Differential effects on apoptosis induction in hepatocyte lines by stable expression of hepatitis B virus X protein

Abstract

AIM: Hepatitis B virus protein X (HBx) has been shown to be weakly oncogenic in vitro. The transforming activities of HBx have been linked with the inhibition of several functions of the tumor suppressor p53. We have studied whether HBx may have different effects on p53 depending on the cell type.

METHODS: We used the human hepatoma cell line HepG2 and the immortalized murine hepatocyte line AML12 and analyzed stably transfected clones which expressed physiological amounts of HBx. P53 was induced by UV irradiation.

RESULTS: The p53 induction by UV irradiation was unaffected by stable expression of HBx. However, the expression of the cyclin kinase inhibitor p21^waf/cip/sdi which gets activated by p53 was affected in the HBx transformed cell line AML12-HBx9, but not in HepG2. In AML-HBx9 cells, p21^waf/cip/sdi-protein expression and p21^waf/cip/sdi transcription were deregulated. Furthermore, the process of apoptosis was affected in opposite ways in the two cell lines investigated. While stable expression of HBx enhanced apoptosis induced by UV irradiation in HepG2-cells, apoptosis was decreased in HBx transformed AML12-HBx9. P53 repressed transcription from the HBV enhancer I, when expressed from expression vectors or after induction of endogenous p53 by UV irradiation. Repression by endogenous p53 was partially reversible by stably expressed HBx in both cell lines.

CONCLUSION: Stable expression of HBx leads to deregulation of apoptosis induced by UV irradiation depending on the cell line used. In an immortalized hepatocyte line HBx acted anti-apoptotic whereas expression in a carcinoma derived hepatocyte line HBx enhanced apoptosis.

Key Words: Apoptosis, Hepatitis B virus, Hepatocyte lines

INTRODUCTION

The epidemiological association between chronic HBV infection and the development of hepatocellular carcinoma is well established[1]. However, the molecular mechanism of transformation by HBV is unsolved. Among many potential factors, a viral regulatory protein named Hepatitis B virus protein X (HBx) is believed to contribute to oncogenesis in conjunction with other mechanisms. Stable expression of HBx showed a weak transforming effect in cell lines of liver origin[2-5], in murine fibroblasts[6,7] and in rat fibroblasts[8-12]. Neither the molecular actions set in motion by HBx during the process of transformation, nor the role of HBx for HBV replication are well understood. HBx is dispensable for replication in vitro[13], but under certain conditions HBx can enhance replication in vitro[14-16]. In the woodchuck, a related virus of the same family of viruses, the hepadnaviridae[17], the analog of HBx is essential for the establishment of infection[18,19]. A plethora of in vitro activities and interactions with cellular partners by HBx have been reported[22,23].

In the development of cancer, the delicate balance between cell proliferation and programmed cell death is often disturbed[24]. Several functions of HBx are linked to transformation: HBx transactivates the transcription of cellular proto-oncogenes[25-27], and is able to override cell cycle check points[28,29]. As the oncoproteins of many tumor viruses influence the course of apoptosis in order to transform their target cells[30-32], several reports described the influence of stable and transient transfection of HBV products on apoptosis. Transfection of replication competent constructs of HBV[33] and the closely related hepadnavirus woodchuck hepatitis virus[34] enhanced apoptosis induction. This pro-apoptotic activity of hepadnavirus genomes was dependent on the integrity of the HBx-ORF[33,34]. Several groups described a pro-apoptotic activity of HBx after transient and stable transfection[33,34-48] whereas other groups described anti-apoptotic effects of HBx[49-59].

The cellular tumor suppressor protein p53 is one of the key players in the induction of apoptosis after genotoxic events[60,61]. Thus, besides influencing programmed cell death the transforming proteins of DNA tumor viruses target and inactivate p53 in quite diverse and elaborate ways[31,32]. Several lines of evidence connect HBx with a disturbance of p53 functions. HBx binds to p53 in vitro[62,63], the intracellular distribution of p53 appears to be affected by HBx[4,64,65], p53-dependent and independent DNA repair functions are affected by HBx[66-70]. Furthermore, HBx relieves the repression of transcription by the tumor suppressor p53 in transient reporter gene assays[71-73], and influences the expression of the p53 induced cyclin kinase inhibitor p21^{waf/cip/sdi[74]}.

