated DNA added to the PCR (no DNA). An unmodified anti-histone H3 antibody (H3) served as a positive control. (B) ChIP analysis of the AFP start site region. (Top) Graph of multiple ChIP assays, as described for panel A. Primers specific for the AFP start site region were used in PCR amplification. (Bottom) Representative PCR analysis, as described for panel A. (C) PCR analysis of DiMetH3K4 and TriMetH3K4, in comparison to DiMetH3K9, associated with chromatin present at the AFP start site region in 8-day and 2-month WT mouse liver tissue.

(Fig. 2C). Active transcription at 8 days pp was correlated with trimethylated H3K4 even more than with DiMetH3K4. Loss of both forms of methylated H3K4 occurred during developmental repression, but TriMetH3K4 decreased more dramatically than DiMetH3K4 (6.9- versus 2.3-fold, respectively), and proportionately more than the increase measured in DiMetH3K9 (2.7-fold). Histone modifications of silenced versus developmentally repressed chromatin. Although AFP expression is undetectable by 2 months of development, AFP chromatin may differ between adult liver tissue, which has a developmental history of expression and the possibility of reactivation during liver

regeneration or tumorigenesis, and a tissue source where there is neither a history nor the potential for expression. Brain tissue from mice at the age of 2 months was processed for ChIP analysis of AFP chromatin, as described for liver tissue. We also determined specific histone modifications present at the Brn-3b gene locus, as a control gene actively expressed within brain tissue at this stage of development (Fig. 1). Comparison of modifications of histones H3 and H4 within brain tissue shows that patterns of AFP and Brn-3b expression correlate with alterations in histone methylation rather than acetylation (Fig. 3A). Silenced AFP chromatin has low dimethylation of H3K4 and abundant DiMetH3K9, while Brn-3b

VOL. 25, 2005


FIG. 3. Analysis of histone modifications present in permanently silenced and constitutively expressed chromatin isolated from brain and liver tissues. (A) ChIP analysis of silenced AFP chromatin at the SBE/p53RE and expressed Brn-3b chromatin isolated from brain tissue. (Top) Graph summarizing the results of multiple ChIP determinations, as described for Fig. 2A. Histone modifications associated with AFP chromatin (gray bars) and Brn-3b chromatin (white bars) in the brain were determined as described in Materials and Methods. (Bottom) Representative PCR analysis, as described for Fig. 2. Crosslinked chromatin lysates from 2-month WT mouse brain tissue were prepared as described. (B) ChIP determinations of histone modifications present in permanently silenced Brn-3b chromatin and constitutively activated ALB chromatin isolated from liver tissue at 8 days and 2 months of development. (Top) Graph summarizing the results of multiple ChIP determinations. Histone modifications were associated with specific genes as indicated. Each gene set is presented in the following order: ALB, 8-day (gray bar) and 2-month (white bar) liver; Brn-3b, 8-day (horizontal gradient bar) and 2-month (vertical gradient bar) liver. (Bottom) Representative PCR analysis. Cross-linked chromatin lysates from 2-month WT mouse liver tissue were prepared as described.

2151

chromatin exhibits high levels of DiMetH3K4 and low levels of DiMetH3K9. However, both silenced AFP chromatin and expressed Brn-3b chromatin display relatively low levels of histone H3 and H4 acetylation. The comparable levels of histone acetylation within expressed and silenced chromatin in the brain are unexpected, supporting the existence of tissue-specific and combinatorial patterns of histone modifications at individual genes and the complexity of histone code interpretation. Developmentally active and silenced chromatin in liver tissue. We performed a reciprocal study, comparing a tissuespecific activated gene (ALB) with a silenced gene (Brn-3b) in 2-month mouse liver tissue, as a control for potential modifications in chromatin due to tissue-specific or aging-related factors. During fetal liver development, AFP and ALB genes are activated at embryonic days 7.5 and 8, respectively (reviewed in references 72 and 76). These highly similar hepatic proteins are thought to perform similar functions, though AFP protein is abundant only in fetal liver. ALB gene expression is driven by a distal enhancer located 11 kb upstream of the transcription start site. Developmentally ordered binding of transcription factors and chromatin structure alterations at the albumin enhancer have been well defined by Zaret and colleagues (7, 40, 45). We analyzed ALB enhancer chromatin as the functional converse of the AFP repressor. At 2 months in liver tissue, DiMetH3K9 is present in abundance within silenced Brn-3b chromatin and is nearly absent at the ALB enhancer (Fig. 3B). Levels of DiMetH3K9 in Brn-3b are six- to eightfold greater than DiMetH3K9 levels in ALB in the liver, proving to be the best indicator of chromatin silencing. DiMetH3K4 is detectable in both active ALB and silenced Brn-3b chromatin, but it is present at higher levels in ALB chromatin. However, DiMetH3K4 is not correlated as tightly with active expression in the liver as in brain tissue. We have not tested whether this is due to regional localization of DiMetH3K4 at upstream distal regulatory regions (ALB enhancer) versus a 3⬘ open reading frame (Brn-3b). Locus-wide analyses of ␤-type globin genes support localized concentration of DiMetH3K4 modification within the gene body and core promoter (58). There are limited developmental changes in the histone modification states of ALB and Brn-3b between 8 days and 2 months pp in the liver. DiMetH3K9 is nearly undetectable in ALB chromatin at 8 days and remains so in the 2-month pp liver. In a mirror image of the DiMetH3K9 marker, AcH4 levels correlate closely with active ALB expression and remain high throughout development. Abundant AcH3K9 and DiMetH3K4 levels decrease, and AcH4 levels increase, all approximately 30%, at the ALB enhancer, although ALB is constitutively expressed in the adult liver (Fig. 1). Silenced Brn-3b chromatin exhibits a similar rise in AcH3K9 levels but remains relatively consistent otherwise (Fig. 3B). Interestingly, recent studies reveal that the capacity for liver regeneration decreases greatly with age (reviewed in references 42 and 69). Whether global or gene-specific chromatin alterations track these and other aging-related changes in the liver is not known. When the AFP repressor (Fig. 2A) is compared with the ALB enhancer in 8-day pp liver (Fig. 3B), where both genes are actively transcribed, there is little difference at the level of histone modifications. Both genes lack significant H3K9 meth-

