SMAR1-derived P44 Peptide Retains Its Tumor Suppressor Function through Modulation of p53*

The use of pharmacologically active short peptide sequences is a better option in cancer therapeutics than the full-length protein. Here we report one such 44-mer peptide sequence of SMAR1 (TAT-SMAR1 wild type, P44) that retains the tumor suppressor activity of the full-length protein. The protein transduction domain of human immunodeficiency virus, type 1, Tat protein was used here to deliver the 33-mer peptide of SMAR1 into the cells. P44 peptide could efficiently activate p53 by mediating its phosphorylation at serine 15, resulting in the activation of p21 and in effect regulating cell cycle checkpoint. In vitro phosphorylation assays with point-mutated P44-derived peptides suggested that serine 347 of SMAR1 was indispensable for its activity and represented the substrate motif for the protein kinase C family of proteins. Using xenograft nude mice models, we further demonstrate that P44 was capable of inhibiting tumor growth by preventing cellular proliferation. P44 treatment to tumor-bearing mice prevented the formation of poorly organized tumor vasculature and an increase in hypoxia-inducible factor-1α expression, both being signatures of tumor progression. The chimeric TAT-SMAR1-derived peptide, P44, thus has a strong therapeutic potential as an anticancer drug.

The integrity of the eukaryotic genome requires several layers of control to ensure that replication of DNA occurs only once during a cell cycle (1). For normal cellular functions and tissue homeostasis, accurate transmission of genetic information between generations is required. Dysregulation of the cell cycle control is a hallmark of cancer (2). Neoplastic progression has been demonstrated to involve increased genetic instability (3)(4)(5), and there are enough reports revealing that the disruption of multiple pathways is required for the development of cancer (6,7). Inactivation or loss of p53 is a common event associated with the development of ϳ60% of all human cancers (8 -11). p53 has been shown to participate in the regulation of several processes, which might inhibit tumor growth, including differentiation, senescence, and angiogenesis (12)(13)(14). However, central to the function of p53 appears to be the ability to induce both cell cycle arrest and/or apoptosis in stressed cells, partly by activating expression of p53-responsive target genes that mediate these responses (15,16). The precise mechanism of p53 activation by cellular stress is of intense interest and may involve both an increase in p53 protein level and in the specific activity of p53 by covalent modifications (17)(18)(19).
Despite tremendous efforts in molecular, biochemical, and cell biological research toward understanding the intra-and extracellular mechanisms involved in the transformation of a normal cell into a cancerous one, the number of successful treatments against cancer is few. A major limitation of cancer therapeutics is the problem of delivering pharmacologically relevant compounds, peptidyl mimetics, antisense oligonucleotides, and proteins into cells (20,21). Peptide-based drugs have limitations in the form of the poor permeability and selectivity of the cell membrane. These problems are now circumvented by attaching protein translocation domains (PTDs) 3 to the peptides that can cross the biological membranes efficiently without any dependence on transporters or specific receptors and mediate the intracellular delivery of a range of biological cargoes (22)(23)(24). The PTD of HIV-1 Tat protein is well known to mediate transduction of heterologous peptides and biologically active proteins in vitro and in vivo (25), and thus it has been shown to be of considerable interest for protein therapeutics (26 -28).
The present investigations were aimed at using the PTD of Tat protein to deliver short peptide sequences of tumor suppressor protein SMAR1 both in vitro and in vivo. SMAR1 (Scaffold Matrix Attachment Region binding protein 1), a 68-kDa protein, has been shown previously to interact with p53 and to regulate the cell cycle (29). It has been further reported that the arginine-serine-rich domain (RS domain) of SMAR1, upon phosphorylation by the PKC family of proteins, phosphorylates p53 specifically at its serine 15 residue (30). The minimal domain of SMAR1 required for its interaction with p53 resides within the 160 -350-aa region. In this study, we show that compared with the full-length protein, a 33-mer SMAR1 peptide (corresponding to its RS domain) conjugated to an 11-mer PTD of TAT protein (TAT-SMAR1 WT; designated as P44) is suffi-* This work was supported in part by the Department of Biotechnology, New cient enough to reduce the tumor growth in nude mice. Exposure of cells to the P44 peptide resulted in increased p53 phosphorylation at its serine 15 residue, in turn activating the p53-mediated cell cycle control. Point mutation studies of the P44 peptide further revealed that the serine 347 residue within the serine-rich motif of SMAR1 plays a pivotal role in mediating the tumor suppressor effect of SMAR1. The 347-serine residue represents the substrate motif for the PKC family of proteins, and its phosphorylation is necessary for activating the p53-dependent pathway. We also observed that tumors excised from mice treated with the SM mutant peptide showed leaky vascular architecture compared with P44-treated tumors. Interestingly, there was no detectable level of hypoxia-inducible factor-1␣ (HIF-1␣) in tumors from mice treated with P44 peptide that is a hallmark of tumor hypoxia. Thus, our results implicate that P44 SMAR1 peptide can be used as an alternative drug for cancer therapy.

