Cancer Cell Biology
Regulation of P53 stability in p53 mutated human and mouse hepatoma cells
Article first published online: 4 JAN 2007
Copyright © 2006 Wiley-Liss, Inc.
International Journal of Cancer
Volume 120, Issue 7, pages 1459–1464, 1 April 2007
How to Cite
Hailfinger, S., Jaworski, M., Marx-Stoelting, P., Wanke, I. and Schwarz, M. (2007), Regulation of P53 stability in p53 mutated human and mouse hepatoma cells. Int. J. Cancer, 120: 1459–1464. doi: 10.1002/ijc.22519
- Issue published online: 30 JAN 2007
- Article first published online: 4 JAN 2007
- Manuscript Accepted: 10 NOV 2006
- Manuscript Received: 20 DEC 2005
- p53 mutation;
- hepatoma cells;
- UV irradiation;
The tumor suppressor p53 is frequently mutated in cancer. We have investigated the regulation of P53 in p53 wild type mouse hepatoma cells (line 55.1c), in p53 heterozygeously mutated cells (56.1b) and in p53 defective cells (lines 56.1d, 70.4 and HUH7) under various experimental settings. The basal levels of P53 were low in 55.1c cells, but nuclear accumulation occurred upon UV-irradiation. Similarly, UV-exposure induced stabilization of P53 in the heterozygeously p53 mutated 56.1b hepatoma cells. By contrast, the 3 hepatoma lines, which lack transcriptionally active P53, demonstrated high basal nuclear concentrations of P53 protein and, unexpectedly, showed loss of P53 upon UV-irradiation. Expression of p53 mRNA was also decreased in p53 defective cells after 24 hr post UV-irradiation, which may be linked to induction of apoptosis of the irradiated cells under these conditions. Other stressors like H2O2 also mediated a decrease in P53 concentration in p53 defective cells. This effect occurred at very low concentrations and was already detectable 1–2 hr after exposure of cells. There were no signs of apoptosis of H2O2-exposed cells at this time point and no significant changes in p53 mRNA or MDM2 level. These unexpected findings indicate a new aspect related to regulation of P53 stability in cells with a defect in the tumor suppressor protein. © 2006 Wiley-Liss, Inc.
The p53 tumor suppressor protein is a powerful cellular caretaker, which protects cells from malignant transformation by means of transcriptional upregulation of proapoptotic, DNA repair and cell cycle arrest related proteins. In unstressed cells the P53 level is kept low by the E3 ubiquitin ligase MDM2, which transfers activated ubiquitin residues to P53.1 Polyubiquitination of P53 leads to a rapid proteasomal degradation of the tumor suppressor, preventing P53 from affecting the cells' viability. P53 itself transcriptionally upregulates MDM2 expression, forming a negative feedback loop.
Stress signals, caused by UV-light, DNA-damaging agents or hypoxia lead to an activation of several kinases, of which the most prominent members are the ATM (ataxia–telangiectasia mutated), ATR (ataxia–telangiectasia and Rad3-related), DNA-dependent protein kinase and casein kinase II, which phosphorylate the P53 protein at different sites (for review see Ref.2). Polyphosphorylation at the N-terminus of P53 reduces MDM2-dependent degradation, leading to an accumulation of the P53 protein in the cell followed by upregulation of apoptosis or cell cycle arrest related proteins. Exposure of cells to UV light correlates with post-translational modifications of P53, such as phosphorylation of serine 15 (mouse serine 18) and serine 20 (mouse serine 23). Desferrioxamine (DFX) treatment, which mimics hypoxia, results in phosphorylation of serine 15. Both sites are known to reduce MDM2-binding when phosphorylated.3
Human hepatocellular carcinomas often exhibit a defect in the p53 signaling pathway resulting from single base substitutions in p53 occurring with a prevalence of ∼25% (see IARC TP53 mutation database4). The mutations affect almost exclusively the DNA-binding domain of the P53 protein, often resulting in a lowered affinity to its sequence-specific DNA binding sites (for a review see Ref.5). This disables its transcriptional activity to induce MDM2-expression, which finally results in elevated levels of mutated P53 in many tumors cells.6 In contrast to human hepatocellular cancers mouse liver tumors are generally devoid of p53 mutations. However, mutations in the tumor suppressor gene become frequently apparent upon brief in vitro culturing of tumor cells, and inactivation of both p53 alleles has been observed in certain mouse hepatoma cell lines.7
In this study, we investigated the stability of P53 in stressed p53 wild-type and p53 defective hepatoma cells, and detected P53 degradation in cell lines with defective P53. This finding was unexpected and may have implications for tumor therapy.
