¹Department of Experimental Oncology, Regina Elena Cancer Institute, Rome, Italy

Correspondence: Professor A Sacchi or Dr G Bossi, Regina Elena Cancer Institute, Molecular Oncogenesis Laboratory, Department of Experimental Oncology, Via delle Messi d'Oro 156, 00158 Rome, Italy. E-mail: sacchi@ifo.it; bossi@ifo.it

Received 10 March 2005; Revised 21 July 2005; Accepted 21 July 2005; Published online 19 September 2005.

Abstract

Mutations in the TP53 tumor suppressor gene are the most frequent genetic alteration in human cancers. These alterations are mostly missense point mutations that cluster in the DNA binding domain. There is growing evidence that many of these mutations generate mutant p53 proteins that have acquired new biochemical and biological properties. Through this gain of function activity, mutant p53 is believed to contribute to tumor malignancy. The purpose of our study was to explore mutant p53 as a target for novel anticancer treatments. To this aim, we inhibited mutant p53 expression by RNA interference in three different cancer cell lines endogenously expressing mutant p53 proteins, and evaluated the effects on the biological activities through which mutant p53 exerts gain of function. We found that depletion of mutant p53 reduces cell proliferation, in vitro and in vivo tumorigenicity, and resistance to anticancer drugs. Our results demonstrate that mutant p53 knocking down weakens the aggressiveness of human cancer cells, and provides further insight into the comprehension of mutant p53 gain of function activity in human tumor.

The TP53 gene is the most frequent target for genetic alterations in tumors, being mutated in over 50% of human cancers. The primary outcome of these mutations is the loss of tumor-suppressing functions of the wild-type p53 (wtp53) protein. However, the high frequency of missense mutations and the high expression levels of mutant p53 (mp53) proteins in cancer cells has led to the conjecture that mp53 proteins have acquired novel functions ('gain of function' (GOF)) that actively contribute to cancer development and progression. In recent years, a growing number of studies have provided compelling evidence supporting the existence of mp53 GOF activities in tumor cells (reviewed in Sigal and Rotter, 2000). These activities range from enhanced proliferation in culture, to increased tumorigenicity in vivo, and enhanced resistance to a variety of anticancer drugs commonly used in clinical practice (Aas et al., 1996; Blandino et al., 1999). At the molecular levels, GOF effects were shown to be linked to the ability of mp53 to modulate the expression of several genes, such as MDR1 (multi-drug resistance gp180 protein) (Sampath et al., 2001), c-myc (Frazier et al., 1998), CD95 (Fas/APO-1) (Zalcenstein et al., 2003), and EGR1 (Weisz et al., 2004), supporting the hypothesis that mp53-specific transcriptional activity is also involved in some of the mp53 GOF effects. However, a wealth of information remains to be uncovered with regard to how mp53 contributes to cancer aggressiveness. These information have the potential for yielding important leads toward the utilization of mp53 as a target for novel cancer therapies.

Here, we evaluated whether depletion of mp53 by RNA interference (RNAi), compromises mp53 GOF activities, including enhanced proliferation, in vitro and in vivo tumorigenicity, and resistance to anticancer drugs. For this study, three human cancer cell lines endogenously expressing different forms of mp53 (i.e., SKBR3 breast cancer cells, HT29 and SW480 colon cancer cells) were employed, because in previous studies mainly p53-null cells (i.e., H1299, SaOS-2, PC3) ectopically expressing mutant p53 have been used.

Oligonucleotide sequences specific to either the human p53 mRNA (p53i) (Brummelkamp et al., 2002b) or the bacterial -galactosidase mRNA (LAcZi, RNAi control) cloned into the pRetroSuper vector were transfected into SKBR3 cells, carrying the p53R175H mutant, and into HT29 and SW480, carrying the p53R273H or the p53R273H/P309S mutants, respectively. The efficiency of mp53 knockdown was evaluated by Western blot and immunohistochemical analyses in puromycin-selected polyclonal populations. Albeit to different extents, the p53i-carrying vector efficiently depleted the expression of mp53 proteins in all cell lines (Figure 1a–c). A maximum reduction was detected in HT29 cells (82%) compared with that observed in SKBR3 (75%) and SW480 (50%) cells. The knocking down of mutant p53 proteins was specific, as no mp53 protein reduction was observed in cells transduced with the control vectors pRetroSuper or LacZi (Figure 1a–c).

