Regulation of the leucocyte chemoattractant receptor FPR in glioblastoma cells by cell differentiation

Abstract

The G protein-coupled formylpeptide receptor (FPR), known to mediate phagocytic leucocyte chemotaxis in reponse to bacterial- and host-derived agonists, was expressed by tumor cells in specimens of surgically removed more highly malignant human gliomas. In human glioblastoma cell lines, FPR activation increased cell motility, tumorigenicity and production of angiogenic factors. In studies of the mechanistic basis for the selective expression of FPR in more highly malignant gliomas, we found that the DNA methyltransferase inhibitor 5-Aza-2′-deoxycytidine (Aza), while promoting the differentiation of human glioblastoma cells, downregulated FPR expression. Aza also reduced the global methylation levels in glioblastoma cells and activated the pathway of p53 tumor suppressor. Methylation-specific polymerase chain reaction revealed that Aza treatment of tumor cells reduced the methylation of p53 promoter, which was accompanied by increased expression of p53 gene and protein. In addition, overexpression of p53 in glioblastoma cells mimicked the effect of Aza treatment as shown by increased cell differentiation but reduction in FPR expression, the capacity of tumor sphere formation in soft agar and tumorigenesis in nude mice. Furthermore, Aza treatment or overexpression of the wild-type p53 in glioblastoma cells increased the binding of p53 to FPR promoter region shown by chromatin immunoprecipitation. These results indicate that increased methylation of p53 gene retains human glioblastoma cells at a more poorly differentiated phase associated with the aberrant expression of FPR as a tumor-promoting cell surface receptor.

Introduction

Formylpeptide receptor (FPR) is a G protein-coupled receptor, originally identified in phagocytic leukocytes, that mediates cell chemotaxis and activation in response to the bacterial chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine (fMLF). Agonist binding to FPR elicits a cascade of signal transduction pathways involving phosphatidylinositol 3-kinase, protein kinase C, mitogen-activated protein kinases and the transcription factor nuclear factor (NF)-κB (1,2). Because of its expression in cells of the immune system and its interaction with bacterial chemotactic peptides, FPR was thought to participate in host defense against microbial infection. In fact, mice depleted of the FPR analog FPR1 were more susceptible to infection by Listeria monocytogenes (3). Recently, a number of novel host-derived chemotactic agonists of FPR have been identified, including formyl peptides potentially released by mitochondria of ruptured cells (4), annexin I produced by activated epithelia (5) and a neutrophil granule protein, cathepsin G (6). In addition, functional FPR has been detected in cells of non-hematopoietic origin, such as lung epithelial cells (7) and hepatocytes (8). These findings suggest that FPR may be involved in a broader spectrum of pathophysiologic processes.

Gliomas are the most common tumor type in the brain, characterized by progressive expansion and resistance to conventional therapy. The capacity of glioma growth and invasion is closely correlated with the expression of cell surface receptors that sense the signals present in the tumor microenvironment (9–11). FPR is one of such receptors that is selectively expressed by glioma cells with a more highly malignant phenotype (12,13). In specimens derived from surgically removed gliomas, FPR expression was detected in all specimens of grade IV glioblastoma multiforme and a majority of grade III anaplastic astrocytoma. In contrast, only two of 13 less aggressive grade II astrocytoma specimens showed positive FPR staining (12). In previous studies of the biological significance of FPR in glioma cells, we found that a human glioblastoma cell line U-87 expresses high levels of FPR, which upon activation by its cognate agonist fMLF or by an agonist activity released by necrotic tumor cells, promotes the directional migration, survival and production of angiogenic factors by tumor cells (12,14). Activation of FPR also transactivates the receptor for epidermal growth factor to exacerbate the tumor cell malignant behaviors of the glioma cells (15).

The present study was aimed at elucidating the mechanisms that regulate FPR expression in selected glioma cells in order to identify novel molecular targets for glioma therapy. We found that p53 plays an important role in controlling the levels of FPR and the degree of differentiation of glioblastoma cells.

Materials and methods

Cells and reagents

Human glioblastoma cell line U-87 was obtained from the American Type Culture Collection (Manassas, VA). The cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal calf serum (FCS) and antibiotics. fMLF, 5-Aza-2′-deoxycytidine (Aza) and dexamethasone were purchased from Sigma–Aldrich (St Louis, MO). Antibodies against p53, NFκB and β-actin were from Cell Signaling Technology (Beverly, MA). Antibodies against vimentin, glial fibrillary acidic protein (GFAP) and FPR were from BD Biosciences (San Jose, CA).

Immunocytochemical staining

For subcellular distribution of p53 and NFκB, cells were grown on chamber slides (Nunc, Naperville, IL) and fixed with 4% paraformaldehyde for 30 min at 4°C. After washing with phosphate-buffered saline (PBS), the cells were permeabilized with ice-cold 0.2% Triton X-100 for 5 min. The slides were washed with PBS, blocked with 0.5% bovine serum albumin in PBS for 30 min and then incubated with the indicated primary antibodies at 4°C overnight. After washing with PBS, the slides were incubated with secondary antibodies and phosphatidylinositol (Sigma–Aldrich) for 1 h. The slides were then mounted and examined under a confocal fluorescent microscope (Carl Zeiss, Oberkochen, Germany).