EXPERIMENTAL PROCEDURES
Plasmid Construction and Transfection-Various truncated forms of SMAR1 were subcloned in-frame using enhanced GFP vector. For the 160 -350-aa truncation, PCR of pBKCMV-SMAR1 (29) was performed using forward and reverse primers that generated EcoRI and BamHI sites at the 5Ј and 3Ј ends, respectively. The PCR product was first cloned into Topo-TA cloning vector (Invitrogen), and the EcoRI-BamHI fragments from in-frame clones were then subcloned into EGFP-C1. For cloning of 160 -288-and 288 -350-aa regions, the PvuII site present at 288-aa junction of SMAR1 (30) was used. The EcoRI-PvuII fragment was cloned into the EcoRI-SmaI site of EGFP-C1. For cloning of the 288 -350-aa region, the insert fragment was isolated by using PvuII-BamHI. The ligated products in the EGFP-C1 vector were then transfected for visualization under microscope. The protein expression of the SMAR1 truncations was further confirmed by immunoblotting with anti-GFP.
Labeling of SMAR1 Peptides with TRITC Fluorochrome-Twenty five g of the TAT-SMAR1-derived peptides dissolved in PBS was mixed with equimolar amount of 0.1 M di-sodium tetraborate buffer, pH 9.0. This mixture was then incubated for 45 min in the dark at room temperature with 10 g of TRITC (5 mg/ml stock solution) dissolved in Me 2 SO. The reaction was stopped by adding 1 M Tris-glycine. The peptide-TRITC conjugate was loaded on the top of a PD-10 column (Bio-Rad) that was previously equilibrated with PBS. The column was eluted with 10 ml of PBS, and the first few fractions of the fluorescent material were collected. The UV absorbance for the labeled peptide was measured at 550 (TRITC) and 280 nm (peptide).
Western Blotting and Immunoprecipitation-HEK 293 cells (5 ϫ 10 5 ) were cultured as exponentially growing subconfluent monolayer on 35-mm plates in DMEM (Invitrogen) supplemented with 10% (v/v) fetal calf serum. After 24 h, cells were treated with varying concentrations of either of the TAT-SMAR1-derived peptides (Table 1). Cells were then incubated at 37°C for 12 h followed by preparation of whole cell protein extracts. For Western blotting, an equal amount of the protein was separated on 10% SDS-PAGE and subsequently transferred to polyvinylidene difluoride membrane (Amersham Biosciences). It was finally probed with the following antibodies: anti-p53 (DO-1); anti-phospho-p53 Ser-15, and anti-pCdc-2. The detailed protocol for the same is discussed in our earlier publication (30).
Immunocytochemistry and Confocal Imaging-For direct detection of TRITC-labeled peptides, HEK 293 cells were plated directly on a glass coverslip and cultured overnight prior to their treatment to TRITC-conjugated various TAT-SMAR1 derived peptides. After 12 h of incubation, three washings with cold PBS were given, and the cells were fixed with 3.7% paraformaldehyde before being mounted in PBS/glycerol (1:1) containing antifading agent. For indirect immunodetection, 2 ϫ 10 5 HEK 293 cells were plated and cultured overnight in 35-mm plates on glass coverslips. The cell monolayer was then treated with either of the various TAT-SMAR1-derived peptides dissolved directly in complete DMEM at the appropriate concentration (final concentration 10 M). After 12 h of incubation, the cells were washed twice with cold PBS and fixed with 3.7% paraformaldehyde. Subsequently, fixed cells were stained for total p53 using anti-p53 (DO-1) (Santa Cruz Biotechnology) for an hour at room temperature. For detection, cells were incubated with a secondary antibody mixture containing fluorescein isothiocyanate-conjugated anti-mouse IgG antibodies (Sigma) for 1 h at room temperature. Slides were then mounted in antifade mounting medium (Dako) and analyzed with a confocal laser scanning microscope (CLSM 510, version 2.01; Zeiss, Thornwood, NY).
Cell Cycle Analysis by Flow Cytometry-HEK 293 cells were transiently transfected with GFP-tagged SMAR1 truncations (160 -288 and 288 -350 aa). After 48 h of transfection, the cells were trypsinized, washed with 1ϫ PBS, and fixed in 70% icecold ethanol. After incubating at Ϫ20°C for 20 min, the cells were spun at 1000 rpm for 5 min at room temperature. The cells were washed with PBS, treated with RNase A (75 units/ml) for 30 min at 37°C, washed again in PBS, and resuspended in PBS containing 50 g/ml propidium iodide. After staining the cells with propidium iodide, they were analyzed by FACSVantage (BD Biosciences) using the cell quest program (Verity Software) for cell cycle profiles.