However, most of the experiments in these reports were done in transient transfection or even cell free in vitro systems. Because the number of HBx molecules in naturally infected cells is assumed to be low[75-77] and transient transfections of HBx expression constructs yield very high amounts of HBx protein, we established cell lines stably expressing low amounts of HBx. Next, we studied the effect of HBx expression on several functions of p53 induced by stimuli that were designed to yield physiological p53 amounts in our cell lines. Using these bona fide natural conditions we were able to find that even low levels of HBx altered or even antagonized p53 functions.

MATERIALS AND METHODS

Cell lines

AML12: The highly differentiated murine hepatocyte line AML12[78] derived from mice transgenic for TGFα can be transformed by HBx[3]. AML12 clones stably transfected with and transformed by HBx and the control cell lines stably transfected with pSV2neo were obtained from N. Fausto. AML12 clones were cultured in Dulbecco’s MEM/Nutrient Mix F12 supplemented with 10% FCS, 0.1 μM Dexamethason, and ITS (Insulin, Transferrin, and Selene, 5 μg/mL each).

HepG2: The differentiated but already transformed human hepatoblastoma line HepG2[79] was cultivated in Mixed Medium/10% FCS.

CAT assay

For CAT assays, the cells per well were transfected with Lipofectamine (Life Technologies, Karlsruhe). Detection of CAT protein was done with the CAT-ELISA Kit (Roche, Mannheim) according to the instructions of the manufacturer.

UV irradiation

All cell lines were plated out one day before irradiation. Directly before irradiation, the medium was removed and the cultures were washed with PBS. UV irradiation was done in the UV-Stratalinker (Stratagene, Leimen, Germany) at 254 nm. The cells were exposed to UV after removing the cover of the dishes.

For induction of apoptosis, 1000 cells were plated in triplicates on culture dishes with a diameter of 95 mm. After irradiation with the indicated energy in the stratalinker, the cell cultures were grown for 5-15 d, until colonies visible by eye were detected. After washing with PBS, the colonies were stained and fixed with 1% crystal violet in 20% methanol for 5 min. After washing with H₂O, the surviving colonies were counted. Each experiment was done in triplicate and repeated at least thrice.

Western blot

For the lysis of cells, 450 μL of ice cold lysis buffer were added to each well of a six-well plate for 30 min on ice. The lysis was based on RIPA-buffer (0.01 mol/L Tris-HCl, pH 7.4; 0.15 mol/L NaCl; 1% (w/v) sodium desoxycholate; 0.1% (w/v) SDS; 1% Triton X-100) supplemented with Leupeptin and Aprotinin. To digest disturbing chromosomal DNA 250 U/mL Benzonase (Novagen, Merck, Darmstadt) was added to the lysis buffer. After the incubation period, the lysate was cleared by centrifugation at 20 000 g for 5 min. The protein content of the lysates was determined using the BCA assay and equal amounts of protein (20-50 μg) were separated on an SDS gel. After blotting and incubation with the appropriate primary and secondary antibodies, the membranes were developed using Chemiluminescence Blotting Substrate from Boehringer Mannheim, Germany according to the manufacturer’s instructions.

Antibodies

The following antibodies were used for the detection of the indicated proteins: murine p53: AB7 (biotinylated polyclonal sheep antibodies from Oncogene Science, via Dianova, Hamburg, Germany); human p53: mAb PAb1801 (Pharmingen, Hamburg, Germany); murine p21: mAb AB4 (OP76 Oncogene, via Dianova, Hamburg, Germany); human p21: p21 (187) (sc-817), St Cruz Biotechnology.

As secondary antibody, we used POD-labeled anti-murine IgG from donkey (Jackson ImmunoResearch, via Dianova, Hamburg, Germany) or POD-strepatividin (Calbiochem, Bad Soden, Germany) for murine p53.

Northern blot

For Northern blot, total cellular RNA was extracted from the cells using TRIzol (Gibco). Twenty micrograms of DNase I digested RNA was separated on a glyoxal agarose gel (1% in 10 mmol/L Na₂HPO₄) and blotted onto positively charged nylon membrane. The blot was hybridized with a ³²P labeled p21-specific probe derived from pCMVp21 (murine).

Ribonucleotide protection assay (RPA)

For RPA, an RNA probe complementary to human p21 was transcribed with Maxiscript, Ambion from the construct pCMVsdi[80] according to the instructions of the manufacturer. As an internal control, we used actin transcribed from pTRI-β-Actin (Ambion, Austin, TX, USA).