Material and methods
The mouse hepatoma cell lines 55.1c, 56.1b, 70.4, 56.1d, and the human line HUH7 were cultured in Dulbecco's modified Eagle's medium supplemented with heat-inactivated 10% FCS (fetal calf serum), 100 U/ml penicillin and 100 μg/ml streptomycin. Cells were grown in a humidified incubator containing 5% CO2 at 37°C. The mouse hepatoma lines have been described in detail in Ref. 7; their p53 genotyope is as follows: 55.1c, wildtype; 56.1b (R277T/wildtype), 70.4 (A135P/C138W) and 56.1d (C132W/del). In addition, HUH7 (Y220C/del) were used. If indicated, cells were UV-irradiated with doses between 1 and 100 mJ/cm2, using a UV-crosslinker (Biometra, Göttingen, Germany).
Cells were washed in ice-cold phosphate-buffered saline (PBS), scraped into PBS, centrifuged at 1500g for 2 min and lysed in a 50 mM Hepes, 150 mM NaCl and 1% Triton X-100 buffer, supplemented with a protease inhibitor cocktail (Complete Mini, Roche, München, Germany). The samples were incubated for 30 min on ice. After centrifugation of the lysate (10 min at 16,000g and 4°C), the protein content of the supernatants was determined by the Bradford protein assay (Biorad, München, Germany). Loading buffer (125 mM Tris/HCl, 10% 2-mercaptoethanol, 4% sodium dodecyl sulfate, 20% glycerol, 0.004% bromophenol blue, pH 6.8) was added to 100 μg of samples, which were incubated for 5 min at 95°C and then separated on a 10% polyacrylamide gel. After protein transfer to a polyvinylidene difluoride membrane (Immobilon, Millipore, Schwalbach, Germany), the membrane was blocked for 1.5 h, with either I-Block (Tropix, PE Applied Biosystems, Weiterstadt, Germany) for P53-detection, or 5% dry milk solution for MDM2. The following primary antibodies dilutions were used: anti-P53 (Ab-1, clone 421, Dianova, Hamburg, Germany) 1:1,000, anti-P53 (Ab-3, clone 240, Dianova) 1:1,000, anti-MDM2 (SMP14, Santa Cruz, Heidelberg, Germany) 1:500 and anti-GAPDH (Chemicon, Ochsenhausen, Germany) 1:1,000. Antibody binding was visualized after incubation of the membranes with alkaline phosphatase coupled secondary antibodies (goat-antimouse, Tropix, 1:10,000) and CDP-Star as substrate (Tropix).
Cells were grown in 3.5 cm culture plates. At a density of 70–80% they were UV-irradiated (1 mJ/cm2) and fixed 24 hr later with acetone:methanol (1:1, 3 min, 4°C). Fixed cells were incubated with normal goat serum (1:20 dilution in PBS) for 20 min followed by incubation with primary antibody (anti p53 Ab-1, clone 421, Dianova) at a dilution of 1:50 for 1 h. After subsequent washes with PBS, cells were incubated with biotinylated antimouse antibody (Biospa, Wedel, Germany; 1:200 in PBS containing 1% albumin and 0.35 M NaCl) for 1 hr and subsequently exposed to streptavidin/alkaline phosphatase complex (Biospa, 1:200 dilution) for 30 min. For staining, Fast Red (Kementec, Copenhagen, Denmark) was used as substrate.
