It is known that p53 alterations are commonly found in tumour cells. Another marker of tumorigenesis is FAK (focal adhesion kinase), a non-receptor kinase that is overexpressed in many types of tumours. Previously we determined that the N-terminal domain of FAK physically interacted with the N-terminal domain of p53. In the present study, using phage display, site-directed mutagenesis, pulldown and immunoprecipitation assays we localized the site of FAK binding to a 7-amino-acid region (amino acids 65–71) in the N-terminal proline-rich domain of human p53. Mutation of the binding site in p53 reversed the suppressive effect of FAK on p53-mediated transactivation of p21, BAX (Bcl-2-associated X protein) and Mdm2 (murine double minute 2) promoters. In addition, to functionally test this p53 site, we conjugated p53 peptides [wild-type (containing the wild-type binding site) and mutant (with a mutated 7-amino-acid binding site)] to a TAT peptide sequence to penetrate the cells, and demonstrated that the wild-type p53 peptide disrupted binding of FAK and p53 proteins and significantly inhibited cell viability of HCT116 p53+/+ cells compared with the control mutant peptide and HCT116 p53−/− cells. Furthermore, the TAT–p53 peptide decreased the viability of MCF-7 cells, whereas the mutant peptide did not cause this effect. Normal fibroblast p53+/+ and p53−/− MEF (murine embryonic fibroblast) cells and breast MCF10A cells were not sensitive to p53 peptide. Thus, for the first time, we have identified the binding site of the p53 and FAK interaction and have demonstrated that mutating this site and targeting the site with peptides affects p53 functioning and viability in the cells.
FAK (focal adhesion kinase) is a non-receptor 125 kDa protein tyrosine kinase that controls cellular proliferation, cell spreading, motility and survival [1–3]. FAK is overexpressed in many types of tumours [4–7]. Previously we have shown that FAK up-regulation occured in early stages of tumorigenesis . Previously, real-time PCR analysis of colorectal carcinoma and liver metastases has been used to demonstrate increased FAK mRNA and protein levels in tumour and metastatic tissues compared with normal tissues . Cloning and characterization of the FAK promoter demonstrated different transcription-factor-binding sites, including p53 that had potential importance for regulation of FAK transcription .
The structure of the FAK protein includes the N-terminal [FERM (4.1/ezrin/radixin/moesin)] domain with a primary auto-phosphorylation site, Tyr397, that directly interacts with the Src SH2 domain and PI3K (phosphoinositide 3-kinase), a central catalytic domain with two major sites of phosphorylation Tyr576/577, and a C-terminal domain containing two proline-rich segments and a FAT (focal adhesion targeting) subdomain that binds paxillin , talin , VEGFR-3 (vascular endothelial growth factor receptor 3)  and other proteins . The N-terminal domain of FAK associates with the EGFR (epidermal growth factor receptor), PDGFR (platelet-derived growth factor receptor) and c-Met receptors [13–15], with cytoplasmic tails of integrins, and with the death-receptor complex binding protein, RIP (receptor-interacting protein) . Previously, the N-terminal domain of FAK had been demonstrated in a complex with PIAS1 [protein inhibitor of activated STAT (signal transducer and activator of transcription) 1], causing sumoylation of FAK, increasing its autophosphorylation activity, and suggesting a novel role for FAK in nucleo-cytoplasmic transport signalling .
Previously we have demonstrated a direct interaction of p53 with the N-terminal domain of FAK [18,19]. We have based our study on the indirect link of FAK and p53 [9,20–23]. We performed mapping analysis and have shown that the FAK N-terminal domain (amino acids 206–422) binds the N-terminal transactivation domain of p53 (amino acids 1–92) . The binding of FAK and p53 has been demonstrated in different cell lines, both cancer cell lines and normal human fibroblasts . In addition, we showed that overexpression of FAK inhibited p53-induced apoptosis in SAOS-2 cells and decreased p53-mediated activation of p21, BAX (Bcl-2-associated X protein) and Mdm2 (murine double minute 2) targets in HCT116 p53+/+ cells .