The RNA probe was purified by electrophoresis in 5% acrylamide/8 mol/L urea gel. The probe (0.5-2 μg) was labeled with biotin using the BrightStar^TM Psoralen-Biotin kit (Ambion, Austin, TX, USA).

The RPA was performed with the HybSpeed^TM RPA kit from (Ambion, Austin, TX, USA). After digestion of the samples as described in the manual of the kit, the samples were run on 5% PAGE/8 mol/L urea in TBE and blotted onto a positively charged nylon membrane (Amersham, Braunschweig). Signal detection was performed with streptavidin labeled with alkaline phosphatase using the BrightStar^TM BioDetect^TM kit from Ambion, Austin, TX, USA.

RT-PCR

For RT-PCR, 2 μg of total cellular RNA extracted with Qiagen RNeasy was digested with DNase I for 45 min. One microgram of DNase I digested RNA was reverse transcribed with superscript (Gibco). One quarter of the reverse transcribed RNA was amplified with Taq polymerase (1´ 94°C, 1´ 54°C, 1´ 72°C: × 35 cycles) using HBx-specific primers, which had been used for cloning the complete HBx ORF into pRcCMV.

Constructs

pRcCMVX: The complete HBx ORF of an HBV genotype A genome (serotype adw2; isolate 991; EMBL accession no.: X51970) was cloned with the help of PCR into the EcoRI and XbaI-site of pCDNA3, Invitrogen, and controlled by sequencing.

pRcCMVXM: Identical to pRcCMVX except for two stop codons introduced by a G to A exchange of nt 1443 (GAA→TAA)[2] and a C to A exchange of nt 1684 (TCA→TAA). Thus, translation of full-length HBx is stopped by exchange of nt 1443 after 23 aa and expression of presumed small HBx proteins is made impossible by exchange of nt 1684, which introduces a stop directly after the third start codon in frame in the HBx-ORF.

PS9: HBV enhancer I (nt 1040-1374 from EMBL accession no.: X51970) was cloned into pBLCAT3[81].

pC53SN3: Expression vector for wt-p53[82]. P53 is expressed from the CMV enhancer in the vector pCMV.

pcDNA3p21 (murine): Murine p21 was cloned by RT-PCR with primers waf-AS and waf-S[33] from total cellular RNA derived from AML12 cells. After EcoRI digestion, the amplicon was cloned into pCDNA3. Sequencing gave complete identity with murine p21 from GenBank.

RESULTS

Establishing HBx-expressing HepG2-cell lines

HepG2 cells-a completely transformed hepatoma cell line- were stably transfected with the HBx-expression construct pRcCMVX. As a negative control, HepG2 cells were transfected with the isogenic HBx protein expression-negative construct pRcCMVXM. Clones obtained after G418 treatment were analyzed for the expression of HBx-RNA expression by RT-PCR (Figure 1). 11/13 wt-HBx transfected and 11/11 HBxM transfected clones expressed HBx RNA detectable by RT-PCR. While all HBxM transfected clones expressed similar amounts of HBx RNA as judged by RT-PCR, it appeared as if HBx RNA expression in wt-HBx transfected clones varied significantly (Figure 1). This reminds of the findings in stably HBx transfected clones of rat fibroblast origin, where HBx expression was downregulated in most clones[33].

Figure 1 RT-PCR for the detection of HBx-RNA in HepG2 cells stably transfected HBx-expression constructs. Total cellular RNA from clones obtained after stable transfection with expression vector pRCCMVX for wt-HBx (X1, X2, X3, X8, X10, and X18) or for inactive HBx (pRCCMVXM, XM1, XM2, XM6, XM12, XM14, and XM20) was amplified by RT-PCR with primers used for cloning the complete HBx-ORF.

Clone HepG2X8 expressing wt-HBx and control clone HepG2 XM2 transfected with the HBx-expression minus construct pRcCMVXM expressed similar amounts of HBx-RNA and were chosen to be studied in detail.

In comparison, we analyzed the immortalized murine hepatocyte line AML12[78] and the AML12-clone HBx9 transformed by HBx[3].