Figure 1.

Knockdown of mp53 proteins. Cancer cell lines (a) HT29, (b) SKBR3, and (c) SW480 were transfected with either pRetroSuper (Brummelkamp et al., 2002a) or pRetroSuper-based vectors (p53i or LacZi) and knockdown efficiency evaluated by Western blot (upper), band intensity quantification (middle), and immunohistochemistry (bottom). Methods: cells obtained from ATCC were maintained in DMEM (Eurobio) supplemented with 10% FBS (GIBCO) and 2 mM L-glutamine (Life Technologies Inc.). Transfections were performed following lipofectamine-plus procedure (Life Technologies Inc.). Western blot: 30 g portion of protein lysate was resolved and probed with monoclonal p53 antibody (DOI; Blandino et al., 1999) or -tubulin-antibody (Sigma). Bands intensities were quantified by Scion-image software and tubulin-normalization. Immunohistochemistry: transduced cells were fixed with 2% formaldehyde, permeabilized with 0.1% Triton, incubated with DOI and FITC-conjugated secondary antibody (Jackson Immuno-Research), Hoechst stained, and analysed under fluorescence microscope.

Full figure and legend (187K)

Ectopic expression of mp53 in p53-null cells was shown to bestow growth advantage (Cadwell and Zambetti 2001; Scian et al., 2004). Thus, we evaluated whether the mp53 knockdown modifies the growth of our cancer cell lines. Cells from early-transduced polyclonal populations were plated and subsequently analysed daily for cell number, percent of viability, and percent of cells in the S-phase of the cell cycle (BrdU incorporation). Figure 2 shows that mp53 depletion reduces the growth rate and the replication rate, as number of BrdU-positive cells, in all three tested cell lines (Figure 2a–c and g–i, respectively), and induces loss of viability in HT29 and SKBR3 cell lines (Figure 2d and e). These data indicate that abrogation of mutant p53 proteins reduces significantly both cell survival and proliferation rate of tested cancer cells. To explain these results, we speculate that depletion of mp53 protein prevents its functional and physical interaction with p53 family members, restoring p73 and p63 transcriptional activity and subsequent antiproliferative effects.

Figure 2.

Depletion of mp53 proteins reduces growth advantage. Early-transduced p53i cells along with their relative controls were plated in triplicate (2.5 10⁴/60 mm plates). Cells were daily analysed as (i) total cell numbers (a) HT29, (b) SKBR3, (c) SW480; (ii) percent of viability (d) HT29, (e) SKBR3, (f) SW480; and (iii) percent of BrdU-positive cells: (g) HT29, (h) SKBR3, (i) SW480. Results of three independent experiments with meanss.d. are reported. Methods: total cell number and viability were evaluated by Trypan blue exclusion counting. Replication rate was estimated by 1 h BrdU incubation (20 M; Sigma). Cells were fixed (MetoH), DNA denaturated (2 M HCl), and incubated with anti-BrdU antibody (Roche). After secondary FITC-conjugated antibody and Hoechst staining, cells were analysed under fluorescence microscope to determine percent of BrdU-positive nuclei.