In the present study, we have identified the 7-amino-acid binding site of FAK in p53 protein and found that it is located in the proline-rich region of human p53 protein, using phage display, site-directed mutagenesis, pulldown and immunoprecipitation assays. We demonstrate that mutating this site reversed the suppressive effect of FAK on the p53-transcriptional activity of p21, BAX and Mdm2 targets. In addition, p53 peptides containing this binding site were able to disrupt the binding of FAK and p53, to activate p53 and to inhibit the viability of HCT116 p53+/+ cells compared with HCT116 p53−/− cells, thus affecting p53-mediated signalling. The same effect was observed in MCF-7 cells. In normal MEF p53+/+ and MEF p53−/− (where MEF is murine embryonic fibroblast) cells and normal breast MCF10A cells, p53 peptides did not have a significant effect on cell viability. In summary, identification of this 7-amino-acid binding site is important for understanding FAK–p53 signalling.
Cell lines and culture
Human colon carcinoma cell lines, HCT116, p53+/+ (wild-type) and p53−/− were provided by Dr Bert Vogelstein (Johns Hopkins University, Baltimore, MD, U.S.A.). HCT116 cells were maintained in McCoy's 5A medium with 10% (v/v) foetal bovine serum and 1 μg/ml penicillin/streptomycin. The MCF-7 cell line was obtained from A.T.C.C. and maintained according to the manufacturer's protocol. MEF p53−/− and MEF p53+/+ cells were a gift from Dr S. May (UF Shands Cancer Center, University of Florida, Gainesville, FL, U.S.A.). MCF10A cells were maintained according to the A.T.C.C. protocol.
Plasmids and reagents
Full-length p53-pcDNA3 DNA and mutant p53-pcDNA3 was used for site-directed mutagenesis and then for a dual-luciferase assay. GST (glutathione transferase) recombinant pGEX4T1-p53 and pGEX4.1-Np53 plasmids were constructed as described in  and were used for site-directed mutagenesis and for the pulldown assay. Isolated purified recombinant baculoviral FAK protein was used for the pulldown assay and has been described previously . The p21-pGL3, BAX-pGL3 and Mdm2-pGL2 promoter luciferase constructs, described in , were provided by Dr Daiqing Liao (UF Shands Cancer Center, University of Florida, Gainsville, FL, U.S.A.). Monoclonal anti-FAK (4.47) antibody was obtained from Upstate Biotechnology, Y397 FAK antibody was obtained from Biosource, monoclonal p53 antibody (Ab-6), clone DO-1, was obtained from Oncogene Research Products and the β-actin antibody was obtained from Sigma.
Peptides used for testing their effect on cell viability were synthesized with the assistance of Dr Alfred Chung and the Protein Chemistry Core Facility (UF Shands Cancer Center, University of Florida, Gainesville, FL, U.S.A.). The TAT (YGRKKRRQRRR) sequence from the HIV sequence was conjugated to the p53-peptide sequence (RMPEAAP) and used for efficient cell penetration . As a control, the mutant TAT-conjugated p53 peptide with a mutated binding site (ΔRMPE)-VPL, GRKKRRQRRRVPL was used. Rhodamine-labelled TAT peptide was used to confirm the effective delivery of peptides under fluorescent microscopy in the cell lines.
For the phage display assay we used a degenerate 7-amino-acid phage display library, consisting of >2×109 sequences that was obtained from (New England Biolabs). Human recombinant GST–FAK-NT protein was used as a target for phage display. The GST–FAK-NT fusion protein was coated to a plate. The phage library was pre-cleared with GST protein and used for experiments. After three rounds of panning, recovered phages were sequenced to identify the potential peptides that bind FAK. All sequences common for both the experiment and control assays with GST protein were excluded from analysis. In two independent experiments we isolated two identical peptides that bind FAK. Using a BLAST (NCBI) search we identified a 7-amino-acid peptide homologous with Woodchuck (Marmota monax) p53 protein.