2152

NGUYEN ET AL.

FIG. 4. Inverse relationship between activators and repressors of transcription during developmental regulation of AFP. (A) ChIP analyses of developmentally staged liver tissue from 8-day-old and 2-month-old mice. Representative PCR analyses of chromatin fragments associated with the transactivator protein FoxA, RNA polymerase II [Pol II (ser5)], and the corepressor protein mSin3A. The presence of bound ALB enhancer, AFP SBE/p53RE, and the AFP start site region was determined and quantified relative to input (% bound). (B) TBP binding in relationship to AcH4 at the AFP transcription start site region during liver development, determined by ChIP.

ylation, and both have relatively high levels of H3 and H4 acetylation and DiMetH3K4. Major differences between AFP and ALB, which occur from 8 days to 2 months of development, lie in the appearance of abundant methylation of H3K9 and loss of H3K4 methylation at the AFP repressor, concomitant with transcriptional repression of AFP. Developmentally regulated transcription factor interactions at the AFP repressor. The enzymatic complexes mediating histone modifications are targeted by proteins that interact directly or indirectly with DNA, potentially nucleating alterations in chromatin that can serve as platforms for additional chromatin modifiers or cofactors (17, 29, 43). We investigated whether specific activating and repressing transcription factor interactions occurred at AFP regulatory regions and compared them to those at the ALB enhancer (Fig. 4). FoxA (HNF-3) is an important mediator of liver-specific gene activation and, as a winged-helix/forkhead family member, may assume a function similar to that of linker histone and position nucleosomes within the ALB enhancer (7, 10–12, 45). FoxA has been proposed as a “pioneer transcription factor,” which binds its chromatin-assembled regulatory element in fetal liver and gener-

MOL. CELL. BIOL.

ates flanking chromatin accessible to downstream interactions between transcription activators, expressed later in development, and the ALB enhancer. We showed previously that FoxA also activated AFP expression in opposition to p53-mediated repression in vitro. We proposed that mutually exclusive occupation of the AFP developmental repressor (SBE/p53RE, at ⫺850) by FoxA and p53 was, in part, responsible for their opposing effects on AFP transcription (14, 39). We used ChIP analysis of liver tissue to determine FoxA interactions at its binding site within the AFP SBE/p53RE and to test the validity of our in vitro-based model (Fig. 4A). At 8 days pp, FoxA is present at the ALB enhancer as well as the AFP SBE/p53RE. The presence of FoxA correlates with RNA polymerase II, phosphorylated at Ser5 of the C-terminal domain, binding within the start site region of AFP. This modified form of RNA polymerase II is strongly associated with transcription initiation (15, 60). FoxA interaction at the AFP repressor element is lost at the age of 2 months, but it is maintained at the ALB enhancer. In the absence of FoxA binding, RNA polymerase II and TBP are no longer bound at the AFP core promoter (Fig. 4), which correlates with repression of AFP transcription. Recruitment of chromatin-silencing complexes. AFP chromatin is altered in parallel with transcription repression, exhibiting deacetylation of H3 and H4, methylation of H3K9, and loss of H3K4 methylation (Fig. 2). We investigated whether protein complexes known to mediate some of these changes in histone modification could be detected at the AFP repressor element and/or at the transcription start site in 2-month postnatal liver. Interestingly, mSin3A, a member of class I HDAC complexes, which include HDAC1 and HDAC2 (reviewed in references 18 and 24), is detected at the SBE/p53RE but not at the AFP transcription start site (Fig. 4A), though decreased acetylation is observed at both sites (Fig. 2). Further support for a biased association of mSin3A with the SBE/p53RE and not the core promoter is provided by analysis of HDAC1 interactions (Fig. 5A). There is no detectable binding of HDAC1 at either SBE/p53RE or the start site region at 8 days pp, when AFP is highly transcribed and chromatin is acetylated (Fig. 4A and 5A). HDAC1 binding occurs at the distal repressor, SBE/ p53RE, when AFP is repressed at 2 months, but is not detectable at the core promoter, confirming the mSin3A distribution. The alterations in histone acetylation found at the AFP start site region may be due to spreading of deacetylation by HDAC complexes targeted along with Sin3A to the distal repressor region (SBE/p53RE) or by deacetylase activity directed to the core promoter by HDAC complexes lacking an mSin3A component. Interestingly, HDAC1 binding to the silenced Brn-3b locus is readily detectable in 8-day pp liver (Fig. 5A) but is considerably diminished in 2-month liver, coincident with the observed increase in AcH3K9 levels in Brn-3b chromatin (Fig. 3B). Whether this represents a transient association of HDAC complexes with silenced chromatin in tissues that are still undergoing differentiation, such as the liver at 8 days pp, remains to be determined. Repressed euchromatin and heterochromatin are composed not only of deacetylated histones but also of methylated H3K9, as we find for developmentally repressed AFP. Chromatin methylated at H3K9 serves as a platform for binding to HP1, which can further interact with histone methyltransferases

