University of Cincinnati

UNIVERSITY OF CINCINNATI

February 6 , 2002

I, Stacey K. Ogden, hereby submit this as part of the requirements for the degree of:

Doctor of Philosophy in: Molecular Genetics, Biochemistry and Microbiology It is entitled: HBx-mediated disruption of p53 tumor suppressor function leading to reactivation of a silenced tumor- marker gene.

Approved by: Dr. Michelle Barton, Chair Dr. Tony Capobianco Dr. Iain Cartwright Dr. Michael Lieberman Dr. Jun Ma

HBx-mediated disruption of p53 tumor suppressor protein function leading to re-activation of a silenced tumor marker gene

A dissertation submitted to the

Division of Research and Advanced Studies of the University of Cincinnati

in partial fulfillment of the requirements for the degree of

DOCTORATE OF PHILOSOPHY (Ph. D.)

in the Department of Molecular Genetics, Biochemistry and Microbiology of the College of Medicine

2002

Stacey K. Ogden

B.A., Miami University, 1997

Committee Chair: Michelle Craig Barton, Ph. D.

Abstract

Chronic infection with Hepatitis B Virus (HBV) is a predominant risk factor associated with development of hepatocellular carcinoma (HCC). Individuals who are chronically infected with

HBV are over 200-times more likely to develop HCC than those who are not infected. Multiple studies have implicated the virally encoded X protein (HBx) as the candidate oncoprotein responsible for cellular transformation. HBx forms a complex with cellular p53 tumor suppressor pr otein to result in modification of p53-mediated gene regulation, DNA damage detection and modulation of apoptosis. The developmentally silenced a-fetoprotein (AFP) tumor marker gene, which is transcriptionally repressed by p53, is tightly correlated with HCC development: it is reactivated in over 80% of all liver carcinomas. p53 mediates transcriptional repression of AFP through an overlapping HNF-3/Smad4/p53 binding element located within the developmental repressor domain of the AFP promoter. Here, us ing AFP as a model gene, we have examined the mechanism by which p53 facilitates transcriptional repression, and how this repression is disrupted upon p53-HBx interaction. In vitro chromatin assembled transcription analysis and enzyme accessibility studie s demonstrate that p53 association at the overlapping binding element is required during chromatin assembly for reorganization of AFP promoter chromatin structure to result in occlusion of restriction enzymes and general transcription factors from the transcription start site. Protein-DNA binding assays show that p53 association at this element is required to recruit mSin3A co-repressor and stabilize association of a putative Smad4 and SnoN containing co-repressor complex with AFP chromatin templates. HBx-mediated reactivation of AFP is achieved through direct p53-HBx interaction resulting in disruption of SnoN co-repressor binding to AFP chromatin templates.

Table of contents

Chapter 1: Introduction ...... 3

Chapter 2: p53 targets chromatin structure alteration to repress α-fetoprotein gene expression.19

Figure 1. p53 represses AFP chromatin transcription in vitro...... 32

Figure 2. p53 organizes chromatin structure...... 33

Figure 3. DNA binding is required for p53 function...... 35

Figure 4. p53 mediated repression of AFP transcription occurs even in the presence of

hyperacetylated histones at the core promoter...... 37

Chapter 3: Hepatitis B viral transactivator HBx alleviates p53-mediated repression of α- fetoprotein gene expression ...... 39

Figure 1: HBx and p53 interaction results in loss of p53-mediated repression of AFP

transcription...... 62

Figure 2: p53-mediated squelching of β-globin transcription is not alleviated by HBx...... 63

Figure 3: HBx alleviates p53 mediated repression of chromatin assembled AFP DNA...... 64

Figure 4: p53 repression and HBx re-activation of AFP transcription are dependent upon

p53-DNA binding...... 66

Figure 5: Tissue specificity of p53-HBx effect on AFP transcription is maintained in

chromatin...... 68

Figure 6: HBx associates with DNA-bound p53. (A) p53-DNA binding is maintained upon

HBx association...... 70

Chapter 4: A p53, Smad4 and SnoN-containing repressor complex is disrupted by the virally encoded HBx protein...... 72

Figure 1: SnoN associates at the AFP p53 regulatory element...... 97

Figure 2: Putative co-repressor complex associates at the p53 regulatory element...... 100

Figure 3: Smad4 and SnoN association with p53 are DNA-dependent...... 101

Figure 4: SnoN is involved in AFP transcription repression...... 103

Figure 5: p53 DNA binding is required for maximal SnoN-mediated AFP repression...... 105

Figure 6: Functional p53 and SnoN are both required for maximal AFP repression...... 107

Figure 7: HBx disrupts p53-stabilized SnoN binding to chromatin DNA templates...... 109

Chapter 5: Summary and Conclusions...... 110

Figure Summary...... 115

References...... 116

Chapter 1: Introduction p53 and Cancer

It has been suggested that in order for a cell to be considered to have a cancer phenotype, it must possess six distinctive traits: self-sufficiency in growth signals, insensitivity to antigrowth signals, evasion of apoptosis, limitless replication potential, sustained angiogenesis, and tissue invasion and metastasis (Hanahan and Weinberg, 2000). The processes by which a cell acquires these traits are complex and, at present, not clearly understood. However, it is clear from multiple lines of evidence that loss of genomic integrity within tumor cells can contribute to development of these cancer traits. Safeguarding genomic integrity is normally controlled by a host of cellular proteins; the most prominent being the p53 tumor suppressor protein. Functional p53 protein has a vast array of roles involved in detection and repair of DNA damage, cell cycle arrest and modulation of apoptosis (reviewed in Ko and Prives, 1996). The importance of p53 in controlling proper cellular growth is directly evidenced by the fact that it is the single most commonly mutated or deleted gene in human cancer. p53 exhibits a mutation rate well over 80% in human tumors (reviewed in Fisher, 2001; Ko and Prives, 1996). Tumor cells with inactive p53 protein typically exhibit increased aneuploidy, DNA mutation and gene amplification in combination with decreased apoptotic potential (reviewed in Kastan, et al., 1995).

Central to p53’s ability to regulate cell cycle control and safeguard against mutation is its ability to activate and repress appropriate target genes upon detection of DNA damage. Among the genes activated by p53 protein in response to cellular stress are p21, mdm-2, GADD45, cyclin

G, bax, and IGF-BP3. The ability of p53 to regulate its target genes requires direct DNA binding. This function is lost in the bulk of p53-null tumors. Most p53 loss-of-function

mutations occur within the DNA binding domain (reviewed in Ko and Prives, 1996). Because most of the above-mentioned p53 regulated genes are either directly or indirectly involved in

DNA damage repair and/or cell cycle arrest, loss of p53 DNA binding and transcription control affords a growth advantage to tumor cells. In fact, p53 DNA binding mutants can act as

“dominant negative” factors in some cell lines. DNA binding mutants form oligomeric complexes with wild type p53 protein and/or transcription co-activators to prevent their binding and activating target genes. This is believed to be one mechanism by which mutant p53 protein transforms cultured cells in tumorigenesis assays (reviewed in Levine, et al., 1991).

In addition to activation of target genes, p53 is also involved in both direct and indirect transcription repression. Multiple viral genes and a handful of cellular genes have been demonstrated to be indirectly repressed by p53 protein. Among the cellular genes indirectly repressed by p53 are c-fos, c-jun, IL-6, Rb and Bcl-2. These genes lack direct p53-DNA binding sites, and are likely repressed by over-expressed p53 through its ability to sequester general transcription factors and co-activators of transcription. This “broad-range” ability of p53 to regulate transcription repression is likely important to p53 tumor suppressor function, as these cellular gene products are also involved either directly or indirectly in cell cycle control

(reviewed in Ko and Prives, 1996).

Initially, it was believed that p53 induced transcriptional repression solely through interactions with cellular DNA-binding factors, rather than by direct DNA binding at p53 response elements. However, as the search for genes specifically repressed by p53 ensued, a handful of cellular genes were revealed to be repressed via direct p53-DNA binding. The first gene demonstrated to be directly repressed by p53 was the liver-specific α-fetoprotein gene

(AFP), as will be discussed below (Lee et al, 1999). Additional genes repressed by direct p53-

DNA binding include cdc25 (a cdc2 phosphatase), p202 (interferon-induced cell growth

regulator), POLD1 (DNA pol delta subunit), Map4, stathmin (involved in microtubule

polymerization) and MDR1 (multi drug resistance gene) (D'Souza, et al., 2001; Johnson, et al.,

2001; Krause, et al., 2001; Li and Lee, 2001; Murphy, et al., 1999).

Of these genes, the liver-specific, developmentally regulated AFP gene is most interesting with regards to expression pattern modification during tumorigenesis. During liver development,

AFP is robustly expressed, as is the tandemly linked albumin gene. AFP is the predominant serum protein of the developing fetus, but it’s exact role in development remains unclear. It has been suggested that AFP may act as a carrier protein, a growth promoter, or may be involved in fetal immuno-protection (reviewed in Crandall, 1981; Taketa, 1990). Both AFP and albumin retain high levels of expression throughout gestation. However, after birth, the genes demonstrate divergent regulation patterns where AFP is silenced while albumin remains expressed at relatively consistent levels. This specific post-natal silencing of AFP is maintained throughout adult life, except in cases of liver regeneration, liver disease and hepatocellular carcinoma (HCC). During these processes AFP is frequently re-activated to high levels of expression by unknown mechanisms (reviewed in Abelev and Eraiser, 1999). Transgenic mouse studies involving the AFP and albumin promoters and enhancers demonstrated that post-natal silencing, and likely inappropriate re-activation, of AFP are controlled by an 800 base-pair element located within the AFP proximal promoter (Camper, et al., 1989). This control region, dubbed the “developmental repressor region”, contains putative binding sites for a handful of transcription factors including CCAT Enhancer Binding Protein (C/EBP), Ets-2, Glucocorticoid

Receptor, and an overlapping Hepatic Nuclear Factor-3 (HNF-3)/p53/Smad4 DNA binding

element.

Studies focusing on the overlapping HNF-3/p53/Smad4 site, centered at -850 with respect

to the transcription start site, demonstrated that p53 association at the binding element disrupts

association of the powerful activator HNF-3, thereby triggering transcription repression (Lee, et

al., 1999). Further studies revealed that the p53 binding site was only partially responsible for

the post-natal repression of AFP. AFP deletion templates lacking the p53 binding site did not

lose their ability to be transcriptionally silenced by adult mouse liver extracts in cell-free in vitro

assays (Crowe, et al., 2000). However, silencing of the AFP gene was delayed as Western blot analysis of liver extracts prepared from 4-month old, p53 null mice did contain AFP (M. Barton, unpublished data). These data suggest that p53 involvement is necessary for proper timing of

AFP silencing, and that in its absence, other factors can compensate for this function. As will be discussed in detail (see Chapter 2), p53 association at the -850 binding element leads to modifications in chromatin structure at the minimal promoter, resulting in a generalized inaccessibility of the transcription start site. This inaccessibility is independent of histone acetylation status, suggesting a role for p53 in nucleosome positioning (Ogden, et al., 2001).

p53, chronic viral infection and cancer

Mutation within the p53 gene is the most common cause of p53 dysfunction in tumor

cells; however, predisposition to a cancer phenotype can also be achieved through functional

inactivation of wild-type p53 protein. In cases of chronic viral infection, p53 protein is frequently inactivated through direct interaction with virally encoded proteins. p53 plays a critical role in generalized disruption of viral replication. As a result many viruses have evolved means to inactivate, sequester or trigger degradation of cellular p53 protein. For example, HIV-1

Tat disrupts p53-mediated repression of HIV replication (Duan, et al., 1994); human papillomavirus E6 protein targets p53 for degradation (Seavey, et al., 1999); and Hepatitis B

Virus (HBV) HBx protein binds and functionally inactivates and/or leads to cytoplasmic sequestration of p53 (reviewed in Andrisani and Barnabas, 1999; Feitelson and Duan, 1997).

Additionally, p53 forms a complex with large T antigen of SV40, BZLF1 of Epstein-Barr virus,

E1b protein of adenovirus and co-localizes with ICP8 of herpes simplex virus-1 (Lane and

Crawford, 1979; Lechner, et al., 1992; Wilcock and Lane, 1991; Yew and Berk, 1992; Zhang, et al., 1994).

Chronic infection with HBV is tightly correlated to development of one of the more devastating and deadly types of cancer, hepatocellular carcinoma (HCC). HCC is the fifth most common malignancy worldwide and is the fourth most common cause of cancer death, with the bulk of these cases occurring in chronic HBV carriers (Arbuthnot and Kew, 2001; Stuver, 1998).

Currently, there are no effective treatments for patients with HCC, resulting in a five-year survival rate of lower than 5% (El-Serag and Mason, 1999). The yearly incidence of HCC is nearly equal to the annual HCC-related mortality rate, attesting to the poor prognosis of HCC patients (reviewed in Arbuthnot and Kew, 2001). Individuals who are chronically infected with

HBV are more than 200-times more likely to develop HCC than those who are not infected

(Feitelson and Duan, 1997; Hildt and Hofschneider, 1998). On a global scale, this is a major

health concern. Nearly 400 million people are chronically infected with HBV, with most cases

occurring in sub-Saharan Africa, South-East Asia and China (reviewed in Arbuthnot and Kew,

2001).

The exact mechanism by which chronic infection with HBV contributes to development

of HCC is not entirely clear. However, recent advances in HBV-HCC research have shed light

on the HBV-encoded transcription trans-activating protein HBx as being the most likely

oncoprotein (reviewed in Andrisani and Barnabas, 1999; Arbuthnot and Kew, 2001). HBx does

not bind double-stranded DNA, but activates transcription of both viral and cellular genes

through interactions with a wide range of cellular DNA binding proteins including general

transcription factors TFIID and TFIIH, transcriptional activators C/EBPα and β, Smad4, and a wide range of bZIP family proteins including IL-6, CREB, ATF-1, ATF-2, jun and GCN4 (Lee, et al., 2001; Ohno, et al., 1999; Palmer, et al., 1997; Perini, et al., 1999; Williams and Andrisani,

1995). In addition, HBx acts through SRE, NF-κB, AP-1, and AP-2 response elements by both direct (protein-protein) and indirect (stimulation of signal transduction cascades via stabilizing

GTP-ras) methods (Palmer, et al., 1997).

While substantial evidence exists regarding HBx modification of transcription activator function, little is known about HBx modification of transcription repressor function. Previous reports reveal that HBx association with bZIP trans-repressors ATF-3 and ICERIIγ augments their repressive potential (Barnabas and Andrisani, 2000). This effect likely modifies hepatocyte physiology, potentially contributing to development or progression of liver cancer in chronically infected patients.

p53 and HBx

In general HBx association with cellular factors enhances the activity of the factor, resulting in a push toward cell growth and replication. This augmentation creates a rich environment for factors assisting in viral replication and survival, as most viruses require cellular growth-promoting factors in order to replicate their own genomes. Conversely, the interaction between HBx and p53 results in a complete reversal of p53 function in both cellular transcription control and modulation of apoptosis.

HBx interaction with p53 likely evolved in response to p53-mediated disruption of HBV replication in hepatic cells (Lee, et al., 1998; Lee, et al., 1995; Ori, et al., 1998). In conjunction with a complex of liver specific co-factors, p53 can bind and repress activation from HBV

Enhancer II, the enhancer responsible for the tissue-specific replication of the virus. The p53- containing complex acts on HBV Enhancer II to block transcription from promoters under its control, resulting in loss of active HBV replication (Lee, et al., 1998; Ori, et al., 1998). HBV overcomes this hindrance by HBx binding to p53, thereby decreasing the negative effects of the p53-containing complex on Enhancer II (Ori, et al., 1998).

Strong support for HBx disruption of p53 function as a contributing factor in development of HCC was first demonstrated by a transgenic mouse study in which mice expressing HBx under control of its own regulatory elements developed liver tumors with 80 to

90% penetrance within four months of birth (Kim, et al., 1991; Ueda, et al., 1995). In these studies, HBx protein bound murine p53 protein to sequester p53 in the cytoplasm. While p53 mutation or inactivation is detected in over 80% of human cancers, p53 mutations are rare in the early stages of HCC. However, cytoplasmic sequestration of p53 by HBx may mediate loss of

p53 function, leading to genomic instability and accumulation of p53 mutations. Although

multiple studies support HBx-mediated p53 dysfunction, the exact mechanism by which HBx

inactivates p53 remains unclear. While the transgenic mouse study suggests that the mechanism

by which HBx functions is to prevent p53 entry into the nucleus, more recent studies suggest that

HBx may disrupt p53’s ability to bind DNA (Wang, et al., 1994), communicate with general

transcription factors, modify the stability of p53 tetramers (Truant, et al., 1995) and as will be

discussed, destabilize p53 co-repressor complexes.

TGF-β signaling and HCC

During HCC development, control of liver cell growth and division is lost. In

hepatocytes, a primary growth control pathway important during liver development, regeneration,

senescence and disease is the TGF-β pathway. The TGF-β family encompasses a large number

of signaling molecules including bone morphogenic proteins (BMP), inhibins/activins, Mullerian

inhibiting substance (MIS) and several isoforms of TGF-β. These molecules are involved in a

range of cellular activities including proliferation, terminal differentiation, lineage determination,

motility, adhesion, embryogenesis, fibrosis, immuno-reactivity, immuno-suppression and

apoptosis. Whether these processes are activated or repressed by TGF-β family signaling is

dependent upon cell type and state of differentiation (reviewed in Wong and Lai, 2001). The

primary TGF-β signaling molecules involved in control of liver growth, differentiation and disease are the BMPs and TGF-β1, 2 and 3. BMPs are required early in embryogenesis for

proper initiation of liver development. BMP2 and BMP4 expression is required in the

mesenchyme for initial hepatogenesis and expression of the early liver marker gene albumin.

BMP4 null mouse embryos exhibit a delay in liver bud formation, greatly decreased albumin

expression and increased pancreatic gene expression, demonstrating the importance of BMP

signaling in early liver development (Rossi, et al., 2001). Of the TGF-βs, TGF-β1 is expressed

at approximately 10-times higher levels than β2 and β3 in rat liver tissue, and is believed to play a role in terminal differentiation of normal hepatocytes (Fausto and Webber, 1993).

TGF-β1 is not usually produced by hepatocytes themselves. Instead, nonparenchymal cells in the liver tissue produce and release TGF-β1 which then acts through TGF-βI and βII receptors located on nearby hepatocytes. TGF-β signaling in the liver involves a complex and variable series of events, as stimulation by TGF-β1 can have differing effects depending on overall status of the surrounding liver tissue. Normally, stimulation by TGF-β1 results in inhibition of hepatocyte DNA replication and cell division. TGF-β signals through intracellular

Smad proteins to activate genes required for cell cycle arrest during late G1. Because of this, it is somewhat confusing to note that TGF-β1 mRNA levels increase significantly within 24 hours of partial hepatectomy, suggesting a role in liver regeneration (Fausto and Webber, 1993).

With regards to liver disease, TGF-β1 has been demonstrated to be involved in both liver cirrhosis and in liver cancer. TGF-β plays a predominant role in progression of liver cirrhosis in that it is a potent fibrogenic agent. Although is has been demonstrated that TGF-β levels increase substantially during cirrhosis, the exact role of TGF-β in the process is not entirely clear.

During cirrhosis, complex changes occur in the liver tissue involving cell proliferation, cell death and fibrogenesis. It has been suggested that a direct effect by TGF-β on fibrogenesis does not exist, but rather that fibrogenesis is a secondary effect in response to TGF-β mediated cell cycle

arrest. Another possibility is that direct stimulation of extracellular matrix proteins by TGF-β

may inhibit cell proliferation (reviewed in Fausto and Webber, 1993). Regardless of the exact

role TGF-β plays in cirrhosis, one can speculate that disruption in TGF-β signaling in fibrogenesis could contribute to tumorigenesis, as cirrhosis is commonly an early indication of

HCC.

The TGF-β pathway involves a number of proteins that have been implicated in cancer development. Most of the protein components in the pathway have been found to be mutated or functionally inactivated in a number of tumor types. In a normal cell, ligand stimulation of TGF-

β RII triggers the receptor to auto-phosphorylate and dimerize with its partner receptor, TGF-β

RI. Receptor dimerization triggers phosphorylation of membrane-anchored receptor activated

Smads (R-Smad) 1, 2 or 3, depending on ligand. Smad1 is typically used in BMP signaling, while Smads 2 and 3 are activated following TGF-β1-3 stimulation. R-Smad phosphorylation allows for association with the co-Smad, Smad4, which is believed to stimulate the DNA binding affinity and specificity of the R-Smads. The R-Smad/Smad4 complex can then enter the nucleus and activate genes containing a Smad response element via recruitment of p300 and other transcription co-activators (reviewed in Wong and Lai, 2001).

Because uncontrolled TGF-β signaling can have dire consequences on cell survival, multiple mechanisms exist to ensure tight regulation of the TGF-β pathway. Negative feedback loops involve inhibitory Smads (I-Smads) 6 and 7 and Ski family proteins Ski and SnoN. Smad6 are 7 are up-regulated in response to BMP and TGF-β/activin ligand stimulation to shorten the duration of signaling within the cell. When present at high concentrations, Smad6 and 7 associate with the membrane receptors to block phosphorylation and activation of R-Smads. At

lower concentrations Smad6 competes with Smad1 for association with Smad4, thereby blocking

DNA binding and gene activation (Massague, 1998). The role of the Ski family proteins is

slightly different in that their primary function is to keep TGF-β regulated genes turned off in the

absence of ligand stimulation. In the absence of TGF-β, Smad response elements are held in a repressed state via association of Ski or SnoN homo- or heterodimers with DNA bound Smad4 or

Smad4 containing complexes. Once associated at a Smad response element, Ski and SnoN recruit nuclear co-repressor N-CoR to the site, which in turn recruits mSin3A and histone deacetylases (HDACs). Assembly of this repressor complex triggers hypo-acetylation of the gene promoter, leading to transcriptional silencing. Stimulation of the cell by ligand triggers a rapid degradation of nuclear Ski and SnoN, allowing for binding of Smad activator complexes and gene activation (Sun, et al., 1999). However, after a short lag, TGF-β signaling also triggers up-regulation of the SnoN gene, allowing for an additional means of feedback inhibition

(reviewed in Miyazono, 2000; Miyazono, 2000).

A common means by which tumor cells lose responsiveness to TGF-β, thus losing partial control of cell cycle arrest, is through loss or down-regulation of TGF-β receptors on the cell surface. Loss of receptors is observed in a wide range of cancers including gastric and gut carcinomas, breast cancer, colon cancer, retinoblastomas, squamous cell carcinoma and head and neck cancer. Loss of receptor function is most frequently attributed to protein truncation or mutation within the conserved serine/threonine kinase domain. These mutations can lead to loss of the receptor extracellular domain in the case of protein truncation or loss of auto- phosphorylation, receptor dimerization and R-Smad activation in the case of kinase domain mutation (Rooke and Crosier, 2001; Wong and Lai, 2001).

Smad mutations are also implicated in cancer progression. Smad4 was first identified as a potential “deleted in pancreatic cancer” (DCP4) tumor suppressor gene (Massague and Wotton,

2000). Multiple Smad4 and Smad2 mutations have been detected in breast, gastric and head and neck carcinomas, and work is currently underway to determine if Smad4 has “global” tumor suppressor activity (reviewed in Rooke and Crosier, 2001). Because Smad4 is the common

Smad involved in multiple TGF-β family signaling pathways, its loss of function is likely to affect a large number of genes involved in cell growth control pathways. Inactivation of Smad4 and Smad2 typically occurs through missense mutations, nonsense mutations, deletions, frameshift mutations, or loss of the chromosomal region. Hot spots for mutation exist in the coding sequence of each gene within the MH2 domain, which is responsible for Smad-Smad interactions. These mutations frequently result in an increased affinity of the MH2 domain for the MH1 domain (DNA binding domain), leading to inappropriate protein folding and loss of function (Massague, 1998).

