Development of p53 Protein Transduction Therapy Using Membrane-permeable Peptides and the Application to Oral Cancer Cells¹

Viral vector-mediated delivery of tumor suppressor genes such as p53 represents a powerful strategy for cancer therapy (1). The viral vector is useful to transduce exogenous genes into various human cells in vitro and in vivo (1). However, previous studies have indicated that virus-mediated gene therapy has some safety problems, such as inflammatory response, toxicity, and random integration of viral vector DNA into the host chromosomes (1). Recent studies have shown that liposomes are able to deliver exogenous genes with minimal toxicity in vivo (2–4). However, the efficiency of gene transduction is worse than that of the viral-mediated gene transfer method at present (1).

The HIV type 1 Tat protein can enter cells when added to culture media (5, 6). The PTD³ of the Tat protein, which contains a high proportion of arginine and lysine residues, has been identified as being responsible for the ability to penetrate the plasma membrane (7). It has been shown to serve as a carrier to direct the uptake of heterologous proteins into cells by generating genetic in-frame PTD fusion proteins (8–10). Interestingly, the PTD can deliver the biologically active form of β-galactosidase, which consists of >1000 amino acids, to all tissues including the brain in vivo (11). Recent studies have shown that poly-arginine, 6–12 residues, also has the same transduction activity as the PTD of the Tat protein (12–14). Moreover, we have shown that a protein transduction method using 11 poly-arginine peptides (11R) is useful for protein delivery in cells and brain slices and in the control of the localization of the delivered protein in cells (14). The protein transduction method has some advantages over viral vector-mediated gene transduction therapy in terms of safety, low toxicity, and random integration of vector DNA. In this study, we showed that p53 protein-fused 11R (11R-p53) was efficiently delivered in oral cancer cell lines and inhibited the proliferation of these cells.

The oral cancer cell line NOS-1 was a gift from Drs. Oku and Komori (Kobe University, Kobe, Japan). The oral cancer cell lines SAS, HSC-3, and HSC-4 were provided by the Cell Resource Center for Biomedical Research (Tohoku University, Sendai, Japan). These cells were maintained in GIT medium (Wako Pure Chemical Industries, Osaka, Japan) supplemented with 100 units/ml penicillin and 100 units/ml streptomycin. SAOS-2 cells were provided by Riken Gene Bank (Riken, Wako, Japan). The cells were maintained in DMEM with 10% fetal bovine serum.

We constructed a bacterial expression vector, p11R-HA, to produce genetic in-frame 11R fusion proteins. The vector p11R-HA was constructed by modification of pET-21a(+) vector (Novagen, Madison, WI). Briefly, the vector was tagged with a 6-histidine leader, followed by an 11-arginine flanked by glycine and glutamic acid residues (for free bond rotation of the domain) in the COOH terminal. Full-length human wild-type p53 cDNA was subcloned into BamHI/EcoRI-digested p11R-HA vector.

The constructed plasmids were transformed into BL21-DE3 Escherichia coli cells. The proteins were expressed in the cells by induction with 0.1 mm isopropyl-1-thio-βd-galactopyranoside. The expressed proteins were purified as described previously (14). After we collected the bacterial pellets, the pellets were sonicated and denatured in lysis buffer containing 20 mm HEPES (pH 8.0), 100 mm NaCl, 2 m urea, and 20 mm imidazole. The denatured soluble fraction was applied to a column of Ni-NTA agarose (Invitrogen, San Diego, CA) to purify His-tagged 11R-p53 proteins.

The recombinant adenovirus containing wild-type human p53 cDNA, chicken β-actin promoter, and SV40 polyadenylation signal was produced using the Adenovirus Expression Vector kit (Takara, Tokyo, Japan). The recombinant adenoviruses containing p53 cDNA (pAdex-p53 or virus-p53) were purified and concentrated using the CsCl step-gradient method as described previously (15). The concentrated adenoviruses were dialyzed with PBS and 10% glycerol and stored at −80°C until use. Recombinant adenovirus of lacZ (pAdex-lacZ) was also produced following the manufacturer’s protocol.

Western blot analysis for p53 was carried out at high stringency, essentially as described previously (16). Briefly, the harvested cells were lysed in a boiled buffer containing 1% SDS with a sonicator. Samples containing 100 μg of total protein were run on SDS-PAGE gels, and then transferred to a nitrocellulose membrane (Hybond ECL; Amersham). The blots were probed with primary antibodies against p53 (1:1000; Pab 1801; Santa Cruz Biotechnology, Inc., Santa Cruz, CA).