VOL. 25, 2005


2153

FIG. 5. HP1 associates with the AFP repressor and promoter regions during developmental repression and tissue-specific silencing, but HDAC1 binding is restricted to the developmental repressor. (A to C) Representative PCR analyses of ChIP studies for the association of HDAC1 and HP1 with the developmental repressor (SBE/p53RE) and start site regions of AFP in comparison to Brn-3b. Results of assays of 8-day (A) and 2-month (B) liver tissue and of 2-month brain tissue(C) are shown. Unmodified anti-histone H3 antibody is shown as an internal control for chromatin recovery in each lysate. Binding is quantified (% bound) in comparison to each input lysate; titrations were analyzed separately (data not shown). (D) Bar graph summarizing the developmental changes in H3K9 modifications and HP1 interactions during liver development and in brain tissue at specific regions of the ALB/AFP gene locus: the AFP start site, centered at ⫹1; the SBE/p53RE, centered at ⫺850; AFP enhancer (enh) III, at ⫺7 kb; and the enhancer of ALB, at ⫺11 kb relative to the ALB start site of transcription and approximately 27 kb upstream of AFP enhancer III. Data derived from ChIP analyses for which results are shown here and in Fig. 2 (also data not shown) were compared by calculation of the ratio between average bound protein and average chromatin recovery (percent unmodified histone H3) for each lysate. Averages of these ratios are shown as bars in the bar graph: light gray and dark gray, 8-day and 2-month AcH3K9 levels; light gray gradient and dark gray gradient, 8-day and 2-month diMetH3K9 levels; faceted gray and solid white, 8-day and 2-month HP1 levels; stippled gray, 2-month brain HP1 levels. Data from 2-month liver tissue only are provided for AFP enhancer III.

(HMTs). The process of H3K9 deacetylation by HDACs, methylation by HMTs, and interaction with HP1, followed by additional HMT binding and histone methylation, has been proposed as a means of spreading a heterochromatic structure (reviewed in references 23, 49, and 57). Developmentally repressed AFP chromatin shows substantial levels of HP1 interaction at the SBE/p53RE and transcription start site regions,

but active AFP chromatin at 8 days pp does not (Fig. 5A). In contrast to the spreading of HP1 in heterochromatin, we find that HP1 protein is localized to the repressed AFP gene and is not present at the upstream enhancers of AFP (Fig. 5B) and ALB (Fig. 5A). Enhancer III of AFP lies approximately 6 kb upstream of SBE/p53RE (Fig. 5D) and is the most distal of three 5⬘ enhancers of AFP expression (5). When present as a

2154

NGUYEN ET AL.

transgene, enhancer III can direct low-level, heterologous gene expression in brain tissue, unlike the downstream, liver-specific enhancers of AFP expression (56). Transgenic models of expression showed that developmental repression of AFP is independent of these enhancers of gene activation (5). The lack of both HP1 and methylation of H3K9 at enhancer III, when AFP is developmentally repressed, supports separation of this region from the repressive domain of AFP chromatin. A summary of these data, as well as developmentally regulated and tissue-specific H3K9 modifications and HP1 levels within regions of AFP and ALB chromatin, is presented in Fig. 5D. We asked if permanently silenced chromatin would have larger amounts of associated HP1 protein than repressed chromatin that is established over a time course of development. We compared HP1 levels in silenced Brn-3b to those in developmentally regulated AFP. In both 8-day pp and 2-month liver tissue, Brn-3b chromatin is associated with detectable HP1 protein (Fig. 5A). HP1 levels decreased approximately twofold over the time course of development for Brn-3b. Contrary to our expectations, levels of HP1 at silenced Brn-3b were somewhat lower than those observed at the repressed AFP locus (2 months). Our determinations included only analysis of HP1␣, and there may be tissue- and developmental-stage-specific differences between HP1 isoforms not apparent in these studies. Specific protein complexes of HP1 isoforms and the Su(var)3-9 HMT, which associate at different chromosomal and subnuclear sites during development, have been identified in Drosophila melanogaster (28). In a reciprocal determination using brain tissue, where AFP and ALB are permanently silenced, HP1 was abundant at both gene loci but was absent from actively expressed Brn-3b (Fig. 5C and D). There are likely several nonspecific factors that come into play when lysates from two different tissues, such as brain and liver, are compared. However, using an antibody specific for total histone H3 (unmodified C-terminal) as a measure of chromatin recovery, we estimated that HP1 levels at the AFP distal repressor SBE/p53RE were equivalent in the liver and the brain but higher at the silenced AFP core promoter in the brain compared with the liver (Fig. 5D). The p53 tumor suppressor protein targets chromatin repression. Mice genetically null for p53 exhibit a delay in repression of AFP expression during liver development, while constitutively expressed ALB and GAPDH genes are unaffected (Fig. 1). Therefore, there must be compensatory mechanisms for the loss of p53 or additive repressor proteins at work to effect robust repression of AFP during hepatic development. We determined a specific end point for AFP expression in p53-null mouse liver by increasing the number of PCR amplification cycles and the sensitivity of AFP RNA detection (Fig. 6A). These analyses showed that AFP is transcribed at low levels in both WT and p53-null mice at the age of 1 month. The low level of AFP transcription continues through early adulthood in p53-null mice but ceases before the age of 2 months in the WT. AFP expression is repressed in both p53null and WT mice within 4 months. The differences between fully repressed transcription and transcription at a detectable but low level are readily apparent when chromatin structure is analyzed (Fig. 6B). Sustained expression of AFP in the 2-month p53-null mouse liver is distinguished by distinct differences in chromatin modification and