Full figure and legend (55K)

It has been reported that overexpression of mp53 proteins in p53-null cells resulted in enhancement of plating efficiency and tumorigenicity (Shaulsky et al., 1991; Dittmer et al., 1993; Lanyi et al., 1998; Cadwell and Zambetti, 2001). To evaluate whether the depletion of endogenous mp53 modifies the aforementioned biological capacity in our cancer cells, we performed in vitro and in vivo assays. For in vitro assays, early-transduced p53i polyclonal populations were plated along with their relative controls, and the colony forming ability was tested 14 days later. In spite of a milder effect observed in SW480 cells, probably due to a moderate mp53 knockdown (Figure 1b), the data in Figure 3(a–c) clearly indicate that mp53 depletion reduces the clonogenicity of all three tested cell lines. Tumorigenic capacity of HT29 and SW480 cell lines, upon RNAi, was tested in nude mice, whereas SKBR3 tumorigenicity was tested by soft-agar assay. Indeed, SKBR3 cell line does not grow in nude mice (Care et al., 2001). Xenograft experiments were performed by subcutaneous injection of either 5.0 10⁵ cells/mouse of HT29p53i or 5.0 10⁶ cells/mouse of SW480p53i, or their relative controls. As reported in Figure 3, mp53 knockdown reduces the in vivo tumorigenicity of HT29 and SW480 cells, showing significant reduction in both tumor growth and tumor uptake in mice bearing HT29p53i (Figure 3d), and SW480p53i cells (Figure 3e).

Figure 3.

The mp53 knockdown reduces tumor malignancy. Plating efficiency: (a) SKBR3. (b) HT29. (c) SW480. Method: early-transduced p53i cells were plated in triplicate along with their control, and 14 days later colonies formed were stained with crystal violet and counted. The percent of clonogenicity as meanss.d. of three independent experiments is reported at the bottom of each panel. In vivo tumorigenicity: (d) HT29, (e) SW480. Method: groups of eight mice each (nu/nu CD1, Charles River) were injected subcutaneously with either HT29 p53i or SW480 p53i along with their relative controls. Tumor growth and percent of tumor uptake were monitored weekly till the fifth. Caliper measurements of perpendicular measures of the tumors and the size in cm³ were estimated by the formula tumor size=a(b²)/2, where a and b are the tumor length and width, respectively, in cm. All procedures involving animals and their care were conducted in conformity with the institutional guidelines. Soft-agar assay: (f) SKBR3 p53i cells along with relative controls were plated in 60 mm dishes as described elsewhere, and 3 weeks later colonies formed were stained with neutral red and scored. Meanss.d. of three independent experiments are reported.

Full figure and legend (167K)

Soft-agar assays were performed seeding 5.0 10⁴ SKBR3p53i or control cells in 60 mm dishes and their colony forming ability under semisolid conditions was tested 3 weeks later. Data showed that mp53 knockdown reduces significantly, almost 50%, the ability of SKBR3 cells to form colonies in soft agar (Figure 3f).

A growing number of studies have provided compelling evidence that mp53 overexpression can exert GOF effects in tumor cells by increasing chemoresistance to a variety of anticancer drugs (Aas et al., 1996; Li et al., 1998; Wang et al., 1998; Blandino et al., 1999; Sigal and Rotter, 2000; Tsang et al., 2005). To evaluate whether in our experimental conditions the depletion of mp53 expression affects the chemosensitivity of tumor cells, their colony forming ability was tested upon treatment with different drugs (i.e., cisplatin, adriamycin, and etoposide). Colony assays were performed challenging early-transduced SKBR3, HT29, or SW480 cells and analysing effects on chemoresistance 2 weeks later. Results of these experiments showed that mp53 knockdown reduces chemoresistance of all three tested cell lines (Figure 4a). To evaluate whether the reduced chemoresistance might be attributed to a higher apoptotic response, the cleaved PARP isoform was monitored after drug treatments. Results of these experiments revealed that, in all tested cell lines mp53 depletion allows a higher apoptotic response to drugs treatment, albeit to different extents (Figure 4b). These data confirm the notion that endogenous mp53 proteins contribute to chemoresistance of tumor cells. The differences in drug response suggest that the protective effects promoted by the mp53 proteins might depend on specific cellular contexts and/or the type of p53 mutation. Moreover, the relevance of cellular context has been recently outlined by the generation of mp53 knock-in mice indicating that, when introduced into different strains of mice, single p53 mutations exert an apparent strain-dependent GOF activity (Lang et al., 2004; Olive et al., 2004).

