Background
The p53 tumour suppressor protein, encoded by
TP53 gene (OMIM 191170), integrates endogenous and exogenous signals to modulate cell fate in response to multiple forms of environmental and cellular stresses [
1]. In response to such stresses, the p53 protein escapes from down-regulation by Hdm2/Hdmx, which bind in the N-terminal region of p53, and acts as E3 ubiquitin ligase to induce p53 nuclear export and degradation by the proteasome [
2,
3]. The resulting nuclear accumulation, combined with multiple steps of post-translational modifications, leads to the activation of sequence-specific DNA-binding to response elements located in a wide panel of target genes, resulting in transcriptional regulation of genes involved in cell-cycle arrest, differentiation, DNA repair, autophagy, apoptosis, senescence and control of oxidative metabolism [
4,
5]. These multiple effects contribute to an extensive repertoire of anti-proliferative biological responses. The type and degree of responses depend upon which specific components of this repertoire are activated in a manner that differs according to tissue, cell type, metabolic context and nature of inducing stress [
5,
6]. Given these multiple, complex effects, it is likely that p53 activity is under extremely tight control and that several, overlapping mechanisms may concur to set tissue- and cell-specific thresholds for p53 activation in response to different types of stimuli.
In recent years, the identification of isoforms of the p53 protein has provided a new mechanism that may contribute to the fine-tuning of p53 activity. Isoforms are produced by alternative splicing, alternative promoter or codon initiation usage, or combinations thereof [
7]. The resulting proteins differ from canonical, full-length p53 protein, by truncation of a variable portion of the N-terminus (ΔN isoforms) and by alternative C-terminal portions (C-terminal isoforms). So far, up to 4 distinct N-terminal and 3 C-terminal variants have been identified, leading in theory to 12 isoforms (including full-length p53; [
8]). These isoforms retain at least part of the DNA binding and oligomerization capabilities, but differ through regulatory domains in the N- and C-terminus, supporting the notion that their main biological effect is to modulate p53 protein functions. However, the existence, expression patterns and detailed biological function of each particular isoform is still poorly documented.
The Δ40p53 isoform is a form of the protein that lacks the first 39 residues containing the main transactivation domain (residues 1–42), as well as major activating phosphorylation sites and the binding site for Hdm2, the main regulator of p53 degradation [
9,
10]. Δ40p53 is produced by two complementary mechanisms, alternative codon initiation usage at AUG 40 in fully-spliced p53 mRNA, and alternative splicing that retains intron 2, which introduces stop codons downstream of the +1 AUG and leads to the synthesis of a truncated protein using AUG 40 in exon 4 as initiation codon [
11,
12].
In vitro studies have shown that Δ40p53 interferes with p53 transcriptional activity, acting as concentration-dependent dominant inhibitor when artificially expressed in excess to full-length p53 [
9]. Two animal models overexpressing Δ40p53 have been reported, one in the mouse [
13,
14] and the other in Zebrafish [
15]. Overexpression of a transgene encoding a p44 protein corresponding to Δ40p53 (MD41p53) did not induce any specific phenotype in p53-deficient mice. However, when expressed in a wild-type Trp53 background, increased dosage of MD41p53 led to reduced size, accelerated aging and a shorter lifespan associated with hypo-insulinemia and glucose insufficiency [
13,
14]. Compatible effects were observed in Zebrafish, in which expression of Δ40p53 in a p53-null background did not lead to a specific phenotype although expression in a p53-competent background resulted in impaired growth and development [
15]. Overall, these results suggest that Δ40p53 exerts its main biological effects by modulating the activity of full-length p53. Furthermore, they suggest that Δ40p53 exerts effects other than simple dominant-negative inhibition of p53.
In this study, we have used biochemical approaches to assess the effects of co-expression of Δ40p53 and full-length p53 at different ratio into p53-null human cancer cell lines and we have analyzed the effects on p53 protein expression, DNA-binding capacity and transcriptional activity towards a p53-dependent reporter gene. Our results show that Δ40p53 exerts an inhibitory effect when expressed in excess over full-length p53. However, when expressed at levels inferior or equal to full-length p53, Δ40p53 appears to exert more complex effects, ranging from inhibition to activation of p53 transcriptional activity, depending upon cellular context. These results support the hypothesis that Δ40p53 is a critical regulator of wild-type p53 function and provide a basis to interpret the complex biological phenotypes induced by this isoform in animal models.