The site-directed mutation of the 7-amino-acid binding site in GST–p53 and GST–N-p53 plasmids, cloned and described in , was generated using the QuikChange® SL site-directed mutagenesis kit (Stratagene), according to the manufacturer's protocol. To generate a 4-bp deletion (ΔRMPE), we used the following oligonucleotides: forward, 5′-CCAGGTCCAGATGAAGCTCCCGCTGCTCCCCGCGTGGCCCCT-3′ and reverse, 5′-AGGGGCCACGCGGGGAGCAGCGGGAGCTTCATCTGGACCTGG-3′. To change the next three amino acids in the 7-amino-acid binding site (AAP→VPL), we used the following oligonucleotides: forward, 5′-CCAGGTCCAGATGAAGCTCCCGTTCCTCTCCGCGTGGCCCCTGCACCAGCAGCT-3′ and reverse, 5′-AGCTGCTGGTGCAGGGGCCACGCGGAGAGGAACGGGAGCTTCATCTGGACCTGG-3′. All mutant plasmids were sequenced at the Automated DNA Sequencing Facility at the University of Florida, Gainsville, FL, U.S.A. Expression of GST-proteins was confirmed by Coomassie Blue staining.
For the luciferase assay, 2×105 cells were plated on 6-well plates, cultured overnight and transfected with the pGL3 plasmids (1 μg/well) using Lipofectamine™ (Invitrogen) transfection agent according to the manufacturer's protocol. For normalization of the luciferase activity, pRL-TK control vector containing the herpes simplex virus thymidine kinase promoter encoding Renilla luciferase was used, resulting in its constitutive expression in a variety of cell types (Promega). The pRL-TK vector was used (0.1 μg/well) together with the pGL3 plasmids for co-transfection. The level of firefly luciferase activity was normalized to that of the Renilla luciferase activity in each experiment. For all experiments, cells were cultured for 24–48 h after transfection and lysed with 1×passive lysis buffer (Promega). Lysates were analysed using a dual-luciferase reporter assay system kit (Promega). Luminescence was measured on a Turner TD 20/20 luminometer (Promega). All experiments were performed at least three times.
Expression of recombinant GST fusion proteins
GST fusion proteins were engineered by PCR. The fusion proteins were expressed in Escherichia coli by incubation with 0.2 mM IPTG (isopropyl β-D-thiogalactoside) for 6 h at 37 °C. The bacteria were lysed by sonication, and the fusion proteins were purified with glutathione–agarose beads. The purified human FAK-NT protein was isolated from the FAK-NT–GST fusion protein by thrombin cleavage according to the manufacturer's protocol (Amersham Biosciences).
For the pulldown binding assay, purified recombinant baculoviral FAK protein (0.1 μg)  was pre-cleared with GST immobilized on glutathione–agarose beads by rocking for 1 h at 4 °C. The washed pre-cleared lysates were incubated with 2–4 μg of GST fusion protein immobilized on the glutathione–agarose beads for 1 h at 4 °C and then washed three times with RIPA lysis buffer [buffer A: 50 mM Tris/HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, 0.5% NaDOC, 0.1% SDS, 5 mM EDTA, 50 mM NaF, 1 mM NaVO3, 10% glycerol and protease inhibitors (10 μg/ml leupeptin, 10 μg/ml PMSF and 1 μg/ml aprotinin)]. Equal amounts of GST fusion proteins were used for each binding assay. Bound proteins were boiled in 2× Laemmli buffer and analysed by Western blotting.
Western blot analysis
Cells were washed twice with cold 1×PBS and lysed on ice for 30 min in a buffer containing 50 mM Tris/HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, 0.5% NaDOC, 0.1% SDS, 5 mM EDTA, 50 mM NaF, 1 mM NaVO3, 10% glycerol and protease inhibitors (10 μg/ml leupeptin, 10 μg/ml PMSF and 1 μg/ml aprotinin). The lysates were cleared by centrifugation at 5000 g for 30 min at 4 °C. Protein concentrations were determined using a Bio-Rad kit. The boiled samples were loaded on Ready SDS/10% PAGE gels (Bio-Rad) and used for Western blot analysis with the protein-specific antibodies. Immunoblots were developed with chemiluminescence Renaissance reagent (NEN Life Science Products).
Immunoprecipitation was performed as described in . Briefly, the pre-cleared lysates with equal amounts of protein were incubated with 1 μg of primary antibody and 30 μl of A/G agarose beads overnight at 4 °C. The precipitates were washed with lysis buffer (buffer A) three times and resuspended in 2× Laemmli buffer. The boiled samples were used for Western blot analysis as described above.