MOL. CELL. BIOL.

FIG. 6. p53-dependent targeting of repression-associated modifications of histones and mSin3A/HDAC1 to AFP chromatin. (A) In the absence of p53, developmental repression of AFP is delayed. RNAs were isolated from 2- and 12-day pp and 1-, 2-, and 4-month livers extracted from WT and p53-null mice. RT-PCR was performed to determine relative levels of AFP and GAPDH expression. PCR amplification at 24 cycles was used to assay GAPDH levels (upper panel, GAPDH), and 30 cycles were used with the same RNA samples to determine AFP (lower panel, AFP). (B) Representative PCR determinations of ChIP assays of specific histone modifications and p53/ mSin3A present at the SBE/p53RE site in 2-month WT and p53-null liver tissue. The amount of DNA associated with each specific protein is quantified (% bound) in comparison to each input lysate; titrations were analyzed separately. (C) ChIP assays of FoxA and HDAC1 association with the SBE/p53RE and start site regions of AFP in p53-null liver tissue. Representative data are shown. (D) Bar graph summarizing the p53-dependent patterns of FoxA, p53, mSin3A, and HDAC1 proteins bound to SBE/p53RE during developmental repression of AFP. Analyses of 8-day WT (light gray bars), 2-month WT (dark gray bars), and 2-month p53-null (white bars) livers were performed two to four times each. Data from these analyses are presented here and in Fig. 4 and 5 (also data not shown). Ratios of bound proteins to unmodified histone H3 were calculated for each lysate and averaged, as described for Fig. 5D.

VOL. 25, 2005


transcription factor interactions from the fully repressed WT. ChIP analyses of p53-null and WT mouse liver chromatin in the region of SBE/p53RE reveal p53-dependent targeting of histone deacetylation and DiMetH3K9, as well as loss of DiMetH3K4. p53-null mice show a fourfold increase in H3K9 acetylation at SBE/p53RE and a concomitant fourfold decrease in H3K9 dimethylation. Both di- and trimethylated H3K4 levels were increased (4.4- and 3.9-fold, respectively) as well. The alterations in histone modifications correlate with loss of mSin3A and HDAC1 interactions in the p53-null mouse liver (Fig. 6B and C). Consistent with our model of mutually exclusive binding of p53 and FoxA at their overlapping sites within SBE/p53RE (39), and the pattern of FoxA binding during development (Fig. 4), we see sustained FoxA interaction, specifically at SBE/p53RE, in the absence of p53 (Fig. 6C). A graph summarizing the regulated interactions of transcription factors and repressor proteins emphasizes the inverse relationship between the FoxA transactivator and complexes of repressor proteins, which are targeted to the AFP developmental repressor region during normal hepatic development (Fig. 6D). p53 is required to establish this pattern of repressor protein binding, FoxA exclusion, and repressive chromatin modification. DISCUSSION The functions of the p53 tumor suppressor during development and as a repressor of transcription are relatively poorly understood (reviewed in references 9, 21, and 31). A role for p53 in the regulation of AFP transcription during development is supported by comparison of RNA levels and chromatin modifications in p53-null and WT mouse livers. Previous studies have shown that p53-mediated repression of transcription is correlated with the association of p53 and mSin3A/HDAC1 in response to hypoxic or stress activation of p53 or its forced nuclear localization (32, 36, 47, 77). Not only can p53-targeted HDAC complexes deacetylate histones, as shown here for AFP during development; they can also alter p53 function by deacetylation of p53 itself, indirectly interfering with p53-dependent transcription activation (41). These studies, in addition to genome-wide profiling of p53-regulated gene expression (68), support the function of p53 as a repressor of transcription; however, only a subset of these genes have been identified as direct targets of p53. The observed increase in histone acetylation of AFP chromatin in adult liver, in the absence of p53 expression, is likely due to loss of p53-HDAC interaction in parallel with increased histone acetylase activity, potentially targeted by FoxA. We find that robust AFP expression in the liver is coincident with the presence of acetylated histones H3 and H4 as well as methylated H3K4. Recruitment of mSin3A/HDAC1, decreases in methylation at H3K4, and acetylation of H4 and H3K9 follow this period of active expression to effect developmentally regulated repression of AFP transcription. Hepatic differentiation is accompanied by decreased cell cycle progression and DNA replication, which precede AFP developmental repression (reviewed in reference 66). Previous studies of DNA replication and cell cycle status in p53-null liver showed a twofold increase in the number of cells undergoing S phase in adult p53-null versus WT liver (22). The increase in replication