Negative regulators of the TGF-β pathway are also potential points of signaling disruption when mutated. Up-regulation or inappropriate over-expression of the I-Smad6 or 7 causes resistance to TGF-β growth inhibition. Over-expression of I-Smads can “swamp out” or squelch receptor activated Smads, thereby preventing association with Smad4 and activation of target genes. Both Smad6 and 7 have been found to be highly expressed in pancreatic cancer, and are believed to contribute to tumor progression (Miyazono, 2000). More notable, perhaps, are conversions of SnoN and Ski to proto-oncogenes. Initially, ski was identified as an oncogene carried by Sloan Kettering retroviruses. Over-expression of Ski transforms avian fibroblasts by mediating complete resistance to TGF-β ligand. Over-expressed Ski inhibits binding of

p300/CBP co-activator to Smad response elements, thus silencing target genes (Miyazono,

2000). snoN is not as potent an oncogene as ski in that is has to be over-expressed at much higher levels to transform avian fibroblasts (Boyer, et al., 1993). This is likely a result of SnoN having a lower metabolic stability than Ski. SnoN protein is degraded much more rapidly, and in response to lower levels of TGF-β than are required to degrade Ski (Sun, et al., 1999). Thus, it is likely that in order for SnoN over-expression to transform a cell, it must occur in combination with additional mutations in the TGF-β and/or other growth control signaling pathways.

Chromatin Structure and Gene Regulation

Regulation of individual genes is highly dependent on ordering and modification of chromatin structure within the gene promoter. The basic building block of chromatin, the core nucleosome, is composed of an (H3-H4)2 tetramer flanked on each side by an H2A-H2B dimer.

Nearly 160 base-pairs of DNA wrap around each core nucleosome in approximately 1.65 super- helical turns (Luger, et al., 1997; Wolffe and Guschin, 2000). Formation of higher order chromatin involves association of linker histone H1 between individual nucleosome cores. This association is facilitated by the N-and C-terminal tails of H1 interacting with DNA within and between the nucleosome cores (reviewed in Wolffe and Guschin, 2000). Histone-DNA and histone-histone interactions are regulated through a range of histone modifications including phosphorylation, methylation, ubiquitination, ADP-ribosylation and, as will be discussed in detail, acetylation and/or deacetylation of core histone tails (reviewed in Strahl and Allis, 2000).

Regulation of genes by transcription factors such as p53 or Smad4 typically involves recruitment of histone deacetylases (HDACs) or histone acetyl transferases (HATs) to DNA.

Both HATs and HDACs exist in multi-factor, enzymatic complexes which appear to have been highly conserved through evolution. Generally speaking, the targets of HATs and HDACs are positively charged lysine residues located in the N-terminal tails of histones H3 and H4.

Acetylation and deacetylation of histone tails are very dynamic processes. Lysine residues in transcriptionally active chromatin are modified at a rapid pace. Acetylated lysines in regions of hyper-acetylated chromatin maintain half-lives of only minutes (reviewed in Wolffe and

Guschin, 2000).

Numerous transcription co-activators including yeast and mammalian GCN5 and mammalian p300/CBP, P/CAF, TAFII230/250 and ACTR possess histone acetyltransferase activity. These proteins typically exist in multi-subunit complexes, which are recruited to promoters via DNA-bound transcriptional activators. Acetylation of N-terminal tails can affect transcription in a number of ways: hyper-acetylation of local chromatin neutralizes charged lysine residues resulting in decreased affinity of the histone tails for nucleosomal DNA and/or neighboring nucleosomes. This allows for increased binding of local transcriptional activators to their recognition elements (Wolffe and Guschin, 2000). Multiple studies have revealed that HAT function is required to facilitate activation of TGF-β responsive genes. Upon ligand stimulation, transcription activation is achieved through recruitment of p300/CBP to target genes by DNA- bound Smad complexes. p300/CBP catalyzes the hyper-acetylation of local chromatin, leading to robust transcription activation (Attisano, 2000; Derynck, et al., 2001). p53 has also been demonstrated to directly utilize p300/CBP in activation of its target genes bax and p21 (Espinosa and Emerson, 2001; Shikama, et al., 1999). In vitro assays demonstrated that p53-mediated targeting of p300/CBP to the p21 promoter triggers a localized cascade of histone hyper-

acetylation beginning at the p53 binding site and spreading to the transcription start site.

Conversion of hyper-acetylated chromatin to hypo-acetylated chromatin through the

action of histone deacetylase complexes is critical in establishing a transcriptionally silenced

gene. The first human HDAC cloned, HDAC1, was found to be a human homologue of the yeast

RPD3 protein, a known transcription co-repressor (Taunton, et al., 1996). Additional human

homologues were subsequently cloned, including HDAC2, HDAC3, HDAC3A, HDAC3B and

HDACs 4-10 with the list continuing to grow (reviewed in Johnson and Turner, 1999).

Recruitment of HDACs to a DNA element results in a local, generalized de-acetylation of histone

tails, triggering chromatin condensation and occlusion of basal transcription factors from the

minimal promoter. Histone deacetylation is a requisite process by which many factors achieve

transcriptional silencing. Nuclear hormone receptors are perhaps the most thoroughly studied

group of factors utilizing HDACs to facilitate transcription repression. In the case of thyroid

hormone receptor, un-liganded heterodimers of TR/RXR (retanoid X receptor) bind

constitutively to DNA target elements and recruit the co-repressor N-CoR. N-CoR binding

facilitates association of mSin3 and recruitment of HDAC, allowing for transcriptional silencing.

Inhibition of the HDAC catalytic activity by a specific inhibitor, trichostatin A (TSA),

completely alleviates repression of the target gene, demonstrating the direct importance of

histone deacetylation in transcriptional repression (reviewed in Wolffe and Guschin, 2000). The

same is true for promoters regulated by downstream TGF-β effectors. In the absence of TGF-β ligand stimulation, DNA- bound Smad proteins associate with nuclear Ski and/or SnoN proteins.

Association of Ski or SnoN at the DNA element facilitates recruitment of N-CoR, followed by mSin3 and HDAC (reviewed in Miyazono, 2000; Miyazono, 2000). mSin3 has also been shown

to associate with p53, and may be involved in modulation of hypo-acetylation at some p53

repressed promoters. In fact, p53 has been demonstrated to specifically recruit mSin3A and

HDAC to repress both the Map4 and stathmin promoters. This repression is absolutely

dependent on HDAC function, as treatment of cells with TSA completely inhibits the p53-

mediated repression (Murphy, et al., 1999).

Conclusion

Clearly, multiple processes are involved in proper regulation of target genes. Appropriate binding of transcriptional activators or repressors to DNA elements are required to target co- factors that facilitate chromatin modifications necessary for gene regulation. These studies were designed to examine how established gene regulation patterns are disrupted during chronic HBV infection and liver tumorigenesis. The α-fetoprotein gene, which is re-activated in over 80% of liver carcinomas, was utilized as a model system to examine three processes potentially involved in development of HCC: How does p53 mediate direct transcription repression? Does the HBV- encoded HBx protein disrupt this process through direct association with p53? If so, what is the mechanism behind HBx-mediated disruption of p53 transcription regulation?

Chapter 2: p53 targets chromatin structure alteration to repress α-fetoprotein gene expression

Summary

Many of the functions ascribed to p53 tumor suppressor protein are mediated through

transcription regulation. We have shown that p53 represses hepatic-specific alpha-fetoprotein (AFP)

gene expression by direct interaction with a composite HNF-3/p53 DNA binding element. Using

solid-phase, chromatin-assembled AFP DNA templates and analysis of chromatin structure and

transcription in vitro, we find that p53 binds DNA and alters chromatin structure at the AFP core promoter to regulate transcription. Chromatin assembled in the presence of hepatoma extracts is activated for AFP transcription with an open, accessible core promoter structure. Distal (-850) binding of p53 during chromatin assembly, but not post-assembly, reverses transcription activation concomitant with promoter inaccessibility to restriction enzyme digestion. Inhibition of histone deacetylase activity by trichostatin-A (TSA) addition, prior to and during chromatin assembly, activated chromatin transcription in parallel with increased core promoter accessibility. Chromatin immunoprecipitation (ChIP) analyses showed increased H3 and H4 acetylated histones at the core promoter in the presence of TSA, while histone acetylation remained unchanged at the site of distal p53 binding. Our data reveal that p53 targets chromatin structure alteration at the core promoter, independently of effects on histone acetylation, to establish repressed AFP gene expression.

Introduction

The tumor suppressor protein p53 plays central roles in the regulation of cell growth, cell cycle

arrest and apoptosis. It is activated in response to a variety of DNA damaging agents and has been

shown to interact with a number of cellular proteins of both mammalian and viral origins. In

general, the functional consequence of p53 DNA binding is transcription activation of target genes

with role(s) in cellular stress response, as well as certain developmental pathways (reviewed in

Greenblatt, et al., 1994; Kastan, et al., 1995; Ko and Prives, 1996; Lane, 1994; Weiss, et al., 1995).

More recently, however, examples of p53-mediated transcription repression through sequence-

specific DNA binding have been reported (Budhram-Mahadeo, et al., 1999; Lee, et al., 1999; Ori, et

al., 1998).

We have established that p53 binds within the AFP distal developmental repressor region,

displacing bound trans-activator HNF-3 (FoxA) protein, and contributes to post-natal, tissue-specific repression of AFP (Lee, et al., 1999). AFP is normally expressed at high levels in the liver of the developing fetus and silenced after birth. Adult expression of AFP is monitored as a tumor marker: aberrant reactivation occurs in up to 85% of all hepatocellular carcinoma cases (reviewed in

Tilghman, 1985).

In this study, we find that p53 binds to DNA during chromatin structure organization as an obligate step in transcription repression. DNA binding of p53 mediates distal regulation of AFP transcription through alteration of chromatin structure at the core promoter. The ability of p53 to regulate nucleosome positioning at the core promoter is independent of histone modification. We show that histone hyperacetylation has direct consequences for core promoter chromatin accessibility and gene activation. These effects are overridden by addition of p53, which represses transcription

by restricting chromatin accessibility even in the presence of hyperacetylated histone H3 and H4 N- terminal tails at the core promoter.

Experimental procedures

Plasmids and solid-phase DNA templates

AFP/lacZ contains 3.8 kb upstream DNA including proximal and distal promoter and enhancer I, fused to the coding region of β-galactosidase (Spear, et al., 1995). DelA/lacZ is identical except that it contains a 10 base pair deletion in one p53 binding half site between -850 and -840, as well as a 4 bp mutation in the other half site. It was constructed by PCR mutation of the previously described

AFP/mut5 template (Lee, et al., 1999). Solid-phase AFP/lacZ and DelA/lacZ templates were coupled to streptavidin-coated paramagnetic beads as described (Crowe and Barton, 1999). The chick adult

β-globin plasmid, pUC18ABC/ contains the entire promoter, coding sequence and 3' enhancer

(Emerson, et al., 1989).

Protein expression and cellular extracts

Cell extracts were prepared from HeLa and HepG2 (AFP positive, ATCC catalog number HB-

8065) cells as described by Dignam et al. (Dignam, et al., 1983) with minor modifications (Lee, et al., 1999). All extracts contained total proteins in concentrations of 5-10 mg/ml. Xenopus egg extract high-speed supernatant (HSS) was prepared as described previously (Barton and Emerson,

1996). Constitutively activated, recombinant p53 protein was expressed from p53 30his as previously detailed (Zaret et al, 1995).

Chromatin assembly and in vitro transcription

In vitro chromatin assembly and transcription reactions were performed as reported (Crowe and

Barton, 1999). When trichostatin-A (TSA, Sigma Chemical Co.) was added to inhibit endogenous

histone deacetylases in the Xenopus egg HSS, the 10 mM DMSO stock solution was diluted to the desired final concentration and incubated with HSS on ice for 10 minutes prior to bead/DNA addition and chromatin assembly. For post-chromatin assembly additions, proteins were added after the one-hour chromatin assembly period and incubated for an additional 30 minutes. Chick β-globin

DNA template was added as a control for RNA recovery. Results were quantified by ImageQuant analysis of scanned autoradiograms.

Restriction enzyme accessibility

Hinc II restriction enzyme accessibility experiments were performed as described (Crowe, et al.,

1999). All restriction enzyme digests were run on a 2% agarose gel and Southern blotted. A 23 bp

32P-endlabeled oligomer corresponding to promoter sequence at +4 to +26, 5'-

CCCACTTCCAGCACTGCCTGCGG-3', was used as probe.

Acetylated histones H3 and H4 antibodies

The N-terminal 24 amino acids of human N-acetyl-lysine modified (4, 9, 14, 18, 23)-H3 and

(5, 8, 12, 16, 20)-H4 were synthesized as tetrameric multiply antigenic peptides (MAP) by

Research Genetics, inc. (Birmingham, AL). Non-acetyl immunoreactive antibodies were removed by subtracting with a synthetic, N-terminal non-acetylated mixture of H3 or H4 peptides. Finally, antibodies were affinity purified using AcH3 and AcH4 MAP. Specificity was confirmed by Western blot analysis, under standard conditions, of histones purified by sodium butyrate treatment and fractionated by acid gel electrophoresis.

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) assays were performed on in vitro chromatin assembled

DNA templates. Protein/DNA complexes were cross-linked by exposure to 1% formaldehyde (final

concentration) for ten minutes at room temperature, followed by 50 minutes on ice. These reactions were diluted three-fold in Xenopus egg extract buffer (Crowe and Barton, 1999) and mixed gently.

The supernatant was removed after magnetic concentration, and bead/DNA was resuspended in 1X

SM2 buffer (500 mM Sucrose, 80 mM KCl, 20 mM HEPES, pH 7.5, 3.5 mM ATP, 6 mM CaCl2) plus 1.5 U micrococcal nuclease (MNase, Boehringer Mannheim) (Crowe and Barton, 1999). A five minute incubation with MNase digested cross-linked chromatin into 200-500 nucleotide size fragments, as determined empirically by agarose gel electrophoresis and Southern blotting. After stopping digestions in 20 mM EDTA/2 mM EGTA (final concentration), 9 volumes of ChIP dilution buffer (16.7 mM Tris-HCl, pH 8.1, 167 mM NaCl, 1.1% Triton X-100, 0.01% SDS, 1.2 mM EDTA) were added to each sample. This diluted sample of chromatin/protein/DNA fragments was removed as a supernatant from the immobilized paramagnetic beads and divided among several reaction tubes for incubation with control and specific antibodies as described (Kuo and Allis, 1999). The presence of immunoprecipitated DNA sequences was determined by Southern blotting using a slot-blot manifold for binding DNA to GeneScreen Plus membrane (Beckman Chemical Co.). Hybridization with 32P-endlabeled, double-stranded oligomers encompassing AFP/lacZ sequences from +4 to +26

(promoter probe), from -860 to -830 (p53 binding site) and random-primed labeled full-length template was performed as described (Crowe, et al., 1999). Results were quantified by ImageQuant analysis of scanned autoradiograms. Values are expressed as a ratio of bound to input, corrected by comparison to no antibody and full-length template controls.

Results

p53 binding mediates transcriptional repression and chromatin structure alteration

We employed in vitro chromatin assembly of solid-phase AFP DNA templates to reconstitute repression of AFP transcription through a distal p53-regulatory element. We showed previously that chromatin assembly by this method establishes physiologically spaced nucleosomes over the entire

DNA template and renders in vitro transcription tissue-specific, in contrast to nucleosome-free DNA

(Crowe and Barton, 1999; Crowe and Barton, 1999). These templates were either transcribed for functional analysis or structurally analyzed for promoter accessibility or histone modification status

(Fig. 1A). Nucleosome assembly silences AFP transcription compared to unassembled DNA templates (Fig 1B, lane 1 compared to β-globin transcripts). Addition of HepG2 extract, prepared

from cultured human hepatoma cells that express AFP, during chromatin assembly derepresses

and/or activates AFP transcription an average of five-fold (lane 2). Titration of increasing amounts

of recombinant p53, in addition to HepG2 extract, during chromatin assembly represses AFP

chromatin transcription up to 3-fold (lanes 6 and 7). However, p53 introduced after chromatin

structure was established had no significant effect on transcription (p53post, average of 1.1-fold

increase, lanes 3-5).

We have utilized restriction enzyme accessibility to monitor changes in promoter chromatin

structure induced by p53 DNA binding (Fig. 2). Hinc II restriction digestion at sites that flank the

AFP promoter at -55 and +29, Fig. 2A, correlates with a core promoter region relatively free of a

bound nucleosome and open to transcription complex assembly (Crowe and Barton, 1999; Crowe

and Barton, 1999). Structural analysis mirrored the functional effects of p53 addition both during

and post-chromatin assembly (Fig. 2B). Compared to chromatin assembly in buffer only, HepG2

extract established an open core promoter (lane 3 compared to lane 2). Addition of p53 protein during chromatin assembly revealed a dose-dependent chromatin closure (lanes 4-6, compared to lane 3). The core promoter chromatin structure remained relatively inert to p53 protein addition post-assembly (lanes 7 and 8). A quantitative average of these data and similar experiments are graphed in Fig. 2C. Comparison of both transcription and chromatin structural consequences of p53 addition pre- and post-chromatin assembly, support the view that p53 action is primarily at the level of chromatin structure organization rather than targeting modification of an established chromatin structure.

The p53 DNA binding site within the AFP upstream region (-860/-830, Fig. 3A) fits the consensus binding sequence for a p53 tetramer at the 5' half (PuPuPuCA/TA/TGPyPyPy, (El-Deiry, et al., 1992)). However, this p53 repressor element has a 3' half site following a 3 base-pair nucleotide spacer that deviates from consensus (PyPyPyCTAGPuPyPu). The influence of DNA binding sequence on p53 conformation and function has been previously described (Thornborrow and Manfredi, 1999), but the specific way in which this response element dictates p53 regulation of

AFP is not known. Deletion and mutation of the p53-binding site (DelA) abolishes both p53 and

HNF-3 binding at this site (data not shown). Consistent with lack of p53 binding, repression of chromatin transcription (Fig. 3B) and structure accessibility at the core promoter (Fig. 3C) are lost.

Histone modification does not alter p53-mediated restriction of promoter accessibility

Nucleosome positioning, which restricts core promoter access to restriction enzymes and transcription preinitiation complexes, may be affected by histone N-terminal tail interactions with

DNA. Chromatin composed of highly acetylated histones, in general, may be more dynamic and

readily activated for transcription; whereas, the opposite may be true for under-acetylated

nucleosomes. Multiple transcription factors interact with protein complexes displaying intrinsic

enzymatic activity, such as histone acetyl transferases (HATs) or histone deacetylases (HDACs), and

target modification of chromatin by virtue of their DNA binding ability (recently reviewed in Howe,

et al., 1999; Sterner and Berger, 2000). Several studies have shown interactions between both HAT

and HDAC complexes and p53, interactions that may promote histone modification, p53

modification or p53 stabilization (Bayle, et al., 1995; Gu and Roeder, 1997; Koumenis, et al., 2001;

Liu, et al., 1999; Murphy, et al., 1999; Woods and Vousden, 2001; Zilfou, et al., 2001). Our baseline

study of histone modification at the core promoter and the p53 binding site at -850 revealed no

significant variation in histone H3 acetylation between these regions, in the presence or absence of

p53. Four separate ChIP experiments were performed and the percent of acetylated histone H3

present at these regions averaged: at the core promoter, 27.8% +/- 9.3 (in the absence of p53) and

27.6% +/-9.2 (in the presence of p53); and at the p53 binding site, 35.1% +/- 5.7 (-p53) and 32.2%

+/- 5.9 (+p53).

Although there were no significant variances between histone modification states localized at the

core promoter or p53-binding site, there were striking differences in acetylation states when the

equilibrium between acetylases and deacetylases was shifted by addition of TSA to Xenopus egg extract in the absence of hepatoma proteins (Fig. 4). In response to increasing concentration of TSA, nucleosomes present at the core promoter region increased in both H3 and H4 acetylation 2-3 fold.

However, in the local region of p53 binding, histone modification was maintained at a baseline level over the concentration range of TSA (Fig. 4A). Somewhat surprisingly, the presence or absence of p53 protein made little difference to the state of histone modification at all levels of TSA. Thus, we

find no evidence that p53 alone targets modification of histone tails at the p53-binding site or at the core promoter.

The specific increases in histone H3 and H4 acetylation at the core promoter led directly to derepression of transcription in the absence of any exogenous activator proteins, Fig. 4B.

Functionally, the 2-3 fold changes in histone acetylation at the core promoter mediated a two-fold increase in transcription of chromatin assembled in Xenopus egg extract treated with TSA (lanes 1-

5). The increase in transcription was not dependent on the p53/HNF3 binding site, as similar transcription of the DelA AFP template in TSA-treated egg extract also showed transcription increase with increasing TSA (Fig. 4C, lanes 2 and 3 compared to 1).

Despite the presence of highly acetylated histones at the core promoter, p53 was able to repress transcription fully when added during chromatin assembly (Fig. 4B, lane 7 compared to lanes 1 and

6). The ability to repress chromatin transcription was absolutely dependent on DNA binding of p53 as the DelA AFP template remained transcriptionally active in the presence of p53 (Fig. 4C, lanes 4 and 5). Restriction accessibility analysis of core promoter chromatin (Fig. 4D) showed a parallel increase of two-fold in the presence of TSA (lane 3), and repression to baseline levels in the presence of added p53 (lanes 4 and 5), both in parallel with effects on transcription function (Fig. 4B). The ability of p53 to modify nucleosome/DNA interactions at the core promoter, resulting in chromatin structure closure and repression of transcription, is not dependent on tissue-specific factors or targeted modification of histone acetylation, and occurs even in the presence of highly acetylated nucleosomes.

Discussion

During hepatic development and post-natal silencing of AFP expression, specific changes in chromatin structure of the AFP gene occur, which likely play a role in regulation (Chou and Wan,

1989; Godbout and Tilghman, 1988). Our present study reveals that the functional outcome of p53

DNA binding within the AFP distal repressor region is dictated by chromatin structure organization.

Chromatin structure can influence transcription regulation by obstructing transcription factor access to DNA or by facilitating interactions between distal regulatory factors and proximal promoter elements to repress or activate transcription (Agarwal and Rao, 1998; Langst, et al., 1998; Shachter, et al., 1993; Shild, et al., 1993; Wallrath, et al., 1994; Wijgerde, et al., 1995). Our in vitro chromatin transcription system recapitulates distal regulation of AFP transcription by p53 bound to DNA 850 base pairs 5' of the transcription start site within the AFP developmental repressor domain. As best studied in Drosphila, distal repressors, like distal enhancers, are essential in regulation of development and differentiation (Gray and Levine, 1996).

Proteins that interact with DNA regulatory elements dictate chromatin activation or repression, as well as the consequences of p53 regulation. AFP chromatin is derepressed or activated in the presence of hepatoma extract, but remains transcriptionally repressed when assembled in HeLa extract (Crowe, et al., 1999). As shown here, p53 mediates repression of hepatoma-activated chromatin, but in the presence of HeLa extract a low level of transcription activation is observed instead (Ogden, et al., 2000). Thus, the interpretation of p53 protein interaction with DNA, whether activating or repressing chromatin structure, is influenced by multiple trans-acting factors. The timing of these interactions relative to chromatin assembly is important as well. We find that p53 lacks the ability to alter an established, activated chromatin structure. HNF-3 (FoxA) protein,

present in the HepG2 extract, mediates core accessibility to Hinc II restriction enzyme and basal transcription factors in chromatin transcription (Crowe, et al., 1999). Zaret and coworkers established that HNF-3 is an architectural transcription factor that can position nucleosomes along the albumin enhancer, rendering it competent for later trans-activator binding (McPherson, et al.,

1993; Shim, et al., 1998). Thus, HNF-3 acts as primary effector of chromatin modification and derepression, and this established chromatin structure cannot be altered by post-assembly addition of p53. Previous investigations revealed that p53 binding to DNA induces considerable bending and twisting at its binding site, and the inherent sequence-dependent form that DNA assumes greatly affects p53 DNA binding properties (Nagaich, et al., 1999). The manipulation of DNA and chromatin structure by proteins such as HNF-3 and/or p53 may establish a requisite order of transcription factor binding to induce specific chromatin-repressed or -activated forms. Studies of

Swi5p-mediated recruitment of Swi/Snf and SAGA complexes at the yeast HO endonuclease gene promoter (Cosma, et al., 1999), as well as temporal recruitment of chromatin remodeling and histone modifying complexes by nuclear receptors (Dilworth, et al., 2000), support the idea that only specific transcription factors can interact with chromatin to initiate a series of regulatory steps. This temporal order is likely influenced by flanking DNA sequence, the complexity of the regulatory element(s) and interacting proteins.