Analysis of HBx functions in HBx-expressing AML12 and HepG2 clones

One well-characterized function of HBx is its potential to transactivate virtually any promoter or enhancer. We were unable to detect HBx protein in stably transfected AML12 and HepG2 clones (data not shown and Ref. [3])-a problem encountered by many other groups. In order to show the presence of active HBx, we analyzed whether transcriptional activation by HBx could be detected in cell lines stably expressing HBx RNA. Indeed, the expression from HBV enhancer I was approximately sevenfold stronger in the HBx expressing cell line HepG2-X8 than in the control cell line HepG2-XM2 (Figure 2B). In AML12 cell transactivation by HBx was lower (Figure 2A).

Figure 2 HBx activates the HBV-Enhancer I in vivo and can partially counteract a repression by p53 induced by UV irradiation. CAT expression driven by Enhancer I of HBV is indicated as bars. AML12 (A) or HepG2 (B) derived cell lines expressing HBx (HepG2-X8 or AML12-HBx9) are indicated in black and the controls which did not express functional HBx (AML12 or HepG2-XM2) are shown in white bars. The cell lines were transfected with a CAT-expression vector driven by enhancer I of HBV in triplicate. Four hours after transfection, the cultures were irradiated with the indicated dose of UV. The lysis of the cells was done 18 h post irradiation.

HBx acts pro-apoptotic in HepG2- and anti-apoptotic in AML12-cell lines

To analyze the extent of apoptosis quantitatively, we investigated the survival of the cell lines under investigation after UV irradiation. Five hundred cells in triplicate were plated and irradiated with the dose indicated. Surviving colonies were counted after about 14 d. Table 1 shows that stable expression of HBx acts pro-apoptotic in HepG2-derived and anti-apoptotic in AML12-derived cell lines. The pro-apoptotic effect of HBx in HepG2-X8 was most obvious in cell lines irradiated with 20 J/m². For example in one experiment 17.7% ± 8.0 cells of the control cell line HepG2-XM2 survived irradiation with 20 J/m², whereas only 1.1% ± 0.6 of identically treated HBx expressing HepG2-X8 cells were able to grow out to visible colonies. AML12 cells were more sensitive against UV irradiation. Expression of HBx acted anti-apoptotic in the HBx expressing AML12 cell line and enhanced the survival rates to similar levels as in HepG2XM control cells.

Table 1 Survival of HepG2- and AML12-derived HBx-expressing and control cell lines after irradiation with UV.

	0 J/m2	20 J/m2	50 J/m2
HepG2-XM2	100 ± 1.4	17.7 ± 8.0	1.0 ± 0.9
HepG2-X8	100 ± 20.9	1.1 ± 0.6	0.2 ± 0.1
HepG2-XM2	100 ± 11.8	15.8 ± 7.1	1.38 ± 1.5
HepG2-X8	100 ± 20.3	2.7 ± 1.5	0.0
HepG2-XM2	100 ± 15.6	13.5 ± 6.9	3.7 ± 0.8
HepG2-X8	100 ± 7.4	6.8 ± 0.5	1.7 ± 1.2
AML12	100 ± 5.1	4.3 ± 0.6	1.1 ± 1.0
AML12-HBx9	100 ± 14.2	13.5 ± 8.9	6.0 ± 2.2
AML12	100 ± 27.2	17.0 ± 6.3	0.6 ± 0.04
AML12-HBx9	100 ± 3.8	28.0 ± 7.0	6.8 ± 2.4
AML12	100 ± 4.4	11.0 ± 4.1	0.4 ± 0.5
AML12-HBx9	100 ± 4.6	17 ± 5.1	3.4 ± 1.8

The percentage of cells surviving UV-treatment of three independent experiments. For higher reproducibility the HBx-expressing and the control cell lines were irradiated in one batch.

No influence of HBx on induction of p53 in HepG2- and in AML12-cell lines

p53 is a central protein in the induction of apoptosis[83]. After many sublethal stimuli, a complex set of events takes place that results in an increased expression of p53 in the nucleus. Thus, we analyzed by immune blots of whole cell extracts whether HBx was able to modulate expression of p53 after UV irradiation of the cell lines. Irradiation with increasing doses (0-50 J/m²) of UV led to enhanced expression of p53 in control and HBx expressing cell lines (Figures 3 and 4). HBx had no effect on the induction of p53 in cell lines of AML12 and HepG2 origin. p53 expression was easily detectable after irradiation with 20 J/m² and maximal expression was reached after irradiation with 40-50 J/m² in the HepG2 and AML12 cell lines investigated (Figures 3 and 4). These data are in accordance with reports on p53 expression in HepG2-cells[84], human[85], and murine[85,86] fibroblasts and primary rat keratinocytes[87].