In Vitro Phosphorylation Assay-HEK 293 cells were seeded at a density of 5 ϫ 10 5 per well and harvested after 24 h either for preparation of whole cell extracts or treatment with 20 nM concentration of staurosporine (a PKC inhibitor) (31,32). After 48 h of treatment, lysates were prepared using the kinase lysis buffer (20 mM Tris, pH 8.0, 500 mM NaCl, 1 mM EDTA, 1 mM EGTA, 10 mM ␤-glycerophosphate, 10 mM NaF, 10 mM p-nitrophenyl phosphate, 30 mM Na 3 VO 4 , 1 mM benzamide, 2 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 0.25% Nonidet P-40, and protease inhibitor mixture). After incubating on ice for 20 min, lysates were spun at 14,000 rpm for 10 min. For the kinase reaction, various TAT-SMAR1-derived peptides (Table 1), were incubated with 2 g of either the whole cell kinase extract or with staurosporine-treated kinase extracts along with 10 M [␥-32 P]ATP, 2 mM MgCl 2 , and the kinase assay buffer (20 mM Tris, 2 mM MgCl 2 , 2 mM CaCl 2 , 10 mM ␤-glycerophosphate, 10 mM NaF, 10 mM p-nitrophenyl phosphate, 30 mM Na 3 VO 4 , 1 mM benzamide, 2 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, and protease inhibitor mixture). The reaction mixture was incubated at 30°C for 30 min, and the reaction was stopped by adding 0.5 mM EDTA. Bovine serum albumin was added at a final concentration of 1 mg/ml along with 10% trichloroacetic acid. Trichloroacetic acid-precipitated peptides were then resuspended in scintillation fluid and checked for 32 P incorporation.
Luciferase Reporter Assay-HEK 293 (p53 ϩ/ϩ ) and K562 (p53 Ϫ/Ϫ ) cells were grown in DMEM supplemented with 10% fetal bovine serum in the presence of 5% CO 2 at 37°C. A total number of 1 ϫ 10 6 cells were plated on a 6-well plate. After 24 h, cells were transiently transfected using Lipofectamine-2000 with either 1 g of p21 expression plasmid having luciferase reporter gene or 5 g of p21 luciferase along with the WT p53 construct in the K562 cell line. Two micrograms of pRL-CMV (Renilla luciferase reporter plasmid) was included in all transfections to normalize the transfection efficiency. Thirty six hours post-transfection, cells were treated with 10 M of either P44 peptide or its various mutants. The cells were harvested 12 h post-treatment, washed with PBS, and lysed in 1ϫ Passive lysis buffer (Promega). After three freeze-thaw cycles, cells were spun at 10,000 rpm at 4°C for 20 min. The supernatants were collected, and protein concentrations were estimated spectrophotometrically using Bradford reagent (Bio-Rad). According to the manufacturer's instructions, luciferase activity was assessed using the dual luciferase assay reporter kit (Promega). The luciferase activity was measured by using Fluoroskan Ascent luminometer (Labsystems). For all the luciferase assays, the data shown are the mean Ϯ S.D. of three independent experiments.
Microarray Analysis-Microarray experiment was commercially performed by Agilent Genotypic Technology, Bangalore, India. Significantly regulated genes are presented with the folds Ͼ0.5 for down-regulation between the mean experiment values of samples treated with either TAT-SMAR1 WT (P44) or TAT-SMAR1 RS mutant (SM) peptide. The experiment was done in duplicate.
Animal Model for Tumorigenesis-B16F1 mouse melanoma cells in the exponential growth phase were trypsinized (Invitrogen) and washed twice with PBS. Cell number and viability were assessed, and cell cultures with viability Ͼ90% were used. Tumors were then established in nude mice by subcutaneous injection of 2 ϫ 10 6 B16F1 cells. Five mice were used in each set of experiments. The mice were maintained under pathogenfree conditions. When the subcutaneous tumor was clearly visible, the TAT-SMAR1 WT (P44) peptide treatment was started. The P44 peptide was subcutaneously injected proximal to the tumor sites at a dose of 200 g/ml mouse three times a week. The treatment was administered for 4 weeks. For control treatment, TAT PTD alone and TAT SMAR1 RS mutant peptide (SM) were injected in the same manner as the P44 peptide.
Immunohistochemical Staining of Tumor Sections-Tumor sections in paraffin-embedded blocks were transferred to poly-L-lysine-coated glass slides and air-dried overnight at 37°C. They were dewaxed in xylene (three changes) and rehydrated in a graded series of decreasing ethanol concentrations. After deparaffinization and rehydration, antigen retrieval was performed by immersing the slides in 10 mM sodium citrate buffer (pH 6.0) and subjected to microwave irradiation for 10 min. After antigen unmasking, a cooling off period of 30 min preceded the incubation of the primary antibody (anti-HIF-1␣; 1/100 dilution; Santa Cruz Biotechnology). Thereafter, HIF-1 was detected by using Cy-3 fluorescence-labeled secondary antibody. All sections were counterstained with hematoxylin and eosin and dehydrated in alcohol and xylene. Tissue samples with nonimmune serum served as negative controls.

SMAR1-(288 -350), Arginine-Serine-rich Motif, Responsible
for Cell Cycle Regulation-Earlier, we have reported that SMAR1 interacts with p53 and modulates its function (29,30). To delineate the minimal domain of SMAR1 that interacts with and activates p53, several truncations of SMAR1 were performed (30). SMAR1 is a 548-aa protein that is alternatively spliced, resulting in a deletion of 39 aa at its N terminus (33). Fig. 1A shows the schematic representation of different domains of SMAR1. It includes an arginine-serine-rich (RS) motif that corresponds to 288 -350 aa and a tetramerization domain located at its C terminus (33). The nuclear localization signal (NLS) of SMAR1 corresponds to 245-288 aa ( Fig. 1A) (30).