Cells were washed in ice-cold PBS, scraped into PBS, centrifuged at 2,000g for 2 min and lysed in RIPA buffer (Sigma, München, Germany), supplemented with a protease inhibitor cocktail (Complete Mini, Roche). The samples were vortexed vigorously 5 times and incubated for 30 min on ice. After centrifugation of the lysate (5 min at 2,000g and 4°C), the protein content of the supernatants was determined with a ND-1000 NanoDrop-Spectrophotometer (peqLab, Erlangen, Germany) at 280 nm. One milligram of protein was added to 500 μl of lysis buffer and incubated with 2 μg of mouse anti-p53 Ab-1 or Ab-3 (Dianova) for 3 hr. Forty μl of protein-G-agarose (Roche) were added and rotated overnight at 4°C. The solution was centrifuged at 1,000g for 1 min at 4°C and the resulting pellet was washed with ice-cold lysis buffer. The centrifugation/washing was repeated 3 times before adding 50 μl of loading buffer, heating samples to 95°C for 10 min and subsequent analysis by SDS-PAGE and Western blotting. For P53 detection a rabbit polyclonal anti-P53 antiserum (CM-1, Dianova) was used at 1:2,000 dilution.
Quantitative real-time RT-PCR
Total cell RNA was isolated by phenol/chloroform extraction using TRIzol (Invitrogen, Paisley, UK) and quantified photometrically at 260 nm. Synthesis of cDNA was performed at 42°C for 60 min with the following reaction conditions: 500 ng of denaturated RNA (5 min at 70°C), 5 mM MgCl2, 100 μM dNTPs (MBI Fermentas, St.Leon-Rot, Germany), 250 ng oligo(dT)15 (Roche), 250 ng random Hexamer primers (Roche) and 12.5 U of AMV Reverse Transcriptase (Promega, Mannheim, Germany). The samples were subsequently heated to 95° for 5 min to terminate the RT reaction. p53 cDNA levels were quantified using the FastStart DNA Master SYBR Green I Kit (Roche) on a LightCycler. The PCR mix contained 3 mM MgCl2 and 0.5 μM of following primer: p53: (Sense: 5′-GGAGACATTTTCAGGCTTATGG-3′; antisense: 5′-AGAAGGGACAAAAGATGACAGG-3′). Denaturation was performed by heating to 95°C for 5 sec followed by annealing at 58°C (p53) or 62°C (Mdm2) for 5 sec and elongation at 72°C for 8 sec (p53). The relative expression ratio was calculated using the crossing points, which were adjusted to 18S RNA (Sense: 5′-CGGCTACCACATCCAAGGAA-3′; antisense: 5′-GCTGGAATTACCGCGGCT-3′).8
P53 levels are high in p53 mutated cell lines and decrease after UV-treatment
Under normal conditions the tumor suppressor protein P53 is kept at very low levels in unstressed p53 wildtype cells, as exemplified by line 55.1c (Fig. 1a). After UV-treatment P53 accumulates in the nucleus (Fig. 1b) and leads to the onset of apoptosis in these cells. In p53 mutated cell lines, however, the P53 protein is frequently detected at constitutively high levels. Intense nuclear staining of P53 was seen in mouse 70.4 (p53 mut/mut), mouse 56.1d (p53 del/mut) and human HUH7 (p53 del/mut) hepatoma cells upon immunohistochemical staining (Figs. 1c, 1e, and 1g). Previous studies using a P53 luciferase-reporter have shown that the transcriptional activity of P53 in the mouse cell lines with defective P53 is strongly impaired.9, 10 Upon treatment of the p53 defective cells with UV-light, we observed a nearly complete disappearance of P53 protein 24 hr after irradiation (Figs. 1d, 1f and 1h). This finding was substantiated by the results of subsequent western blotting analyses using different antibodies against P53. In unstressed p53 wildtype cells (55.1c) the protein level was below the detection limit of our analysis, which was performed without preceding immunoprecipitation. After treatment with very low UV doses (1 mJ/cm2), however, the tumor suppressor protein became detectable (Fig. 2). All investigated cell lines with transcriptionally inactive P53 showed a decrease in P53 protein after UV-irradiation, yet the sensitivity of the cells towards UV-light differed considerably. While an almost complete disappearance of P53 was observed in 70.4 cells at 24 hr after treatment with 1 mJ/cm2, the same dose and time interval were not sufficient to see such an effect in HUH7 cells, where only a moderate reduction was obtained at the high dose of 100 mJ/cm2 (Fig. 2). Since we detected a decrease in P53 staining by 3 different P53-antibodies, Clone PAb 421, PAb 240 and DO-1, which recognize different and nonoverlapping epitopes on P53, it is very unlikely that this effect results from masking of an epitope by post-translational modification of P53. It is therefore much more likely that UV-irradiation induced a decrease in P53 protein in cells with mutated p53.