Figure 3 Induction of p53- and p21^waf/cip/sdi-protein expression in HepG2-derived cell lines after irradiation with UV by immune blot. The dose of UV used for irradiation is indicated. The cells were lysed 18 h after irradiation with UV.

Figure 4 Induction of p53- and p21^waf/cip/sdi-protein expression in AML12-derived cell lines after irradiation with UV by immune blot. The dose of UV used for irradiation is indicated. The cells were lysed 18 h after irradiation with UV.

We also analyzed the kinetics of p53 induction after UV irradiation with 40 J/m². p53 levels increased after 1 h to reach a maximum 18 h after irradiation. Thereafter, p53 expression decreased (Figure 5). Again, HBx had no influence on the kinetics of p53 induction after UV irradiation.

Figure 5 Kinetics of induced p53- and p21^waf/cip/sdi-protein expression after irradiation with UV by immune blot in AML12 cell lines. The time in hours after irradiation with 40 J/m² UV is indicated.

HBx counteracts a repression by UV irradiation in vivo

We next asked whether HBx may influence functions of p53. One extensively investigated activity of p53 is the ability to repress the transcription of many genes[88]. Thus, we analyzed whether expression of the reporter gene CAT from HBV enhancer I is repressed by p53 - or other factors activated after induction of programmed cell death - in vivo. Irradiation of HepG2 and AML12-cell lines with increasing doses of UV caused indeed a strong repression of expression from HBV enhancer I (Figures 2A and B). The repression was highest at 40 J/m² (Figures 2A and B), the UV dose leading to maximum expression of p53 (Figures 2 and 3). However, a strong reduction of CAT-repression could already be observed at the lowest UV dose of 10 J/m² administered. Interestingly, in the tumor cell line HepG2 (Figure 2B), UV irradiation caused a much stronger repression of expression from enhancer I than in AML12 cells (Figure 2A).

In HBx expressing cell lines, the UV-induced repression was partially counteracted. This anti-repressive activity of HBx was strongest in AML12-HBx9 and still significant in HepG2 cells (Figures 2A and B). In the AML-cell line HBx9 irradiated with 40 J/m² the CAT-expression was 4.9-fold higher than in the parental cell line AML12 without HBx. Thus, HBx seems to be able to alleviate transcriptional repression by p53.

Induction of p53 and p21^waf/cip/sdi by UV irradiation

Because we observed that HBx influenced the extent of transcriptional repression by p53 we also wanted to analyze if HBx had any influence on transcriptional activation by p53. After UV irradiation, p53 is stabilized[89] and acts as a transcription factor by binding to the promoter of certain target genes leading to enhanced transcription from these promoters, for example of the cyclin kinase inhibitor p21^{waf/cip/sdi[90]}. Thus, after UV irradiation, the expression of p21^waf/cip/sdi is enhanced[86,91]. We therefore analyzed if stable expression of HBx influenced p53’s transcriptional activation of the endogenous p21^waf/cip/sdi-promoter. UV irradiation of the cell lines with increasing doses of UV led to an increase of p21^waf/cip/sdi protein expression (Figures 3 and 4). As reported by other groups, irradiation with high doses of 40-50 J/m² UV again caused a decline of the induced p21^{waf/cip/sdiwaf} protein expression[86,91]. An analysis of the temporal regulation of p21^{waf/cip/sdiwaf} protein expression (Figure 5) after UV irradiation in AML12 cells showed an undulation of p21^{waf/cip/sdiwaf} protein expression in the parental cell line and the HBx transformed clone 9. A similar phenomenon was described in the mouse liver after 70% partial hepatectomy[92]. However, these authors analyzed RNA expression and different time points. Interestingly, stable expression of HBx in HepG2 led to a lower expression of p21^waf/cip/sdi-protein. However, in the HBx transformed cell line AML12-HBx9, the level of p21^waf/cip/sdi-protein expression after UV irradiation changed from experiment to experiment and no reproducible data for p21 could be obtained, although p53-expression from identical blots (upper lane) gave reproducible results (Figures 4 and 5). It appeared as if the expression of p21^waf/cip/sdi-protein was deregulated in a seemingly random way by HBx because p21^waf/cip/sdi-protein expression was reproducible in the parental cell line (Figures 4 and 5).