Figure 4.

The mp53 knockdown reduces chemoresistance in cancer cells. (a) Colony forming assay: early-transduced p53i cells along with their relative controls were plated in triplicate in six-well plates, the day after medium was replaced with DMEM+2% FBS and 24 h later, cells were challenged with different drugs, for 48 h (SKBR3) or 72 h (HT29, SW480). Then, cells were washed twice in PBS and replenished with fresh medium. After 14 days, cells were stained by crystal violet and colonies formed were scored. Percent of colonies was calculated relative to cells receiving drug vehicle alone. Meanss.d. of three independent experiments are reported. (b) Apoptotic response: p53i cells along with relative controls were plated in 60 mm dishes and treated with different drugs as described. Apoptosis was monitored by detection of cleaved PARP isoform (82 kDa) and quantified by band densitometric analyses, values are reported at the bottom of each Western. Experiments of SKBR3 cells with cisplatin were performed in 10% FBS. Methods: A 50 g portion of total lysate was resolved and probed with anti-82 kDa PARP-isoform antibody (Promega) or anti-actin antibody (Sigma). Densitometric analyses of bands were performed by Scion-image software and actin normalization. We used a dose corresponding to the DL₂₀ for clonogenic assay and DL₅₀ for apoptotic assay. The different lengths of treatments were dependent on the maximum achievable effect. (c) DD miniprotein: SKBR3 cells were transfected with pCDNA-HA-DD vector, and following G418 selection, clones were isolated and the exogenous DD expression was monitored by Western blotting of anti-HA (top panel). Chemoresistance was evaluated by treating isolated clones with cisplatin (2.5 g/ml) and 48 h later viability was monitored by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay (MTT) (bottom panel) following the manufacturer's procedures.

Full figure and legend (149K)

In view of the fact that, as far as concern possible therapeutic purposes, the chemoresistance is the most relevant effect of mp53 GOF activity, we ascertain whether using an alternative approach for mp53 protein inactivation might reproduce restoration of chemosensitivity. To this end a p53 dominant-negative protein (DD mini-protein) was employed. The DD mini-protein (residues 302–390) was shown to interact with wt p53 protein and inhibits its activity in a dominant negative manner (Shaulian et al., 1992). More recently, DD mini-protein was reported to interact with mp53 proteins in living cells (Strano et al., 2000). HA-tagged DDp53 was transfected into SKBR3 cells and, after selection with G418, single-cell clones expressing different levels of the exogenous protein were isolated (Figure 4c, upper panel). Cell clones were challenged with the drug and viability evaluated by MTT assay. Results of these experiments showed that DD mini-protein reduces chemoresistance of SKBR3 cells in a dose-dependent manner; indeed, higher level of exogenous DD mini-protein (clone DD#2) was associated with a higher reduction of viability (Figure 4c, lower panel).

Most of the studies reported thus far on GOF activity of mp53 proteins have been performed upon overexpression of exogenous mutant proteins. However, because of the artificial systems employed, the concept that mp53 proteins are pro-oncogenic has been questioned (Vousden and Prives, 2004). A few studies were performed in a more physiologic context, as cancer cell lines endogenously expressing mp53 proteins (Irwin et al., 2003; Weisz et al., 2004) or generating knock-in mice (Lang et al., 2004; Olive et al., 2004). In agreement with the results obtained by these strategies, our data show that RNAi of mp53 proteins reduces the aggressiveness of human cancer cells in vitro and in vivo, supporting the existence of a GOF activity of some human tumor-derived p53 mutants. Finally, our findings indicate RNAi as an efficient tool to study GOF activity of mp53 proteins and reduce drug resistance in tumors harboring p53 mutations.