Our results showing that p53 cannot bind and repress a previously activated chromatin template are in contrast to p53 action at a chromatin-repressed (assembled in buffer) p21 gene template

(Espinosa and Emerson, 2001). In this case, p53 can bind to chromatin and target p300 to acetylate histone tails at a p53-binding site, which then spreads distally to the core promoter. Together, p53 and p300 activate transcription of chromatin-repressed p21 gene templates. The ability of p53

protein to target histone modification was also suggested by previous studies. Transcription repression of the MAP4 gene is correlated with p53-mediated histone deacetylation and promoter- localized histone modification in cultured cells (Murphy, et al., 1999). Our studies of histone modification revealed little change in histone acetylation mediated by p53 addition during chromatin assembly in untreated Xenopus egg extract. When the equilibrium between histone acetylase and deacetylase activities was shifted toward acetylation by TSA addition, the p53 binding site region was maintained in a state of reduced acetylation, even in the absence of exogenous p53, compared to the core promoter. We propose that local histone modification at the AFP repressive p53-binding site is maintained by multiple protein complexes (S. Ogden and M. Barton, unpublished results).

Regional, regulated shifts in acetylation/deacetylation equilibria may be revealed by disruption of histone acetylase or deacetylase activities (Vogelauer, et al., 2000), which exist endogenously in

Xenopus egg extract.

Despite increased histone acetylation, p53 regulates chromatin structure alteration that creates inaccessibility at the AFP core promoter. The ability of p53 to modify chromatin structure at the core promoter is not dependent on the presence of hepatoma-specific factors, and is inhibited when tissue- specific factors have previously established active chromatin. Our studies reveal a hierarchy of transcription regulation by p53 in which distal alterations in chromatin structure rather than local modification of histone tails play a key role in repressing AFP gene expression. It will be of great interest to compare other p53-repressed and -activated genes and establish if the interpretation of p53 induction as a transcription activation or repression signal is dictated by DNA regulatory site complexity, interacting proteins and consequent chromatin structure alteration.

Figure 1. p53 represses AFP chromatin transcription in vitro.

A. Solid-phase chromatin assembly. Chromatin is assembled in the presence of proteins, extracts and/or buffer for one hour prior to magnetic concentration, washing and removal of unbound proteins. Chromatin templates are then transcribed or analyzed for chromatin structure accessibility. B. Transcription analysis of p53 added during chromatin assembly or after chromatin assembly. 500ng of AFP/lacZ template DNA were incubated in buffer only (lane 1) or with HepG2 extract, lanes 2-7, and 0 (-, lane 2), 150ng (+, lane 3), 300 ng(++, lanes 4 and 6), or 450ng (+++, lanes 5 and 7) p53 and transcribed. p53 was added either prior to chromatin assembly (lanes 6 and 7) or after chromatin assembly was completed (lanes 3-5). Molecular weight standards (MW) were radiolabeled x174 HaeIII digestion reaction (Gibco BRL). β- globin serves as an RNA recovery control.

Figure 2. p53 organizes chromatin structure.

A. Core promoter accessibility assay. The core promoter of AFP has HincII sites at -55 and

+29. Digestion of chromatin with Hinc II releases an 84 bp fragment only when both sites are accessible. The ovals represent nucleosomes bound at the AFP promoter: dashed ovals are

nucleosomes that obstruct enzyme digestion and, presumably, transcription factor binding. B.

Restriction enzyme probing of chromatin structure. Hinc II accessibility analysis was performed on chromatin assembled templates. Chromatin-free DNA (lane 1) and chromatin- assembled DNA (lanes 2-8) were preincubated with p53 (lanes 4, 5 and 6; 75, 150 and 300 ng, respectively) or p53 was added after assembly (lanes 7 and 8, 150 and 300 ng, respectively) in the presence of HepG2 extract (lanes 3-8). Uncut DNA is present is lane 9. C. Graphic representation of transcription activation and promoter accessibility. The quantitative averages of at least two separate assays for p53-regulated transcription (open boxes) and promoter accessibility (shaded boxes), added prior to (pre) or after (post) chromatin assembly, are compared to HepG2 activated chromatin assembled in the absence of p53. Concentrations of p53 are 1x, 150 ng, and 2x, 300 ng.

Figure 3. DNA binding is required for p53 function.

A. A mutated AFP template. The -860 to -830 sense strand of AFP is depicted with the p53 binding site (consensus 5' half site, solid line, and non-consensus 3' half site, dashed line) marked above the normal DNA sequence (AFP) and the HNF-3 site below mutated p53 sequence (DelA).

DelA lacks the nucleotides indicated by hyphens and has altered bases in lower case letters. B. p53 binds to DNA to repress transcription. AFP (lanes 1-5) and DelA (lanes 6-9) bead-DNA

were transcribed in vitro as chromatin-free (lane 1) and -assembled (lanes 2-9) templates.

Chromatin was assembled in the presence of buffer (lanes 2 and 6), HepG2 extract (lanes 3-5 and

7-9), plus p53 protein (lanes 4 and 8, 150 ng; lanes 5 and 9, 300 ng). C. p53 binds DNA to alter core promoter accessibility. AFP (lanes 1-5) and DelA templates (lanes 6-10) were incubated in buffer (lanes 5 and 10) or chromatin assembled (lanes 1-4 and 6-9) as described above. HepG2 extract (lanes 2-5, 7-10) was supplemented with 150 (+, lanes 3 and 8) and 300

(++, lanes 4 and 9) ng of p53 during assembly. Southern blot analysis reveals relative HincII accessibility of chromatin (84 bp HincII band).

Figure 4. p53 mediated repression of AFP transcription occurs even in the presence of hyperacetylated histones at the core promoter.

A. Addition of TSA promotes hyperacetylation of histones H3 and H4 at the AFP core

promoter but not at the p53 binding site. ChIP analyses of H3 (white bars) and H4 (shaded bars)

acetylated histone populations present at the core promoter region (open bars) and the p53 binding

region (hatched bars) were performed on in vitro chromatin assembled AFP templates in the

presence (+, 300 ng) or absence (-) of p53. Chromatin was assembled in Xenopus egg extract, incubated in increasing amounts of TSA: none, 10 nM, 100 nM and 3 µM. Histone acetylation levels were expressed as a ratio compared to baseline values determined in the absence of TSA (see text). These TSA titration experiments were performed twice, thus standard deviation values are not presented. B. Histone hyperacetylation leads directly to derepressed chromatin transcription that is silenced by p53 addition. Xenopus egg extract incubated in the presence of 0, 10 nM, 100 nM, 3 µM and 10 µM TSA was used to assemble AFP bead-DNA into chromatin and transcribed

(lanes 1-5, respectively). p53 (300 ng) was added during chromatin assembly (lanes 6 and 7) in egg extract incubated in the presence of 0 (lane 6) or 3 µM TSA (lane 7). RNA recovery control primer extension analysis of added β-globin DNA template was performed separately on these reactions

(small gel insert). C. p53-mediated silencing of TSA-derepressed transcription requires DNA binding. DelA bead-DNA was assembled in chromatin as described above, using Xenopus egg extract incubated in the presence of 0 (lane 1), 3 µM (lanes 2, 4 and 5) and 10 µM (lane 3). p53 protein was added during chromatin assembly (lane 4, 150 ng; and lane 5, 300 ng). D. p53 mediates core promoter closure even in the presence of hyperacetylated histones. HincII restriction accessibility was performed on AFP templates in the absence of chromatin assembly (lane

1) or when assembled into chromatin (lanes 2-5). Chromatin was assembled in Xenopus egg extract incubated in no TSA (lane 2) or 10 µM TSA (lanes 3-5). Addition of p53 (+, 150 ng, lane 4; ++, 300 ng, lane 5) altered chromatin accessibility to HincII enzyme.

Chapter 3: Hepatitis B viral transactivator HBx alleviates p53-mediated repression of α-

fetoprotein gene expression

Summary

Chronic infection with Hepatitis B Virus (HBV) is associated with development of hepatocellular

carcinoma (HCC). The exact mechanism by which chronic infection with HBV contributes to

onset of HCC is unknown. However, previous studies have implicated the HBV transactivator

protein, HBx, in progression of HCC through its ability to bind the human tumor suppressor

protein, p53. In this study, we have examined the ability of HBx to modify p53 regulation of the

HCC tumor marker gene, alpha-fetoprotein (AFP). By utilizing in vitro chromatin assembly of

DNA templates prior to transcription analysis, we have demonstrated that HBx functionally

disrupts p53-mediated repression of AFP transcription through protein-protein interaction. HBx

modification of p53 gene regulation is both tissue-specific and dependent upon the p53 binding

element. Our data suggest that the mechanism by which HBx alleviates p53 repression of AFP

transcription is through an association with DNA-bound p53 resulting in a loss of p53 interaction

with liver specific transcriptional co-repressors.

Introduction

Chronic infection with Hepatitis B Virus (HBV) is a predominant risk factor associated

with development of hepatocellular carcinoma (HCC). Multiple lines of evidence support the

relationship between chronic HBV infection and HCC: geographic correlation exists between

global distribution of HCC and the prevalence of HBV carrier states; a high incidence of HBV

markers in blood and tissue samples is detected in HCC patients; 30% of all virally-induced

human tumors involve HBV infection (Hildt and Hofschneider, 1998; Kew, 1996). Based on

epidemiological studies involving chronic HBV infection, it is estimated that the relative risk of

developing HCC may be between 100 to 200-fold higher for HBV carriers than for non-carriers

(Feitelson and Duan, 1997; Hildt and Hofschneider, 1998).

The most likely scenario for HBV’s role in HCC predisposition is by modification of host

gene regulation (Andrisani and Barnabas, 1999; Hsu, et al., 1988; Yoo, et al., 1996). Integration

of viral DNA into the host genome can mediate host gene deregulation by a variety of

mechanisms: integration of viral promoters can activate and/or mutate neighboring host genes

(Hsu, et al., 1988); integration of viral DNA encoding the HBV transactivator X protein (HBx)

enhances HBx expression and subsequent interaction with cellular genes and regulatory proteins

(Buendia, 1992; Hildt and Hofschneider, 1998; Stuver, 1998). Although HBx has not been

reported to bind double-stranded DNA, it can activate transcription of both viral and cellular

genes through interaction with a variety of host DNA binding proteins (reviewed in Andrisani

and Barnabas, 1999). HBx association with cellular transcriptional activators and general

transcription factors such as C/EBPα, TBP and TFIIH enhances gene activation. In contrast,

HBx binding to human p53 protein antagonizes p53 mediated transcriptional activation (Truant,

et al., 1995; Wang, et al., 1994), and p53-mediated apoptosis (Elmore, et al., 1997; Wang, et al.,

1995). p53 is a classical tumor suppressor with a diverse range of functions in transcriptional activation, cell cycle arrest, DNA damage repair, apoptosis (Ko and Prives, 1996; Levine, 1997) and, as recently demonstrated, transcriptional repression by sequence specific DNA binding (Lee, et al., 1999; Thornborrow and Manfredi, 1999).

While p53 mutation or inactivation is detected in over 60% of human cancers, p53 mutations are rare in the early stages of HCC (Feitelson, et al., 1993). However, p53-HBx interaction may disrupt p53 function, leading to genomic instability and accumulation of p53 mutations. Multiple studies support HBx-mediated p53 dysfunction, but the exact mechanism by which HBx inactivates p53 remains unclear. Strong support for HBx disruption of p53 function during development of HCC was demonstrated by a transgenic mouse study in which mice expressing HBx developed liver tumors with 80 to 90% penetrance within four months of birth

(Kim, et al., 1991; Ueda, et al., 1995). In these studies, HBx protein bound murine p53 protein, sequestering it in the cytoplasm. These studies suggest one mechanism by which HBx functions is to prevent p53 entry into the nucleus; more recent studies demonstrate nuclear localization of

HBx (reviewed in Murakami, 1999), where it could potentially disrupt the ability of nuclear p53 to bind DNA (Wang, et al., 1994), communicate with general transcription factors or form stable tetramers (Truant, et al., 1995).

We have shown recently that p53 represses transcription of the alpha-fetoprotein (AFP) gene by sequence specific DNA binding (Lee, et al., 1999). This repression is both tissue and developmentally specific, and contributes to the developmental regulation pattern of AFP. AFP is secreted by the visceral endoderm of the yolk sac, and is the predominant fetal serum protein

synthesized by the developing liver (Crandall, 1981; Godbout, et al., 1986). At birth, AFP expression is silenced: its mRNA level is down regulated approximately 10,000-fold. This nearly undetectable expression is maintained throughout adult life, except in cases of liver regeneration and/or HCC where AFP expression is reactivated in 75-80% of hepatocarcinomas

(Belayew and Tilghman, 1982; Camper, et al., 1989).

In these studies we have utilized regulation of AFP expression in a cell free assay system as a marker for HBx-mediated disruption of p53 function. By employing fractionated Xenopus egg extracts to assemble AFP templates into chromatin prior to in vitro transcription analysis, we have examined regulated AFP expression devoid of effects mediated by non-specific transcriptional squelching. Our results demonstrate that HBx can destroy the ability of p53 to regulate AFP in a tissue specific manner. Additionally, the data demonstrate that the mechanism by which HBx alleviates p53-mediated repression of AFP transcription is by physically interacting with p53, blocking interaction with liver specific proteins that may act in transcription repression.

Experimental Procedures

Plasmids. The AFP/lacZ template contains 3.8 kb of the AFP 5’ flanking sequence,

including the previously defined distal and proximal promoter elements and Enhancer I, fused to

the β-galactosidase coding region (Spear, 1994; Spear, et al., 1995). The AFP/DelA template

was prepared from the AFPmut5 plasmid previously described (Lee, et al., 1999) by 3 step PCR

(Nelson and Long, 1989). The 10 base pair deletion within the p53 binding site, located at –853

relative to the AFP transcriptional start site, was created through PCR amplification using the

AFP/mut5 template and the following primers:

A. 5’ CCTCCATTTTATGAGTACACTATA 3’

B. 5’ GTGTCTTAAGCGTTGCTAAGG 3’

C. 5’CGAGGGGAAAATAGGTGGTTGCGCG 3’

D. 5’ CCTTAGCAACGCTTAAGACAC 3’

A and B primers were used in the 5’ amplification step, with C and D primers used in the

3’ amplification step. Primers A and D were used for final amplification. The PCR product

containing the mutated p53 binding site was subcloned into the TA vector, pCR2.1 (Invitrogen)

and recovered by BamHI and HindIII restriction digest and gel isolation. The fragment was

subcloned into an AFP(-1.0)/lacZ template lacking Enhancer I (Spear, 1994; Spear, et al., 1995).

The –3.8 to –1 kb Enhancer I region was subsequently cloned as a BamHI fragment into this

intermediate plasmid to yield AFP/DelA.

The β-globin template contains the chick β-globin upstream promoter, structural gene and enhancer. Its construction has been previously described (Emerson, et al., 1989).

Protein expression. Recombinant histidine-tagged p53∆30 protein (Hupp, et al., 1992),

lacking 30 amino acids from the C-terminus of the protein, was prepared from the

pET23bp53∆30 plasmid as previously described for HNF-3α protein (Zaret and Stevens, 1995).

Briefly, 500 mL cultures were grown at 37° C to an A600 of 0.4. Protein production was induced

by addition of IPTG to a 1mM final concentration. Cultures were grown for an additional two

hours, then collected, pelleted and resuspended in 20 mL 1X binding buffer (5 mM Imidazole,

0.5 M NaCl, 20 mM Tris-HCl, pH 7.9). Cells were sonicated on ice for 30 seconds. The crude

protein extract was collected by centrifugation at 24,000 x g for 15 minutes at 4° C, resuspended

in 2.5 mL 1X binding buffer/6M urea and sonicated for 30 seconds on ice prior to 1 hour

incubation at 4° C. Soluble p53 was separated from insoluble debris by centrifugation at 24,000

x g for 15 minutes at 4° C. Soluble p53 was purified from the crude fraction by affinity

chromatography over a Ni2+-NTA agarose column (Qiagen). Bound protein was washed and

eluted as described (Zaret and Stevens, 1995). Purified p53 was dialyzed against dialysis buffer

(20 mM Hepes pH 6.5, 400 mM KCl, 1 mM MgCl2, 20% glycerol, 0.1% NP-40, 0.5 mM PMSF,

3 mM DTT) with decreasing urea concentrations (5M, 4M, 2M, 1M and 0M urea).

Recombinant histidine-tagged HBx was prepared from the pRSETc::X plasmid as

previously described (Haviv, et al., 1996).

Cell extracts. HepG2 whole cell extracts were prepared as previously described,

(Dignam, et al., 1983) (Crowe, et al., 1999). HeLa cell nuclear extracts were prepared exactly as described in Current Protocols in Molecular Biology (Ausubel, et al., 1987). Xenopus egg

extracts were prepared exactly as previously described (Barton and Emerson, 1996; Crowe and

Barton, 1999). High speed supernatant (HSS) soluble fractions used for chromatin assembly had

protein concentrations ranging from 50 ug/ul to 100 ug/ul.

In vitro transcription. In vitro transcription analysis of templates as chromatin-free

(naked) and chromatin-assembled DNA was performed as previously described (Crowe, et al.,

1999), with minor modifications. For naked DNA transcriptions, recombinant p53 and HBx proteins were added to 500 ng supercoiled DNA templates in transcription reaction mix and allowed to bind for five minutes at room temperature prior to addition of transcribing extract.

RNA products were purified and analyzed by primer extension.

Solid phase DNA templates for chromatin transcriptions were prepared as described

(Crowe and Barton, 1999; Crowe, et al., 1999). Briefly, AFP DNA was digested with EcoRI and

ClaI, then biotinylated with Biotin-21 dUTP and Biotin-14 dATP and Klenow fragment DNA polymerase (Gibco) prior to coupling to streptavidin coated, paramagnetic beads (Dynal). In chromatin transcription reactions, p53 and HBx were added to 500 ng solid-phase DNA templates during a 20 minute pre-incubation in HepG2 or HeLa cellular extracts prior to chromatin assembly. After 1 hour chromatin assembly in fractionated Xenopus egg extract, solid-phase DNA templates were washed three times in modified nuclear dialysis buffer (mNDB)

(20 mM Hepes, pH 7.9, 50 mM KCl, 0.2 mM EDTA, 10% glycerol, 1 mM DTT) plus 0.01% NP-

40, then transcribed in HeLa extract and analyzed as above.

p53 immunoprecipitation and western blot analysis. Immunoprecipitations were performed in HepG2 whole cell extract or HeLa nuclear extract diluted to a total protein concentration of 8 ug/ul in mNDB. Recombinant p53 (approximately 600 ng) and HBx proteins

(approximately 300 ng) were added to 25 ul of cellular extract and incubated at 4° C for 20

minutes. Anti-p53 antibody (Santa Cruz pAB 240) was added to the reaction mix and bound for

1 hour in the presence of 1% NP-40 with gentle rocking at 4° C. Immuno-complexes were

collected with Protein A/Protein G+ agarose beads (Santa Cruz) equilibrated in mNDB plus 1%

NP-40. Immuno-complexes were washed three times with IP wash buffer (100 mM Tris, 1 M

NaCl, 0.3% SDS), then resuspended in sample buffer (0.06 M Tris-HCl pH 6.8, 2% SDS, 10%

glycerol, 5% β-mercaptoethanol, 0.001% Bromophenol Blue). Additional wash buffers used

included: AC wash buffer 1 (100 mM Tris pH 8.0, 1 M NaCl); AC wash buffer 2 (100 mM Tris

pH 8.0, 100 mM NaCl, 1 mM EDTA, 1% NP-40, 0.3% SDS); AC wash buffer 3 (10 mM Tris pH

8.0, 0.1% SDS); mNDB wash buffer (mNDB + 1% NP-40). Samples were analyzed by SDS-

PAGE and Western blot. Western blots were performed as previously described (Crowe and

Hayman, 1993), with minor modifications. Membranes were probed with the appropriate

antibody (anti-p53 pAB 240, Santa Cruz, and anti-HBx 11/121/52, gift of Dr. Claus H. Schroder,

Virus-Host Interactions, German Cancer Center (Su, et al., 1998)). Binding was visualized by

ECL Western Blot Analysis System (Amersham).

p53 electromobility shift assay (EMSA). EMSA was performed using the double- stranded p53 regulatory element from the AFP distal promoter (bases -862 through -830)

5’GATCCTTAGCAAACATGTCTGGACCTCTAGAC as previously described (Lee, et al.,

1999), with protein-DNA binding carried out for 30 minutes at 30° C. Protein binding assays

contained 7 ug of HepG2 or HeLa cell extract and approximately 1 ug purified p53 protein and 1

ug purified HBx protein, except as indicated

Solid-phase DNA-protein purification. Solid-phase DNA oligomers were generated by

annealing 5’ biotinylated p53 regulatory element (5’ Bio-

GATCCTTAGCAAACATGTCTGGACCTCTAGAC) to complementary strand prior to

coupling to streptavidin coated paramagnetic beads (Dynal). Control reactions to assess protein

binding specificity were performed in parallel with -1007 (5’Bio-

GATCCAATATCCTCTTCAG). Approximately 200 ng p53 regulatory element or -1007 oligo were washed in 1x PBS/1% BSA prior to incubation with 1 ug p53 in the presence or absence of

HBx (1 ug) and/or 70 ug (total protein) HepG2 or Hela cell extract. Binding reactions proceeded for 30 minutes at 22°C. DNA bound protein complexes were collected by magnetic concentration and washed two times in 1X PBS/1% NP-40 and once in wash buffer (100 mM

NaCl, 50 mM Bis Tris-HCl, 10 mM MgCl2, 1 mM DTT, pH 6.0) prior to elution. DNA associated proteins were eluted for 10 minutes at 37° C in urea elution buffer (5M urea, 10 mM

Tris pH 8.0, 100 mM NaH2P04, 1% β-mercaptoethanol). Eluted proteins were analyzed by gel electrophoresis and silver stain or western blot.

Quantitation. Image analysis was performed by use of ImageQuant 5.0 software

(Molecular Dynamics).

Results

HBx disrupts p53-mediated repression of AFP transcription. In order to examine the

regulatory consequences of HBx transactivator expression on a hepatic-expressed, cellular gene,

we performed in vitro transcription analysis of AFP DNA templates in the presence of HBx and p53 proteins. Transcription extracts isolated from the human hepatoma cell line HepG2, which actively expresses AFP but does not carry integrated HBV DNA, were used for in vitro

transcription (Crowe, et al., 1999). We utilized a constitutively activated p53 protein harboring a

C-terminal truncation to examine the ability of p53 to regulate AFP transcription independently

of post-translational modifications within the protein C-terminus, which activate p53 for DNA

binding (Ko and Prives, 1996). Addition of recombinant p53 protein resulted in 2- and 3-fold

repression of AFP transcribed as naked DNA in HepG2 extract (Fig 1A, lanes 2 and 3 compared

to lane 1). Addition of recombinant HBx protein to p53-repressed transcription reactions

alleviated p53-mediated repression (Fig 1A, compare lanes 3 and 4). The level of AFP

transcription detected upon addition of HBx was derepressed 4-fold to a level slightly higher than

that of transcription in the absence of p53 (Fig 1A, compare lanes 1 and 4). Addition of HBx in

the absence of p53 resulted in modest activation (less than 2-fold) of AFP transcription (Fig 1A,

compare lanes 1 and 5), suggesting that the observed activation of AFP is dependent primarily

upon HBx effects on p53, rather than HBx association with liver-enriched transcriptional

activators.