In our earlier study, we demonstrated that the SMAR1-(160 -350) domain is the minimal domain of SMAR1 that interacts with p53 and modulates its activity through phosphorylation at its Ser-15 residue (30). The 160 -350-aa core domain of SMAR1 includes both the NLS domain (aa 245-288) as well as the RS domain (aa 288 -350). To further evaluate the functional motif within residues 160 -350, both the NLS and RS domains (GFP-tagged) were analyzed for their role in cell cycle regulation. HEK 293 cells were transiently transfected with GFPtagged 160 -288-and 288 -350-aa truncations of SMAR1, and their protein expression was confirmed using GFP antibody (Fig. 1B). After 48 h, fluorescence-activated cell sorter-based cell cycle analyses were performed. Cells transfected with 288 -350 aa showed a significantly altered cell cycle. Compared with mock-or 160 -288-aa-transfected cells (Fig. 1C, left and middle panels, respectively), overexpression of SMAR1-(288 -350) resulted in 17% of the cells remaining at the G 2 /M phase (Fig.  1C, right panel). Interestingly, 29% of cells were found to be in the S phase. Thus, the 288 -350-aa region is the functional motif within SMAR1-(160 -350) that modulates the cell cycle at S to G 2 /M transition as well as the G 2 /M checkpoint.
Uptake and Nuclear Compartmentalization of Chimeric TAT PTD-SMAR1-derived Peptides-Because the amino acids 288 -350 play a critical role in p53 modulation (30) and in turn cell cycle regulation, we commercially synthesized shorter SMAR1-derived peptides to evaluate their efficacy in vitro and in vivo. Several reports have established that the chemical conjugation of the PTD derived from HIV-1 TAT protein was able to induce cellular internalization of large proteins such as ␤-galactosidase or horseradish peroxidase (34,35). With the pros-pect of using similar tools for drug delivery, it was of interest to design SMAR1-derived peptide and explore its mechanism of internalization. A 33-mer peptide sequence extending from aa residues 324 to 357 was conjugated with 11-mer TAT-PTD and designated as TAT-SMAR1 WT (P44). This short peptide sequence overlapped with the PKC substrate motif (the serine rich motif) of the full-length SMAR1, known to be involved in its phosphorylation and subsequent nuclear accumulation of p53 (30). Various serine mutants, where serine was replaced by alanine, of the P44 peptide (PS347A, PS348A, PS349A, and PS350A) and SMAR1 RS mutant (SM), were also commercially synthesized (Table 1). To understand the mechanism of uptake and intracellular compartmentalization of all these peptides, they were labeled with TRITC fluorochrome and purified using PD-10 column. Upon exposure of cells to either of the peptides, except the control ( Fig. 2A), all others were observed to get localized into the nucleus (Fig. 2, B-G).
Internalization of all TAT-SMAR1 chimeric peptides within the nucleus occurred in a dose-dependent manner (data shown only for P44 peptide) (Fig. 2, H-K) as observed by the intensity of the recorded signal. At a concentration of around 50 M (which seemed to be saturating), TRITC-labeled peptides could also be detected in the cytoplasm (Fig.  2K). There was no overall variation in the uptake and localization between the various SMAR1-derived TAT-conjugated peptides (Fig. 2, B-G). A non-TAT-conjugated SMAR1 peptide labeled with TRITC was used as negative control, which, as expected, was not taken up by the cells as recorded with no fluorescent signal ( Fig. 2A). The results thus confirmed that the chimeric peptides of SMAR1 were efficiently translocated into the nucleus by TAT PTD.