Cell viability assay
Cells were treated with peptides for 24 h at different peptide concentrations. The 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium compound from the Promega viability kit was added, and cells were incubated at 37 °C for 1–2 h. The attenuance on a 96-well plate was analysed with a microplate reader at 490 nm to determine cell viability.
A Student's t test was performed to determine significance. The difference between data with P<0.05 was considered significant.
Identification of the 7-amino-acid binding site in the proline-rich region of the N-terminal domain of p53 protein, interacting with the N-terminal domain of FAK
Previously we have shown that the N-terminal region of p53 interacts with the N-terminal domain of FAK . In order to determine the exact binding site in the p53 protein that interacts with FAK, we performed phage-display assays similar to the experiments with FAK-CD protein, when we identified the VEGFR-3 binding partner . For panning we used the recombinant N-terminal domain of FAK, the FAK-NT protein. The scheme of the phage-display assay is shown in Figure 1. Using phage-display assays with a 7-amino-acid peptide library from New England Biolabs and by BLAST searching we identified the peptide that had 5 out of 7 (71%) amino acids identical with the Woodchuck (M. monax) p53 protein (Figure 2). The peptide had 3 out of 7 (43%) amino acids identical with human p53 protein (Figure 2) and 4 out of 7 (57%) amino acids identical with mouse p53 protein (Figure 2). In humans, the peptide is located from amino acids 65–71 in a conservative proline-rich PXXP region (amino acids 61–95) of p53, important for protein–protein interactions .
The scheme of phage display assay
Sequence of the p53 peptide obtained using the phage display method
To confirm that the p53 peptide found by the phage-display assay is the binding site of human p53 with FAK, we performed site-directed mutagenesis by deleting the first four amino acids (RMPE) of the p53-binding site and mutating the next three amino acids AAP to VPL in human GST–p53 (Figure 3). Thus this mutation changes all amino acids in the 7-amino-acid binding site of human p53. We created GST–p53 and GST–N-terminal domains of p53 fusion proteins that contained these mutant 7-aminoacid regions, termed GST–p53 mut7 and GST–N-p53mut7 respectively (Figure 4A, marked by asterisks). We used these GST-fusion proteins to perform a pulldown assay with purified baculoviral FAK protein  (Figure 4A). Mutation in the 7-amino-acid binding site of p53 in the N-terminal GST–p53 and GST–p53 significantly decreased binding with FAK protein compared with the wild-type p53 and FAK protein binding (Figure 4A). Thus the 7-amino-acid region of p53 detected by phage display is critical for binding of FAK and p53 in vitro.
Site-directed mutagenesis of p53–GST fusion proteins
The binding of the FAK protein to wild-type p53 protein is stronger than to the mutant p53 protein containing the mutated binding site
To detect that mutation of the 7-amino-acid binding site abrogates FAK and p53 binding in vivo, we transfected HCT116 p53−/− cells with either wild-type p53 or mutant p53, containing the 7-amino-acid change in the binding site that was shown in Figure 3. Immunoprecipitation of p53 caused binding of FAK and wild-type p53, whereas mutant p53 did not bind FAK in HCT116 cells (Figure 4B). Thus the 7-amino-acid binding site is critical for binding FAK and p53 in the cells in vivo.
Thus, by phage display, site-directed mutagenesis, pulldown assays and immunoprecipitation we identified the 7-amino-acid binding site inside the p53 protein that interacts with FAK protein.
Mutation of the FAK-binding site in p53 reverses FAK-directed inhibition of p53 transcriptional activity of p21, BAX and Mdm2 targets
Next, we investigated the effect of mutation in the binding site of p53 on function. We have shown previously that FAK inhibited p53-mediated p21, Mdm2 and BAX promoter luciferase activities through physical interaction with the p53 protein . Thus we performed experiments where we co-transfected the FAK plasmid with either the wild-type p53 or the p53 plasmid containing the mutated 7-amino-acid binding site together with p21-luciferase p53 target in HCT116 p53−/− cells (Figure 5A). Previously we have shown that FAK blocked p53-induced activity of p21 . In the present study, FAK was not able to block the p53-induced activity of p21 when p53 contained a mutation at the p53 FAK-binding site (Figure 5A). Thus mutation of the FAK-binding site in p53 reversed FAK-inhibited p53-transcriptional activity of the p21 target. The same result was obtained for the p53-induced target BAX (Figure 5B). FAK inhibited p53-directed activation of the BAX-luciferase promoter, but not with p53 that contained a mutation of the FAK-binding site (Figure 5B). An even more significant effect was observed with the Mdm2 target (Figure 5C). FAK significantly blocked p53-mediated activation of Mdm2 in the case of wild-type p53, but not with the mutant p53 similar to p21 and BAX targets (Figure 5C). Thus mutation of the binding site between FAK and p53 causes a defect in the p53-trancriptional activity of p21, BAX and Mdm2 targets by FAK.