Effects of p53-HBx interaction are tissue specific. AFP is expressed in the fetus by

endoderm-derived cells of the yolk sac, liver and gut (Tilghman, 1985; Tilghman and Belayew,

1982). We have previously shown by cell culture transfection studies that p53 repression of AFP

is tissue specific (Lee, et al., 1999). To determine if the observed effects of HBx on p53- regulated AFP expression were also tissue specific, we performed in vitro transcription analysis in cervical cancer derived Hela cell extract. In contrast to the observed repression of AFP transcription following addition of p53 to HepG2 reactions, p53 introduction to Hela transcription reactions resulted in modest activation of AFP expression (less than 2-fold) (Fig 1B, compare lane 1 with lanes 2 and 3). HBx addition to transcription reactions in the presence of p53 slightly augmented p53 activation to a level of 3-fold over basal (Fig 1B, lanes 4 and 5).

This is in sharp contrast to results observed in the hepatoma extract where HBx addition strongly reversed p53 effects on AFP transcription.

Addition of low concentrations of HBx protein, in the absence of p53, activated AFP expression approximately 2-fold (Fig 1B, compare lane 1 with lanes 6 and 7). This level of activation, in the absence of p53, is comparable to that observed in hepatoma extracts, and supports the hypothesis that HBx-mediated activation of AFP is due primarily to a modification of p53 regulation. Interestingly, increasing concentrations of HBx protein in the absence of p53 did not activate AFP transcription, but diminished transcription to basal levels (Fig 1B, compare lanes 1 and 8). This result could be due apparent squelching by HBx through self-oligomerization or transcription factor binding. HBx has not previously been demonstrated to squelch transcription; however, because of its ability to associate with multiple transcriptional activators and general transcription factors (GTFs) (reviewed in Andrisani and Barnabas, 1999), addition of high concentrations of HBx in the absence of p53 may promote nonfunctional HBx-protein interactions.

To determine if derepression of p53-regulated AFP transcription was due to a direct

interaction between HBx and p53 that occurs only in a hepatoma extract, we performed immunoprecipitations with anti-p53 antibody (pAB 240, Santa Cruz) in both HepG2 and Hela transcription extracts. p53 and HBx proteins incubated in HepG2 whole cell extract formed a complex, as shown by immunoprecipitation and Western blot analysis with anti-p53 (pAB 240) and anti-HBx (11/121/52) antibodies (Fig 1C, lane 3). Additionally, p53 and HBx interacted in both Hela nuclear extract and extract buffer (Figure 1C, lanes 6 and 8), demonstrating that p53 and HBx proteins form a stable complex in the absence of DNA or additional proteins present in hepatoma cell extracts. Taken together, these data demonstrate that the tissue specific effects of

HBx on p53-regulated AFP transcription are not due to an inability of the proteins to form a stable complex in a non-hepatic cell extract, but rather are likely due to tissue specificity of p53 transcriptional repression.

HBx does not disrupt p53-mediated squelching of transcription. The ability of p53 to repress in vitro transcription of AFP templates could be explained by a number of mechanisms.

Multiple regions of p53 protein interact with and bind a wide range of proteins mediating, in part, the pleiotropic functions of the tumor suppressor (Ko and Prives, 1996; Levine, 1997). p53 can interact with GTFs including TFIID subunits TBP, TAF 31 and TAF70, and TFIIH subunits p62,

XPB and XPD. The ability of p53 to squelch transcription through interactions with GTFs, particularly TPB, is well documented (Horikoshi, et al., 1995; Ragimov, et al., 1993; Seto, et al.,

1992). If p53-mediated AFP repression was due in part to p53 squelching of TBP in the hepatoma extract, the apparent derepression upon HBx addition could be due to HBx disruption of p53-TBP or p53-GTF interactions.

To determine if HBx could reverse p53-mediated transcriptional squelching, in vitro

transcription analysis was performed using chick β-globin DNA as template. β-globin DNA has no p53 binding site and is not directly regulated by p53 (data not shown). In the presence of high levels of p53 protein, apparent transcriptional repression can be attributed, most likely, to p53- mediated squelching of basal transcription factors. As demonstrated in Figure 2, addition of increasing amounts of p53 to β-globin in vitro transcription reactions resulted in greater than 5- fold squelching of transcription (compare lane 1 with lanes 2 through 4). Addition of HBx resulted in an approximately 1.5-fold reactivation of squelched transcription (Fig 2, compare lanes 5 and 6 with lane 4). Addition of HBx in the absence of p53 resulted in approximately 2- fold repression of transcription, potentially due to HBx-mediated squelching, as discussed above

(Fig 2, compare lanes 1 and 7). Because HBx reactivation of p53-squelched β-globin transcription was much lower than the reactivation of p53-repressed AFP in vitro transcription, we suspected that HBx alleviation of p53-mediated AFP repression was not due simply to reversal of p53 squelching. This inability of HBx to activate transcription from a promoter lacking a p53 binding site strongly suggested that HBx must be targeted to DNA-bound p53 in order to render its activating effects.

HBx disrupts p53 regulation of AFP chromatin transcription. In order to demonstrate conclusively that HBx-mediated derepression of AFP expression in our in vitro transcription system was not an effect on squelching, we utilized in vitro chromatin assembly of

AFP DNA templates prior to transcription analysis. Solid-phase AFP DNA templates were prepared by coupling biotinylated AFP DNA to paramagnetic beads as previously described

(Crowe and Barton, 1999). Chromatin assembly was achieved by incubating the solid-phase

DNA templates in fractionated Xenopus egg extracts (Crowe, et al., 1999). In this system DNA

templates are pre-incubated with cell extracts to allow activating and/or repressive proteins to bind DNA prior to chromatin assembly. An iso-osmotic wash is performed following assembly and prior to transcription, to remove factors that do not directly or indirectly bind the DNA templates. Washing away excess cellular factors prior to in vitro transcription prevents effects mediated by non-specific squelching.

Chromatin assembly mediates general repression of basal AFP transcription in the absence of activating factors present in hepatoma cell extracts (Fig 3, compare lanes 1 and 2)

(Crowe, et al., 1999). Upon addition of HepG2 whole cell extract, transcription of chromatin assembled AFP templates was activated 50-fold (compare lanes 2 and 3). Titration of p53 into the reaction resulted in greater than 7-fold repression of AFP transcription (lanes 4 and 5), demonstrating that p53 repression of AFP transcription is indeed a specific event, and not simply an effect of p53 squelching GTFs necessary for functional transcription complexes. Addition of

HBx during pre-incubation yielded greater than 6-fold derepression of p53-regulated AFP transcription, almost to levels achieved in the absence of p53 (compare lanes 3 and 6). The ability of HBx to alleviate p53 repression of AFP chromatin transcription, as well as a relative inability to reverse p53-mediated squelching of β-globin transcription, further supports a requirement for p53 DNA binding in order for HBx to render its effects.

p53-DNA binding is required for HBx re-activation of AFP transcription. In order to determine if HBx mediated derepression of AFP transcription is dependent upon p53 DNA binding, we performed in vitro transcription analysis with an AFP DNA template containing a mutation within the p53 binding site located at -853 (DelA), relative to the transcriptional start site (Fig 4A). The DelA template contains a 10 base-pair deletion of the 5’ p53-binding half site,

as well as sequence mutations in the 3’ half site, that prevent binding of p53 to its consensus site

within the AFP promoter (unpublished data, K. Lee). AFP DelA transcription as naked DNA in

both HepG2 and Hela cell extracts (Fig 4B) and as chromatin assembled (Fig 4C) DNA was

unaffected by addition of p53(less than 1.2-fold repression in both naked DNA and chromatin

assembled DNA in vitro transcriptions) (Fig 4B, compare lanes 1 with lanes 2-4 and lane 7 with

lanes 8-10; Fig 4C, compare lane 3 with lanes 4-6). Addition of HBx to p53-containing reactions

in either hepatoma or Hela cell extracts maintained a basal level of transcription (Fig 4B,

compare lane 4 with lanes 5 and 6; lane 10 with 11-13; Fig 4C, compare lanes 6 and 7). Addition

of HBx in the absence of p53 had negligible effects on DelA chromatin transcription (1.2-fold

repression) (Fig 4C, compare lanes 3 and 8). These data further demonstrate that HBx

derepression of p53-silenced AFP in a hepatoma background requires p53 binding to its

regulatory element within the AFP developmental repressor region.

Tissue specificity of HBx-mediated disruption of p53 function is maintained with

chromatin assembly. To further examine the apparent tissue specific effects of the p53-HBx

interaction on regulation of AFP transcription, we chromatin assembled AFP DNA templates in

the presence of Hela cell nuclear extract in the presence or absence of increasing amounts of p53

or p53 plus HBx prior to in vitro transcription.

We have previously shown that chromatin assembly confers tissue-specificity on AFP

transcription in vitro (Crowe et al, 1999). Though capable of robust transcription of chromatin- free AFP DNA templates (Figs. 1 and 4 and Crowe et al, 1999), Hela extract cannot establish efficient chromatin transcription. This is a result of activated AFP chromatin transcription being dependent upon function of liver-enriched transcriptional activators such as HNF3α (Crowe, et

al., 1999). Pre-incubation of AFP DNA templates in Hela cell nuclear extract prior to chromatin assembly does not program the gene for establishment of activated, open chromatin, and transcription is silenced (Figure 5, compare lanes 2 and 3). Addition of p53 to the Hela extract during pre-incubation resulted in up to 7-fold activation of AFP transcription (Figure 5, compare lanes 4-6 with lane 3). As was observed in Hela transcription of chromatin-free AFP DNA templates, addition of HBx in the presence of p53 augmented p53 activation of AFP to that of 8- fold over Hela alone (compare lane 3 with 6 and 7). Interestingly, addition of HBx to Hela pre- incubation in the absence of p53 resulted in 11-fold activation over Hela alone (compare lanes 3 and 8), suggesting that, in the absence of p53, HBx can associate with a non-hepatic derived

DNA binding protein in order to activate AFP expression. This possibility is currently under investigation in our laboratory.

Thus far, our data demonstrated that HBx interference with p53-mediated regulation of transcription was both hepatoma-specific and required targeting by DNA-bound p53. Therefore, we hypothesized that HBx might be interacting with DNA-bound p53 to potentially dislodge putative tissue specific co-repressors of transcription. An association with DNA-bound p53 in a non-hepatic derived cell extract, where p53 acts in the absence of putative co-repressors as an activator of AFP transcription, would then allow augmentation of activation.

HBx associates with DNA-bound p53. In order to assess the effect of HBx on p53

DNA binding and to determine if HBx can associate with DNA-bound p53, we performed a series of electromobility shift assays (EMSA) with both purified proteins and cellular extracts. In examining the ability of purified p53 to bind to its AFP regulatory element in the presence of

HBx, we found that addition of HBx resulted in both a super shift of bound probe (Fig. 6A, lanes

3-8), signaling formation of a potential p53-HBx complex, in addition to an enhancement of p53 association with the regulatory element (lanes 7 and 8). HBx did not associate with the p53 regulatory element in the absence of p53 (lane 9).

To assess the possibility of a tissue-specific effect by HBx on p53 DNA binding in liver and non-liver cell extracts, we performed EMSA with the p53 regulatory element from the AFP distal promoter with both HepG2 and Hela transcription extracts in the presence and absence of p53 and HBx (Fig 6B). Addition of both HepG2 and Hela transcription extracts to binding reactions with the p53 regulatory element resulted in the appearance of a single predominant band, labeled complex 1 (lanes 2 and 11). Addition of p53 to HepG2 and Hela binding reactions resulted in a mobility shift similar to that of p53 alone (lane 7 and compare lanes 2 with 3 and 11 with 12). Addition of HBx in the presence of p53 and extract had little to no effect on either of the shifted bands, but did appear to result in a super shift of probe into the well (SS2) (compare lanes 3 with 4 and 12 with 13), as was observed with the purified recombinant proteins (Fig 6A and 6B lane 8). These results suggest complex formation between HBx and DNA-bound p53.

Addition of p53 antibody (FL 393, Santa Cruz) in the presence of p53 and either the HepG2 or

Hela transcription extracts resulted in slight decrease of the p53 band concomitant with an increase in probe shifted into the well (compare lanes 3 with 5 and 12 with 14). Addition of p53 antibody to purified p53 protein resulted in appearance of a distinct super shift, SS1 (lane 9).

Incubation of HepG2 and Hela extracts with a regulatory element probe harboring a 10 base-pair deletion and 4 point mutations within the p53 binding site (Del A) resulted in a shift of complex

1, but did not display a p53-specific shift (data not shown).

Because it appeared that HBx-p53 interaction did not modify p53’s ability to bind DNA

in either the hepatic or non-hepatic cell extract, we wanted to determine if tissue-specific co- factors binding with p53 to its regulatory site could be affected by HBx addition. To this end, we performed solid-phase DNA-protein pull downs. Incubation of biotinylated p53 regulatory element DNA coupled to streptavidin paramagnetic beads (solid phase DNA) with purified p53 protein in the absence (Fig. 6C, lane 1) or presence of HBx (lanes 2 and 3) resulted in p53 association with the DNA template, as evidenced by silver stain (Fig 6C) and western blot (data not shown). Additionally, HBx co-purified with DNA bound p53 (lanes 2 and 3), confirming that HBx is capable of associating with DNA-bound p53. p53-DNA binding was also detected upon incubation of the solid-phase p53 binding element with HepG2 or Hela extract in the absence (lanes 4 and 6) or presence (lanes 5 and 7) of HBx. Also, as with the purified proteins,

HBx was found to co-purify with DNA bound p53 in both the HepG2 (lane 5) and Hela (lane 7) cell extracts. Interestingly, two proteins co-purifying with p53 from the HepG2 extract were greatly reduced upon addition of HBx (bands indicated by asterisk, compare lanes 4 and 5). This loss of protein binding was not observed in the Hela extract (compare lanes 6 and 7). These protein bands may represent liver-specific p53 co-repressors, and their lack of association with

DNA-bound p53 in the presence of HBx could contribute to the tissue-specific functional effects of p53-HBx interaction. These tissue-specific co-repressors are currently under investigation in our laboratory.

To serve as control, p53, HBx and extract were incubated with a non-specific oligo (-

1007) lacking any defined p53 binding element. p53 and HBx association with the -1007 oligo could not be detected by silver stain (Fig 6C, lanes 8 and 9), demonstrating that p53 and HBx association with the p53 regulatory binding element is specific, and dependent upon p53 DNA

binding.

Discussion

Regulation of AFP gene expression is a complex process mediated, in part, by a number of transcriptional activators and repressors including HNF-1, retinoic acid receptor, C/EBP and other factors binding to sites within the AFP gene (Feuerman, et al., 1989; Wen, et al., 1993;

Wen, et al., 1991; Wen and Locker, 1994). Tissue and developmental specific regulation of AFP expression is controlled in part by a developmental repressor region located between -1000 and -

200 in the AFP promoter (Camper and Tilghman, 1989; Camper and Tilghman, 1991; Jin, et al.,

1998; Ramesh, et al., 1995; Vacher and Tilghman, 1990). We have previously characterized an overlapping p53/HNF-3 binding site within the developmental repressor region that mediates opposing regulatory signals. HNF-3 is a potent activator of AFP transcription, while p53 acts to repress AFP transcription (Crowe, et al., 1999; Lee, et al., 1999).

Aberrant activation of AFP is a hallmark of hepatocellular carcinoma, and exemplifies a loss in regulated gene expression that is common to numerous cancers. The exact mechanism by which AFP is reactivated in the diseased liver is unknown, but likely is due to transcriptional activator and/or repressor dysfunction. One such transcription factor, p53, which is mutated or modified in over 60% of human cancers (Ko and Prives, 1996; Levine, et al., 1991), is a target of the Hepatitis B Virus encoded X protein. Using cell-free transcription systems, we show that

HBx destroys p53-mediated regulation of AFP expression allowing transcriptional derepression of the tumor marker gene. Evidence of HBx transactivation of viral and cellular genes as a result of its binding to cellular transcriptional activators has been well documented (reviewed in

Andrisani and Barnabas, 1999). However, to our knowledge, this is the first account of HBx binding specifically to a transcriptional repressor to derepress a silenced, cellular gene.

p53 association with viral proteins is widespread: p53 forms complexes with the E6 protein from human papillomavirus, the large T antigen of SV40, the BZLF1 protein of Epstein-

Barr virus, the E1b protein of adenovirus, and co-localizes with the ICP8 protein of Herpes

Simplex Virus-1 (Lane and Crawford, 1979; Lechner, et al., 1992; Wilcock and Lane, 1991; Yew and Berk, 1992; Zhang, et al., 1994). In most cases, binding of p53 to viral proteins or DNA has a repressive effect on viral replication. As a result, viral protein-mediated antagonism of p53 function has evolved to overcome the general repressive effects of p53. For example, HIV-1 Tat protein disrupts p53-mediated repression of HIV-1 LTR promoter activity and HIV replication

(Duan, et al., 1994); human papillomavirus E6 protein interaction targets p53 for degradation

(Seavey, et al., 1999); and, as demonstrated here and elsewhere, HBx association with p53 can inhibit p53 function and/or lead to cytoplasmic sequestration of the p53 protein (reviewed in

Andrisani and Barnabas, 1999; Feitelson and Duan, 1997).

HBx binding to p53 likely evolved as a consequence of p53-mediated disruption of HBV replication in hepatic cells (Lee, et al., 1995; Lee and Yun, 1998; Ori, et al., 1998). In conjunction with a complex of liver specific co-factors, p53 can bind and repress activation from

HBV Enhancer II, the enhancer responsible for the tissue specific replication of the virus. The p53-containing complex acts on HBV Enhancer II to block transcription from promoters under its control, resulting in loss of active HBV replication. HBx expression overcomes this hindrance by HBx binding to p53, decreasing the negative effects of the p53-containing complex on Enhancer II (Lee, et al., 1998). The exact mechanism by which HBx overcomes p53- mediated repression of HBV replication has not been established, but could be due to HBx disruption of p53 association with liver-specific cofactors, as we believe to be the case with

derepression of AFP transcription.

An understanding of how HBV modifies the expression pattern of AFP is significant in that changes in AFP expression can be detected before onset of HCC in HBV-infected individuals. A study examining AFP expression in hepatocytes and oval cells in HBV-infected patients demonstrated high levels of AFP reactivation prior to the onset of HCC (Hsia, et al.,

1992). Recent evidence also indicates that AFP secretion stimulates human hepatoma tumor cell growth and proliferation (Wang and Xie, 1999). Because the majority of cells losing the ability to silence AFP have a disease phenotype, it is likely that the mechanism by which AFP is reactivated is one way that HBV contributes to the development and progression of HCC. Our demonstration that HBx can derepress AFP through its interaction with p53 lends additional support to the hypothesis that HBx is the primary factor by which HBV contributes to the development of HCC, and that its interaction with p53 is an integral step in the process.

In conclusion, we have shown that HBx overcomes p53-mediated repression of a liver- specific, tumor marker gene by disrupting DNA-bound p53 interaction with potential liver- specific co-repressors. Modification of p53 interaction or communication with protein partners by HBx may be a global mechanism affecting the ability of p53 to regulate multiple genes, potentially contributing to development of HCC in infected individuals. Future studies will include examination of p53 protein partners and additional liver-specific transcriptional repressors and activators of AFP that may contribute to development of HBV-associated HCC.

Fig 1 p53 - + ++ ++ - HBX --- ++ MW 1 2 3 4 5 A

AFP

B p53 - + ++ ++ ++ - - - HBX ---+++++++ +++ 12345678 MW

AFP

C Hela------+ + HepG2 + + + + - - -- p53 - + + + + + ++ HBX - -- + + - + -+ anti-p53 + + + - + + ++ 12 3 4 5 6 7 8 anti-p53 probe anti-HBx probe

Figure 1: HBx and p53 interaction results in loss of p53-mediated repression of AFP transcription.

(A) p53 mediated AFP transcriptional repression is lost upon addition of HBx. AFP templates were in vitro transcribed in HepG2 whole cell extract (25 ul) in the presence of increasing amounts of p53 (600 ng, lane 2; 1.4 ug, lanes 3 and 4) and/or HBx (450 ng, lanes 4 and 5). Transcripts were detected by primer extension. Radiolabeled φX 174 DNA digested with

Hae III (Gibco/BRL) was used as a molecular weight marker (MW). The 84 base pair AFP primer extension product is indicated by an arrow. (B) Effects of p53 and HBx on AFP transcription are tissue specific. AFP templates were in vitro transcribed in Hela nuclear extract (15 ul) in the presence of increasing amounts of p53 (700 ng, lane 2; 1.8 ug, lanes 3, 4 and 5) and/or HBx (250 ng, lanes 4 and 6; 1 ug, lanes 5 and 8; and 625 ng, lane 7). Transcripts were detected by primer extension. Radiolabeled φX 174 DNA digested with Hae III

(Gibco/BRL) was used as a molecular weight marker (MW). AFP primer extension products are indicated. (C) p53 and HBx interact in HepG2 and Hela transcription extracts.

Recombinant p53 ((600 ng), lanes 2-8) and HBx ((300 ng), lanes 3, 4, 6 and 8) were added to 25 ul of HepG2 whole cell extract (approximately 200 ug total protein, lanes 1-4), buffer (lanes 5 and 6) or Hela extract (approximately 200 ug total protein, lanes 7 and 8). p53 complexes were immunoprecipitated with anti-p53 (Santa Cruz pAB 240), subjected to SDS-PAGE, and immuno-blotted with anti-p53 and anti-HBx, as indicated.

p53 - + ++ +++ ++++++ - HBX ----+++++ 1 2 34567MW

β-globin

Figure 2: p53-mediated squelching of β-globin transcription is not alleviated by HBx.

β-globin DNA templates were in vitro transcribed in 15 ul of HepG2 whole cell extract (approx.

100 ug total protein) with increasing amounts of p53 protein (450 ng, lane 2; 1.1 ug, lane 3; 2.25 ug, lanes 4-6) in the absence or presence of HBx (500 ng, lane 5; 1.0 ug, lanes 6 and 7).

Transcripts were detected by primer extension. β-globin primer extension products are indicated.

p53 - - - + ++ ++ HBX - - - - - + HepG2 - - + + + + chromatin - + + + + + 123456MW

AFP

extract or buffer ± p53 ± HBx wash

pre-inc. chromatin assembly bead-DNA transcription 20 min 1 hr HSS

Figure 3: HBx alleviates p53 mediated repression of chromatin assembled AFP DNA.

Immobilized AFP templates were incubated with nuclear extract buffer (lanes 1 and 2) or HepG2 whole cell extract ( 20 ul: approx. 200 ug total protein; lanes 3-6) prior to 1h. chromatin assembly in fractionated Xenopus egg extract. Reactions were supplemented with recombinant p53 protein (370 ng, lane 4; 1.8 ug, lane 5) or recombinant p53 (1.8 ug) plus recombinant HBx

(470 ng, lane 6). Chromatin-assembled templates were washed in nuclear extract buffer and in vitro transcribed in Hela nuclear extract. AFP primer extension products are indicated.

A Developmental Enhancer I Repressor Region lacZ

-2.5 kb -1.0 kb -.2 kb

p53 p53 GCCTTAGCAAACATGTCTGGACCTCTAGACA -860 -830 AFP CGGAATCGTTTGTACAGACCTGGAGATCTGT p53 p53 GCCTTAGCAAAC ATGTCTGGACCTCTAG AC A CGGAATCGTT GCGAATTCTGT DelA

B HepG2 HeLa - + ++ +++ +++ ++ p53 - ++++++++++++ +++ - --- -++HBx ----+++++++++ 123 4 5 6MW 7981011121314

DelA

p53 - - - + ++ +++ +++ - HBX ------+ + HepG2 - - + + + + + + chromatin - + + + + + + + 123 4 5 6 78 MW

DelA

Figure 4: p53 repression and HBx re-activation of AFP transcription are dependent upon p53-DNA binding.

(A) AFP-DelA template contains a 10 base-pair deletion within the p53 binding site. AFP and DelA templates are diagramed to show the modifications existing within the DelA template.