Flow cytometry

For detection of FPR, tumor cells in single-cell suspension were stained by phycoerythrin-conjugated anti-FPR antibody (BD Biosciences) in buffer with 20% FcR-Blocking Reagent (Miltenyi, Bergisch Gladbach, Germany). For detection of vimentin and GFAP, paraformaldehyde-fixed (4%) tumor cells were permeablized by 1% Triton X-100 and blocked by 5% bovine serum albumin and then were incubated with fluorescein isothiocyanate-conjugated anti-vimentin and anti-GFAP antibodies. Fluorescence was collected by flowcytometer (BD Biosciences) and quantified by Cell Quest Acquisition and Analysis software (BD Biosciences).

Chemotaxis

Tumor cell chemotaxis was measured with 48-well chemotaxis chambers (NeuroProbe, Gaithersburg, MD) (12,16). A 27 μl aliquot of chemoattractants was placed in the wells of the lower compartment of the chamber, and 50 μl of tumor cells (at 3 × 10⁶/ml) were placed in the wells of the upper compartment. Two compartments of the chemotaxis chamber were separated by a 10 μm pore-sized polycarbonate filter (GE Osmonics Labstore, Minnetonka, MN) coated with collagen type I (BD Biosciences) at 50 μg/ml. After incubation at 37°C for 270 min, the filters were removed and stained, and cells that migrated across the filters were counted under light microscopy. The results were expressed as the mean number (±standard error) of migrated cells in three high-powered fields (×400 magnification) in triplicate samples or as chemotaxis index representing the fold increase in cell migration in response to stimulants over medium control.

Reverse transcription–polymerase chain reaction and real-time reverse transcription–polymerase chain reaction

Total cellular RNA was extracted with RNeasy (Qiagen, Valencia, CA). Real-time reverse transcription (RT²)–polymerase chain reaction (PCR) was performed with the Applied Biosystems 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA). Total cellular RNA was reverse transcribed with Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA) and oligo-(dT)12. PCRs containing complementary DNA, PCR master Mix (Fermentas, Glen Burnie, MD) for reverse transcription–PCR or Power SYBR Green PCR Master Mix (Applied Biosystems) for RT²–PCR and primers for FPR, p53 or glyceraldehyde-3-phosphate dehydrogenase were performed for 30 (reverse transcription–PCR) or 40 (RT²–PCR) cycles of 95°C for 15 s, 60°C for 15 s and 72°C for 20 s. Primers for FPR: 5′-TCCAGTTGGACTAGCCACA-3′ and 5′-CCATCACCCAGGGCCCAATG-3′. Primers for glyceraldehyde-3-phosphate dehydrogenase: 5′-TCTAGACGGCAGGTCAGGTCCACC-3′ and 5′-CCACCCATGGCAAATTCCATGGCA-3′. Primers for p53: 5′-GGAGCCGCAGTCAGATCCTAG-3′ and 5′-CAAGGGGGACAGAACGTTG-3′. Reverse transcription–PCR products were electrophoresed on 2% agarose gels and visualized with ethidium bromide staining.

Immunoblot

Tumor cells were lysed in 150 μl of 1× sodium dodecyl sulfate sample buffer (62.5 mM Tris–HCl at pH 6.8, 2% sodium dodecyl sulfate, 10% glycerol and 50 mM dithiothreitol, sonicated for 3 s and then boiled for 5 min. The cell lysate was centrifuged at 10 000g at 4°C for 10 min. Total protein (40 μg) was electrophoresed on 4–12% gradient Tris–Glycine precast gels (Invitrogen) and transferred onto Immunoblot P membranes (Millipore, Billerica, MA). The membranes were blocked by incubation in 3% non-fat dry milk for 1 h at room temperature and then incubated with primary antibodies in PBS containing 0.01% Tween-20 overnight at 4°C. After incubation with a horseradish peroxidase-conjugated secondary antibody, the protein bands were detected with Super Signal Chemoluminescent Substrate Stable Peroxide Solution (Pierce, Rockford, IL) and BIOMAX-MR film (Eastman Kodak, Rochester, NY). When necessary, the membranes were stripped with Restore Western Blot Stripping Buffer (Pierce) and reprobed with antibodies against various cellular proteins.

Cell proliferation

According to the instructions of the AlamarBlue™ assay kit (Biotium, Hayward, CA), tumor cells were cultured in 96-well tissue culture plates at 1000 cells per well in 10% FCS/DMEM containing 10% AlamarBlue that incorporates a fluorometric/colorimetric cell growth indicator based on detection of metabolic activity. After indicated hours, the absorbance at 570 and 600 nm was measured and calculated according to the instructions. The results are expressed as ‘growth index’ representing fold increase in cell numbers during culture.