Decrease of P53 levels after UV-treatment is time and dose dependent but independent of MDM2
The kinetics of UV-irradiation-induced changes in P53 levels were also studied (Fig. 3). In p53 wild-type 55.1c cells, P53 protein became detectable 6 hr after UV-treatment; maximal induction was reached after 12 hr and the P53 levels stayed high for at least 24 hr after irradiation. We also detected an increase in MDM2 in these cells, which paralleled that of P53 but already reached a maximum level at 6 hr. In contrast, there was no accumulation of P53 protein in the p53 mutated lines 70.4, 56.1d and HUH7 at 6 hr post irradiation. The P53 level in the mouse hepatoma cell lines remained constant for the first 12 hr, while a large proportion of the P53 protein had disappeared in these cells 24 hr after irradiation. In the human HUH7 line, which was irradiated with a 100-fold higher dose in comparison to the mouse cells, a decrease in P53 was already detectable at 6–12 hr after UV treatment. MDM2 levels were much lower in the treated p53-mutated lines than in the treated p53 wild-type cell line, and there was no detectable change in the concentration of the protein following UV-irradiation of the cells.
UV-irradiation induces stabilization of P53 in hepatoma cells with heterozygous mutation of p53
To investigate the effect of UV-exposure on hepatoma cells with a heterozygous mutation of p53, we used 56.1b mouse hepatoma cells, which harbor both a wild-type and a mutated allele of the tumor suppressor gene and show some residual transcriptional activity of P53.9 Western blot analysis of P53 (Fig. 4a) clearly indicated that the concentration of total P53 increases in 56.1b cells 6–24 hr after UV-irradiation, as seen in p53 wildtype cells. To discriminate between the wild-type and mutant forms of P53 we used the monoclonal antibodies Ab-1 (clone 421) and Ab-3 (clone 240), which detect the mutation- and wild-type-specific conformations of P53, respectively. Upon immunopreciptitation of P53 with either of the 2 antibodies, P53 was detected to be increased in concentration upon UV-irradiation (Fig. 4b). This indicates that the mutant form of P53 is stabilized in irradiated cells that also express the wild-type protein in contrast to cells that only carry mutant p53 alleles.
Other stressors also mediate P53 degradation in p53 mutated hepatoma cells
The stability of P53 in p53-wild-type and p53-defective cells was also studied after treatment of cells with desferrioxamine (DFX), which mimics hypoxia and H2O2, to expose the cells to reactive oxygen species. Lack of oxygen is probably the most physiological inducer of P53, but the mechanisms are not yet entirely understood. Interaction of P53 with the hypoxia-inducible factor 1α11 or hypoxia-induced phosphorylation of loci at the N-terminus of P533 might be involved in physiological stabilization. In accordance with An et al.11 we observed maximal P53 accumulation 6 hr after start of DFX treatment in p53 wild-type hepatoma cells, while the effect vanished at 24 hr after treatment (Fig. 5). The p53 mutated cell lines, by contrast, did not exhibit any alteration in P53 concentration 6 hr after DFX-treatment (Fig. 5), although DFX induced HIF-1α expression (data not shown). Interestingly, however, a decrease in P53 level was observed in 70.4 and 56.1d cells 24 hr after DFX-treatment (Fig. 5), suggesting that the effect on P53 degradation is not restricted to UV-irradiation. The human HUH7 cell line showed no alteration in P53 level after treatment of cells with DFX at a concentration of 300 μM.