TAT-SMAR1 WT Peptide (P44) Modulates p53 Function-To examine whether P44 peptide could alone activate p53, luciferase reporter assays were performed following treatment of cells with the chimeric TAT-SMAR1-derived peptides. HEK 293 (p53 ϩ/ϩ ) cells were transfected with p21 expression plasmid having a luciferase reporter gene, and K562 (p53 Ϫ/Ϫ ) cells were co-transfected with p21 luciferase reporter and p53 constructs. Treatment of cells with 10 M of P44 peptide resulted in a 2-fold activation of p53 in p53 ϩ/ϩ (Fig. 3A, bar 3) and about 1.5-fold activation in p53 Ϫ/Ϫ cell line (Fig. 3B, bar 4). In the presence of P44 peptide alone, there was no activity in p53 Ϫ/Ϫ cells indicating that peptide itself does not transactivate the p53-responsive p21 promoter (Fig. 3B, bar 3). However, point mutation at serine 347 (PS347A) as well as the SMAR1 RS mutant peptide (SM) resulted in complete loss of p21 promoterdriven luciferase activity in both p53 WT and p53 null cell lines (Fig. 3, A, bars 4 and 8, and B, bars 5 and 9, respectively). No significant alteration in activation of p53 was observed by peptides with point mutations at serine 348 (PS348A), serine 349 (PS349A), or serine 350 of SMAR1 (PS350A) (Fig. 3, A, bars 5-7, and B, bars 6 -8). Thus, these results indicate that serine 347   within the RS domain (288 -350 aa) of SMAR1 is necessary for modulating p53 activity. P44 Peptide Controls the Activity of Cell Cycle Regulatory Proteins- Fig. 4A represents a schematic of TAT-SMAR1-derived chimeric peptide demonstrating the fusion of PTD domain of the TAT protein with the RS domain of SMAR1. The minimal dose of P44 peptide that was functionally effective in activating p53 was evaluated by incubating HEK 293 cells with varying concentrations of P44 peptide (1-100 M) (Fig. 4B), and after 12 h of incubation, protein lysates were prepared and processed for immunoblotting with total p53 (DO-1) and p53 Ser-15 phospho-specific antibodies. A slight increase in the total p53 expression was observed. However, there was a 4.5fold increase in the expression of phosphorylated p53 (phosphoserine 15 p53) in P44 peptide-treated cells (Fig. 4B, lanes  2-6). To exclude the possibility that p53 activation is an effect of the protein transduction domain of HIV-1 TAT protein, a similar set of experiments was performed using TAT-PTD peptide alone (consisting of only the 11-mer TAT sequence). As expected, no phosphorylation of the serine 15 residue of p53 was observed in TAT peptide-treated cells even at much higher concentrations (50 -100 M) (Fig. 4C, lanes 1 and 2) with respect to P44 peptide (5-10 M) (Fig. 4C, lanes 3 and 4). This observation reconfirmed that the RS domain (288 -350 aa) of SMAR1 could exclusively activate p53 by mediating its phosphorylation specifically at the serine 15 residue. Ser-15 phosphorylation of p53 is associated with its increased transcription efficiency, decreased affinity for MDM2, and its increased nuclear retention (10,36). One of the target genes activated by p53 is p21, an inhibitor of a subset of the cyclin-dependent kinases including Cdc-2 (37,38). To evaluate the significance of peptide-mediated p53 activation, P44 peptide-treated lysates were further checked for tyrosine 15 phosphorylation of Cdc-2. Membrane immunoblotted with pCdc-2 antibody showed upregulation of pCdc-2 (Fig. 4D, lanes 2-5) in comparison with untreated cells (Fig. 4D, lanes 1), again confirming that P44 peptide was capable of mediating the effects of full-length SMAR1, and thus it may possess the entire functional activity to regulate the cell cycle. No changes were observed in the total ERK levels that were used as a loading control. TAT-SMAR1 RS-mutant peptide (SM) was further used to demonstrate that the serine motif (SSSSYS) of the SMAR1 minimal domain (arginine-serine rich) is essential for mediating the effects of fulllength SMAR1 in p53 activation signaling. HEK 293 cells were treated with SM peptide and checked for p53 as well as pCdc-2 levels. The SM peptide-treated lysates showed no increase both in the expression levels of p53 phosphoserine 15 or pCdc-2 levels when compared with untreated cells (Fig. 4E).
Serine 347 Residue of SMAR1 Is Critical for Mediating the Effects of Full-length SMAR1-Once it was established that the serine motif (SSSSYS) plays a pivotal role in SMAR1-mediated p53 activation, it was important to identify the serine residue within this motif that might be necessary in mediating these effects. To identify the critical serine residue within the P44 peptide, various point-mutant peptides were custom-synthesized, which varied with respect to the position of serine residue that was mutated (Table 1). To evaluate the functionality of mutant chimeric TAT-SMAR1 peptides (PS347A, PS348A, PS349A, PS350A, and RS mutant (SM)) in comparison with the TAT-SMAR1 WT peptide (P44), HEK 293 cells were treated with 10 M concentration of either of the peptides, and protein lysates were prepared after a 12-h incubation period. The lysates were then processed for immunoblotting with Ser(P)-15 p53 and pCdc-2 antibodies. We observed that when the first serine residue of P44 peptide was replaced with alanine (PS347A), there was no significant activation of p53 (Fig. 5A,  lane 2), thus demonstrating the significance of the serine 347 residue of SMAR1. The increase in the pCdc-2 levels that was observed in the case of P44 (Fig. 5A, lane 6) was decreased in PS347A-treated cells (Fig. 5A, lane 2). However, this was not the case with other mutated peptides. When the alanine was replaced back to serine at the 347 residue (Fig. 5A, lanes 3-5), the functionality of the peptide was restored as observed with an increase in pCdc-2 levels. No significant decrease in p53 activation and subsequent increase in pCdc-2 was observed between various P44 mutant peptides as follows: PS348A, PS349A, and PS350A (Fig. 5A, lanes 3-5). However, in the case of SM-treated cells (where all serine residues from 347 to 350 were replaced by alanines) (Fig. 5A, lane 7), there was no difference in the Ser(P)-15 p53 as well as pCdc-2 levels compared with either untreated or PS347A-treated cells (Fig. 5A, lanes 1  and 2, respectively). These observations strongly suggest that the first serine residue of the SMAR1 serine motif is most critical and essential for SMAR1-mediated p53 activation. Similar sets of experiments were performed in HCT 116 cells (Fig. 5B) and MCF-7 cells (Fig. 5C). In corroboration with our results obtained in HEK 293 cells, we again establish that the serine 347 residue is indispensable for full-length SMAR1 in activating the p53-mediated pathway (Fig. 5, B and C, lanes 3 and 4, respectively). Treatment of either of the cell lines (HCT 116 or MCF-7) with the P44 peptide demonstrated a 3.5-4-fold increase in p53 serine 15 phosphorylation (Fig. 5, B and C, lane 7, respectively) and a 3.2-4-fold up-regulation in pCdc-2 expression (Fig. 5, B and C, lane 7, respectively), in comparison with the untreated (Fig. 5, B and C, lane 1, respectively) or SM-treated cells (Fig. 5, B and C, lane 3, respectively) signifying activation of the p53-mediated cell cycle pathway.