FAK blocks p53 transcriptional activity in the case of wild-type p53, but not with the mutant p53 containing a mutated binding site
p53 peptides containing the binding site of FAK and p53 cause loss of cancer cell viability
To determine the biological effect of FAK and the p53-binding site, we synthesized the TAT-conjugated 7-amino-acid p53 peptide, containing the binding site with FAK (p53 peptide) and the TAT-conjugated control peptide with the same 7-amino-acid mutation, as shown in Figure 3. In the present study the TAT-peptide containing the 7-amino-acid mutation of the FAK-binding site is termed the control peptide. The wild-type p53 peptide, but not the control peptide, caused a significant decrease in cell viability in HCT116 p53+/+ cells compared with HCT116 p53−/− cells, demonstrating specificity of the p53 peptides for HCT116 p53+/+ cells (Figure 6A).
p53 peptides containing the binding site with FAK inhibit HCT116 p53+/+ colon cancer cell viability
To demonstrate that p53 peptide affected p53 function, we analysed the effect of p53 peptide on p53-transcriptional activity of the p21 target (Figure 6B). We co-transfected p53 and FAK plasmids together with the p21 plasmid in the HCT116 p53−/− cells treated with either the wild-type p53 or control p53 peptide (Figure 6B). The wild-type p53 peptides significantly increased and reversed FAK-inhibited p53 transcriptional activity of the p21-target compared with untreated cells or cells treated with the control peptide (Figure 6B). Thus p53 peptides containing the binding site with FAK significantly increased transactivation activity of p53 consistent with decreased viability in HCT116 p53+/+ cells.
The same effect of p53 peptides on viability was observed in the wild-type p53 breast cancer cell line MCF-7 (Figure 7, upper panel). The wild-type p53 peptide caused a significant decrease in cell viability compared with the control peptide in MCF-7 cells (Figure 7). In addition, we performed immunostaining of p53 in both HCT116 p53+/+ and MCF-7 cells (Figure 7, lower panel). Both of these cells treated with the wild-type p53 peptides demonstrated increased nuclear staining of p53 compared with control peptides (Figure 7, lower panel). In normal mouse fibroblast MEF p53+/+ and MEF p53−/− cells, p53 peptides did not significantly decrease cell viability (Figure 8A). In MEF p53−/− cells transfected with the human p53 plasmid, p53 peptides also did not cause a significant decrease in the viability of the cells compared with untreated MEF p53−/− cells (Figure 8A). No difference in cell viability was observed in normal breast cancer cells MCF10A (Figure 8B, upper panel). Immunostaining of p53 detected did not increase the intensity of nuclear p53 staining in MCF10A cells (Figure 8B, lower panel). Thus p53 peptides containing the FAK–p53 binding site specifically decreased viability in human cancer cells.
p53 peptides containing the binding site with FAK inhibit MCF-7 breast cancer cell viability
p53 peptides containing the binding site with FAK do not inhibit the viability of normal fibroblast MEF and breast MCF10A cells
To show the differences between the cancer and normal cells, we compared the levels of FAK, active FAK (Tyr397-FAK) and p53 proteins in the cells (Figure 9A). Normal MEF and MCF10A cells had lower levels of activated FAK than cancer HCT116 and MCF-7 cell lines (Figure 9A), which is consistent with the differences in viability in response of these cell lines to p53 peptides.