2155

levels, combined with loss of p53-targeted HDACs and increased acetylation of H3K9, may contribute to depletion of methylated histones and decreased DiMetH3K9 levels in p53null liver. Decreased Suv39h activity in cycling versus differentiated cells, as suggested by studies of Suv39h-depleted muscle cells, may also contribute to a specific loss of DiMetH3K9 relative to MetH3K4 (2). Previous work showing interactions of transcription factors, KRAB/KAP-1 proteins, and tumor suppressor Rb with HMT enzymes and HP1 established that methylated H3K9 and HP1 proteins were involved in targeted repression of euchromatic genes in addition to heterochromatic silencing (50, 61). Similarly, we find a p53-dependent association between DiMetH3K9 and HP1 at AFP chromatin during developmental repression in vivo. DiMetH3K9 and HP1 association and repression of transcription do not spread to the distal enhancer of AFP or to the highly homologous, but constitutively expressed, ALB gene, which lies upstream of AFP. The considerable homology between the ALB and AFP genes, which arose by duplication from a common precursor and exhibit more than 30% primary sequence identity within their coding regions (27), does not extend to the AFP repressor region. This lends further support to localized targeting of HP1 by p53-mediated enrichment of DiMetH3K9 within AFP chromatin, and lack of heterochromatic spreading throughout the gene locus. Recently, a pathway of heterochromatin silencing, which is nucleated by ATF/CREB family members binding to their sequence-specific binding elements, was defined (34). In contrast with repression of euchromatin targeted by p53, Rb, or KRAB/KAP1, this transcription factor-dependent means of silencing the mating-type locus of Schizosaccharomyces pombe induced heterochromatin formation and spreading, like the parallel RNA interference-dependent mechanism. We compared histone modifications and HP1 association during developmental repression with silencing of AFP in a nonexpressing tissue. Unlike the situation in developmental repression, HP1 protein was abundant at both the AFP and ALB gene loci in brain tissue. The relative levels of HP1 interaction at the AFP core promoter were higher in the brain but did not differ between the liver and brain at the distal repressor region. Interestingly, sites at nucleotide ⫹3 and the first intronic region of the AFP gene exhibit increased CpG methylation during developmental repression in the liver and silenced expression in fibroblasts (38, 53). DNA methylation and H3K9 methylation have been linked with HP1 in chromatin silencing (reviewed in references 19, 23, 26, and 33), and more recently a direct association between enzymes that methylate DNA and HP1␤ protein has been reported (25). Our analyses of p53-null mice reveal that relatively small differences in transcription levels are marked by more obvious distinctions in chromatin modifications and bound transcription factors. As ChIP analyses of solid tissues develop, and analytical systems become more sensitive, their use in early detection of changes in tumor marker gene regulation may be more direct than assays of RNA or protein levels, which are also altered by posttranscriptional and posttranslational mechanisms. The DiMetH3K9 histone modification, which is abundant in silenced AFP chromatin within brain tissue and is easily detected in developmentally repressed AFP chromatin in WT liver tissue, may be especially useful in establishing a normal

2156

NGUYEN ET AL.

signature for tissue-specific chromatin structure and as a marker of aberrant gene activation or repression when altered.

MOL. CELL. BIOL.

25.

ACKNOWLEDGMENTS We are grateful to A. J. Crowe, J. Park, X. Mu, C. Smith, J. ParkerThornburg, G. Lozano, R. Kori, S. Y. Dent, and members of the Dent and Barton laboratories for technical assistance, materials, and/or helpful discussions. This work was supported by grants GM53683 and GM60213 from the National Institutes of Health to M.C.B. and, in part, by an NCI Cancer Center support grant to the U.T. M. D. Anderson Cancer Center.

26. 27. 28.