In a similar approach, immunofluorescence studies were performed in HEK 293 cells, to demonstrate p53 stabilization upon peptide treatment. HEK 293 cells were treated with 10 M of various TAT-SMAR1-derived peptides. After 12 h of incubation, cells were indirectly stained for p53 and counterstained with fluorescein isothiocyanate, and the expression of p53 was observed with confocal imaging. As demonstrated in our earlier report that SMAR1 overexpression results in increased retention of activated p53 within the nucleus (30), treatment of cells with P44 peptide also showed a similar effect. The peptide could activate and stabilize p53 within the nucleus (Fig. 5D). However, both PS347A and SM peptide were unable to activate p53, as evident by negligible expression of p53 within the nucleus (Fig. 5, E and I, respectively). On the other hand, p53 stabilization was observed in cells treated with PS348A, PS349A, or PS350A (Fig. 5, F-H, respectively), thereby confirming that the serine 347 residue of SMAR1 was indispensable for SMAR1-mediated p53 activation and hence stabilization.
Serine 347, Substrate for PKC Family of Proteins-Because the P44 peptide could result in a significant increase in p53 Ser-15 phosphorylation, an in vitro kinase assay was performed to determine the phosphorylation status of P44 peptide. We observed that in the presence of whole cell extract from 293 cells enriched with cellular kinases, P44 but not the SM peptide gets phosphorylated. There was about 7-fold increased [␥-32 P]ATP incorporation in the wild-type peptide (P44) in comparison with its mutant counterpart (SM) (Fig. 6, dark bars 2 and 7, respectively). P44 phosphorylation was abolished as soon as the serine 347 residue was mutated to alanine (PS347A) (Fig. 6, dark bar  3). However, phosphorylation was restored with other pointmutated peptides (PS348A/PS350A) (Fig. 6, dark bars 4 -6, respectively), strongly suggesting that the serine 347 residue of SMAR1 serves as the kinase-targeted molecule. Previously, we have reported that the PKC family of proteins is responsible for post-transcriptional modification of SMAR1 at its arginine-serine-rich domain (30). To further analyze whether the serine 347 residue serves as a substrate motif, specifically for PKC, whole cell extract from staurosporine (STS) (a PKC inhibitor)-treated 293 cells was used for performing in vitro phosphorylation assays with various TAT-SMAR1 peptides. As expected, due to inhibition of PKC, cellular extracts prepared post-STS treatment failed to phosphorylate P44 peptide along with PS348A, PS349A, and PS350A peptides (Fig. 6, light bars 2 and 4 -6, respectively). On the other hand, in case of PS347A and SM peptides, there was no difference in their phosphorylation status with or without STS treatment (Fig. 6, light and dark bars 3  and 7, respectively), thereby confirming that serine 347 residue of SMAR1 serves as the substrate for PKC family.
P44 Peptide, a Potent Tumor Regressor-To test whether the differences in SMAR1-derived peptide-induced activation of p53 translated to differences in drug sensitivity in vivo, B16F1 mouse melanoma cells were subcutaneously grafted into athymic nude mice, and tumor growth together with therapeutic sensitivity was monitored. Once the tumor nodule was established into the mouse, P44 peptide was injected in the tumorlocalized areas at a physiological dose of 200 g/mouse three times a week. The treatment was continued for 4 weeks. In a parallel experiment, either TAT PTD or TAT-SMAR1 RS . P44 peptide activates p53-mediated cell cycle pathway. A, schematic representation of the TAT PTD-SMAR1 chimeric peptide along with its sequence. B, HEK 293 cells were seeded at a density of 5 ϫ 10 5 and after 24 h were incubated with increasing concentrations of the TAT-SMAR1 WT (P44) peptide. Protein lysates were prepared and then immunoblotted with antibodies to total p53, phosphoserine 15. C, same set of experiments was repeated with TAT PTD peptide representing the PTD carrier sequence alone and immunoblotted with phosphoserine 15, in comparison with P44-treated cells. D, status of phospho-Cdc-2 was further analyzed in P44 peptide-treated cells. Total ERK was used as an internal control. E, same set of experiments was again repeated using TAT-SMAR1 RS mutant (SM) peptide, and whole cell lysates were probed for phosphoserine 15-p53 and pCdc-2 antibodies. Actin was used as a loading control. mutant peptide SM (Fig. 7, A and B, respectively) was injected in tumor-bearing mice to be used as control. Interestingly, there was a marked difference in the response of the xenograft to the P44 peptide treatment. Almost a 5-7-fold regression in tumor was observed in the mice injected with P44 peptide (Fig.  7C) compared with either TAT-injected or SM-injected mice. Fig. 7, D-F, corresponds to the magnified tumor images of Fig.  7, A-C, respectively. After excision, the tumor weight was found to be 0.2-0.8 g for mice treated with the P44 peptide which was significantly less when compared with the mice treated with either the TAT PTD (1.5-3 g) or SM peptide (1.2-   2.5 g) (Fig. 7G). Thus, the P44 peptide mimics the function of full-length SMAR1 in drastically reducing the tumor growth.