Disruption of FAK and p53 binding with p53 peptides containing the binding site with FAK
Disruption of FAK–p53 protein binding by p53 peptides containing the binding sites of FAK and p53
To demonstrate the disruption of FAK and p53 binding by p53 peptides in different cell lines, we analysed the association of FAK and p53 proteins in cells by immunoprecipitation performed either with wild-type p53 or control p53 peptides (Figure 9B). We performed immunoprecipitation of FAK and p53 in p53-positive cells (HCT116 p53+/+, MCF-7, MEF p53+/+ and MCF10A cells; Figure 9B). These results clearly show that p53 peptide disrupted the binding of FAK and p53 in HCT116 p53+/+ cells, whereas control peptide did not affect the binding (Figure 9B, upper left-hand panels). The p53 peptides also disrupted the binding of FAK and p53 in MCF-7 cells, whereas control peptides did not (Figure 9B, upper right-hand panels). In contrast, p53 peptides did not disrupt binding of FAK and p53 in mouse fibroblast MEF p53+/+ cells (Figure 9B, lower left-hand panels). In human MCF10A cells, wild-type p53 peptide also disrupted the binding of FAK and p53, whereas the control peptide with the mutated protein-binding site did not affect the FAK and p53 binding (Figure 9B, lower right-hand panels). Thus p53 peptides specifically and effectively disrupted binding of human p53 and FAK in human cell lines.
Thus p53 peptides containing the binding site of FAK and p53 caused disruption of FAK and p53 binding and p53-dependent loss of viability in human cancer cells.
In the present study, we identified the exact 7-amino-acid binding site in the p53 protein that interacts with FAK protein. Phage display and pulldown and immunoprecipitation assays demonstrated that this binding site is located in the proline-rich region of human p53. In addition, we demonstrated that mutating this binding site reversed the suppressive effect of FAK on p53 transcriptional activity with its targets (p21, BAX and Mdm2). Also, synthetic peptides containing the binding site of p53 with FAK caused disruption of FAK and p53 binding and loss of p53+/+cancer cell viability in a p53-dependent manner in human cancer HCT116 cells. The same was observed in MCF-7 cells. In addition, in normal mouse MEF cells and human MCF10A cells there was no significant difference in viability between the p53 peptide and the control mutant peptide. The results suggest that therapeutics can be developed that target the FAK–p53 interaction in cancer cells with overexpressed FAK.
We have shown previously that the N-terminal domain of FAK protein interacts with the N-terminal domain of p53 . In the present study we localized the site of this binding to seven amino acids of the N-terminal transactivation domain of p53 located in its proline-rich region from amino acids 65–72. This finding is critical for understanding intracellular signalling. It is well-known that the proline-rich region of human p53, localized between amino acids 64 and 92, is necessary for efficient growth suppression . The proline-rich region of p53 has been shown to be required for inducing cell-cycle arrest and apoptotic signalling through inducing pro-apoptotic genes and inhibiting anti-apoptotic genes . The proline-rich region of p53 PXXP has been shown to co-operate with anti-neoplastic agents to cause tumour cell apoptosis .
Furthermore, the proline-rich region of p53 has been shown to be important for binding to the nuclear matrix (the non-chromatin sub-structure of the nucleus important for DNA repair, transcription, recombination and replication) . We have shown that FAK and p53 interacted in both the cytoplasm and nucleus . It will be important to analyse the subnuclear fractions of this interaction in future studies. Interestingly, another tyrosine kinase protein–Etk/BMX, a member of the Btk (Bruton's tyrosine kinase) family that is important for migration and survival functioning, and is activated by the N-terminal domain of FAK  also binds to the same proline-rich region of p53, in a similar manner to FAK . The BMX kinase and p53 localized in both the cytoplasm and nucleus, and the deletion of the proline-rich region of p53 resulted in nuclear localization of p53 . Since BMX inhibited p53 transcriptional activity and functioning, the authors suggest that cytoplasmic retention of p53 by BMX may result from masking the nuclear localization signal of p53 by bound BMX and thus inhibiting p53 functioning. Thus BMX and FAK proteins can represent novel signalling with survival kinases that sequester p53 from apoptotic functioning, impairing its intracellular localization and cytoplasmic–nuclear shuttling. In addition, the proline-rich region of p53 can create binding sites for SH3 (Src homology 3) domains and can represent a site of protein–protein interaction . The proline-rich region of p53 has been proposed to be a molecular switch regulating either apoptosis or cell-cycle arrest . The proline-rich domain of p53 has been shown to be important for the transcription-independent apoptotic function of p53 through binding and activation of the pro-apoptotic Bcl-2 protein BAX in the cytoplasm . Another important feature of the proline-rich region of p53 is that it contains the polymorphic amino acid, Pro72, which frequently occured and changed to an arginine residue in different tumours. Pro72 decreases p53 apoptotic functioning, affecting tumorigenesis and survival signalling . Thus FAK binding to p53 inside its proline-rich region domain can influence binding of p53 with other proteins and change its intracellular localization and intracellular functioning.