Immunocytochemistry was performed as described previously (14). Briefly, the cells were fixed with 4% paraformaldehyde, and then incubated with monoclonal anti-p53 antibodies (1:100; Pab 1801; Santa Cruz Biotechnology) and polyclonal anti-α-tubulin antibodies (1:200; C-20; Santa Cruz Biotechnology). After the antibodies were washed off, the cells were incubated with FITC-conjugated secondary antibodies (1:200; Sigma Japan) and rhodamine-conjugated secondary antibodies (1:100; Sigma Japan) or rhodamine phalloidin (1:10,000; Molecular Probes, Eugene, OR) instead of the rhodamine-conjugated secondary antibodies.

The reporter assay was performed as described previously (17). The luciferase reporter vector pGL2-basic (Promega) that inserted a 2.4-kbp fragment of human p21/WAF1 promoter into the HindIII site was a gift from Drs. T. Akiyama (Tokyo University, Tokyo, Japan) and K. Yoshikawa (Osaka University, Osaka, Japan). The luciferase reporter vector was transfected into ∼70% confluent SAOS-2 cells in 35-mm dishes by the calcium phosphate method. The cells were then either infected with each concentration of virus-p53 adenovirus or incubated with each concentration of 11R-p53. After 36 h, the cells were harvested, and the luciferase activities were measured with a luminometer using a reagent kit (Tokyo Ink, Tokyo, Japan). The background luciferase activity was subtracted from all experiments.

Cell proliferation was assessed at each time point after infection by pAdex-p53 or the addition of 11R-p53 by measuring the conversion of the triazolium salt WST-1 to formazan according to the manufacturer’s instructions (Roche).

Each of the cancer cells (∼70% confluent) cultured on coverslips in 35-mm dishes was either transfected with virus-p53 (50 MOI) or added to 11R-p53 (1 μm). After 24 h, the cells were treated with 0.2 μg/ml cisplatin (Pharmacia) for 4 h. The cells were then washed with PBS twice, and new medium was added. One μmol of 11R-p53 was added to the new medium again. The cells were fixed with 4% paraformaldehyde (pH 7.4) 36 h after adding the cisplatin, and then Hoechst dye staining was performed. The fixed cells were treated with 3.3 μm Hoechst 33342 (Sigma Japan) for 15 min at room temperature and observed with a fluorescence microscope. Apoptotic cells were identified by the presence of highly condensed or fragmented nuclei.

To demonstrate whether 11R-p53 proteins were delivered into the cells through the plasma membrane, a p53 gene-deficient osteosarcoma cell line, SAOS-2, was incubated with 11R-p53 (1 μm), and the protein expression level was compared with that in pAdex-p53-infected cells. Western blot analysis revealed that 11R-p53 proteins were strongly expressed in SAOS-2 cells (Fig. 1B). The mobility shift of 11R-p53 signals was seen on SDS-PAGE gel because 11R had a positive charge (Fig. 1B). We also incubated SAOS-2 cells with non-11R-p53 proteins (1 μm), which were not fused with 11R, as a control. Western blotting analysis revealed that the proteins were not expressed in the cells (Fig. 1B). To demonstrate whether 11R-p53 was localized in the nucleus of the cells and to exclude the possibility that the protein was just attached to the surface of the cell membrane, the cells were double-stained with anti-p53 antibody (green) and anti-tubulin antibody (red; Fig. 1,Ca). The p53 signals were localized in the nucleus of SAOS-2 cells. Moreover, the cells were double-stained with anti-p53 antibody (green) and rhodamine phalloidin (red), and serial optical sections of 2-μm steps along the Z-dimension of the immunostained cells were then collected with a confocal microscope (Fig. 1Cb). The p53 signals were not seen on the surface of the cell membrane. Next, we demonstrated whether the delivered 11R-p53 has a transcriptional activity in living cells. The cells were incubated with either 11R-p53 or pAdex-p53, and a reporter assay for p53 was then performed using a luciferase reporter vector driven by the p21/WAF1 promoter, which contains the p53-binding site. The transcriptional activity of 11R-p53 increased in a dose-dependent manner (Fig. 1D). In 1 μm 11R-p53-transduced cells, the transcriptional activity was 32-fold higher than that of control. In non-11R-p53-incubated cells, the activity was not detected (Fig. 1D). These results indicate that the 11R-p53 protein penetrates through the cell membrane and localizes in the nucleus. Moreover, the protein has full transcriptional activity in cells.