Activation of p53-mediated Pathway in Tumors Treated with P44 Peptide-As the P44 treatment in tumor-bearing nude mice demonstrated inhibition of tumor growth, we wanted to evaluate the status of p53 activation in these tumors. There was a marked increase in p53 mRNA levels in P44-treated tumors when compared with SM-treated tumors (Fig. 8A, upper panel). By reverse transcription-PCR as well as immunoblotting, we observed a 2-3-fold increase in the p21 expression when compared with SM-treated tumors (Fig. 8, A, middle panel, and B, 4th lane, respectively) indicating a strong up-regulation of phosphoserine15 p53 in tumors treated with P44 peptide (Fig.  8B, 2nd lane). The increase in pCdc-2 expression (Fig. 8B, 3rd lane) further emphasized the activation of the p53-mediated cell cycle arrest pathway in P44-treated tumors. Immunohistochemical analysis on tumor sections obtained from SM-and P44-treated mice, by using p21 antibody, demonstrated augmented p21 expression and its nuclear localization in P44 tumor sections (Fig. 8C), signifying p53 activation.
Histopathological Changes in the Subcutaneous Tumors-To analyze the tumor vasculature of TAT PTD, SM-treated and P44-treated mice, we stained the tumor sections with hematox-ylin and eosin. Tumors from TAT PTD-(data not shown) and SM-treated mice exhibited poorly organized vascular architecture and compressed blood vessels because of extensive cell proliferation (Fig. 9A, lower panel). On the contrary, P44treated tumors showed intact vasculature wherein the red blood cells within the vessels were observed in healthy condition (Fig. 8A, upper panel). In SM-treated tumors, because of compressed vessels, the shape of red blood cells was also distorted (Fig. 9A, lower right panel, arrow). We further observed that the inter-vessel distances in P44-treated tumors were significantly less compared with SM-treated tumors (Fig. 9A,  upper and lower left panels, respectively). The presence of hypoxic cells is a hallmark of cancer (39,40). Cellular responses to hypoxia are triggered by HIF-1␣, which is known to restore tissue homeostasis in hypoxic conditions (41). To determine the status of HIF-1␣ expression, immunohistochemical analysis was performed on tumor sections obtained from SM-and P44-treated mice by using antibody against HIF-1␣. Compared with P44 tumor sections, SM sections demonstrated increased HIF-1␣ expression (Fig. 9B) thereby resulting in the prolifera- , and ␤-actin (lower panel) mRNA expression from the tumor tissue obtained from nude mice treated either with SM or P44 peptide. B, tumor tissue from SM-and P44-treated nude mice were isolated and homogenized, and whole cell lysates were prepared. The lysates were then immunoblotted with total p53, phosphoserine 15 p53, p21, and pCdc-2 antibodies. Actin was used as the loading control. C, immunohistochemistry of paraffin-embedded tumor sections obtained from SM-and P44-treated nude mice. Left panels show DAPI staining to localize nucleus. Middle panels correspond to staining with anti-p21 followed by detection using Cy-3 conjugated mouse immunoglobulin. Right panels correspond to the merged image. tion of tumor cells even under hypoxic conditions. All these observations implicate the significance of P44 peptide in restoring tissue architecture in tumor cells and thus potentiates its role as a tumor regressor.