29. REFERENCES 1. Agalioti, T., S. Lomvardas, B. Parekh, J. Yie, T. Maniatis, and D. Thanos. 2000. Ordered recruitment of chromatin modifying and general transcription factors to the IFN-␤ promoter. Cell 103:667–678. 2. Ait-Si-Ali, S., V. Guasconi, L. Fritsch, H. Yahi, R. Sekhri, I. Naguibneva, P. Robin, F. Cabon, A. Polesskaya, and A. Harel-Bellan. 2004. A Suv39hdependent mechanism for silencing S-phase genes in differentiating but not in cycling cells. EMBO J. 23:605–615. 3. Camper, S. A., R. Godbout, and S. M. Tilghman. 1989. The developmental regulation of albumin and alpha-fetoprotein gene expression. Prog. Nucleic Acid Res. Mol. Biol. 36:131–143. 4. Camper, S. A., and S. M. Tilghman. 1991. The activation and silencing of gene transcription in the liver. Bio/Technology 16:81–87. 5. Camper, S. A., and S. M. Tilghman. 1989. Postnatal repression of the alpha-fetoprotein gene is enhancer independent. Genes Dev. 3:537–546. 6. Chadee, D. N., M. J. Hendzel, C. P. Tylipski, C. D. Allis, D. P. Bazett-Jones, J. A. Wright, and J. R. Davie. 1999. Increased Ser-10 phosphorylation of histone H3 in mitogen-stimulated and oncogene-transformed mouse fibroblasts. J. Biol. Chem. 274:24914–24920. 7. Chaya, D., T. Hayamizu, M. Bustin, and K. S. Zaret. 2001. Transcription factor FoxA (HNF3) on a nucleosome at an enhancer complex in liver chromatin. J. Biol. Chem. 276:44385–44389. 8. Chen, H., J. O. Egan, and J. F. Chiu. 1997. Regulation and activities of alpha-fetoprotein. Crit. Rev. Eukaryot. Gene Expr. 7:11–41. 9. Choi, J., and L. A. Donehower. 1999. p53 in embryonic development: maintaining a fine balance. Cell. Mol. Life Sci. 55:38–47. 10. Cirillo, L. A., F. R. Lin, I. Cuesta, D. Friedman, M. Jarnik, and K. S. Zaret. 2002. Opening of compacted chromatin by early developmental transcription factors HNF3 (FoxA) and GATA-4. Mol. Cell 9:279–289. 11. Cirillo, L. A., C. E. McPherson, P. Bossard, K. Stevens, S. Cherian, E. Yon Shim, K. L. Clark, S. K. Burley, and K. S. Zaret. 1998. Binding of the winged-helix transcription factor HNF3 to a linker histone site on the nucleosome. EMBO J. 17:244–254. 12. Cirillo, L. A., and K. S. Zaret. 1999. An early developmental transcription factor complex that is more stable on nucleosome core particles than on free DNA. Mol. Cell 4:961–969. 13. Cosma, M. P., T. Tanaka, and K. Nasmyth. 1999. Ordered recruitment of transcription and chromatin remodeling factors to a cell cycle- and developmentally regulated promoter. Cell 97:299–311. 14. Crowe, A. J., L. Sang, K. K. Li, K. C. Lee, B. T. Spear, and M. C. Barton. 1999. Hepatocyte nuclear factor 3 relieves chromatin-mediated repression of the ␣-fetoprotein gene. J. Biol. Chem. 274:25113–25120. 15. Dahmus, M. E. 1996. Reversible phosphorylation of the C-terminal domain of RNA polymerase II. J. Biol. Chem. 271:19009–19012. 16. Darlington, G. J. 1987. Liver cell lines. Methods Enzymol. 151:19–38. 17. Davie, J. K., and S. Y. Dent. 2004. Histone modifications in corepressor functions. Curr. Top. Dev. Biol. 59:145–163. 18. de Ruijter, A. J., A. H. van Gennip, H. N. Caron, S. Kemp, and A. B. van Kuilenburg. 2003. Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem. J. 370:737–749. 19. Dillon, N., and R. Festenstein. 2002. Unravelling heterochromatin: competition between positive and negative factors regulates accessibility. Trends Genet. 18:252–258. 20. Dilworth, F. J., C. Fromental-Ramain, K. Yamamoto, and P. Chambon. 2000. ATP-driven chromatin remodeling activity and histone acetyltransferases act sequentially during transactivation by RAR/RXR in vitro. Mol. Cell 6:1049–1058. 21. Donehower, L. A. 2002. Does p53 affect organismal aging? J. Cell. Physiol. 192:23–33. 22. Dumble, M. L., B. Knight, E. A. Quail, and G. C. Yeoh. 2001. Hepatoblastlike cells populate the adult p53 knockout mouse liver: evidence for a hyperproliferative maturation-arrested stem cell compartment. Cell Growth Differ. 12:223–231. 23. Elgin, S. C., and S. I. Grewal. 2003. Heterochromatin: silence is golden. Curr. Biol. 13:R895–R898. 24. Fleischer, T. C., U. J. Yun, and D. E. Ayer. 2003. Identification and charac-

30. 31. 32. 33. 34. 35. 36.

37. 38. 39. 40.

41. 42. 43. 44. 45. 46.

47.

48.

49. 50.

51.