Next, we demonstrated whether 11R-p53 inhibits the proliferation of NOS-1 cells, which has been established from a patient with oral squamous cell carcinoma and in which the p53 gene was point mutated at site 248 (18). Each concentration of 11R-p53 or 10 μm 11R-GFP (14) was added to the culture medium at day 0. After 6 h, the medium was replaced with new medium excluding 11R-p53 or 11R-GFP. The cells were then cultured for an additional 4 days, and a WST assay was performed to demonstrate the cell proliferation every 24 h. The WST assay revealed that 0.1 μm 11R-p53 did not significantly inhibit the proliferation of NOS-1 cells at any of the time points. The growth curve of 0.1 μm 11R-p53-transduced cells was the same as that in 11R-GFP-transduced cells. In contrast, both 1 and 10 μm 11R-p53 significantly inhibited the proliferation of NOS-1 cells compared with that of 11R-GFP-transduced cells (1 μm 11R-p53, P < 0.01 at day 3, P < 0.05 at day 4 versus 11R-GFP on the same days; 10 μm 11R-p53, P < 0.01 at day 2 and day 3, P < 0.05 at day 4 versus 11R-GFP on the same days). However, the effect of 11R-p53 on the inhibition of the proliferation of the cells was weaker than that of pAdex-p53 (Fig. 2B). The WST assay revealed that infection of pAdex-p53 at MOIs of 20 and 100 more significantly inhibited the cell growth compared with that of 1 and 10 μm 11R-p53 at day 4 (Fig. 2, A and B, pAdex-p53 at 20 and 100 MOI, 1.05 ± 0.08 and 0.76 ± 0.025, respectively; 1 μm 11R-p53, 1.59 ± 0.04; 10 μm 11R-p53, 1.59 ± 0.09, n = 10 in each group; P < 0.01, pAdex-p53 at an MOI 20 or 100 versus 10 μm 11R-p53). Infection of pAdex-p53 at MOI of 300 completely inhibited the cell growth (Fig. 2B). However, the effect on the proliferation inhibition was attributable to virus toxicity because the infection of pAdex-lacZ at MOI of 300 also had the same effect (data not shown).

These data suggested that 11R-p53 inhibited the proliferation of NOS-1 cells, but the effect was weaker than that of the pAdex-p53 virus. One possibility is that the expression level of 11R-p53 was not sustained for a sufficient length of time in NOS-1 cells because the protein was provided only at day 0. It is well known that p53 is rapidly degraded by the ubiquitin-proteasome-mediated pathway (19). We demonstrated time-dependent changes of the expression level of 11R-p53 (Fig. 2C). The protein levels of 11R-p53 were higher between 6 and 12 h after the addition of 11R-p53. However, the level rapidly decreased 36 h after the addition (Fig. 2C). In contrast, the expression of endogenous p53 protein was unchanged at each time point. To maintain the protein level of 11R-p53 in NOS-1 cells, we added 11R-p53 every 24 h. Western blotting analysis revealed that the repeated application of 11R-p53 maintained the expression level of the proteins (Fig. 2D) The WST assay showed that the repeated application of the proteins strongly inhibited the proliferation of NOS-1 cells. The effect of 10 μm 11R-p53 was equivalent to that of pAdex-p53 at an MOI of 100 [10 μm 11R-p53, 0.63 ± 0.06 (n = 10); pAdex-p53 at 100 MOI, 0.75 ± 0.05 (n = 10)]. These data show that p53 protein transduction using 11R inhibits the proliferation of oral cancer cells expressing the mutated p53 gene.

Next, we demonstrated whether 11R-p53 inhibits the proliferation of other types of oral cancer cells. SAS cells contain the wild-type p53 gene (20). HSC-3 cells harbor an insertional mutation between codon 305 and 306 (21), and in HSC-4 cells, a heterozygous missense mutation corresponding to a CAG → CGG transition in codon 248 was seen in the p53 gene (20). Western blotting analysis revealed that the 11R-p53 proteins were effectively delivered into all types of these cells (Fig. 3A). In Fig. 3,B, we compare the effect of the proteins on the inhibition of the proliferation of these four cell lines. The 11R-p53 (1 μm) was applied every 24 h, for a total of four times, and the WST assay was performed 96 h after the first application of the proteins. As a control, each of the cell lines was infected with pAdex-p53 at an MOI of 100. The 11R-p53 proteins significantly inhibited the proliferation of all of the cell lines. In all of the cell lines, moreover, the effect of the inhibition was the same as that of pAdex-p53 (Fig. 3B).