We next investigated the effect of DMSO and H2O2, both of which may induce a conformational change in P53, which could potentially result in degradation of the protein in the p53-mutated cells. The conformation of P53 is redox-regulated12 and may be affected by reactive oxygen species.13 Cells of the p53 double-mutated line 70.4 were incubated with DMSO and H2O2, and the levels of P53 protein were investigated by western blot analysis. In fact, both DMSO and H2O2 lowered the level of P53, and this effect could be entirely blocked by MG-132 (Fig. 6a). The effect was very fast and clearly detectable within 1–2 hr after addition of the chemicals to the medium, being in striking contrast to the corresponding effects produced by UV-light, which occurred much later. HUH7 cells appear to be more resistant to oxidative stress, since no comparable effects could be observed. Noticeably, the concentrations of H2O2 (500 μM) and DMSO (1:1,000) used were not sufficient to induce a detectable accumulation of P53 in 55.1c p53 wild-type cells within the time window of 2 hr. Much lower concentrations of H2O2 were also tested and found to lead to a lowering in P53 in 70.4 cells (Fig. 6b). We also observed a decrease in P53 level following γ-irradiation (8 Gy, 24 hr) of 70.4 cells (data not shown), which led us to conclude that the phenomenon of P53 degradation in p53 defective murine cell lines is not limited to UV-treatment.
The mRNA of p53 is down-regulated in UV-treated 70.4 cells but not in 55.1c cells
The observed decrease in P53 in the UV-irradiated p53-mutated cell lines could be the result of either transcriptional down-regulation or protein degradation. We therefore comparatively analyzed the mRNA levels of p53 by quantitative real time PCR in UV-irradiated 70.4 and 55.1c cells. At 24 hr after UV-irradiation of 70.4 cells there was an ∼12-fold decrease in p53 mRNA level, which was highly significant (p < 0.001, Student's t test, n = 7). By contrast, there was a tendency for an increase rather than a decrease in p53 mRNA in UV-exposed 55.1c cells, but the effect was not statistically significant (n = 4, data not shown). Both 70.4 and 55.1c cells are very similarly triggered into apoptosis by UV-irradiation9 and this effect was also observed under the conditions used in the present series of experiments 24 hr post irradiation. Induction of apoptosis was, among others, indicated by a strong increase in the activity of the executor caspase 3, which is activated in the so-called ‘post-mitochondrial’ phase, about 16–24 hr after UV-irradiation of 70.4 and 55.1c cells (data not shown). Induction of apoptosis in p53 wild-type 55.1c cells therefore does not result in a lowering of their p53 mRNA, in contrast to what happens in p53 mutated 70.4 cells. It is also noteworthy, that treatment of 70.4 cells with H2O2 (at concentrations of 10 and 100 μM for 1 hr, n = 7) did not lead to a any significant diminution of p53 mRNA, while a decrease in P53 protein was already detected after 1–2 hr in the H2O2-exposed cells, without any visible signs of apoptosis.
P53 is one of the most extensively studied proteins today. There are hundreds of reports describing an accumulation of the protein in p53 wildtype cells in response to exposure to stressors, such as UV-light, DNA-damaging agents, γ-radiation, hypoxia or reactive oxygen species generating chemicals. We have developed in our laboratory a series of different mouse hepatoma lines, which were molecularly characterized with respect to their p53 genotype.7 Cells of the p53 wild-type line 55.1c, used in the present study, respond typically to UV-irradiation with an increase in P53 and nuclear accumulation of the protein. The accumulation of P53 in these cells is believed to result from a decrease in interaction of the tumor suppressor protein with its negative regulator MDM2, which is mainly caused by post-translational modifications of P53 (for a review see Ref.2). By contrast, cells of hepatoma lines 70.4, 56.1d and HUH7, which are devoid of functional P53, did not respond with an increase in P53 level to UV-irradiation, in accordance with observations reported by others using comparable cell systems.6 There are several potential explanations for the aberrant P53 response in the 3 p53 mutated hepatoma lines investigated: (i) since P53 is transcriptionally inactive, the negative regulatory loop with MDM2 may be interrupted; (ii) the mechanisms underlying DNA damage-dependent P53 stabilization may be defective; (iii) an increase may not be detectable in cells with constitutively high levels of P53; (iv) stabilization of mutant P53 by binding to heat shock protein 90 (Hsp90), may play some role as has been shown in other systems (for example see Ref.14, 15, 16).