Methods
Cell culture
Human breast cancer 21PT cells (gift of V. Band, [
16]) were grown at 37°C in 5% CO
2 in MEM with Earle’s Salts medium (PAA, Linz, Austria) supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin-glutamine, 10 ng/ml Epidermal Growth Factor, 10 μM hydrocortisone and 1 μg/ml human insulin. Human colorectal HCT116 cells (gift of B. Vogelstein, [
17]) were cultured at 37°C in 5% CO
2 in Mc Coy’s 5A medium modified (Invitrogen, Carlsbad, CA, USA) supplemented with 10% of fetal bovine serum and 1% of penicillin-streptomycin-glutamine. Human lung carcinoma H1299 cells (p53-null) (ATCC N°CRL-5803) were grown at 37°C in 5% CO
2 in RPMI-1640 medium (PAA) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin-glutamine. Human osteosarcoma Saos-2 cells (p53-null) (ATCC N°HTB-85) were grown at 37°C in 5% CO
2 in Mc Coy’s 5A medium modified (Invitrogen) supplemented with 15% fetal bovine serum and 1% penicillin-streptomycin-glutamine. Mouse embryonic fibroblast BALB/c 10.1 cells (p53-null) (ATCC N°CCL-163) were grown at 37°C in 5% CO
2 in Dulbecco’s Modified Eagle’s Medium (PAA) supplemented with 10% of fetal bovine serum and 1% of penicillin-streptomycin-glutamine.
Expression vectors, transfections and reporter gene assays
Expression vectors pcDNA3-TAp53 (mut40) and pcDNA3-Δ40p53, based on the pcDNA3.1 vector (Invitrogen) containing the Cytomegalovirus (CMV) enhancer-promoter, which is constitutively active and drives high levels of mRNA expression in human cells, were described previously [
9,
18]. The pcDNA3-TAp53 (mut40) plasmid contains a mutation at codon ATG 40 (ATG > TTG) to prevent Δ40p53 expression by internal initiation of translation. The plasmid pCMV-Neo-Bam-Hdm2 was kindly provided by J.C. Marine [
19]. The pcDNA3-empty vector was used as negative control as well as adjustment vector in dose-dependent transfection experiments in order to keep the amount of transfected DNA to a constant level. Depending upon the cell lines used, transfections were performed using Fugene (Promega, Fitchburg, WI, USA) (H1299 and Saos-2 cells) or Lipofectamine 2000 (Invitrogen) (BALB/c 10.1 cells).
Cycloheximide treatment
In 60 mm dishes, 4 × 105 HCT116 p53+/+ and 21PT cells were treated with 50 μg/ml of cycloheximide (Sigma Aldrich, St Louis, MO, USA).
Co-immunoprecipitation assay
In 100 mm dishes, 2 × 106 BALB/c 10.1 cells were transfected using Lipofectamine 2000 with 2 μg/ml of pcDNA3-expression vectors (empty, -TAp53, -Δ40p53). Different ratios of the two pcDNA3-expression vectors TAp53/Δ40p53 were transfected (0.5/0.5 and 1.5/0.5 μg/ml). The total amount of plasmid was adjusted by adding empty vector to reach 2 μg/ml. Cells were lysed 24 h post-transfection by adding 250 μl of lysis buffer (10 mM Tris pH 7.5, 140 mM NaCl, 0.5% Triton, 2 mM DTT, 1 mM EDTA and 1 tablet of protease inhibitor cocktail/10 ml Complete Mini, Roche). Total protein extracts (500 μg) were cleared by incubation (4°C-1 h) with 25 μl of anti-mouse sepharose beads (Sigma) and then incubated (4°C-overnight) with p53 mouse monoclonal DO-7 antibody, which binds the epitope at residues 19–26 and detects only TAp53 (DAKO, Glostrup, Denmark). Proteins/antibodies complexes were analyzed by Western blot using rabbit polyclonal anti-p53 CM1 antibody (Novocastra, Newcastle, UK).
Electrophoretic mobility shift assay (EMSA)
EMSA were performed as described in Verhaegh et al., 1997 [
20]. Briefly, nuclear cellular proteins of H1299-transfected cells were extracted and incubated with mixture containing
32P-radio-labeling oligonucleotide with p53 response element consensus p53
con (5
′-GGACATGCCCGGGCATGTCC-3
′) [
18]. Two antibodies were used to shift DNA: p53 complexes, the monoclonal PAb421 antibody recognizing p53 DBD and known to stabilize DNA: p53 complexes; and the monoclonal DO-7 antibody specific for the N-terminal portion of p53 protein, recognizing TAp53 but not Δ40p53.