Previous studies have shown that transduction of the wild-type p53 gene into cancer cells enhances the effect of anticancer drugs on the induction of apoptosis of the cancer cells (22, 23). We demonstrated whether the transduction of 11R-p53 increased the effect of cisplatin on the induction of apoptosis of NOS-1 cells. The transduction of 11R-p53 per se did not induce apoptosis (Fig. 4, E and G). However, the transduction of the protein enhanced the cisplatin-dependent induction of apoptosis of the cells (Fig. 4, F and G). The treatment with cisplatin after 11R-p53 transduction markedly induced apoptosis of the cells compared with the treatment with cisplatin only (Fig. 4G, CDDP + 11R-p53, 57 ± 6.7%; CDDP only, 35.2 ± 7.3%; n = 100 in each group). Moreover, the combination of overexpression of p53 and cisplatin might inhibit the cell viability of the cancer cells because the cell number in Fig. 4, D and F, was fewer than that of other treatments. Therefore, we compared the cell viability among each condition. The transduction of 11R-p53 significantly enhanced the cisplatin-dependent inhibition of the cell viability (Fig. 4F). These effects of 11R-p53 on the induction of chemosensitivity were the same as that of virus-p53.

A large number of studies have attempted to evaluate the potential of tumor suppressor gene delivery in cancer therapy (1). Adenovirus-mediated p53 gene delivery is being tested in human clinical trials to treat non-small cell carcinoma, head and neck cancers (24). However, a number of reports have indicated that virus-mediated gene therapy has some safety problems, such as inflammatory response, toxicity, and random integration of viral vector DNA into the host chromosomes (1). Direct delivery of tumor suppressor proteins such as p53 may reduce these problems. In the present study, we showed that 11R-p53 markedly inhibited the proliferation of oral cancer cells and the effect was at the same level as that of adenovirus-mediated p53 gene transfection. Our data suggest that protein transduction therapy may become a promising cancer therapy. A recent study has shown that a protein fused with the HIV-1 Tat protein-transduction domain (RKKRRQRRR) or arginine-rich peptides (RRRRRRRRR) efficiently penetrated the epidermis to the same extent as the dermis of the s.c. layer in vivo, when the protein was sprayed on animal skin (25). This characteristic is useful for the treatment of oral cancer because we are able to directly transduce 11R-p53 into the tumor with ointment.

We also considered the disadvantage of the protein transduction method in the present study. The expression level of 11R-p53 was not maintained for >36 h. We needed repeated transduction of 11R-p53 for the inhibition of oral cancer cell proliferation. This problem may be solved by the following methods. A previous study showed that 2-methoxyestradiol induces the stabilization of the wild-type p53 protein in human cancer cell lines posttranscriptionally (26). Transduction of 11R-p53 with 2-methoxyestradiol may induce stabilization of 11R-p53. Another possible method is the transduction of the mutated p53 protein, which is resistant to ubiquitin-proteosome-mediated degradation. Simultaneous mutation of lysine residues 370, 372, 373, 381, 382, and 386 of the p53 protein to arginine residues generates a p53 molecule with potent transcriptional activity that is resistant to ubiquitin-proteosome-mediated degradation (19). Studies to develop a transduction method for stable 11R-p53 proteins are currently in progress in our laboratory.

In conclusion, we have provided three important findings: (a) the p53 protein-fused 11 arginine (11R-p53) was delivered and functioned in cancer cells; (b) transduction of 11R-p53 inhibited the proliferation of oral cancer cells; and (c) 11R-p53 enhanced the CDDP-induced apoptosis of oral cancer cells. These effects of 11R-p53 were the same as those of adenovirus-mediated delivery of the p53 gene. Our p53 protein transduction therapy is a novel method of cancer therapy and should be a promising approach.

We thank A. Kemori for technical assistance. We are grateful to Drs. T. Akiyama and K. Yoshikawa for luciferase reporter vector pGL2-basic that inserted a 2.4-kbp fragment of human p21/WAF1 promoter.