DelA contains a 10 base pair deletion within the consensus p53 binding site, and four single base point mutations within the p53 half site, as indicated. (B) Effects of p53-HBx interaction on

AFP transcription require p53-DNA binding. DelA DNA templates were in vitro transcribed in HepG2 whole cell extract (approximately 100 ug total protein, lanes 1-6) or Hela nuclear extract (approximately 100 ug total protein, lanes 7-14) in the presence of increasing amounts of p53 (560 ng, lanes 2 and 8; 1.1 ug, lanes 3, 6 and 9; 2.25 ug, lanes 4, 5 and 10-13) with and

without recombinant HBx (400 ng, lanes 5, 6 and 11; 600 ng, lane 12; 1 ug, lanes 13 and 14 ).

DelA primer extension products are indicated. (C) p53-HBx binding has no effect on chromatin transcription of AFP-DelA. Immobilized DelA templates were incubated with extract buffer (lane 2) or HepG2 whole cell extract (approximately 200 ug total protein) in the presence of increasing amounts of p53 (560 ng, lane 4; 1.1 ug, lane 5; 2.25 ug, lanes 6 and 7) and/or recombinant HBx (400 ng, lanes 7 and 8) prior to 1 hour chromatin assembly in fractionated Xenopus egg extract. Chromatin-assembled templates were washed in nuclear extract buffer and in vitro transcribed in Hela nuclear extract. DelA primer extension products are indicated.

p53 - - - + ++ +++ +++ - HBX ------+ + Hela - - + + + + + + chromatin - + + + + + + + 123456 78MW

AFP

Figure 5: Tissue specificity of p53-HBx effect on AFP transcription is maintained in chromatin.

Immobilized AFP DNA templates were incubated in nuclear extract buffer (lane 2) or Hela nuclear extract (100 ug total protein, lanes 3-8) in the absence (lane 3) or presence of p53 (370 ng, lane 4; 900 ng, lane 5; 1.8 ug, lanes 6 and 7) and HBx (approx. 1 ug, lanes 7 and 8) prior to 1 hour chromatin assembly in fractionated Xenopus egg extract. Chromatin-assembled templates were washed in nuclear extract buffer and in vitro transcribed in Hela nuclear extract. AFP primer extension products are indicated.

n n

i i

e e

t t

o o

r r

p p

A HBx

o o

HB N N p53 123456789

p53

Free probe

3 3

5 5

x x

p p

- -

n B n B

i i

α α α α α α α α

α α α α α α

H H

e B --p53 e

t t

3 α α + H α α + α α α α

+ +

o o

r r

3 3 3 + + 3

3 3 3

p p

--p53 B

5 5 5 5

5 5 5

3 3

p p α α p α α p H α α α α

o p o p p

5 5

+ + + p53 No protein N + + p p N + +

+ p53 + p53 HBx + p + + HepG2 HeLa 12345 6 78 9 10 11 12 13 14 15 SS2 SS1 p53 Complex 1

Free probe

x x

B B

x x

H H

B B

+ +

H H

3 3 3 3 3

H 2 + +

5 5 5 5 5

p p p p p 3 3

+ + a

5 5

p L

+ 3 3 + + + +

p p

e e

5 5

pp53 pp53 + HBx p HepG2 HeLa H + H + 12 3456 7 89

* *

p53

HBx p53 Reg p53 Reg -1007

Figure 6: HBx associates with DNA-bound p53. (A) p53-DNA binding is maintained upon

HBx association.

p53 (approx. 500 ng) was incubated with radio-labeled double stranded p53 regulatory element

probe in the absence (lane 2) or presence of increasing amounts of HBx (approx. 200 ng, lane 3;

400 ng, lane 4; 600 ng, lane 5; 1 ug, lane 6; 1.6 ug, lane 7 and 2 ug, lane 8). Lane 9 contains

HBx (approx. 2 ug) and labeled p53 regulatory element probe. Reactions were incubated for 30

minutes at 30° C. Shifted DNA-protein complexes are indicated. (B) HBx associates with

DNA-bound p53 in both HepG2 and Hela cell extracts. Labeled p53 regulatory element probe

was incubated with purified p53 protein (approx. 1 ug, lanes 7-9) or p53 (approx. 1 ug) plus

HepG2 (approx. 7 ug, lanes 2-6) or Hela (approx 7 ug, lanes 11-15) extract in the absence or presence of HBx (approx 1 ug, lanes 4, 8, 13 and 15). p53 antibody (FL 393, Santa Cruz) was included in extract or purified protein binding reactions as indicated (lanes 6, 9, 14). Binding reactions were incubated for 30 minutes at 30° C. Shifted complexes are indicated. (C) HBx co- purifies with DNA-bound p53. Immobilized p53 regulatory element or –1007 DNA templates were incubated with p53 (approx 1.5 ug, lanes 1-9) in the absence or presence of HBx (approx 1 ug, lanes 2, 5, 7-9; 2 ug, lanes 2) and/or cell extract (HepG2 extract, approx. 70 ug, lanes 4, 5 and

8; Hela extract, approx. 70 ug, lanes 6, 7 and 9). Binding reactions were allowed to proceed for

30 minutes at 22° C. Complexes were eluted and analyzed by SDS-PAGE and silver stain. p53,

HBx and putative co-factors are indicated (asterisk denotes putative p53 co-repressor).

Chapter 4: A p53, Smad4 and SnoN-containing repressor complex is disrupted by the virally encoded HBx protein.

Summary

Chronic infection with Hepatitis B Virus correlates with development of hepatocellular carcinoma (HCC). Here, we have examined the ability of the virally encoded HBx protein to reverse transcriptional silencing of the AFP tumor marker gene through an interaction with

DNA-bound p53 protein. We find that HBx association at the p53 regulatory element in the AFP promoter displaces the SnoN co-repressor from an over-lapping Smad4 binding element.

Additionally, we find that SnoN is a transcriptional repressor of the AFP gene, and that its ability to associate with chromatin assembled AFP templates is stabilized by p53 association at the regulatory element. The ability of p53 to repress transcription of AFP chromatin templates appears to be dependent on SnoN, as p53 can not facilitate transcription repression of chromatin templates assembled in the presence of SnoN immuno-depleted extracts. Taken together, our results suggest a role for the p53/Smad4/SnoN containing repressor complex in the proper post- natal silencing of the AFP gene. Further, we suggest that disruption of AFP silencing during chronic HBV infection is the result of an HBx-mediated disruption of this repressor complex.

Introduction

Modifications in established gene expression patterns and signal transduction pathways

are common during chronic viral infection. Chronic infection with Hepatitis B virus (HBV) can

lead to activation of a wide range of cellular genes including Interleukin-8, EGF receptor, TGF-

β, Interleukin-6, c-fos, c-jun, c-myc, TATA binding protein, α-fetoprotein and many others

(reviewed in Murakami, 1999). Additionally, HBV can trigger modification of multiple cellular

signal transduction pathways including the MAP-kinase, NF-κB, JAK-STAT and TGF-β signaling pathways (reviewed in Arbuthnot and Kew, 2001). We have previously demonstrated that the HBV-encoded transcription transactivator HBx can reverse transcriptional repression of the α-fetoprotein (AFP) gene through a direct interaction with the p53 tumor-suppressor protein

(Ogden, et al., 2000). Here, we further analyze HBx-mediated re-activation of AFP, and find that

HBx disrupts a DNA-mediated repressor complex consisting of p53 and downstream TGF-β

effectors. The α-fetoprotein gene exhibits a tightly controlled, developmental- and tissue-

specific expression pattern of strong activation beginning at embryonic day 7.5 and continuing

throughout gestation, followed by silencing shortly after birth. Transgenic mouse studies

revealed post-natal repression of AFP to be controlled by a portion of the AFP proximal promoter, located between -200 and -1000 with regards to the transcription start site (Camper, et al., 1989). This region, dubbed the “developmental repressor region”, contains consensus binding sites for a hand-full of transcription factors including CCAT Enhancer Binding Protein

(C/EBP), an overlapping Ets-2/Glucocortocoid receptor binding site, and an overlapping Hepatic

Nuclear Factor-3 (HNF-3)/p53/Smad4 binding element. We have previously demonstrated that silencing of AFP is achieved, in part, through direct DNA binding of the p53 tumor suppressor

protein to its binding element located within the developmental repressor region (Lee, et al.,

1999). Association of p53 with this DNA element initiates a series of events involved in AFP repression; these include displacement of liver-specific activator HNF-3 from its overlapping

DNA binding site (Lee, et al., 1999), and re-organization of chromatin structure to create a region of generalized inaccessibility around the transcription start site (Ogden, et al., 2001). Repression of AFP by p53 is tissue-specific, and likely involves the function of liver-enriched co-repressors

(Lee, et al., 1999; Ogden, et al., 2000).

During development, AFP is secreted by the visceral endoderm of the yolk sac, and is the predominant fetal serum protein produced by the developing liver. Its exact function is not entirely clear, but AFP has been suggested to act as a carrier protein, to be involved in fetal immuno-protection and, most recently, to act as a hepatocyte growth factor (Crandall, 1981;

Godbout, et al., 1986). Re-activation of developmentally silenced AFP is linked to re-entry of hepatocytes into active cell cycle during liver regeneration, liver disease and liver cancer. It is unclear how AFP is re-activated during these processes, but modification of regulatory protein function is a likely possibility. One such regulatory protein, p53, is mutated in over 80% of all human carcinomas, and is commonly functionally inactivated during chronic HBV infection.

Chronic infection with HBV is tightly linked to development of hepatocellular carcinoma (HCC).

While p53 mutations are not common in the early stages of liver cancer, functional inactivation of wild-type p53 protein may lead to increased genomic instability and accumulation of mutations. Functional inactivation of p53 during chronic HBV infection is achieved through an interaction with HBx that can trigger a number of changes in hepatocyte physiology. These include modifications in p53 target gene regulation and p53-mediated modulation of apoptosis

(Andrisani and Barnabas, 1999). Re-activation of AFP, potentially through the p53-HBx

interaction, is observed during chronic HBV infection prior to onset of HCC, and may be

involved in HCC development (Hsia, et al., 1992).

Recent evidence demonstrates that HBx is also involved in activation of TGF-β target

genes through an interaction with the Smad4 DNA binding protein. Normally, TGF-β stimulation results in dimerization of TGF-β receptors leading to phosphorylation of receptor activated Smads (R-Smads). Once activated, the R-Smads form heterodimers with the common

DNA binding Smad, Smad4. Association of Smad4 with R-Smads allows for nuclear translocation of the Smad complex and subsequent activation of TGF-β responsive genes

(reviewed in Attisano, 2000; Massague, 1998; Miyazono, 2000). HBx interaction with Smad4 bypasses the need for receptor activation and R-Smad-Smad4 interaction in order to activate target genes (Lee, et al., 2001).

TGF-β-induced changes in hepatocyte gene regulation and cellular proliferation are not well defined. While TGF-β signaling in hepatocytes typically results in decreased DNA replication and G1 arrest, up-regulation of TGFβ expression is observed during liver regeneration, cirrhosis and chronic viral infection. TGF-β expression/exposure in hepatocytes results in liver fibrosis, potentially contributing to cirrhosis and development of HCC (reviewed in Fausto and Webber, 1993).

Our previous studies demonstrated that regulation of AFP by p53 is a tissue specific event, and led us to examine whether liver-enriched p53 co-repressors were involved in AFP gene silencing. Here, we describe our findings that in addition to disruption of HNF-3 binding and organization of AFP promoter chromatin structure, p53 is also involved in stabilization of a Smad4

and SnoN containing transcription repressor complex. Further, we find that HBx disrupts association of SnoN with AFP chromatin templates, and suggest that the mechanism by which the virally-encoded protein re-activates AFP is through disruption of a DNA-mediated repressor complex.

Experimental Procedures

Plasmids and solid-phase DNA templates – AFP/lacZ contains 3.8- kilobase pair upstream

DNA encompassing the proximal and distal promoters and enhancer I, fused to the coding region

of β-galactosidase (Spear, et al., 1995). DelA/lacZ is identical with the exception of a 10-base

pair deletion in the -850 to -840 p53 half site, and four point mutations in the -840 to -830 half

site. Its construction was previously described (Ogden, et al., 2000). For in vitro chromatin

transcription reactions, AFP/lacZ was coupled to streptavidin coated paramagnetic beads as

described previously (Crowe and Barton, 1999; Crowe and Barton, 1999). The p53KH

expression plasmid (LTR-KH) has been previously described (Hinds, et al., 1989). SnoN and

Ski expression vectors were the kind gift of E. Stavnezer.

Protein expression and cellular extracts – Cellular extracts were prepared from HeLa, HepG2

(AFP-positive, ATCC catalog number HB-8065) and adult mouse liver as previously described

(Dignam, et al., 1983; Gorski, et al., 1986) with minor modifications (Lee, et al., 1999). Cellular

extracts had total protein concentrations between 6 -15 mg/mL. Xenopus egg extract high speed

supernatant (HSS) was prepared exactly as described previously (Barton and Emerson, 1996;

Crowe and Barton, 1999; Crowe and Barton, 1999). HSS soluble fractions used for chromatin

assembly contained total protein concentrations ranging from 50-100 mg/mL.

Recombinant histidine-tagged p53FL and ∆30 (Hupp, et al., 1992) proteins were prepared exactly as previously described for soluble p53 protein (Ogden, et al., 2000). Recombinant- histidine tagged HBx protein was prepared as described previously (Haviv, et al., 1996). Casein kinase II (NEB)-mediated in vitro phosphorylation of full-length p53 protein to activate for DNA binding was carried out for thirty minutes at 30°C. Following phosphorylation reactions, p53

phospho-protein was dialyzed into p53 buffer (20 mM Tris-HCl, pH 8.0, 0.5 mM EDTA, pH 8.0,

50 mM KCl, 20% glycerol, 1.0 mM phenylmethylosulfonyl fluoride, 0.5 mM dithiothreitol) for 2

hours at 4° C.

Solid-phase protein purification – Solid-phase DNA oligomers were generated by annealing

5’-biotinylated p53 regulatory element (5’- Bio

GATCCTTAGCAAACATGTCTGGACCTCTAGAC), p53DelA (5’- Bio

GCCTTAGCAACcgCTtaACA), or -1007 (5’- Bio GATCCAATATCCTCTTGAC) (Life

technologies, Inc.) to complementary strand prior to coupling to streptavidin-coated

paramagnetic beads (Dynal). Extracts were pre-cleared with ~ 150 ng non-specific bead-DNA

for 20 minutes at 4° C prior to 30-minute incubation with specific bead-DNA at 25° C.

Approximately 200 ng solid-phase oligomeric bead DNA was washed one time in 1X PBS/1%

NP-40 and equilibrated in extract buffer prior to addition to binding reactions. Following

binding, bead-DNA/protein complexes were concentrated, washed twice in 1X PBS/1% NP-40

and eluted at 37° C for 10 minutes in elution buffer (5 M urea, 10 mM Tris, pH 8.0, 100 mM

NaH2PO4, 1% β-mercaptoethanol). Eluted proteins were separated by denaturing gel

electrophoresis and analyzed by silver stain or western blot with the following antibodies:

p53(pAB 240), mSin3A(K-20), Smad4(B8) (Santa Cruz); SnoN (Upstate Biotechnology); HBx

(ABR). Bound proteins were visualized using ECL Plus Western Blot Analysis System

(Amersham).

MALDI analysis – Solid-phase protein purification was preformed as described above with minor modifications. Prior to incubation in adult mouse liver (ML) or HeLa nuclear extracts, solid-phase p53 regulatory element double stranded oligomers were blocked in mNDB + 1%

BSA. Blocked, solid-phase oligomers were washed twice in reaction buffer, pre-bound with p53 for 30 minutes at room temperature, then incubated in nuclear extract for 30 minutes at room temperature. Proteins co-purifying with recombinant p53 protein on solid-phase DNA templates were digested in gel as follows: Protein-containing gel bands were diced and incubated in 25 mM NH4HCO3/50% acetonitrile for 10 minutes at room temperature. Supernatant was discarded and gel fragments were washed an additional two times with 25 mM NH4HCO3/50% acetonitrile.

Following washes, gel pieces were dried, weighed and vortexed for 10 minutes in 3 volumes trypsin solution (12.5 ng/uL trypsin in 25 mM NH4HCO3). The mixture was then incubated at 4°

C for 30 minutes. Gel pieces were resuspended in 25 mM NH4HCO3 and incubated over-night at

37° C.

Peptides were extracted from gel fragments by sonication for 5 minutes. The aqueous fraction was removed to a fresh tube containing 5 uL 50% acetonitrile/5% formic acid. Gel fragments were vortexed for 10 minutes in 50% acetonitrile/5% formic acid, then sonciated for an additional 5 minutes. Supernatant was collected and pooled with the initial aqueous fraction.

The pooled, aqueous peptide solutions were concentrated and resuspended in fresh 50% acetonitrile/5% formic acid. Peptides were analyzed using a Voyager DE-PRO MALDI-TOF mass spectrometer (Applied Biosystems) in reflector mode. The MALDI spectra were then analyzed using Protein Prospector MS-Fit algorithm against the NCBInr database

(http://prospector.ucsf.edu). p53 immunoprecipitation – Immunoprecipitations were preformed from whole cell HepG2 extract or adult mouse liver nuclear extract as described previously (Ogden, et al., 2000) with minor modifications. Recombinant p53 protein (approximately 500 ng) and HBx protein

(approximately 250 ng) were added to cell extract and pre-cleared with Protein A/G agarose

beads (Pierce) for 30 minutes at 4° C. Agarose beads were collected by centrifugation at 500 x g

for 1 minute at 4° C. Pre-cleared extracts were incubated with anti-p53 antibody (pAB 240,

Santa Cruz) beads or anti-Smad4 (B8, Santa Cruz) beads for 1 hour with gentle rocking at 4° C.

Following binding reactions, antibody beads and associated proteins were collected by centrifugation at 500 x g for 2 minutes at 4° C and washed three times for 1-3 minutes with 1X

PBS/1% NP-40. Protein-antibody complexes were re-suspended in sample buffer (0.06 M Tris-

HCl, pH 6.8, 2% SDS, 10% glycerol, 5% β-mercaptoethanol, 0.001% bromphenol blue), separated by denaturing gel electrophoresis, and analyzed by western blot analysis.

Smad4 and p53 antibody beads were generated by binding antibodies to protein A/G agarose beads (Pierce) at room temperature for 1 hour with gentle agitation. Following binding, antibodies were cross-linked using the Seize X Protein G Immunoprecipitation kit (Pierce, catalogue number 45210) per manufactures instructions.

Transfections and reporter assays – Hepa1-6 murine liver carcinoma cells were transfected using Lipofectamine (Life Technologies, Inc.) per manufacturers instructions. Cells were plated at a density of 3.0 X 105 cells/60-mm plate 16-18 hours prior to transfection. For each plate, a total of 1.5-2 ug of DNA were transfected in serum-free Opti-MEM (Life Technologies, Inc) media. Cells were incubated with DNA for approximately 24 hours, fed with fresh media for an additional 24 hours, and then harvested for extract preparation. Cell lysates were prepared and β-

galactosidase and luciferase activity were analyzed using Sigma luciferase kit (Sigma, Luc-1) per

manufacturers instructions. Luciferase activity was analyzed using an analytical luminescence

laboratory luminometer. The luciferase reporter construct, pGL2 (Promega) was used as an

internal control for transfection efficiency.

TGF-β treatment – Hepa1-6 cells were plated at a density of 3.0 X 105 cells/60-mm plate and

grown for 18-20 hours (~ 90% confluence) prior to treatment with TGF-β1 ligand. Cells were

treated with TGF-β1 ligand (Research Diagnostics, Inc.) (4 ng/mL media) in Opti-MEM (Life

Technologies) serum-free media or vehicle (0.4 mM HCl, 1% BSA) for 4 to 72 hours. At 36

hours growth media was replaced with fresh media containing TGF-β1 ligand or vehicle.

Conditioned media was collected at 4, 12, 24, 36, 48, 60 and 72 hours and analyzed for the

presence of secreted AFP by gel electrophoresis and western blot analysis (anti-AFP, Santa

Cruz). Total secreted protein was visualized by silver stain.

Chromatin Immunoprecipitation – Chromatin immunoprecipitation (ChIP) assays were

performed on in vitro chromatin assembled DNA templates exactly as previously described

(Ogden, et al., 2001). Southern blots were preformed using 32P-end-labeled, single-stranded

oligomers encompassing AFP/lacZ sequences from -860 to -830 (p53 RE probe) and +752 to

+782 (5’ CTCGGCGTTTCATCTGTGGTG) vector probe. Results were quantified using

ImageQuant analysis of phosphor images (Molecular Dynamics). Values are expressed as

percent bound normalized to antibody control.

HepG2 extract immuno-depletion and chromatin transcription – HepG2 whole cell extract

was immuno-depleted using anti-SnoN (Upstate) cross-linked to Protein A/G agarose beads

(Pierce, as described above). Extract was incubated with antibody beads or control beads (mock)

for 1 hour at 4° C. Antibody bead protein complexes were collected by centrifugation at 500 x g for 2 minutes at 4° C. Following depletion, control, immuno-depleted and mock-depleted extracts were examined along with SnoN-bound antibody beads and control (mock) beads by

denaturing gel electrophoresis and Western blot analysis.

Solid-phase AFP DNA templates were incubated with control, SnoN immuno-depleted or

mock-depleted HepG2 extract in the presence or absence of recombinant p53 protein for 20

minutes prior to the addition of fractionated Xenopus egg extract high speed supernatant (HSS).

Following1 hour chromatin assembly in HSS, DNA templates were washed three times in modified nuclear dialysis buffer (mNDB) (20 mM Hepes, pH 7.9, 50 mM KCl, 0.2 mM EDTA,

10% glycerol, 1 mM dithiothreitol) plus 0.01% NP-40, then transcribed in HeLa extract. RNA products were purified and analyzed by primer extension.

Results

SnoN co-purifies with DNA-bound p53. We have previously demonstrated that HBx

association with DNA-bound p53 disrupts the binding of two liver-enriched factors of

approximately 80- and 65-kilodaltons (kDa) from the AFP p53-DNA binding element (Ogden, et

al., 2000). Because p53-mediated repression and HBx re-activation of AFP transcription are tissue-specific events, we wanted to determine if the two factors were liver-enriched p53 co- repressors. In attempts to identify these factors, we pre-bound recombinant p53 protein to p53 regulatory element oligomers coupled to paramagnetic streptavidin beads (p53RE beads), and incubated the protein-bead DNA complexes in adult mouse liver or HeLa nuclear extract (Fig

1A). Proteins associating with p53 RE beads were eluted, separated by denaturing gel electrophoresis and visualized by silver staining. Analysis of proteins co-purifying with p53 from adult mouse liver and HeLa nuclear extracts revealed an apparent liver-enriched, p53- stabilized protein of approximately 80 kDa associating with the p53RE DNA (Fig 1B and 1C, denoted by ∗, compare lanes 5 and 9, Fig 1B and lanes 4 and 10, Fig 1C). The 80 kDa protein and p53 bands were excised from silver stained SDS-PAGE gels, trypsinized and subjected to matrix-assisted laser desorption/ionization time-of-flight mass spectrophotometry (MALDI) analysis. Spectra were analyzed using Protein Prospector MS-Fit algorithm against the NCBInr database, which revealed the likely identity of the 80-kDa protein to be the transcriptional co- repressor SnoN.

SnoN is a negative regulator of TGF-β family signaling pathways, and is responsible for transcriptional repression of TGF-β target genes. SnoN does not bind DNA, but does associate at Smad binding elements (SBE) through an interaction with DNA-bound Smad proteins

(reviewed in Miyazono, 2000; Wotton, et al., 1999). In the absence of TGF-β ligand stimulation,

SnoN associates at SBEs via an association with Smad2/Smad4 and/or Smad3/Smad4

heterodimers to hold target genes in a repressed state. TGF-β stimulation triggers degradation of nuclear SnoN, allowing for conversion of the SBE from a repressor binding element to an activator binding element (Stroschein, et al., 2001; Stroschein, et al., 1999; Sun, et al., 1999).