Butz et al had described an uncoupling of the regularly inducible p21^waf/cip/sdi-RNA expression from the deregulated p21^waf/cip/sdi-protein expression upon genotoxic stress in some cell lines[109]. Thus, we analyzed the induction of p21^waf/cip/sdi-RNA in the HBx transfected cell lines. Indeed, analysis of p21^waf/cip/sdi-RNA expression in the hepatocyte lines under investigation gave reproducible results. Irradiation with UV increased the expression of p21^waf/cip/sdi-RNA in the HBx-expressing cell lines HepG2-X8 and the control cell lines HepG2-XM2 and AML12 to reach a peak at 20-30 J/m² to decrease thereafter (Figures 6 and 7). Analysis of p21^waf/cip/sdi-RNA expression in the HBx expressing clone AML12-HBx9 - in contrast to analysis of p21^waf/cip/sdi-protein expression - gave reproducible results. However, no decline of p21^waf/cip/sdi-RNA expression was observed at higher doses of UV irradiation as with the other three cell lines under investigation (Figure 7). Quite in contrast, in the HBx-expressing clone AML12-HBx9 the amount of p21^waf/cip/sdi-RNA increased continuously even at the highest dose of UV irradiation of 50 UV J/m². Thus, HBx strongly influenced the regulation of p21^waf/cip/sdi expression in the HBx-transformed cell line AML12-HBx9.

Figure 6 Detection of p21^waf/cip/sdi-RNA by ribonuclease protection assay in the cell line HepG2-XM2 (A) and -HBx8 (B). Total cellular RNA was isolated from cell cultures irradiated with the indicated dose of UV 18 h after irradiation. As internal control actin was detected by RPA.

Figure 7 Detection of p21^waf/cip/sdi-RNA by northern blot in the cell line AML12 (A) and the HBx-transformed clone AML12-HBx9 (B). Total cellular RNA was isolated from cell cultures irradiated with the indicated dose of UV 18 h after irradiation.

CONCLUSION

One of the most important functions of p53 is the induction of programmed cell death in order to maintain the integrity of the genome by inhibiting the survival of cells with damaged DNA[93]. To analyze the integrity of p53 function in cell lines stably expressing HBx RNA, we investigated several p53 activities during induction of apoptosis by UV. Irradiation with 20 J/m² quite uniformly induced massive cell death in all cell lines analyzed. The cell death was due to apoptosis because we observed nucleosomal DNA fragmentation in all cell lines under investigation (data not shown).

However, we observed that induction of apoptosis was enhanced in HBx expressing cell lines of HepG2-origin and reduced in HBx expressing cell lines of AML12-origin. These opposing effects of HBx are in agreement with the published results on the influence of HBx on programmed cell death. Several groups described a pro-apoptotic activity of HBx after transient and stable transfection[33,34-48] whereas other groups described anti-apoptotic effects of HBx[49-59].

We thus set out to study key events in the induction of apoptosis after UV irradiation. We investigated whether HBx had an effect on the induction of p53 in our cell lines. No evidence was found that HBx influenced the induction of p53 by UV. Neither p53’s expression by stabilization of p53 as reported for adenovirus E1A or the large T-Antigen of SV40[94] was increased, nor was p53’s expression decreased in the presence of HBx as described for E6 of the papilloma viruses[95]. Our results agree with reports from Refs. [85] and [97] who also did not find altered p53 induction or expression in HBV expressing HepG2 cells.

A widely accepted hypothesis assumes that a higher expression of the cyclin kinase inhibitor p21^waf/cip/sdi after DNA damage gives an advantage for survival because cells arrest in G₁ and gain time to repair DNA damage instead of succumbing to apoptosis[88]. DNA damage induced in cell lines with a targeted deletion of the p21^waf/cip/sdi-gene led to programmed cell death, whereas the wt-cell line survived in cell cycle arrest[97]. After induction, p53 binds as a transcription factor to the promoter of p21^waf/cip/sdi and increases its transcription[90]. Our results on induction of p21^waf/cip/sdi-protein and RNA by UV irradiation in the control cell lines (AML12 and HepG2-XM2) are in accordance with data from literature[86,87,91]. Interestingly, the HBx-expressing clone HepG2-clone X8 expressed less p21^waf/cip/sdi than the control. A decrease as well as an increase of p21^waf/cip/sdi expression in response to HBV infection has been described in literature. In microarrays, an enhanced expression of p21^waf/cip/sdi was seen in Hep3B and HepG2 lines stably expressing HBx[96,98] and in the HBV expressing cell line HepG2.2.15[99], decreased expression of p21^waf/cip/sdi was seen in HBV associated HCC[100], in transiently transfected primary human hepatocytes[101] and in two liver samples from patients with chronic HBV infection[101].