β-galactosidase assay
In 6-well plates, 3 × 105 H1299 and Saos-2 cells were transfected in duplicate using Fugene transfection reagent with 0.5 μg/ml of pRGCΔFosLacZ plasmid (pRGC), containing the Ribosomal Gene Cluster (RGC), which includes three p53-binding sites with the same consensus sequence as the one of p53con, upstream β-galactosidase gene, and 2 μg/ml of pcDNA3-expression vector (empty, -TAp53, -Δ40p53). The following TAp53/Δp53 vector ratios were used: 0.5/0.5 and 0.5/1.5 µ/ml. The total amount of plasmid has been adjusted by adding pcDNA3-empty vector to reach 2.5 μg/ml of total transfected DNA in each condition. In addition, cell transfection efficiency was assessed by transfecting 1 μg/ml of a GFP-expressing vector. Twenty four hours post-transfection, cells were harvested and β-galactosidase activity was measured using β-galactosidase Enzyme Assay System kit (Promega).
Western blot
Proteins were extracted using RIPA-like buffer (50 mM Tris–HCl pH 7.4, 250mM NaCl, 0.1% SDS, 0.5% NP-40, 2 mM DTT, 1 mg/ml protease inhibitors [500 mM phenyl-methyl-sulfonyl-fluoride (PMSF), 0.5 mg/ml leupeptin, 2 mg/ml aprotinin, 1.4 mg/ml pepstatin A]) and 30 μg of cleared extracts were loaded on mini protean TGX 4-15% gels (Bio-Rad, Hercules, CA, USA) and transferred using TransBlot Turbo Transfer System (Bio-Rad). Membranes were hybridized with several antibodies: monoclonal anti-p53 DO-7 antibody; rabbit polyclonal anti-p53 CM1 antibody and monoclonal anti-p53 PAb1801 (Oncogene Research Products, Cambridge, MA, USA), which detects both TAp53 and Δ40p53; p53 anti-phospho-Serine 15 antibody (Cell Signaling Technology, Beverly, MA, USA); p21WAF1 antibody (Calbiochem, Billerica, MA, USA); Hdm2 (Calbiochem) and Ku80 (Abcam, Cambridge, UK) or anti-actin (Santa Cruz, CA, USA) antibodies were used as a loading control. Secondary peroxidase coupled goat anti-mouse or -rabbit antibodies were used, followed ECL detection according to the manufacturer’s instructions (Amersham Biosciences AB, Buckinghamshire, UK).
Statistical analyzes
Statistical analyzes were performed using the software GraphPad Prism (GraphPad Software, Inc) and P-values were indicated by (*) when P < 0.05, (**) when P < 0.01 and (***) when P < 0.001.
Discussion
Δ40p53 was the first identified N-terminal isoform of p53 and there is evidence from animal model systems that it exerts a strong regulatory effect on the biological activity of the p53 protein [
13‐
15]. In this study, we have used
in vitro approaches to show that Δ40p53 can form mixed oligomers with TAp53 that can bind to specific DNA and modulate the capacity of p53 to activate a generic reporter gene driven by a standard p53 response element. This modulation did not follow a proportional dose–response according to the relative expression of TAp53 and Δ40p53. In H1299 cells, co-expression of the two proteins induced a decrease in the transcriptional activity, but the amplitude of the effect varied depending upon the predicted composition of the hetero-tetramer. In Saos-2, a paradoxical effect was observed, with no decrease and possibly a small increase in transcriptional activity for hetero-tetramers predicted to contain 1 or 2 monomers of Δ40p53, and a strong decrease for hetero-tetramers predicted to contain 3 monomers of Δ40p53. These different effects suggest that Δ40p53 may exert subtle modulation on p53 activity, in particular in the range of expression levels that are generally observed in non-transfected cells, where Δ40p53 is expressed at lower or, at the most, equal levels to TAp53.
The differences between the two cell lines can be rationalized by considering that Δ40p53 may alter the dynamics of hetero-tetramers in two opposite ways. First, it may alter the binding to component of the basal transcription machinery, thus decreasing transcriptional activation. Second, it may interfere with the binding of Hdm2 to the hetero-tetramer, thus modulating its subsequent post-translational modification, its nuclear export and its degradation through the proteasome machinery. Supporting this view, we show that, in Saos-2 cells, Δ40p53 could at least partially protect TAp53 from Hdm2-mediated degradation. The differences between the two cell lines with respect to the effect of Δ40p53 on the transcriptional activity of TAp53 may be due to multiple cell-specific mechanisms, including levels of expression of Hdmx, which co-regulates p53 stability together with Hdm2, as well as Hdm2-independent effects. Of note, p53 monomers are resistant to Hdm2-mediated degradation
in vitro and in intact cells, despite the presence of Hdm2 binding sites on each p53 monomer. Dimerization, but not tetramerization, is required for Hdm2-dependent degradation [
22].