In examining the p53RE, we found high affinity Smad3 and Smad4 binding sites within the over-lapping HNF-3/p53 binding element (Fig 2A). To test for binding of endogenous Smad proteins to the p53 regulatory element, we repeated bead-DNA protein purification experiments using adult mouse liver nuclear and HepG2 whole cell extracts (Fig 1A). Western blot analysis of proteins bound to the p53RE revealed association of endogenous Smad4 in both the adult mouse liver nuclear (ML) and HepG2 whole cell extracts (Fig 2B). We did not detect association of endogenous Smad3 protein with the p53RE in either the adult mouse liver or HepG2 extract, but have not ruled out its binding (data not shown). Addition of recombinant p53 protein to adult mouse liver extract during the incubation period modestly stabilized Smad4 association (Fig 2B, compare lanes 1 and 2, upper panel, and lanes 1 and 6, lower panel), while recombinant HBx protein appeared to destabilize Smad4 association (Fig 2B, compare lanes 2 with 3-5, upper panel and lane 1 with 2-5, lower panel, Fig 2C, compare lanes 2 and 3). Because Smad proteins may exhibit a moderately relaxed DNA binding specificity (Zawel, et al., 1998), we wanted to determine if Smad4 association with the p53RE was specific. To this end we coupled a double- stranded AFP oligomer from 150 basepairs upstream of the p53 binding element (-1007) to paramagnetic beads and performed similar binding assays. Binding of Smad4 to the p53RE was specific, as the protein did not associate with control DNA (Fig 2B, lanes 7-9, lower panel).

Alteration of Smad4-DNA binding in response to recombinant p53 and HBx addition in whole cell HepG2 extract was not as evident as in the adult mouse liver extract (Fig 2B, compare lane 6 with 7 and 8). This is likely due to the use of a whole cell extract, rather than nuclear extract. Smad4 activation and nuclear translocation are mediated through phosphorylation by receptor activated Smads. Presumably, Smad4 present in adult mouse liver nuclear extract is activated and present at lower concentrations than in HepG2, thereby allowing for more sensitive detection of p53-induced binding modifications. In contrast, both activated and inactivated

Smad4 is present and able to associate with exogenous DNA incubated in the HepG2 extract.

Western blot analysis of proteins co-purifying from HepG2 whole cell extract confirmed our earlier MALDI data that SnoN was binding at the p53 regulatory element (Fig 2C, lanes 1-3).

This association did not appear to be modified in response to recombinant p53 (Fig 2C, compare lanes 1 and 2) or HBx (compare lanes 2 with 3) for chromatin-free oligomers, but as will be discussed below, modification was evident upon chromatin assembly of AFP templates. SnoN binding was specific to the p53RE, as it did not associate with control DNA (lane 4).

SnoN mediated transcription repression is likely achieved through recruitment of co- repressors including N-CoR, mSin3 and histone deacetylases (HDAC) which maintain chromatin in an under-acetylated state associated with silenced target promoters (reviewed in Cohen, et al.,

1999; Wotton, et al., 1999). In examining binding of mSin3 and N-CoR at the p53 RE, we find that mSin3A associates only when p53 is bound, and that this binding is not decreased upon association of HBx (Fig. 2C, compare lanes 1-3). Strong association of N-CoR with the p53RE could not be detected, although modest binding was observed occasionally (data not shown).

This may be due to lack of chromatin assembly in the binding assays, as some factors

preferentially associate with chromatin assembled DNA templates.

SnoN-p53 interaction is DNA dependent. To determine if the apparent SnoN/p53 association at the p53 regulatory element was through protein-protein interaction or mediated through DNA, we preformed co-immunoprecipitation assays using p53 and Smad4 antibodies. Recombinant p53 protein was added to adult mouse liver nuclear or HepG2 whole cell extract and briefly incubated. Following incubation, p53-containing complexes were immuno-precipitated with p53 antibody beads, washed, eluted, separated by gel electrophoresis and analyzed by Western blot

(Fig. 3). We could not detect association of soluble p53 with Smad4 or SnoN by immunoprecipitation with either p53 or Smad4 antibodies (Fig 3 and data not shown). While endogenous

(end.) levels of Smad4 and SnoN proteins were detectable in both extracts (lanes 4 and 9), immuno-precipitation with p53 antibody did not pull down Smad4 or SnoN protein from either the adult mouse liver (lanes 1-2) or HepG2 extracts (lanes 6-7). These data suggest that association and/or co-repressor function between p53 and Smad4/SnoN in AFP transcription repression are mediated predominantly through their common DNA binding element.

SnoN is involved in transcriptional repression of AFP. To determine if SnoN was involved in repression of the AFP gene we carried out a series of co-transfections in Hepa1-6 murine liver carcinoma cells (Fig 4A). We have previously demonstrated that co-transfection of an AFP-lacZ reporter along with a p53 expression construct results in a 3-5-fold repression of reporter gene expression (Lee, et al., 1999). As expected, p53 was able to repress the AFP reporter gene approximately 3-fold. Interestingly, co-transfection of SnoN along with p53 increased the level of AFP repression to 5-fold. Transfection of SnoN in the absence of exogenous p53 was also able to mediate repression of the AFP reporter construct approximately 5-fold. While this may

suggest that SnoN does not require p53 as a co-repressor for AFP transcription, it should be

noted that Hepa1-6 cells do express functional, endogenous p53 protein (Lee, et al., 1999) that

may assist over-expressed SnoN in repression of the AFP reporter gene.

Oncoprotein family members Ski and SnoN, which show high levels of homology, frequently associate at Smad binding elements as heterodimers with Smad proteins to elicit similar co-repressor functions (Cohen, et al., 1999; Wotton, et al., 1999). We wanted to determine if Ski functioned equally well in AFP repression. Interestingly, over-expression of Ski protein in the absence of co-transfected SnoN or p53 had little effect on AFP reporter construct

expression (less than 1.3-fold repression). However, co-transfection of Ski with p53 resulted in

robust activation (5.4-fold) of the AFP reporter construct when compared to p53 alone. Co- transfection of Ski along with p53 and SnoN resulted in a significant decrease in the level of AFP repression when compared to SnoN plus p53 or SnoN alone (5-fold repression vs. approximately

1.3-fold repression). These results are surprising, as SnoN and Ski are highly related proteins.

Over-expressed Ski could possibly mediate apparent transcription activation, rather than repression of AFP through squelching of endogenous factors such as histone deacetylase complex members. Over-expression of Ski has been demonstrated to repress TGF-β signaling through tight association with DNA binding factors such as endogenous SnoN, Smads and perhaps N-CoR (Zheng, et al., 1997). Similar mechanisms may be responsible for the apparent activation of AFP upon over-expression of Ski in Hepa1-6 cells.

TGF-β regulates endogenous AFP expression. Because our transfection data suggested that down-stream TGF-β effectors were involved in regulation of the AFP gene, we wanted to determine if endogenous AFP expression was regulated by TGF-β treatment of Hepa1-6 cells in

culture. Because AFP is a secreted protein, levels of AFP protein expression can be detected

through Western blot analysis of conditioned Hepa1-6 media. Treatment of cultured cells with

TGF-β1 ligand resulted in an initial increase in AFP secreted into the media, evident beginning at

12 hours and continuing through 36 hours post-treatment (Fig 4B, 12, 24 and 36 hours compare

vehicle (V) vs. TGF-β (T) treatment). Beginning at 48 hours post-treatment and continuing

through the rest of the time course, a significant decrease in secreted AFP was evident (48, 60

and 72 hours, compare vehicle with TGF-β treatment). The initial increase and subsequent decrease of AFP secretion is supportive of SnoN-mediated regulation of the endogenous AFP gene. Treatment of cells with TGF-β triggers a rapid degradation of nuclear SnoN protein, thereby allowing activation of TGF-β regulated genes (Stroschein, et al., 2001; Sun, et al., 1999;

Sun, et al., 1999). After a lag, SnoN expression is stimulated by TGF-β as a means of feed back inhibition that triggers re-silencing of target genes.

The p53 regulatory element is required for SnoN-mediated transcriptional repression of

AFP. To determine if p53 and/or Smad4-DNA binding was required for SnoN to render its repressive effects on the AFP promoter, we performed co-transfections with an AFP reporter,

AFP-DelA, which has a 10 bp deletion and 4 bp point mutations within the p53RE (Fig 5A).

These mutations disrupt p53 DNA binding to result in loss of p53 mediated transcriptional repression (Fig 5C and (Ogden, et al., 2000)). Upon comparing the effects of p53 and/or SnoN on regulation of AFP reporter constructs, we found that both p53 and SnoN-mediated repression were significantly disrupted upon mutation of the p53 binding site (Fig 5B, compare AFP (gray) and DelA (black) expression levels to baseline). Though disruption of the p53-DNA binding site converted p53 from a potent repressor to an activator of AFP, SnoN was still able to mediate a

modest level of repression (~ 1.25-fold repression). This could potentially be the result of

deletion of a portion of the p53 and Smad4 binding sites creating a new, cryptic Smad4 binding

element (Fig 5A, indicated in blue).

To examine association of Smad4 and SnoN with the wild type and mutated p53 RE, we

carried out bead-DNA protein binding assays using control (p53 RE) and mutated (DelA) oligos

(Fig 5C). Incubation of bead-DNAs with HepG2 extract plus and minus recombinant p53 protein

revealed that while p53-DNA binding was restricted to wild type oligomers, endogenous Smad4

and SnoN could associate with both the wild type and mutant p53RE’s (compare lanes 1-2 with

lanes 4-5). Taken together with the transfection data, these finding suggest that although SnoN is

able to bind the AFP promoter and trigger a decreased level of AFP repression in the absence of

p53, SnoN requires p53-DNA binding in order to achieve maximal transcription repression of

reporter constructs.

SnoN is required for AFP transcription repression. In order to examine the potential

interaction between p53 and SnoN-mediated regulation we carried out a series of co-transfection

assays utilizing a dominant negative, DNA binding mutant p53 protein, p53-KH (Hinds, et al.,

1989). Because p53 binds DNA as a tetramer, p53-KH can disrupt the function of endogenous

p53 protein by squelching it away from DNA binding. We have previously demonstrated that

Hepa1-6 cells express functional, endogenous p53 protein that can be squelched by p53-KH to

activate, rather than repress, AFP reporter expression (Lee, et al., 1999). In contrast to wild type

p53, p53-KH triggered modest activation of an AFP reporter construct when co-transfected into

Hepa1-6 cells (1.0 vs. 1.2 relative expression level, Fig 6A). Co-transfection of SnoN with p53-

KH resulted in a 3.4-fold decrease in AFP reporter expression when compared to p53-KH, 3.1-

fold when compared to baseline (Fig 6A, compare KH with KH + SnoN).

To determine to what extent SnoN was required for AFP transcription repression we immuno-depleted whole cell HepG2 extract for endogenous SnoN protein, then utilized this depleted extract for in vitro chromatin transcription analysis (Fig 6B, top panel). AFP DNA templates were pre-incubated in control, SnoN immuno-depleted, or mock depleted extracts in the absence or presence of recombinant p53 protein. Chromatin assembly in buffer only mediates transcription repression (lane 2 compared to lane 3). As we previously demonstrated, incubation of AFP DNA templates in HepG2 extract prior to chromatin assembly programs AFP for activated chromatin transcription (Fig 6B lower panel, lane 3 and (Ogden, et al., 2000)).

Addition of recombinant p53 protein to chromatin assembly reactions in control HepG2 extract resulted in an approximate 3-fold repression of AFP transcription (Fig 6B lower panel, compare lanes 4 and 5 with 3). Conversely, addition of p53 to AFP chromatin assembly reactions in SnoN

ID HepG2 extract did not trigger transcription repression (compare lanes 7 and 8 with 6), suggesting that p53 can not function to repress AFP chromatin templates in the absence of SnoN protein. Addition of p53 to mock depleted reactions revealed a decrease(~ 2-fold, compare to lanes 5 and 8) in p53’s ability to repress AFP transcription that correlated with the modest amount of SnoN protein removed during the mock-depletion procedure (Fig 6B, mock bead bound). p53 is required to stabilize SnoN chromatin-DNA binding. To examine the effects of p53 on

SnoN DNA-binding under more physiological conditions, we turned to a system of in vitro chromatin assembly followed by chromatin immunoprecipitation (ChIP) analysis. Full-length

AFP DNA templates were pre-incubated with HepG2 extract in the presence or absence of

recombinant p53 and HBx proteins prior to addition of chromatin assembly extract (Fig 7). After chromatin assembly, DNA and associated proteins were formaldehyde cross-linked and processed by micrococcal nuclease digestion to mono-and di-nucleosomal fragments. The percent of SnoN protein bound at the p53 regulatory element when chromatin assembled was determined by Southern blot analysis of SnoN antibody and control precipitated nucleosomal fragments. ChIP analysis revealed that in the absence of exogenous p53 protein, approximately

4% (3.89 +/- .39%) of the p53RE DNA was occupied by SnoN (Fig 7A and B). Interestingly, addition of recombinant p53 protein nearly doubled the amount of SnoN associated at the p53

RE to that of 7.19 +/- 1.28%, suggesting that p53 stabilizes SnoN binding to chromatin templates. SnoN association with the AFP distal repressor in a chromatin context is specific to the p53RE, as association was not detected with vector control probe (Fig 7A).

We have previously shown that binding of an approximately 80-kDa molecular weight protein to the p53RE was disrupted in the presence of p53 and HBx (Ogden, et al., 2000). In examining the effect of HBx on SnoN binding to AFP chromatin templates, we found that addition of recombinant HBx protein with p53 protein to chromatin assembly reactions in HepG2 extract resulted in a dramatic 22-fold decrease (0.32 +/- 0.4%) in the amount of SnoN bound at the p53RE (Fig 7B and C). From these data, we conclude that HBx is capable of disrupting assembly of the putative repressor complex, and suggest that this may be the mechanism by which AFP is reactivated in response to HBx-p53 interaction.

Discussion

The exact regulatory processes responsible for the tightly controlled expression pattern of

the AFP tumor marker gene have yet to be clearly defined. Proper function of a range of

transcription activators and repressors including HNF-3, C/EBP, p53, Smad4 and others are

required for the appropriate developmental activation and post-natal silencing of the AFP gene.

We previously identified an over-lapping p53/HNF-3 regulatory element centered at -850 in the

AFP distal repressor necessary for direct p53-mediated transcription repression. Repression of

AFP by p53 is a tissue specific event (Lee, et al., 1999; Ogden, et al., 2000), and likely involves

the function of liver-enriched transcription co-repressors. Here we demonstrate the existence of

a putative chromatin-enhanced p53, Smad4 and SnoN-containing complex that mediates

repression of the AFP gene at a composite, distal (-850) regulatory element. Association of p53

and Smad4 with over-lapping DNA binding elements results in the recruitment and stabilization

of SnoN and additional co-repressors to AFP chromatin templates. To our knowledge, this is the

first report of a DNA-mediated cooperation between p53 and downstream TGF-β effectors to regulate a target gene.

A role for TGF-β and its down-stream effectors in repression of the AFP gene was first

suggested by the observation that treatment of cultured hepatocytes with TGF-β1 ligand resulted in a decrease in AFP secretion into growth media (Beauchamp, et al., 1992; Nakao, et al., 1991).

The regulatory mechanism responsible for the down-regulation of AFP in these studies appeared to differ between cell lines. Studies involving human liver cell lines supported a direct transcriptional regulation of the AFP gene, while studies of cultured murine hepatocytes suggested that the level of regulation was post-transcriptional. Our results support a direct role

for Smad4/SnoN heterodimers in repression of the AFP gene, and suggest that Smad4/Smad3

heterodimers may be involved in activation of the AFP gene during liver development. Support for involvement of Smad3 and an additional down-stream TGF-β effector, Smad2, in transcription activation of the AFP gene come from studies involving Smad2+/-; Smad3+/-

transgenic mice. Analysis of livers from these animals revealed improper liver architecture and

decreased AFP expression in response to decreased Smad2 and Smad3 protein expression

(Weinstein, et al., 2001). If, in fact, Smad proteins are involved in both the activation and

repression of AFP in response to differential ligand stimulation during development, it is

possible that the over-lapping HNF-3/p53/Smad binding element may be the AFP master control element, or “developmental switch”. Changes in co-repressor or co-activator association at the developmental switch in response to changing regulatory signals may be responsible for re- activation of the gene during liver regeneration or liver disease.

Aberrant re-activation of the AFP gene tightly correlates with chronic Hepatitis B virus

(HBV) infection, liver disease and liver cancer. Its over-expression is observed during chronic

HBV infection prior to the onset of liver cancer (Hsia, et al., 1992), and may contribute to inappropriate cell cycle progression and cancer development. AFP has been suggested to act as a hepatocyte growth factor in that exposure to AFP triggers replication of liver cells in culture

(Wang and Xie, 1999). We have previously demonstrated that interaction of the HBV transactivator protein HBx with p53 results in re-activation of AFP gene expression. Here we suggest that the mechanism by which HBx alleviates p53-mediated repression of AFP is through an association with DNA-bound p53 to block recruitment of SnoN and additional co-repressors to AFP chromatin templates.

HBV- and HBx-mediated modification of TGF-β target gene regulation in the liver has not been studied as thoroughly as HBx-mediated inactivation of p53 function. However, a role for HBx in disruption of TGF-β signaling was first suggested by studies involving HBx transgenic animals. Initial studies of HBx transgenics led to the discovery of p53-HBx association during liver cancer development (Ueda, et al., 1995). Subsequent examination of the animals revealed co-localization of HBx with TGF-β1 ligand in altered foci of the diseased livers. Because TGF-β1 lignad was not detected in normal, adult livers, regulation of the gene in transgenic livers was closely examined. HBx up-regulated the TGF-β1 promoter through an

interaction with the Egr-1 transcription factor, thereby allowing down-stream activation of

multiple TGF-β target genes (Yoo, et al., 1996). More recent studies support direct activation of

TGF-β target genes by HBx through an association with DNA-bound Smad4 protein (Lee, et al.,

2001). This interaction resulted in dramatic up-regulation of transcription driven by Smad4

binding elements, and by-passed the need for ligand stimulation normally required for TGF-β

target gene activation. Both direct and indirect modification of TGF-β signaling by HBx may

contribute to liver cancer in that multiple genes regulated by TGF-β family pathways are

involved in extra cellular matrix production, liver fibrogenesis, cirrhosis and growth control

(reviewed in Fausto and Mead, 1989; Fausto and Webber, 1993).

Re-activation of AFP during chronic HBV infection may occur through a series of events

induced by the HBx protein. Our data support the HBx-mediated disruption of a putative

repressor complex through an association with DNA-bound p53 resulting in loss of SnoN

association with AFP chromatin templates. Additionally, we speculate that HBx may work

through DNA-bound Smad4 to trigger robust activation of the AFP gene. HBx-mediated

disruption of p53 repressor complexes may be a recurring theme in HBV physiology. Previous studies identified a liver-enriched, p53-containing complex necessary for repression of HBV

Enhancer II. This complex was disrupted upon p53-HBx interaction, thereby alleviating all repressive effects of p53 on Enhancer II (Lee, et al., 1998). The individual components of the complex were not identified, so it is unclear if similar components exist in the AFP complex.

Regardless, the ability of HBx to destroy p53- and/or Smad4/SnoN-containing repressor complexes is likely a key process by which HBV infection can dramatically alter the physiology of infected cells.

1A p53 RE

Liver, hepatoma or HeLa extract +/- r-p53/HBx Wash

Western blot Silver stain MALDI 1B + p53-p53 + p53 3 L sh sh sh p5 M UB Wa E1 E2 UB Wa E1 E2 UB Wa E1 E2 1 2 3 4 5 67 8 9 1011121314MW 200

∗ ∗ 68 BSA

p53 † 43

p53 Reg Control † = p53, * = SnoN 1C h h h h 3 L as as as as p5 M W E1 E2 W E1 E2 W E1 E2 W E1 E2 1 23 4567891011121314MW 200

* BSA 68

p53 † † 43

p53 RegControl p53 Reg Control ML + p53 HeLa + p53

Figure 1: SnoN associates at the AFP p53 regulatory element.

(A) Solid phase DNA-protein purification scheme. Immobilized p53 regulatory element (p53

RE) DNA or -1007 (control) DNA oligomers were pre-bound with recombinant p53 protein and incubated in adult mouse liver, hepatoma or HeLa nuclear extract at room temperature for 30 minutes. After incubation, DNA-protein complexes were washed and bound proteins were eluted and analyzed by silver stain for p53 and putative co-repressor association. Proteins appearing to be bound or stabilized in response to p53 were then purified, trypsinized and subjected to MALDI analysis. (B) p53 stabilizes the DNA binding of an apparent 80 kDa protein. Immobilized p53 regulatory element or control (-1007) DNA were pre-incubated with approximately 500 ng recombinant p53 protein (lanes 3-6 and 11-14) or buffer plus BSA (lanes

7-10) then incubated in approximately 200 ug adult mouse liver nuclear extract (ML). Bound fractions were washed 2 times in PBS/1% NP-40 and eluted at 37° C for 30 minutes (E1; lanes 5,

9 and 13) then at 37° for 10 minutes (E2; lanes 6, 10 and 14). Eluted proteins were separated by gel electrophoresis along with recombinant p53 (lane 1), adult mouse liver extract (ML, lane 2), unbound fractions (UB, lanes 3, 7 and 11) and combined wash fractions (lanes 4, 8 and 12).

Proteins were visualized by silver stain analysis. Recombinant p53 protein (denoted by †) and a putative p53-stabilized DNA binding factor (denoted by *) were excised, trypsinized and subjected to MALDI analysis. (C) A putative p53 co-factor is liver enriched. Immobilized p53 regulatory element or control (-1007) DNA was pre-incubated with approximately 500 ng recombinant p53 protein then incubated in approximately 200 ug adult mouse liver (ML, lanes 3-

8) or HeLa (lanes 9-14) nuclear extract as described. Eluted proteins were separated by gel electrophoresis along with recombinant p53 (lane 1), adult mouse liver extract (lane 2), second

wash fractions (lanes 3, 6, 9 and 12) and visualized by silver stain analysis. Recombinant p53 protein (denoted by †) and the putative p53-stabilized DNA binding factor (denoted by *) were subjected to MALDI analysis as described

2A p53 p53

-860 GCCTTAGCAAACATGTCTGGACCTCTAGACA-830 AFP CGGAATCGTTTGTACAGACCTGGAGATCTGT HNF 3 Smad 4 Smad 3

ML HepG2 p53 - +++--++ - HBx --1x 2x 2x -- 2x 2x 12345 678 9

p53

HBx

Smad4

p53 RE

ML p53 - --1x 2x 2x - - 2x HBx -1x2x2x2x- -2x 2x 123456789 Smad4

p53 RE -1007

HepG2 p53 - ++ + HBx - - ++ 1234 SnoN Smad4 mS in3 A p53 HBx p53 RE -1007

Figure 2: Putative co-repressor complex associates at the p53 regulatory element.

(A) p53 regulatory element contains Smad4 and Smad3 binding elements. The p53/HNF-3 regulatory element centered at -850 in the AFP promoter (Lee, et al., 1999) contains consensus

Smad4 (7/8) and Smad3 (5/7) binding elements (indicated in red)(Zawel, et al., 1998). (B)

Smad4 binding at the p53 RE is moderately weakened by HBx. Immobilized p53 RE (upper and lower panels) or control (-1007, lower panel, lanes 7-9) DNA oligomers were incubated in approximately 200 ug of adult mouse liver nuclear (ML, lanes 1-5 upper panel and 1-9 lower panel) or HepG2 whole cell (lanes 6-9 upper panel) extract in the presence of recombinant p53

(approximately 500 ng, lanes 2-4 and 7-8, upper panel and 250 ng, lane 4 and 500 ng, lanes 5, 6 and 9, lower panel) and HBx (approximately 350 ng, lane 3 upper panel and lane 2 lower panel, or 700 ng, lanes 4-5 and 8-9, upper and lower panels). Bound proteins were washed twice in

PBS/1% NP-40, eluted, separated by gel electrophoresis and analyzed by western blot. p53, HBx and Smad4 proteins are indicated. (C) SnoN and mSin3A associate at the p53 regulatory element. Immobilized p53 RE (lanes 1-3) or control (-1007, lane 4) DNA oligomers were incubated in approximately 200 ug HepG2 whole cell extract in the presence of recombinant p53

(approximately 500 ng, lanes 2-4) and HBx (approximately 700 ng, lanes 3 and 4). Bound proteins were concentrated, washed twice in PBS/1% NP-40, separated by gel electrophoresis and analyzed by western blot as indicated.