DISCUSSION
Interest in peptides and proteins is becoming increasingly important, not only as molecular tools for the understanding of protein-protein interactions but also as therapeutic compounds. Several oligopeptides such as the p53 C-terminal peptide (42,43), the BH3 domain of Bak (47), and p21 WAF1-(44), p16 INK4A- (45,46), Sos- (48), and c-Myc-derived peptides (49) have been developed as a cargo and proved to function against cancer cells. In vitro studies involving systematic screening of panels of human tumor-derived cell lines for sensitivity to therapeutic agents have revealed associations with p53 status and drug sensitivity (50). Recently, we identified the RS-rich domain as the minimal core region of SMAR1 that is responsible for activating the p53-mediated pathway (30). In this study, we investigated the antitumorigenic activity of a 33-amino acid peptide sequence corresponding to the RS domain of SMAR1. The peptide was conjugated to the PTD of HIV-1 Tat protein (TAT-SMAR1 WT peptide; P44) because PTD-TAT protein has been known to deliver bioactive peptides into tissues and across the blood-brain barrier (22,26,35,51). We found that almost 100% of cells were efficiently transduced by various SMAR1-derived peptides as observed through confocal imaging. All the chimeric TAT-SMAR1-derived peptides demonstrated efficient nuclear compartmentalization irrespective of the point mutations (PS347A/PS350A and SM) in the wild-type SMAR1 peptide (P44), thereby suggesting that TAT PTD fusion with the SMAR1 peptide worked as an efficient peptide-delivery system. The P44 minimal peptide sequence of SMAR1 retained the functional activity of the fulllength SMAR1 as it was capable of activating p53 as well as retaining into the nucleus. Accumulation of p53 in the nucleus resulted in the arrest of the cell cycle at G 2 /M phase, which is in accord with the known growth inhibitory properties of high levels of wild-type p53. Interestingly, the microarray data also demonstrated down-regulation of important cell cycle regulatory proteins ( Table 2). Genes involved in regulating mitosis, mitogen-activated protein kinase (MAPK) signaling, and cell cycle checkpoints showed significant decreased expression upon P44 treatment, emphasizing its role as a cell cycle modulator. Furthermore, reduced expression of proteins involved in ubiquitin-proteasome signaling ( Table 2) may serve as an alternative mechanism to support P44-mediated stabilization and increased nuclear retention of p53.
Recent studies have proposed that phosphorylation of N-terminal amino acids of p53 contribute to its regulation by affecting the binding of co-activators and the negative regulator MDM2 (52,53). These studies emphasize the significance of phosphorylation at the serine 15, serine 20, or serine 37 residue of p53 in maintaining protein stability as well as transactivation properties (54). Interestingly, P44 peptide alone could mediate the phosphorylation of p53 at its serine 15 residue and in effect result in up-regulation of the phospho-Cdc2, indicating that p53-modulating activity of full-length SMAR1 resided entirely within the P44 peptide. Results obtained from luciferase reporter assays further confirmed the bioactivity of the SMAR1-derived peptide and demonstrated that it followed a p53-dependent p21-growth suppression pathway. We have previously demonstrated that the substrate motif for protein kinase C family of serine/threonine kinases resides within the arginine-serine-rich domain of SMAR1 (30). In this study, by using various point-mutated peptides of P44, we demonstrated that the serine 347 residue of SMAR1 is critical for its function and upon mutation to alanine (in case of PS347A as well as SM mutant peptide) results in loss of its phosphorylation and hence reduced functional activity. However, staurosporine treatment resulted in a complete inhibition of phosphorylation of peptides containing intact serine 347 residue. Peptides where serine 347 was replaced with alanine (that included PS347A as well as SM) demonstrated no difference in their phosphorylation status irrespective of the presence or absence of STS, thereby emphasizing the importance of serine 347 in phosphorylation of P44 peptide. Moreover, in the SM peptide wherein serine 347-350 residues were mutated to alanine, we observed further reduction in its phosphorylation compared with the PS347A peptide, thus suggesting the significance of other serine residues (348 -350) in the phosphorylation of the P44 peptide.
In this study using xenograft tumor nude mice model, we demonstrate that the TAT-SMAR1 WT (P44) peptide strongly regressed tumors, and the anti-tumorigenic activity of the P44 peptide was significantly reduced when the serine residues were mutated to alanine (TAT-SMAR1 RS mutant (SM)). Histopathological analysis of tumor section from control tumors (TAT PTD and SM-treated) revealed increased cellular proliferation resulting into blood vessel condensation. However, there was pronounced destruction of the tumor architecture upon treatment with P44 peptide. It prevented vascular damage and maintained cellular integrity. Protection against hypoxia in solid tumors is an important step in tumor development and progression. A multifaceted adaptive response is triggered by hypoxia, which is primarily mediated by the HIF-1 system, which plays a crucial role especially in angiogenesis and carcinogenesis (41,55,56). Alteration and overexpression of HIF-1␣ has been detected in a variety of solid tumors, including breast, lung, ovarian, and oral cancer with varying staining patterns (57)(58)(59). We also observed an increased expression of HIF-1␣ in cells treated with either TAT PTD or SM peptide compared with those treated with P44 peptide. Thus, P44 peptide restores normoxia in tumor cells that may be responsible for decreasing HIF-1␣ expression, even though the mechanism is yet not clear. Major research efforts are aimed at the discovery of molecular targets that are specific as well as toxic to cancer cells. Identification of potential targets for therapeutic intervention thus fuels a hope for curing cancer. Peptide-mediated molecular therapeutic delivery systems have currently emerged as an alternative means to effectively substitute or augment the present gene therapy technologies, e.g. TAT, VP22, engineered peptides (22). Our findings potentiate the use of P44 peptide of SMAR1 for peptidometic cancer drug design so as to allow therapeutic intervention in the target cell biochemistry without the need to alter its genome.