terization of three new components of the mSin3A corepressor complex. Mol. Cell. Biol. 23:3456–3467. Fuks, F., P. J. Hurd, R. Deplus, and T. Kouzarides. 2003. The DNA methyltransferases associate with HP1 and the SUV39H1 histone methyltransferase. Nucleic Acids Res. 31:2305–2312. Fuks, F., P. J. Hurd, D. Wolf, X. Nan, A. P. Bird, and T. Kouzarides. 2003. The methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J. Biol. Chem. 278:4035–4040. Gorin, A. G., D. L. Cooper, F. Eiferman, P. van de Rijn, and S. Tilghman. 1981. The evolution of ␣-fetoprotein and albumin. J. Biol. Chem. 256:1954– 1959. Greil, F., I. van der Kraan, J. Delrow, J. F. Smothers, E. de Wit, H. J. Bussemaker, R. van Driel, S. Henikoff, and B. van Steensel. 2003. Distinct HP1 and Su(var)3–9 complexes bind to sets of developmentally coexpressed genes depending on chromosomal location. Genes Dev. 17:2825–2838. Grewal, S. I., and S. C. Elgin. 2002. Heterochromatin: new possibilities for the inheritance of structure. Curr. Opin. Genet. Dev. 12:178–187. Hatzis, P., and I. Talianidis. 2002. Dynamics of enhancer-promoter communication during differentiation-induced gene activation. Mol. Cell 10:1467– 1477. Ho, J., and S. Benchimol. 2003. Transcriptional repression mediated by the p53 tumour suppressor. Cell Death Differ. 10:404–408. Hoffman, W. H., S. Biade, J. T. Zilfou, J. Chen, and M. Murphy. 2002. Transcriptional repression of the anti-apoptotic survivin gene by wild type p53. J. Biol. Chem. 277:3247–3257. Jaenisch, R., and A. Bird. 2003. Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat. Genet. 33(Suppl.):245–254. Jia, S., K. Noma, and S. I. Grewal. 2004. RNAi-independent heterochromatin nucleation by the stress-activated ATF/CREB family proteins. Science 304:1971–1976. Jin, D. K., J. Vacher, and M. H. Feuerman. 1998. ␣-Fetoprotein gene sequences mediating Afr2 regulation during liver regeneration. Proc. Natl. Acad. Sci. USA 95:8767–8772. Koumenis, C., R. Alarcon, E. Hammond, P. Sutphin, W. Hoffman, M. Murphy, J. Derr, Y. Taya, S. W. Lowe, M. Kastan, and A. Giaccia. 2001. Regulation of p53 by hypoxia: dissociation of transcriptional repression and apoptosis from p53-dependent transactivation. Mol. Cell. Biol. 21:1297–1310. Krumlauf, R., R. E. Hammer, S. M. Tilghman, and R. L. Brinster. 1985. Developmental regulation of alpha-fetoprotein genes in transgenic mice. Mol. Cell. Biol. 5:1639–1648. Kunnath, L., and J. Locker. 1983. Developmental changes in the methylation of the rat albumin and alpha-fetoprotein genes. EMBO J. 2:317–324. Lee, K. C., A. J. Crowe, and M. C. Barton. 1999. p53-mediated repression of alpha-fetoprotein gene expression by specific DNA binding. Mol. Cell. Biol. 19:1279–1288. Liu, J. K., Y. Bergman, and K. S. Zaret. 1988. The mouse albumin promoter and a distal upstream site are simultaneously DNase I hypersensitive in liver chromatin and bind similar liver-abundant factors in vitro. Genes Dev. 2:528–541. Luo, J., F. Su, D. Chen, A. Shiloh, and W. Gu. 2000. Deacetylation of p53 modulates its effect on cell growth and apoptosis. Nature 408:377–381. Mackey, S., P. Singh, and G. J. Darlington. 2003. Making the liver young again. Hepatology 38:1349–1352. Maison, C., and G. Almouzni. 2004. HP1 and the dynamics of heterochromatin maintenance. Nat. Rev. Mol. Cell Biol. 5:296–304. McKinsey, T. A., C. L. Zhang, and E. N. Olson. 2001. Control of muscle development by dueling HATs and HDACs. Curr. Opin. Genet. Dev. 11: 497–504. McPherson, C. E., R. Horowitz, C. L. Woodcock, C. Jiang, and K. S. Zaret. 1996. Nucleosome positioning properties of the albumin transcriptional enhancer. Nucleic Acids Res. 24:397–404. Mu, X., P. D. Beremand, S. Zhao, R. Pershad, H. Sun, A. Scarpa, S. Liang, T. L. Thomas, and W. H. Klein. 2004. Discrete gene sets depend on POU domain transcription factor Brn3b/Brn-3.2/POU4f2 for their expression in the mouse embryonic retina. Development 131:1197–1210. Murphy, M., J. Ahn, K. K. Walker, W. H. Hoffman, R. M. Evans, A. J. Levine, and D. L. George. 1999. Transcriptional repression by wild-type p53 utilizes histone deacetylases, mediated by interaction with mSin3a. Genes Dev. 13:2490–2501. Nakabayashi, H., T. Hashimoto, Y. Miyao, K. K. Tjong, J. Chan, and T. Tamaoki. 1991. A position-dependent silencer plays a major role in repressing alpha-fetoprotein expression in human hepatoma. Mol. Cell. Biol. 11: 5885–5893. Nakayama, J., J. Rice, B. D. Strahl, C. D. Allis, and S. Grewal. 2001. Role of histone H3 lysine 9 methylation in epigenetic control of heterochromatin assembly. Science 292:110–113. Nielsen, S. J., R. Schneider, U. M. Bauer, A. J. Bannister, A. Morrison, D. O’Carroll, R. Firestein, M. Cleary, T. Jenuwein, R. E. Herrera, and T. Kouzarides. 2001. Rb targets histone H3 methylation and HP1 to promoters. Nature 412:561–565. Ninkina, N. N., G. E. Stevens, J. N. Wood, and W. D. Richardson. 1993. A

VOL. 25, 2005

52. 53. 54. 55.

56. 57. 58. 59. 60. 61.