100

IP End. UB IP End. p53 +++-++++- HBx -++-+-++- α-p53 + + -- - + + - - 1 23456789

Smad 4

SnoN

p53

ML HepG2

Figure 3: Smad4 and SnoN association with p53 are DNA-dependent.

Recombinant p53 (approximately 500 ng, lanes 1-3 and 6-8) and HBx (approximately 750 ng, lanes 2-3 and 7-8) proteins were incubated in adult mouse liver nuclear (approximately 200 ug, lanes 1-3 ) or whole cell HepG2 (approximately 200 ug, lanes 6-8) extracts for 30 minutes at 4°

C followed by 1 hour incubation with p53 antibody (Santa Cruz pAB 240, lanes 1-2 and 6-7).

Antibody bound complexes (IP, lanes 1-3 and 6-8) were collected by incubation with Protein

A/G agarose beads (Pierce), washed three times with PBS/1% NP-40 and separated by gel electrophoresis along with antibody control reactions (lanes 3 and 8), adult mouse liver extract

(endogenous (end.), lane 4), unbound mouse liver fraction (UB, lane 5) or HepG2 whole cell extract (end., lane 9). Proteins were analyzed by western blot as indicated.

101

4A 3

2.5

1.5

1 Relative AFP expression Relative

0.5

0 Baseline p53 SnoN p53, SnoN Ski p53, Ski SnoN, Ski p53, SnoN, Ski Transfected DNA

TGF-β 4 Hr. 12 Hr. 24 Hr. 36 Hr. 48 Hr. 60 Hr. 72 Hr. Treatment VT VT VT VT VT VT VT

AFP

Load

102

Figure 4: SnoN is involved in AFP transcription repression.

(A) AFP reporter constructs are repressed by SnoN in transient transfection assays.

Cultured Hepa 1-6 cells were co-transfected with the AFP/lacZ reporter construct (500 ng/plate)

and indicated expression vectors (p53, SnoN or Ski, 500 ng/plate each). Each plate was also co-

transfected with the pGL2 luciferase vector (50 ng/plate) to standardize and control for

transfection efficiency. Expression levels relative to baseline are indicated. Each experiment

was performed in duplicate a minimum of three times. (B) TGF-β treatment regulates secretion of AFP by cultured Hepa 1-6 cells. Hepa 1-6 cells were cultured out and grown to confluency in 60 mm tissue culture plates. Cells were treated with TGF-β1 ligand (4 ng/mL serum-free media (Gibco, Opti-MEM)) or vehicle (4 mL HCl, 1% BSA) for up to 72 hours, with media change at 36 hours post treatment. Media was collected at indicated time points and analyzed for AFP secretion by gel electrophoresis and western blot (upper panel). AFP is indicated. Loading control (Load) was visualized by silver stain.

103

p53 p53 -860 GCCTTAGCAAACATGTCTGGACCTCTAGACA -830 -860 GCCTTAGCAACcgCTtaACA -830 CGGAATCGTTTGTACAGACCTGGAGATCTGT CGGAATCGTTGgcGAatTGT Smad 4 Smad 3 Smad 4 Smad 3

Wild type p53 RE p53 DelA RE

5B AFP vs. AFP-DelA

1.6 1.4 1.2 1 AFP 0.8 DelA 0.6 0.4 Relative AFP 0.2

expression (B-gal/luc) 0 Baseline p53 sno p53 + sno Transfected DNA

5C HepG2 p53 -+++- 12 34 5 p53

Smad4

SnoN

p53 RE -1007 DelA

104

Figure 5: p53 DNA binding is required for maximal SnoN-mediated AFP repression.

(A) Wild type and mutated p53 regulatory element oligos. Putative high and low affinity

Smad4 binding sites are indicated for wild type (p53RE) and mutant (DelA) oligomers. (B)

SnoN transcription repression is compromised by disruption of p53 DNA binding. Hepa 1-

6 cells were transfected with AFP/lacZ or AFP DelA/LacZ reporter constructs (500 ng/plate) along with indicated expression vectors (p53 and SnoN, 500 ng/plate each). Each plate was also co-transfected with the pGL2 luciferase vector (50 ng/plate) to standardize and control for transfection efficiency. Expression levels relative to baseline are indicated. Each experiment was performed twice in duplicate, so standard error is not expressed. (C) Smad4 and SnoN can bind mutant p53 regulatory element. Immobilized p53 RE or p53 RE DelA oligos were incubated in approximately 250 ug HepG2 whole cell extract in the presence of recombinant p53 protein (approximately 1 ug, lanes 2, 3 and 5) or buffer (lanes 1 and 4). DNA-protein complexes were washed twice in PBS/1% NP-40, eluted, separated by gel electrophoresis and western blotted as indicated.

105

1.4

1.2

0.8

0.6

0.4

Relative AFP expression AFP Relative 0.2

0 AFP p53KH SnoN p53KH + SnoN Transfected DNA

6B Bead Bead Con. SnoN ID Mock ID bound bound

SnoN

Pre-inc. Buff. Buff. Con. HepG2 SnoN ID Mock ID p53 --- - -+ Chromatin -+++++++++ MW12345678910

AFP

106

Figure 6: Functional p53 and SnoN are both required for maximal AFP repression.

(A) Dominant negative p53 hinders SnoN-mediated AFP transcriptional repression. Hepa

1-6 cells were co-transfected with AFP/lacZ reporter construct (500 ng/plate) along with the

indicated expression vectors (p53 and SnoN, 500 ng/plate each). Each plate was also co-

transfected with the pGL2 luciferase vector (50 ng/plate) to standardize and control for

transfection efficiency. Expression levels relative to baseline are indicated. Each experiment

was performed a minimum of three times in duplicate. (B) SnoN is required for p53-mediated

repression of in vitro AFP chromatin transcription. SnoN was immuno-depleted from HepG2

whole cell extract (upper panel). Immobilized AFP DNA templates were incubated in control

(lanes 3-5; approximately 200 ug total protein), immuno-depleted (lanes 6-8; approximately 200

ug total protein) or mock-depleted (lanes 9-10; approximately 200 ug total protein) extract or

buffer (lanes 1-2) in the presence of p53 (approximately 1 ug, lanes 4 and 7; 2 ug, lanes 5, 8 and

10) or buffer (lanes 1-3, 6 and 9) prior to 1 hour chromatin assembly in fractionated Xenopus egg extract. Chromatin assembled DNA templates were washed in nuclear extract buffer and in vitro transcribed in HeLa nuclear extract. AFP primer extension products are indicated.

107

7A d d n n u d u d t o n t o n u b u u b u p n p n In U Bo In U Bo

Control HepG2 + p53 -HBx α -SnoN

Control HepG2 + p53 + HBx α -SnoN

p53 Reg. Probe Vector Probe

7B SnoN bound at p53 RE

4 % bound 3

0 HepG2 HepG2 + p53 HepG2 + p53 + HBx Pre-incubation

108

Figure 7: HBx disrupts p53-stabilized SnoN binding to chromatin DNA templates.

(A) HBx disrupts SnoN from binding chromatin assembled AFP DNA templates.

Immobilized AFP DNA templates were incubated in HepG2 whole cell extract (approximately

200 ug total protein) plus recombinant p53 (approximately 200 ng) in the presence of HBx

(approximately 400 ng, as indicated) or buffer for 20 minutes prior to 1 hour chromatin assembly

in fractionated Xenopus egg extract. Chromatin assembled DNA templates were washed in

extract buffer prior to formaldehyde cross-linking and MNase digestion to 200-500 base-pair

fragments. Chromatin immunoprecipitation of protein-DNA complexes was performed with

anti-SnoN antibodies (Upstate Biotechnology). DNA fragments were purified from immune

complexes after reversal of crosslinks and subjected to Southern analysis with the p53 regulatory

element probe or vector probe (control). (B) HBx disrupts p53-mediated stabilization of

SnoN association with AFP chromatin templates. A series of chromatin

immunoprecipitations (≥ 3 for each data point) were performed as described above. The amount of p53RE occupied by SnoN protein in the presence or absence of recombinant p53 or p53 plus

HBx is indicated. Values for percent SnoN associating at the p53 RE were determined by phosphor image analysis (Molecular Dynamics) and normalized to no antibody control reactions.

109

Chapter 5: Summary and Conclusions

Disruption of established gene expression patterns is common during cellular

transformation and tumor development. Using the α-fetoprotein tumor marker gene as a model system, these studies establish one mechanism by which p53 mediates direct transcriptional repression, and how repression is modified during chronic viral infection and cancer progression.

The p53 tumor suppressor protein, which plays critical roles in gene regulation, cell cycle arrest and apoptosis, is a direct transcriptional repressor of the AFP gene (Lee, et al., 1999). Its

association with an over-lapping HNF-3/p53/Smad4 regulatory element in the distal repressor of

the AFP promoter is required during chromatin organization to ensure transcriptional repression

(Chapter 2).

Disruption of p53 function leading to changes in gene expression and modulation of

apoptosis is common during viral infection. Chronic infection with Hepatitis B virus (HBV) is

tightly correlated to development of hepatocellular carcinoma (HCC). Studies involving

transgenic mice revealed the likely HBV-encoded oncoprotein to be the transcription

transactivator, HBx. Diseased livers of transgenic animals revealed association of HBx with

murine p53 protein, resulting in cytoplasmic sequestration and functional inactivation of p53

(Ueda, et al., 1995). Here, we have examined the effect of p53-HBx association on AFP

expression in attempts to understand how HBV modifies p53 gene regulation to contribute to

liver cancer development. HBx association with DNA-bound p53 reverses p53-mediated

repression of both chromatin-free and in vitro chromatin assembled AFP DNA templates. The mechanism by which HBx reverses p53 regulation of the AFP gene appears to be through disruption of a putative Smad4 and SnoN-containing co-repressor complex from associating at

110

the p53 regulatory element.

Concomitant binding of p53 and Smad4 to their over-lapping DNA binding elements

allows for recruitment of mSin3A and SnoN co-repressors. Recruitment of mSin3A is dependent

upon p53 binding and, while it was not examined here, it is well documented that SnoN requires

the DNA binding function of Smad proteins (reviewed in Attisano, 2000; Miyazono, 2000;

Miyazono, 2000; Wotton, et al., 1999). The co-repressor functions of both SnoN and mSin3A

involve modification of local chromatin to result in transcription repression. SnoN recruits the

nuclear hormone co-repressor, N-CoR, which in turn recruits HDAC proteins. mSin3A directly

recruits HDAC proteins to facilitate transcription repression. Interestingly, effects of p53 and

SnoN on AFP transcription repression appear to be enhanced and increasingly dependent upon each other with chromatin assembly of AFP templates. This is evidenced by transient transfections, in vitro chromatin transcription and chromatin immunoprecipitation (ChIP) analysis. In transient transfection assays, the ability of SnoN to repress AFP reporter genes is largely independent of p53. However, when examining binding of the two factors to AFP chromatin templates, the ability of SnoN to associate at the element is dramatically increased by p53 DNA binding (Chapter 4). This apparent enhancement of SnoN association with the element in the presence of p53 may be a result of p53’s ability to organize AFP chromatin templates.

Binding of p53 to the element may increase the accessibility of the site to facilitate Smad4 binding and SnoN recruitment. Similarly, p53 appears to require SnoN in order to repress chromatin DNA templates, but is capable of repressing reporter constructs in the absence of over- expressed SnoN protein. Further studies examining the interdependence of the two proteins on each other to mediate chromatin-based transcriptional repression are needed. ChIP assays

111

examining the binding of Smad4 and SnoN to AFP-DelA (disrupted for p53 binding) chromatin templates would be quite revealing. While we have shown that both Smad4 and SnoN are able to bind AFP-DelA double-stranded oligomers in the absence of p53, we have not examined if they can gain access to the site on chromatin templates. Additionally, while the ChIP assays demonstrate that SnoN binding to chromatin templates is hindered by the absence of p53, a direct functional consequence was not determined. The data strongly suggest that AFP repression would be disrupted, and could be confirmed by in vitro chromatin transcription utilizing p53 immuno-depleted extracts in conjunction with recombinant SnoN protein. Further studies involving the acetylation status of the AFP promoter in HepG2 or Hepa1-6 cells would also be interesting. While we demonstrate that histone tail deacetylation is not required for p53- mediated repression of AFP chromatin templates assembled in the absence of HepG2 or Hepa 1-

6 extract (Chapter 2), the role of histone tail modification in the liver-specific repression of AFP has not been established. Targeted deacetylation of histone tails by p53 and mSin3A is likely required to over-come the function of numerous tissue-specific AFP transactivating factors present in liver-derived extracts.

The basis of HBx-mediated alleviation of p53 transcription repression appears to be its ability to disrupt the formation of a putative Smad4 and SnoN containing co-repressor complex

(Chapter 3 and 4). Association of HBx with DNA-bound p53 results in a substantial loss of

SnoN binding at the AFP p53 regulatory element, which likely hinders the recruitment of N-CoR, mSin3A and HDAC containing complexes. As with the SnoN and p53 transcriptional effects,

HBx-mediated effects appear to be enhanced for chromatin templates, underscoring the importance of chromatin on AFP gene regulation. Recent evidence supports a role for HBx in

112

enhancing the activator function of Smad4 through direct protein-protein interaction (Lee, et al.,

2001). Smad2/Smad4 and Smad3/Smad4 complexes achieve transcriptional activation of target

genes through recruitment of p300/CBP co-activator to target local chromatin for hyper-

acetylation resulting in gene activation (Attisano, 2000; Pouponnot, et al., 1998). Although it has

yet to be demonstrated, it is interesting to speculate that HBx association with DNA-bound

Smad4 may increase its affinity for p300/CBP. Therefore, if HBx works through both p53 and

Smad4 to reactivate AFP, one would expect HBx-mediated conversion of the AFP p53 regulatory element from a repressor site to an activator site to involve dramatic changes in the acetylation state of the local and distal chromatin. Interestingly, preliminary ChIP analysis suggests that HBx association at the p53 regulatory element increases the acetylation of histone

H3 and H4 tails at the AFP transcription start site (S.Ogden, unpublished data). Additional analysis is needed to confirm theses results, and to examine the acetylation state at the p53 regulatory element in liver-derived extracts both plus and minus recombinant HBx and p53 proteins. Also, while solid-phase protein purification experiments suggest that Smad4 may be displaced from the p53 regulatory element upon HBx association, we have yet to determine

Smad4 DNA binding status of chromatin templates plus and minus p53 and HBx.

Evidence of p53 involvement in the stabilization and function of a Smad4/SnoN- containing repressor complex implicated in the developmental silencing of the AFP gene connects two pathways commonly disrupted during cancer progression: TGF-β signaling and p53 gene regulation/tumor suppressor activity. Because theses pathways are involved in the regulation of multiple genes controlling cell growth and/or cell cycle progression, dysfunction of proteins involved in the pathways can have disastrous consequences. Here we have described

113

studies examining p53-mediated transcriptional repression of the AFP gene, and the ability of the virally encoded HBx protein to reverse the function of a p53/Smad4/SnoN-containing repressor complex to trigger AFP reactivation. Effects of the repressor complex and reversal of these effects by HBx appear to be enhanced for AFP chromatin templates, suggesting a direct role for chromatin remodeling in AFP gene expression (see model). Further studies that more closely examine regulation of AFP chromatin structure in response to p53/Smad4/SnoN and HBx would complement these studies to yield a complete understanding of how gene expression is manipulated during chronic viral infection leading to cancer development.

114

Summary

P Nuclear membrane Smad 3 Smad4

P Smad 3 Smad4 SnoN

Smad4 SnoN

N-CoR AFP HDAC6 AFP

XXmS in3 A SnoN X p53 p53Smad4

-850 -850

HBx

P Nuclear membrane Smad 3 Smad4

mS in3 A P Smad 3 Smad4 SnoN N-CoR

Figure Summary HDAC6 Smad4 SnoN SnoN

HBx

AFP AFP HBx HBx 115 p53 p53Smad4 -850 -850

References

Abelev, G. I.and Eraiser, T. L. (1999). Cellular aspects of alpha-fetoprotein reexpression in tumors.

Sem. Cancer Biol. 9, 95-107.

Agarwal, S.and Rao, A. (1998). Long-range transcriptional regulation of cytokine gene expression.

Curr Opin Immunol 10, 345-352.

Andrisani, O. M.and Barnabas, S. (1999). The transcriptional function of the hepatitis B virus X

protein and its role in hepatocarcinogenesis. Intl. J. Cancer 15, 373-379.

Arbuthnot, P.and Kew, M. (2001). Hepatitis B virus and hepatocellular carcinoma. Int J Exp Pathol.

82, 77-100.

Attisano, L. (2000). Smads as transcriptional co-modulators. Curr. Opin. Cell Biol. 12, 235-243.

Barnabas, S.and Andrisani, O. M. (2000). Different regions of Hepatitis B virus X protein are

required for enhancement of bZip-mediated transactivation versus transrepression. J. Virol. 74,

83-90.

Barton, M. C.and Emerson, B. M. (1996). Regulated gene expression in reconstituted chromatin and

synthetic nuclei. Meth. Enz. 274, 299-312.

Bayle, J. H., Elenbaas, B.and Levine, A. J. (1995). The carboxy-terminal domain of the p53 protein

regulates sequence-specific DNA binding through its nonspecific nucleic acid-binding activity.

Proc. Natl. Acad. Sci. 92, 5729-5733.

Beauchamp, R. D., Sheng, H.-M., Alam, T., Townsend, C. M.and Papaconstantinou, J. (1992).

Posttranscriptional regulation of albumin and alpha-fetoprotein messenger RNA by transforming

116

growth factor beta-1 requires de novo RNA and protein synthesis. Mol. Endocrinol. 6, 1789-

1796.

Belayew, A.and Tilghman, S. M. (1982). Genetic analysis of alpha-fetoprotein synthesis in mice.

Mol. Cell. Biol. 2, 1427-1435.

Boyer, P. L., Colmenares, C., Stavnezer, E.and Hughes, S. H. (1993). Sequence and biological

activity of chicken snoN cDNA clones. Oncogene 8, 457-66.

Budhram-Mahadeo, V., Morris, P. J., Smith, M. D., Midgley, C. A., Boxer, L. M.and Latchman, D.

S. (1999). p53 suppresses the activation of the Bcl-2 promoter by the Brn-3a POU family

transcription factor. J. Biol. Chem. 274, 15237-15244.

Buendia, M. A. (1992). Mammalian hepatitis B viruses and primary liver cancer. Semin. Cancer

Biol. 3, 309-20.

Camper, S. A., Godbout, R.and Tilghman, S. M. (1989). The developmental regulation of albumin

and alpha-fetoprotein gene expression. Prog Nucleic Acid Res Mol Biol 36, 131-43.

Camper, S. A.and Tilghman, S. M. (1989). Postnatal repression of the alpha-fetoprotein gene is

enhancer independent. Genes Dev 3, 537-46.

Camper, S. A.and Tilghman, S. M. (1991). The activation and silencing of gene transcription in the

liver. Biotechnology 16, 81-7.

Chou, J. Y.and Wan, Y.-J. Y. (1989). Regulation of expression of the rat a-fetoprotein gene. Oxf

Surv Eukaryot Genes 6, 1-31.

Cohen, S. B., Zheng, G., Heyman, H. C.and Stavnezer, E. (1999). Heterodimers of the SnoN and Ski

oncoproteins form preferentially over homodimers and are more potent thranforming agents.

Nuc. Acid Res. 27, 1006-1014.

117

Cosma, M. P., Tanaka, T.and Nasmyth, K. (1999). Ordered recruitment of transcription and

chromatin remodeling factors to a cell cycle- and developmentally regulated promoter. Cell 97,

299-311.

Crandall, B. F. (1981). Alpha-fetoprotein: a review. Crit. Rev. Clin. Lab. Sci. 15, 127-185.

Crowe, A. J.and Barton, M. C. (1999). Functional analysis of chromatin assembled in synthetic

nuclei. Methods 17, 173-187.

Crowe, A. J.and Barton, M. C. (1999). In vitro reconstitution of nuclei for replication and

transcription. Methods Enz. 304, 63-76.

Crowe, A. J.and Hayman, M. J. (1993). Post-translational modificiations of the env-sea oncogene

product: The role of proteolytic processing in transformation. Oncogene 8, 181-189.

Crowe, A. J., Piechan, J. L., Sang, L.and Barton, M. C. (2000). S-Phase progression mediates

activation of a silenced gene in synthetic nuclei. Mol. Cell. Biol. 4169-80.

Crowe, A. J., Sang, L., Lee, K. C., Spear, B. T.and Barton, M. C. (1999). Hepatocyte nuclear factor 3

relieves chromatin-mediated repression of the alpha-fetoprotein gene. J. Biol. Chem. 274,

25113-25120.

D'Souza, S., Xin, H., Walter, S.and Choubey, D. (2001). The gene encoding p202, an interferon-

inducible negative regulator of the p53 tumor suppressor, is a target of p53-mediated

transcriptional repression. J. Biol. Chem. 276, 298-305.

Derynck, R., Akhurst, R. J.and Balmain, A. (2001). TGF-B signaling in tumor suppression and

cancer progression. Nature Gen. 29, 117-129.

Dignam, J. D., Lebovitz, R. M.and Roeder, R. G. (1983). Accurate transcription initiation by RNA

polymerase II in a soluble extract from isolated mammalian nuclei. Nucl. Acids Res. 11, 1475-

118

1489.

Dilworth, F. J., Fromental-Ramain, C., Yamamoto, K.and Chambon, P. (2000). ATP-Driven

chromatin remodeling activity and histone acetyltransferases act sequentially during

transactivation by RAR/RXR In vitro. Mol. Cell. 6, 1049-1058.

Duan, L., Ozaki, I., Oakes, J. W., Taylor, J. P., Khalili, K.and Pomerantz, R. J. (1994). The tumor

suppressor protein p53 strongly alters human immunodeficiency virus type I replication. J. Virol.

68, 4302-4313.

El-Deiry, W. S., Kern, S. E., Pietenpol, J. A., Kinzler, K. W.and Vogelstein, B. (1992). Definition of

a consensus binding site for p53. Nature Gen. 1, 45-49.

El-Serag, H. B.and Mason, A. C. (1999). Rising incidence of hepatocellular carcinoma in the United

States. 340, 745-50.

Elmore, L. W., Hancock, A. R., Chang, S. F., Wang, X. W., Chang, S., Callahan, C. P., Geller, D.

A., Will, H.and Harris, C. C. (1997). Hepatitis B virus X protein and p53 tumor suppressor

interactions in the modulation of apoptosis. Proc. Natl. Acad. Sci., USA 94, 14707-14712.

Emerson, B. M., Nickol, J. M.and Fong, T. C. (1989). Erythroid-specific activation and derepression

of the chick ß-globin promoter in vitro. Cell 57, 1189-1200.

Espinosa, J. M.and Emerson, B. M. (2001). Transcriptional regulation by p53 through intrinsic

DNA/chromatin binding and site-directed cofactor recruitment. Mol. Cell 8, 57-69.

Fausto, N.and Mead, J. E. (1989). Regulation of liver growth: protooncogenes and transforming

growth factors. Laboratory Investigation 60, 4-13.

Fausto, N.and Webber, E. M. (1993). Control of liver growth. Crit. Rev. Eukaryot. Gene Expr. 3,

117-35.