While the P53 response to irradiation in p53 wild-type cells has been repeatedly described by many groups, we now observed that cells with defective P53, which accumulate high concentrations of P53 protein in their nucleus, tend to degrade the protein upon UV-irradiation or exposure to other cellular stressors. In principle, this effect could result from transcriptional down-regulation of the gene, protein destabilization caused by elevated activities of the ubiquitin-proteasome machinery, or other unknown processes. Transcriptional down-regulation of p53 mediated by an autoregulatory loop has been described.17 We did indeed detect a decrease in p53 mRNA in UV-exposed 70.4 cells, but this effect is probably unrelated to an autoregulatory mechanism, since cells of line 70.4 lack functional P53. The observed decrease in p53 mRNA, however, could well serve as an explanation for the lowered levels of P53 protein in UV-exposed 70.4 cells. Degradation of P53 occurred within 12–24 hr after UV irradiation of mouse hepatoma cells and with similar kinetics after treatment of the p53 mutated cells with DFX. Both UV-light and DFX induce apoptosis of mouse hepatoma cells, an effect which is independent of functional P53.9 The decrease in P53 in UV-exposed 70.4 cells may therefore be linked to induction of apoptosis in the irradiated cells. Apoptotic cells may slow down RNA synthesis, which would lead to a disappearance of short-lived proteins in the cells. There are arguments, however, against this simple explanation for the observed disappearance of P53 in 70.4 cells: (i) the levels of 2 other short-lived proteins, known to be stringently controlled by proteasomal degradation, namely β-catenin and hypoxia inducible factor 1α, remained unchanged after UV-irradiation of 70.4 cells (data not shown). (ii) Cells of the p53 wildtype 55.1c line are at least sensitive towards UV-induced apoptosis when compared to 70.4 cells.9 One would therefore expect a comparable decrease in P53 protein in UV-irradiated 55.1c cells, in particular since the half-life of the wild-type protein is much shorter than that of the mutant form. This effect, however, was clearly not observed. The same holds true for 55.1b cells, which harbor both a mutated and a wild-type form of P53. 55.1b cells strongly respond to the apoptotic stimulus of UV-irradiation,9 but the level of P53 protein was increased rather than decreased in these cells, and both the wild-type and the “mutated” forms became stabilized in the UV-irradiated cells (Fig. 4). (iii) Exposure of p53-mutated cells to H2O2 or DMSO induced disappearance of P53 protein like UV-irradiation. In contrast to the latter, however, the decrease in P53 protein level occurred already within 1–2 hr of incubation and at very low concentrations of the stressors, when there was no indication of apoptotic processes in the cell cultures. Reactive oxygen plays a role in regulation of wild-type P53 by MDM2 independent pathways.18 Reactive oxygen may also play some role in degradation of mutant P53, since there is often a massive production of reactive oxygen species during the late phase of the apoptotic process (for a review see Ref.19), which occurred in 70.4 cells at the 24 hr time point post-irradiation. It remains unclear, however, why there was no comparable effect in UV-exposed 55.1c cells where the wild-type form of the P53 protein was still present at high levels 24 hr after irradiation of cells, in particular, since MDM2, which initiates degradation of P53, was increased in 55.1c cells following UV-irradiation while it remained unchanged in UV-exposed p53 defective cells.
Stabilization of P53 in response to cellular stress is a well documented phenomenon, which has been investigated intensively in p53 wild-type cells. Unexpectedly, we now observed an inverse effect in hepatoma cells that entirely lack a wild-type allele of the p53 gene, where the protein was found to be degraded response to cellular stress. In hepatoma cells with a heterozygous mutation of p53, however, P53 protein became stabilized upon irradiation, like in p53 wild-type cells. Deletion of the remaining wild-type allele in these cells, a condition often found in cancer cells, may then represent a strong selective advantage because the function of the tumor suppressor as transcriptional regulator may become totally lost.
The excellent technical assistance of Ms. Silvia Vetter and Ms. Elke Zabinsky is gratefully acknowledged.