The reason for the differential effect of HBx on the expression of p21^waf/cip/sdi in AML12 and HepG2 cell lines is unknown. However, transduction of HBx into primary human hepatocytes and stably transformed hepatocyte lines had reciprocal effects on the expression of several target genes[101]. It appears possible that stable expression of HBx in the immortalized AML12 cell line mimics the situation found in acutely infected liver with deregulated and in some circumstances enhanced p21^waf/cip/sdi expression[100], whereas stable expression of HBx in hepatoma cell line HepG2 with slightly decreased expression of p21^waf/cip/sdi mirrors the situation found in HBV associated HCC[100], i.e. in the endpoint of hepatocarcinogenesis.

It appears possible that HBx transactivates the enhancer of p21^waf/cip/sdi in AML12 cells directly, as has been reported for other cell lines[74,96]. In addition, HBx may enhance the physiological activation of p21^waf/cip/sdi by p53. Haviv et al[102] have shown that HBx can drastically enhance p53 mediated transcriptional activation. In consequence, HBx expressing AML12 clones would express more p21^waf/cip/sdi than controls. Furthermore, HBx may have a stronger effect on the activation of the transcription factor ets in AML12 cells than in HepG2 cells, as ets has been suggested to play a crucial role in the transactivation of the p21^waf/cip/sdi enhancer by HBx[96]. These assumptions would be in agreement with the higher and deregulated expression of p21^waf/cip/sdi-RNA in the HBx expressing AML12 cell line. An enhanced expression of p21^waf/cip/sdi in the HBx transformed cell line AML12-HBx9 reminds of findings in HTLV-1 transformed cell lines[103,104]. Tax, the transforming protein of HTLV-1 virus[105] has many functional similarities with HBx[106]. The oncoprotein E7 of the papilloma viruses also enhances the expression of p21^{waf/cip/sdi[107,108]}. As described for otherwise unrelated cell lines[109], we also observed that the expression of p21^waf/cip/sdi-protein was deregulated and did not seem to correlate with p21^waf/cip/sdi-RNA-expression in HBx-expressing cell line AML12-HBx9 (Figures 4, 5 and 7).

Our results indicate that HBx is able to partially relieve repression of expression from the HBV enhancer I by physiological amounts of p53 as observed after induction by UV irradiation (Figure 2). Other groups reported complete relief of transcriptional repression by p53 after transient transfection of HBx[71] or HBV-dimers[72] or in in vitro transcription assays using in vitro translated p53 and HBx proteins[73]. However, transient transfection of HBx expression constructs leads to high expression of HBx, whereas stable expression of HBx as in our experiments only leads to low level expression of HBx that may be more comparable to the natural situation in tissue of chronically infected individuals where detection of HBx is rather difficult because of its low level of expression[75-77]. Moreover, in transient assays using expression vectors for p53 and HBx, a complete relief of transcriptional repression by p53 was only seen when 100-fold more HBx expression vector than p53 was used[71]. When equal amounts of expression constructs were used, HBx was not able to relieve transcriptional repression by p53[71,110]. Our results indicate that stable expression of HBx in hepatocyte lines antagonizes functions of physiological doses of p53 induced by UV and confirm the results of transient transfections[71,72].

In summary, we found several p53 linked activities that were altered in cell lines stably expressing HBx. By use of conditions which closely match the natural situation, we were able to show that HBx is able to alleviate some of the p53 induced effects taking place after the induction of apoptosis by UV irradiation. The differences we observe between the HBx expressing immortalized hepatocyte line AML12 and the hepatoma cell line HepG2 are due to a differential activation of intracellular signal transduction pathways - as found in inducible HBx expressing AML12-clones[5,111] - and will be the subject of further investigations.

ACKNOWLEDGMENTS

This work was performed in partial fulfillment of the requirements for the Ph.D. thesis of N.F. and for the MD. thesis of E.Q.

We thank Dr. N. Fausto for the AML12 and AML12-HB9 cell line. We thank Dr. J. R. Smith, Baylor College of Medicine for the gift of the construct pCMVsdi and Dr. B. Vogelstein for pC53SN3.