The observation that Δ40p53 may exert, in defined condition and cell context, a positive effect on p53 transcriptional activity, may account for some of the biological characteristics of animal models expressing this isoform. Both in the mouse expressing p44, an equivalent to Δ40p53, and in Zebrafish, expression of Δ40p53 alone in a p53-null background does not significantly affect the phenotype [
13‐
15]. In contrast, dramatic effects are seen when both TAp53 and Δ40p53 are co-expressed, and these effects are consistent with an increase and modulation of p53-dependent suppressive effects rather than an inhibition of these effects. Mice co-expressing TAp53 and Δ40p53 do not show a tumor-prone phenotype attributable to inhibition of p53 activity [
13,
14]. In contrast, they show increased organismal and tissular senescence attributable to decline in stem cell renewal and in general fitness. In Zebrafish, co-expression of the two proteins lead to developmental defects attributable to hypoplasia, malformation of the head, eyes and somites, a phenotype which is also compatible with enhanced, rather than decrease, p53 suppressive activity [
15]. These effects are rescued by co-expression of a dominant-negative p53 mutant, suggesting that they are indeed dependent upon increased p53 suppressive activity. Overall, our observations shed a new light on a possible role of Δ40p53 in the regulation of p53 activity in stem and progenitor cells, contributing to either enhance or inhibit p53 function in a manner which is exquisitely regulated by the relative levels of each isoform and their interrelations with p53 regulatory systems such as Hdm2/Hdmx.
A recent study by Ungewitter and Scrable (2010) has determined that Δ40p53 was the major p53 isoform expressed in mouse embryonic stem cells (ESC) as well as during the early stages of embryogenesis [
23]. They further showed that altering the dose of Δ40p53 had a strong impact on the maintenance of the ESC state. Haploinsufficiency for Δ40p53 (resulting in lower levels of this isoform) caused loss of pluripotency and changes in cell-cycle with the acquisition of a cell-cycle profile characteristic of the one of somatic, differentiated cells. By contrast, increased dosage of Δ40p53 caused an extension of pluripotency and prevented progression to a more differentiated state. These observations suggest that in mouse embryonic cells, high levels of expression of Δ40p53 may prevent the activation of a mechanism of exit from ESC status into a somatic cell status. Within such a context, our results of a dual effect of Δ40p53 on TAp53 activity suggest that Δ40p53 might play a subtle role in regulating the transition from ESC to somatic cell status. At high levels, Δ40p53 may maintain pluripotency by neutralizing p53. As cells shift towards a somatic status, decreasing levels of Δ40p53 may contribute to activate TAp53 and to switch on a mechanism of progression out of ESC status towards differentiation. Further studies are needed to determine whether presence of defined levels of Δ40p53 may be required to adequately drive the exit from ESC status during embryogenesis.
It is recognized that this
in vitro study has a number of limitations. First, the effects have been studied using only one type of consensus p53 RE and reporter gene. Given the wide repertoire of p53 RE and their capacity to respond in a different way to p53 activation, it is plausible that modulation of transactivation by Δ40p53 may have different effects on different p53 targets. Second, although implicating Hdm2 as an essential regulator of the activity of TAp53/Δ40p53 hetero-oligomers, our study has not quantified variations in the binding of Hdm2 to different types of hetero-oligomers and has not clearly identified the consequences on p53 protein stability. Third, Δ40p53 is not the sole N-terminal isoform which may regulate p53 activity. In particular, Δ133p53 (which lacks the N-terminal transactivation domain, the proline-rich domain and the proximal part of the DNA-binding domain) also retains the oligomerization domain and may interfere as third partner in hetero-oligomers containing TAp53 and Δ40p53. We have previously shown that, in contrast to Δ40p53, Δ133p53 does not bind to DNA and prevents TAp53 to recognize its cognate response element. Recent studies have proposed that Δ133p53 may play a role as enhancer of p53-dependent cell senescence [
24].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HH carried out transfections and molecular analyses and drafted the manuscript. DS performed background cell culture work and western blot analyses. SCC performed initial experiments on Δ40p53 stability and oligomerization. PH conceived and coordinated the study, and oversaw the writing of the manuscript. All authors read and approved the final manuscript.