In the present study, we have shown that a 4-bp deletion (Δ-RPME) in the binding site and 3-amino-acid mutation (AAP to VPL) changed the p53-transcriptional activity of p53 with p21, BAX and Mdm2 targets. The effect was more dramatic on the Mdm2 target, suggesting that the FAK and p53 interaction can more strongly influence Mdm2–p53 binding and signalling. The proline-rich region of p53 (amino acids 63–97) was also proposed as a spacer between the transactivation domain, that interacts with transcriptional regulators such as Mdm2 and p300, and the DNA-binding central domain (amino acids 102–292) . Thus FAK can affect Mdm2 regulation through its binding to the p53 proline-rich region. It has been show that Mdm2 also binds the N-terminal domain of p53 and negatively regulates p53 transcription function . The p53 and Mdm2 proteins comprise a feedback loop, as p53 regulates Mdm2 transcription, whereas Mdm2 causes proteasome-mediated p53 degradation . It is possible that a feedback regulation loop exists in FAK–p53–Mdm2 activity. We have shown that a p53 peptide disrupted the FAK and p53 complex in HCT116 p53+/+ cancer cells, but not in MEF p53 +/+ normal cells. This can be explained by structural differences of mouse and human p53 proteins. In MEF p53−/− cells transfected with human p53, the binding of p53 with FAK protein was disrupted by p53 peptides (results not shown), but p53 peptides did not change viability in the cells similar to normal MEF p53+/+ and MCF10A cells which can be explained by less-activated FAK in normal cells. Thus the therapeutic effect of p53 peptides can be applied to cancer cells with overexpressed activated FAK .
The growth inhibitory effect of p53 peptides on cancer cells, but not in normal mouse fibroblasts or normal MCF10A breast cancer cells is consistent with the effect of a dominant-negative FAK inhibitor, the C-terminal domain of FAK-CD  or VEGFR-3 AV3 peptides, containing a binding site with FAK . The normal fibroblasts and MCF10A cells are different from the cancer cells. The inhibitor of FAK, dominant-negative FAK-CD, did not cause apoptosis in normal cells, whereas it caused apoptosis in cancer cells . Also, recently we have demonstrated that cancer cells were sensitive to the novel FAK phosphorylation inhibitor, TAE226 (Novartis), whereas normal MCF10A cells were resistant to the inhibitor . Normal and cancer cells have different intracellular signalling, for example an inhibitor of actin polymerization caused apoptosis in normal MCF10A cells, but not in metastatic mammary MDA-MB-453 cells, suggesting that MCF10A cells are more dependent on actin polymerization than mammary carcinoma cells .
In summary, the present study is the first one that has exactly identified the binding site of the FAK and p53 interaction. The binding is located within amino acids 65–71, a proline-rich domain of p53, and is very important for survival/apoptotic signalling in cells. The present study shows functional significance of this binding site in FAK–p53 mediated signalling, as mutating this site affected FAK-inhibited p53 transcriptional activity. In addition, p53 peptides containing the binding site of FAK and p53, disrupted binding of FAK and p53 in human cancer cell lines and decreased cancer cell viability. Thus our results provide a basis for future targeted therapy of this protein interaction.
We would like to thank Dr Alfred Chung and the Protein Chemistry Core Facility at the University of Florida for the synthesis and purification of peptides. This work is supported by an NIH (National Institutes of Health) grant number CA 65910 (to W. G. C.) and Susan G. Komen for the Cure grant number BCTR0707148 (to V. M. G.).