119

Feitelson, M. A.and Duan, L. X. (1997). Hepatitis B virus X antigen in the pathogenesis of chronic

infections and the development of hepatocellular carcinoma. Am. J. Path. 150, 1141-1157.

Feitelson, M. A., Zhu, M., Duan, L. X.and London, W. T. (1993). Hepatitis B x antigen and p53 are

associated in vitro and in liver tissues from patients with primary hepatocellular carcinoma.

Oncogene 8, 1109-17.

Feuerman, M. H., Godbout, R., Ingram, R. S.and Tilghman, S. M. (1989). Tissue-specific

transcription of the mouse alpha-fetoprotein gene promoter is dependent on HNF-1. Mol Cell

Biol 9, 4204-12.

Fisher, D. E. (2001). The p53 tumor suppressor: critical regulator of life and death in cancer.

Apoptosis 6, 7-15.

Godbout, R., Ingram, R.and Tilghman, S. M. (1986). Multiple regulatory elements in the intergenic

region between the alpha-fetoprotein and albumin genes. Mol Cell Biol 6, 477-87.

Godbout, R.and Tilghman, S. M. (1988). Configuration of the alpha-fetoprotein regulatory domain

during development. Genes Dev 2, 949-56.

Gorski, K., Carneiro, M.and Schibler, U. (1986). Tissue-specific in vitro transcription from the

mouse albumin promoter. Cell 47, 767-776.

Gray, S.and Levine, M. (1996). Transcriptional repression in development. Curr. Opin. Cell Biol. 8,

358-364.

Greenblatt, M. S., Bennett, W. P., Hollstein, M.and Harris, C. C. (1994). Mutations in the p53

Tumor Suppressor Gene: Clues to Cancer Etiology and Molecular Pathogenesis. Cancer Res.

54, 4855-4878.

Gu, W.and Roeder, R. G. (1997). Activation of p53 sequence-specific DNA binding by acetylation

120

of the p53 C-terminal domain. Cell 90, 595-606.

Hanahan, D.and Weinberg, R. A. (2000). The hallmarks of cancer. Cell 100, 57-70.

Haviv, I., Vaizel, D.and Shaul, Y. (1996). pX, the HBV-encoded coactivator, interacts with

components of the transcription machinery and stimulates transcription in a TAF-independent

manner. EMBO J. 15, 3413-3420.

Hildt, E.and Hofschneider, P. H. (1998). The preS2 activators of the hepatitis B virus: activators of

tumour promoter pathways. Recent Results Cancer Res. 154, 315-29.

Hinds, P., Finlay, C.and Levine, A. J. (1989). Mutation is required to activate the p53 gene for

cooperation with the ras oncogene and transformation. J. Virol. 63, 739-746.

Horikoshi, N., Usheva, A., Chen, J., Levine, A. J., Weinmann, R.and Shenk, T. (1995). Two

domains of p53 interact with the TATA-binding protein, and the adenovirus 13S E1A protein

disrupts the association, relieving p53-mediated transcriptional repression. Mol. Cell. Biol. 15,

227-234.

Howe, L., Brown, C. E., Lechner, T.and Workman, J. L. (1999). Histone acetyltransferase complexes

and their link to transcription. Crit. Rev. Eukaryot. Gene Expr. 9, 231-243.

Hsia, C. C., Evarts, R. P., Nakatsukasa, H., Marsden, E. R.and Thorgeirsson, S. S. (1992).

Occurrence of oval-type cells in hapatitis B virus-associated human hepatocarcinogenesis.

Hepatology 16, 1327-1333.

Hsia, C. C., Kleiner, D. J., Axiotis, C. A., Di, B. A., Nomura, A. M., Stemmermann, G. N.and Tabor,

E. (1992). Mutations of p53 gene in hepatocellular carcinoma: roles of hepatitis B virus and

aflatoxin contamination in the diet [see comments]. J. Natl. Cancer Inst. 84, 1638-41.

Hsu, T. Y., Moroy, T., Etiemble, J., Louise, A., Trepo, C., Tiollais, P.and Buendia, M. A. (1988).

121

Activation of c-myc by woodchuck hepatitis virus insertion in hepatocellular carcinoma. Cell 55,

627-635.

Hupp, T. R., D.W., M., Midgley, C. A.and Lane, D. P. (1992). Regulation of the specific DNA

binding function of p53. Cell 71, 875-886.

Jin, D. K., Vacher, J.and Feuerman, M. H. (1998). alpha-Fetoprotein gene sequences mediating Afr2

regulation during liver regeneration. Proc. Natl. Acad. Sci. USA 95, 8767-72.

Johnson, C. A.and Turner, B. M. (1999). Histone deacetylases: complex transducers of nuclear

signals. Semin. Cell. Dev. Biol. 10, 179-88.

Johnson, R. A., Ince, T. A.and Scotto, K. W. (2001). Transcriptional repression by p53 through

direct binding to a novel DNA element. J. Biol. Chem. 276, 27716-27720.

Kastan, M. B., Canman, C. E.and Leonard, C. J. (1995). p53, cell cycle control and apoptosis:

Implications for cancer. Cancer and Metastasis Reviews 14, 3-15.

Kew, M. C. (1996). Hepatitis B and C viruses and hepatocellular carcinoma. Clin. Lab. Med. 16,

395-405.

Kim, C., Koike, K., Saito, I., Miyamura, T.and Jay, G. (1991). HBx gene of hepatitis B virus induces

liver cancer in transgenic mice. Nature 351, 317-320.

Ko, L. J.and Prives, C. (1996). p53: puzzle and paradigm. Genes Dev. 10, 1054-1072.

Koumenis, C., Alarcon, R., Hammond, E., Sutphin, P., Hoffman, W., Murphy, M., Derr, J., Taya, Y.,

Lowe, S. W., Kastan, M.and Giaccia, A. (2001). Regulation of p53 by hypoxia: dissociation of

transcriptional repression and apoptosis from p53-dependent transactivation. Mol. Cell. Biol. 21,

1297-1310.

Krause, K., Haugwitz, U., Wasner, M., Wiedmann, M., Mossner, J.and Engeland, K. (2001).

122

Expression of the cell cycle phosphatase cdc25C is down-regulated by the tumor suppressor

protein p53 but not by p73. Biochem. Biophys. Res. Commun. 284, 743-50.

Kuo, M. H.and Allis, C. D. (1999). In vivo cross-linking and immunoprecipitation for studying

dynamic Protein:DNA associations in a chromatin environment. Methods Companion Methods

Enzymol. 19, 425-433.

Lane, D. P. (1994). The regulation of p53 function: Steiner award lecture. Int. J. Cancer 57, 623-

627.

Lane, D. P.and Crawford, L. V. (1979). T antigen is bound to a host protein in SV40-transformed

cells. Nature 278, 261-263.

Langst, G., Becker, P. B.and Grummt, I. (1998). TTF-I determines the chromatin architecture of the

active rDNA promoter. EMBO 17, 3135-3145.

Lechner, M. S., Mack, D. H., Finicle, A. B., Crook, T., Vousden, K. H.and Laimins, L. A. (1992).

Human papillomavirus E6 proteins bind p53 in vivo and aborgate p53-mediated repression of

transcription. EMBO J. 11, 3045-3052.

Lee, D. K., Park, S. H., Yi, Y., Choi, S. G., Lee, C., Parks, W. T., Cho, H., de Caestecker, M. P.,

Shaul, Y., Roberts, A. B.and Kim, S. J. (2001). The hepatitis B virus encoded oncoprotein pX

amplifies TGF-beta family signaling through direct interaction with Smad4: potential mechanism

of hepatitis B virus-induced liver fibrosis. 15, 455-66.

Lee, H., Kim, H. T.and Yun, Y. (1998). Liver-specific enhancer II is the target for the p53-mediated

inhibition of hepatitis B viral gene expression. J. Biol. Chem. 273, 19786-19791.

Lee, H., Lee, Y. H., Huh, Y. S., Moon, H.and Yun, Y. (1995). X-gene product antagonizes the p53-

mediated inhibition of hepatitis B virus replication through regulation of the pregenomic/core

123

promoter. J. Biol. Chem. 270, 31405-31412.

Lee, K. C., Crowe, A. J.and Barton, M. C. (1999). p53-mediated repression of alpha-fetoprotein gene

expression by specific DNA binding. Mol. Cell. Biol. 19, 1279-1288.

Lee, Y. H.and Yun, Y. (1998). HBx protein of hepatitis B virus activates Jak1-STAT signaling. J.

Biol. Chem. 273, 25510-25515.

Levine, A. (1997). p53, the Cellular Gatekeeper for Growth and Division. Cell 88, 323-331.

Levine, A. J., Momand, J.and Finlay, C. A. (1991). The p53 tumor suppressor gene. Nature 351,

453-456.

Li, B.and Lee, M. Y. W. (2001). Transcriptional regulation of the human DNA polymerase delta

catalytic subunit gene POLD1 by p53 tumor suppressor and Sp1. J. Biol. Chem. 276, 29729-

29739.

Liu, L., Scolnick, D. M., Trievel, R. C., Zhang, H. B., Marmorstein, R., Halazonetis, T. D.and

Berger, S. L. (1999). p53 sites acetylated in vitro by PCAF and p300 are acetylated in vivo in

response to DNA damage. Mol Cell Biol 19, 1202-1209.

Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F.and Richmond, T. J. (1997). Crystal

structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251-60.

Massague, J. (1998). TGF-beta signal transduction. Annu. Rev. Biochem. 67, 753-91.

Massague, J.and Wotton, D. (2000). Transcriptional control by the TGF-beta/Smad signaling system.

EMBO J. 19, 1745-1754.

McPherson, C. E., Eun-Yong, S., Friedman, D. S.and Zaret, K. S. (1993). An active tissue-specific

enhancer and bound transcription factors existing in a precisely positioned nucleosomal array.

Cell 75, 387-398.

124

Miyazono, K. (2000). Positive and negative regulation of TGF-beta signaling. J. Cell Sci. 113,

1101-1109.

Miyazono, K. (2000). TGF-beta signaling by Smad proteins. Cyto. Growth Fact. Rev. 11, 15-22.

Murakami, S. (1999). Hepatitis B virus X protien: structure, function and biology. Intervirology 42,

81-99.

Murphy, M., Ahn, J., Walker, K. K., Hoffman, W. H., Evans, R. M., Levine, A. J.and George, D. L.

(1999). Transcriptional repression by wild-type p53 utilizes histone deacetylases, mediated by

interaction with mSin3a. Genes Dev. 13, 2490-2501.

Nagaich, A. K., Zhurkin, V. B., Durell, S. R., Jernigan, R. L., Appella, E.and Harrington, R. E.

(1999). p53-induced DNA bending and twisting: p53 tetramer binds on the outer side of a DNA

loop and increases DNA twisting. Proc Natl Acad Sci USA 96, 1875-1880.

Nakao, K., Nakata, K., Mitsuoka, S., Ohtsuru, A., Ido, A., Hatano, M., Sato, Y., Nakayama, T.,

Shima, M., Kusumoto, Y.and et, a. l. (1991). Transforming growth factor beta 1 differentially

regulates alpha-fetoprotein and albumin in HuH-7 human hepatoma cells [published erratum

appears in Biochem Biophys Res Commun 1991 Mar 29;175(3):1165]. Biochemical &

Biophysical Research Communications 174, 1294-9.

Nelson, R.and Long, G. (1989). A general method of site-specific mutagenesis using a modification

of the Thermus aquaticus polymerase chain reaction. Anal. Biochem. 180, 147-151.

Ogden, S. K., Lee, K. C.and Barton, M. C. (2000). Hepatitis B viral transactivator HBx alleviates

p53-mediated repression of AFP gene expression. J. Biol. Chem. 275, 27806-27814.

Ogden, S. K., Lee, K. C., Wernke-Dollries, K., Stratton, S. A., Aronow, B.and Barton, M. C. (2001).

p53 targets chromatin structure alteration to repress alpha-fetoprotein gene expression. J. Biol.

125

Chem. 276, 42057-42062.

Ohno, H., Kaneko, S., Lin, Y., Kobayashi, K.and Murakami, S. (1999). Human hepatitis B virus X

protein augments the DNA binding of nuclear factor for IL-6 through its basic-leucine zipper

domain. J. Med. Virol. 58, 11-18.

Ori, A., Zauberman, A., Doitsch, G., Paran, N., Oren, M.and Shaul, Y. (1998). p53 binds and

represses the HBV enhancer: an adjacent enhancer element can reverse the transcription effect of

p53. EMBO J. 17, 544-553.

Palmer, C. R., Gegnas, L. D.and Schepartz, A. (1997). Mechanism of DNA binding enhancement by

hepatitis B virus protein pX. Biochemistry 36, 15349-15355.

Perini, G., Oetjen, E.and Green, M. R. (1999). The hepatitis B pX protein promotes dimerization and

DNA binding of cellular basic region/leucine zipper proteins by targeting the conserved basic

region. J. Biol. Chem. 274, 13970-13977.

Pouponnot, C., Jayaraman, L.and Massague, J. (1998). Physical and functional interaction of Smads

and p300/CBP. J. Biol. Chem. 273, 22865-22868.

Ragimov, N., Krauskopf, A., Navot, N., Rotter, V., Oren, M.and Aloni, Y. (1993). Wild-type but not

mutant p53 can repress transcription initiation in vitro by interfering with the binding of basal

transcription factors to the TATA motif. Oncogene 8, 1183-1193.

Ramesh, T. M., Ellis, A. W.and Spear, B. T. (1995). Individual mouse alpha-fetoprotein enhancer

elements exhibit different patterns of tissue-specific and hepatic position-dependent activities.

Mol. Cell. Biol. 15, 4947-55.

Rooke, H. M.and Crosier, K. E. (2001). The smad proteins and TGF-beta signalling: uncovering a

pathway critical in cancer. Pathology 33, 73-84.

126

Rossi, J. M., Dunn, N. R., Hogan, B. L.and Zaret, K. S. (2001) Distinct mesodermal signals,

including BMPs from the septum transversum mesenchyme, are required in combination for

hepatogenesis from the endoderm. Genes Dev. 15, 1998-2009.

Seavey, S. E., Holubar, M., Saucedo, L. J.and Perry, M. E. (1999). The E7 oncoprotein of human

papillomavirus type 16 stabilizes p53 through a mechanism independent of p19(ARF). J. Virol.

73, 7590-7598.

Seto, E., Usheva, A., Zambetti, G. P., Momand, J., Horikoshi, N., Weinmann, R., Levine, A. J.and

Shenk, T. (1992). Wild-type p53 binds to the TATA-binding protein and represses transcription.

Proc. Natl. Acad. Sci. USA 89, 12028-12032.

Shachter, N. S., Zhu, Y., Walsh, A., Breslow, J. L.and Smith, J. D. (1993). Localization of a liver-

specific enhancer in the apolipoprotein E/C-I/C-II gene locus. Journal of Lipid Research 34,

1699-707.

Shikama, N., Lee, C. W., France, S., Delavaine, L., Lyon, J., Krstic-Demonacos, M.and La Thangue,

N. B. (1999). A novel cofactor for p300 that regulates the p53 response. Mol. Cell. 4, 365-76.

Shild, C., Claret, F. X., Wahli, W.and Wolffe, A. P. (1993). A nucleosome-depedent static loop

potentiates estrogen-regulated transcription from the Xenopus vitellogenin B1 promoter in vitro.

EMBO 12, 423-433.

Shim, E. Y., Woodcock, C.and Zaret, K. S. (1998). Nucleosome positioning by the winged helix

transcription facotr HNF3. Genes Dev. 12, 5-10.

Spear, B. T. (1994). Mouse a-fetoprotein gene 5' regulatory elements are required for postnatal

regulation by raf and Rif. Mol. Cell. Biol. 14, 6497-6505.

Spear, B. T., Longley, T., Moulder, S., Wang, S. L.and Peterson, M. L. (1995). A sensitive lacZ-

127

based expression vector for analyzing transcriptional control elements in eukaryotic cells. DNA

Cell Biol. 14, 635-642.

Sterner, D. E.and Berger, S. L. (2000). Acetylation of histones and transcription-related factors.

Microbiol. Mol. Biol. Rev. 64, 435-459.

Strahl, B. D.and Allis, C. D. (2000). The language of covalent histone modifications. Nature 403,

41-5.

Stroschein, S. L., Bonni, S., Wrana, J. L.and Luo, K. (2001). Smad3 recruits the anaphase-promoting

complex for ubiquitination and degradation of SnoN. Genes Dev. 15, 2822-2836.

Stroschein, S. L., Wang, W., Zhou, S., Zhou, Q.and Luo, K. (1999). Negative feedback regulation of

TGF-beta signaling by the SnoN oncoprotein. Science 286, 771-774.

Stuver, S. O. (1998). Towards global control of liver cancer? Semin. Cancer Biol. 8, 299-306.

Su, Q., Schroder, C. H., Hofmann, W. J., Otto, G., Pichlmayr, R.and Bannasch, P. (1998).

Expression of hepatitis B virus X protein in HBV-infected human livers and hepatocellular

carcinomas. Hepatology 27, 1109-1120.

Sun, Y., Liu, X., Eaton, E., Lane, W. S., Lodish, H. F.and Weinberg, R. A. (1999). Interaction of the

Ski oncoprotein with Smad3 regulates TGF-beta signaling. Mol. Cell 4, 499-509.

Sun, Y., Liu, X., Ng-Eaton, E., Lodish, H. F.and A., W. R. (1999). SnoN and Ski protooncogenes are

repidly degraded in response to transforming growth factor beta signaling. Proc. Natl. Acad. Sci.

USA 96, 12442-12447.

Taketa, K. (1990). Alpha-fetoprotein: reevaluation in hepatology. Hepatology 12, 1420-1432.

Taunton, J., Hassig, C. A.and Schreiber, S. L. (1996). A mammalian histone deacetylase related to a

yeast transcriptional regulator Rpd3. Science 272, 408-411.

128

Thornborrow, E. C.and Manfredi, J. J. (1999). One mechanism for cell type-specific regulation of the

bax promoter by the tumor suppressor p53 is dictated by the p53 response element. J. Biol.

Chem. 274, 33747-33756.

Tilghman, S. M. (1985). The structure and regulation of the alpha-fetoprotein and albumin genes.

Oxf Surv Eukaryot Genes 2, 160-206.

Tilghman, S. M.and Belayew, A. (1982). Transcriptional control of the murine albumin/a-fetoprotein

locus during development. Proc Natl Acad Sci USA 79, 5254-5257.

Truant, R., Antunovic, J., Greenblat, J., Prives, C.and Cromlish, J. (1995). Direct interaction of the

hepatitis B virus HBx protein with p53 leads to inhibition by HBx of p53 response element-

directed transactivation. J Virol 69, 1851-1859.

Ueda, H., Ullrich, S. J., Gangemi, J. D., Kappel, C. A., Ngo, L., Feitelson, M. A.and Jay, G. (1995).

Functional inactivation but not structural mutation of p53 causes liver cancer. Nature Gen. 9, 41-

47.

Vacher, J.and Tilghman, S. M. (1990). Dominant negative regulation of the mouse alpha-fetoprotein

gene in adult liver. Science 250, 1732-5.

Vogelauer, M., Wu, J., Suka, N.and Grunstein, M. (2000). Global histone acetylation and

deacetylation in yeast. Nature 408, 495-498.

Wallrath, L. L., Lu, Q., Granok, H.and Elgin, S. C. R. (1994). Architectural variations of inducible

eukaryotic promoters: preset and remodeling chromatin structures. Bioessays 16, 165-170.

Wang, X.-W.and Xie, H. (1999). Alpha-fetoprotein enhances the proliferation of human hepatoma

cells in vitro. Life Sci 64, 17-23.

Wang, X. W., Forrester, K., Yeh, J., Feitelson, M. A., Gu, J.-R.and Harris, C. C. (1994). Hepatitis B

129

virus X protein inhibits p53 sequence-specific DNA binding, transcriptional activity, and

association with transcription factor ERCC3. Proc. Natl. Acad. Sci. USA 91, 2230-2234.

Wang, X. W., Gibson, M. K., Vermeulen, W., Yeh, H., Forrester, K., Sturzbecher, H. W.,

Hoeijmakers, J. H. J.and Harris, C. C. (1995). Abrogation of p53-induced apoptosis by the

hepatitis B virus X gene. Cancer Res. 55, 6012-6016.

Weinstein, M., Monga, S. P., Liu, Y., Brodie, S. G., Tang, Y., Li, C., Mishra, L.and Deng, C. X.

(2001). Smad proteins and hepatocyte growth factor control parallel regulatory pathways that

converge on beta1-integrin to promote normal liver development. Mol. Cell. Biol. 21, 5122-31.

Weiss, J., Heine, M., Arden, K. C., Korner, B., Pilch, H., Herbst, R. A.and Jung, E. G. (1995).

Mutation and expression of TP53 in malignant melanomas. [Review]. Recent Results in Cancer

Research 139, 137-54.

Wen, P., Crawford, N.and Locker, J. (1993). A promoter-linked coupling region required for

stimulation of alpha-fetoprotein transcription by distant enhancers. Nucleic Acids Research 21,

1911-8.

Wen, P., Groupp, E. R., Buzard, G., Crawford, N.and Locker, J. (1991). Enhancer, Repressor, and

Promoter Specificites Combine to Regulate the Rat a-Fetoprotein Gene. DNA Cell Biol. 10,

525-536.

Wen, P.and Locker, J. (1994). A novel hepatocytic transcription factor that binds the a-fetoprotein

promoter-linked coupling element. Mol. Cell. Biol. 14, 6616-6626.

Wijgerde, M., Grosveld, F.and Fraser, P. (1995). Transcription complex stability and chromatin

dynamics in vivo. Nature 377, 209-213.

Wilcock, D.and Lane, D. P. (1991). Localization of p53, retinoblastoma and host replication proteins

130

at sites of viral replication in herpes-infected cells. Nature 349, 429-431.

Williams, J. S.and Andrisani, O. M. (1995). The hepatitis B virus X protein targets the basic region-

leucine zipper domain of CREB. Proc. Natl. Acad. Sci. USA 92, 3819-3823.

Wolffe, A. P.and Guschin, D. (2000). Review: chromatin structural features and targets that regulate

transcription. J Struct Biol 129, 102-22.

Wong, S.-F.and Lai, L. C. (2001). The role of TGF-B in human cancers. Pahtology 33, 85-92.

Woods, D. B.and Vousden, K. H. (2001). Regulation of p53 function. Exp. Cell Res. 264, 56-66.

Wotton, D., Lo, R. S., Lee, S.and Massague, J. (1999). A smad transcriptional co-repressor.

Yew, P. R.and Berk, A. J. (1992). Inhibition of p53 transactivation required for transformation by

adenovirus early 1B protein. Nature 357, 82-85.

Yoo, Y. D., Ueda, H., Park, K., Flanders, K. C., Lee, Y. I., Jay, G.and Kim, S. J. (1996). Regulation

of transforming growth factor-beta 1 expression by the hepatitis B virus (HBV) X transactivator.

J. Clin. Invest. 97, 388-395.

Zaret, K. S.and Stevens, K. (1995). Expression of a highly unstable and insoluble transcription factor

in escherichia coli: Purification and characterization of the fork head homolog HNF3alpha. Pro.

Expr. Purif. 6, 821-825.

Zawel, L., Dai, J. L., Buckhaults, P., Zhou, S., Kinzler, K. W., Vogelstein, B.and Kern, S. E. (1998).

Human Smad3 and Smad4 are sequence-specific transcription activators. Mol. Cell 1, 611-7.

Zhang, Q., Gutsch, D.and Kenney, S. (1994). Functional and physical interaction between p53 and

BZLF1: implications for epstein-barr virus latency. Mol. Cell. Biol. 14, 1929-1938.

Zheng, G., Teumer, J., Colmenares, C., Richmond, C.and Stavnezer, E. (1997). Identification of a

core functional and structural domain of the v-Ski oncoprotein responsible for both

131

transformation and myogenesis. Oncogene 15, 459-71.

Zilfou, J. T., Hoffman, W. H., Sank, M., George, D. L.and Murphy, M. (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.

132