62. 63. 64.


novel Brn3-like POU transcription factor expressed in subsets of rat sensory and spinal cord neurons. Nucleic Acids Res. 21:3175–3182. Ogden, S. K., K. C. Lee, K. Wernke-Dollries, S. A. Stratton, B. Aronow, and M. C. Barton. 2001. p53 targets chromatin structure alteration to repess ␣-fetoprotein gene expression. J. Biol. Chem. 276:42057–42062. Opdecamp, K., M. Riviere, M. Molne, J. Szpirer, and C. Szpirer. 1992. Methylation of an alpha-foetoprotein gene intragenic site modulates gene activity. Nucleic Acids Res. 20:171–178. Pandey, R. S., and M. S. Kanungo. 1984. Developmental changes in chromatin of skeletal muscle of rats: acetylation of chromosomal proteins and transcription. Mol. Biol. Rep. 10:79–81. Parrizas, M., M. A. Maestro, S. F. Boj, A. Paniagua, R. Casamitjana, R. Gomis, F. Rivera, and J. Ferrer. 2001. Hepatic nuclear factor 1-␣ directs nucleosomal hyperacetylation to its tissue-specific transcriptional targets. Mol. Cell. Biol. 21:3234–3243. Ramesh, T. M., A. W. Ellis, and B. T. Spear. 1995. Individual mouse alphafetoprotein enhancer elements exhibit different patterns of tissue-specific and hepatic position-dependent activities. Mol. Cell. Biol. 15:4947–4955. Richards, E. J., and S. C. Elgin. 2002. Epigenetic codes for heterochromatin formation and silencing: rounding up the usual suspects. Cell 108:489–500. Schneider, R., A. J. Bannister, F. A. Myers, A. W. Thorne, C. Crane-Robinson, and T. Kouzarides. 2004. Histone H3 lysine 4 methylation patterns in higher eukaryotic genes. Nat. Cell Biol. 6:73–77. Schreiber, S. L., and B. E. Bernstein. 2002. Signaling network model of chromatin. Cell 111:771–778. Schroeder, S. C., B. Schwer, S. Shuman, and D. Bentley. 2000. Dynamic association of capping enzymes with transcribing RNA polymerase II. Genes Dev. 14:2435–2440. Schultz, D. C., K. Ayyanathan, D. Negorev, G. G. Maul, and F. J. Rauscher III. 2002. SETDB1: a novel KAP-1-associated histone H3, lysine 9-specific methyltransferase that contributes to HP1-mediated silencing of euchromatic genes by KRAB zinc-finger proteins. Genes Dev. 16:919–932. Soutoglou, E., and I. Talianidis. 2002. Coordination of PIC assembly and chromatin remodeling during differentiation-induced gene activation. Science 295:1901–1904. Spear, B. T. 1999. Alpha-fetoprotein gene regulation: lessons from transgenic mice. Semin. Cancer Biol. 9:109–116. Strahl, B. D., and C. D. Allis. 2000. The language of covalent histone modifications. Nature 403:41–45.

2157

65. Su, R. C., K. E. Brown, S. Saaber, A. G. Fisher, M. Merkenschlager, and S. T. Smale. 2004. Dynamic assembly of silent chromatin during thymocyte maturation. Nat. Genet. 36:502–506. 66. Tilghman, S. M. 1985. The structure and regulation of the alpha-fetoprotein and albumin genes. Oxf. Surv. Eukaryot. Genes 2:160–206. 67. Turner, B. M. 2000. Histone acetylation and an epigenetic code. Bioessays 22:836–845. 68. Wang, L., Q. Wu, P. Qiu, A. Mirza, M. McGuirk, P. Kirschmeier, J. R. Greene, Y. Wang, C. B. Pickett, and S. Liu. 2001. Analyses of p53 target genes in the human genome by bioinformatic and microarray approaches. J. Biol. Chem. 276:43604–43610. 69. Wang, X., H. Kiyokawa, M. B. Dennewitz, and R. H. Costa. 2002. The Forkhead Box m1b transcription factor is essential for hepatocyte DNA replication and mitosis during mouse liver regeneration. Proc. Natl. Acad. Sci. USA 99:16881–16886. 70. Wells, J., and P. J. Farnham. 2002. Characterizing transcription factor binding sites using formaldehyde crosslinking and immunoprecipitation. Methods 26:48–56. 71. Wells, J., C. R. Graveel, S. M. Bartley, S. J. Madore, and P. J. Farnham. 2002. The identification of E2F1-specific target genes. Proc. Natl. Acad. Sci. USA 99:3890–3895. 72. Zaret, K. 1998. Early liver differentiation: genetic potentiation and multilevel growth control. Curr. Opin. Genet. Dev. 8:526–531. 73. Zaret, K. 1994. Genes that control the formation of the liver. Hepatology 19:794–796. 74. Zaret, K. S. 1994. Genetic control of hepatocyte differentiation, p. 53–68. In I. M. Arias, J. L. Boyer, N. Fausto, W. B. Jakoby, D. A. Schachter, and D. A. Shafritz (ed.), The liver: biology and pathobiology, 3rd ed. Raven Press, New York, N.Y. 75. Zaret, K. S. 1996. Molecular genetics of early liver development. Annu. Rev. Physiol. 58:231–251. 76. Zaret, K. S. 1995. Nucleoprotein architecture of the albumin transcriptional enhancer. Semin. Cell Biol. 6:209–218. 77. Zilfou, J. T., W. H. Hoffman, M. Sank, D. L. George, and M. Murphy. 2001. The corepressor mSin3a interacts with the proline-rich domain of p53 and protects p53 from proteasome-mediated degradation. Mol. Cell. Biol. 21: 3974–3985.