Ca²⁺ flux

Ca²⁺ mobilization was measured by incubating 2 × 10⁷ tumor cells in 1 ml loading medium (DMEM containing 10% FCS and 2 mM glutamine) with 7 μM Fura-2 acetoxymethyl ester (Molecular Probes, Eugene, OR) for 45 min at room temperature. The dye-loaded cells were washed and resuspended in saline buffer (138 mM NaCl, 6 mM KCl, 1 mM CaCl₂, 10 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, 5 mM glucose and 0.1% bovine serum albumin at pH 7.4) at a density of 0.5 × 10⁶ cells/ml. The cells were then transferred into quartz cuvettes (1 × 10⁶ cells in 2 ml of saline buffer) and were placed in a fluorescence spectrometer (Perkin–Elmer, Beaconsfield, UK). Stimulants were added to the cuvettes in a volume of 20 μl and the intensity of the fluorescence was measured by the use of the ratio of the absorbance at 340 to 380 nm, calculated with an FL WinLab program (Perkin–Elmer).

Generation of U-87 cells stably expressing wild-type p53

Retroviral vector stocks were produced by transient transfection of Phoenix-Ampho cells with the Superfect Transfection Reagent (Qiagen) and 5 μg p53 expression plasmid pBABE-p53 (Cell Biolabs, San Diego, CA) or blank plasmid pBABE (Cell Biolabs). The virus was collected from the culture supernatants on day 2 after transfection, and the cells were infected with the retroviral vectors in the presence of polybrene at 5 μg/ml. U-87 cells containing p53 expression cassettes were selected and maintained by incubation with 2 μg/ml puromycin (BD Biosciences).

Bisulphate treatment of DNA and methylation-specific PCR

As described in the CpGenome™ Fast DNA Modification Kit (Chemicon, Temecula, CA), 1 μg tumor cell DNA in 100 μl H₂O was incubated with 7 μl 3 M NaOH at 37°C for 10 min, followed by a 16 h treatment at 50°C after adding 500 μl freshly prepared DNA modification reagent. The DNA was desalted by column binding–washing–elution protocol. PCR was carried out in a volume of 25 μl containing 50 ng template DNA with FastStart Taq polymerase (Roche, Grenzach-Wyhlen, Germany). After an initial heat denaturation for 4 min treatment at 95°C, 30 cycles of 95°C for 15 s, 60°C for 15 s and 72°C for 20 s were carried out. The PCR products were separated by 1.2% ethidium bromide-containing agarose gel electrophoresis with 1× tris-acetate EDTA and visualized under ultraviolet. To verify the PCR results, representative amplified products of each target gene were purified and sequenced. Primers used for methylated CpG island were 5′-TTCGGTAGGCGGATTATTTG-3′ and 5′-AAATATCCCCGAAACCCAAC-3′ (17,18). Primers for unmethylated CpG island were 5′-TTGGTAGGTGGATTATTTGTTT-3′ and 5′-CCAATCCAAAAAAACATATCAC-3′ (17,18).

Chromatin immunoprecipitation

Chromatin immunoprecipitation (CHIP) was performed using a CHIP assay kit (Upstate Biotechnology, Lake Placid, NY) as per instructions of the manufacturer. Briefly, 10⁷ tumor cells were treated with 1% formaldehyde for 10 min. Cells were lysed, sonicated and one-fourth of the lysate were subjected to immunoprecipitation with 1 μg indicated antibodies at 4°C overnight. After extensive washing, DNA was extracted from immunoprecipitate as well as from one-tenth of lysate without immunoprecipitation. Primers (5′-GGTGACTCATGCTTGTAATCCCA-3′ and 5′-CTCCCAGATTCAAGCAATTCTCC-3′) were used to PCR amplify the segment of FPR gene containing a putative p53-responsive element (PRE).

Electrophoretic mobility shift assay

Nuclear extracts of tumor cells were subjected to electrophoretic mobility shift assay (EMSA) according to the instructions of the EMSA kit (Pierce). Oligonucleotides containing consensus-binding sites for p53 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). FPR probe (5′-TTCGAGACCAGCCTGGCCAACATGTTGAAACCT-3′) containing a putative PRE synthesized by Integrated DNA Technologies (Coralville, IA) was labeled with biotin. For supershift, DNA–protein complexes were incubated with an anti-p53 antibody (Cell Signaling Technology) for 20 min prior to the electrophoresis (6% non-denaturing polyacrylamide gel). The complexes were transferred to Biodyne® B nylon membranes (Pierce) and cross-linked by ultraviolet and then were detected with a streptavidin–horseradish peroxidase-chemiluminescent system.

Tumor microsphere formation

Anchorage-independent growth of tumor cells was examined in a semisolid agar. DMEM containing 10% FCS was warmed to 48°C to dissolve Bacto-Agar into a 0.6% (wt/vol) solution. The solution (300 μl) was poured into 24-well culture plates and allowed to solidify for 15 min at room temperature. A volume of 200 μl 0.3% agar solution containing 2500 tumor cells were then layered over the bottom agar in each well. The plates were incubated in 5% CO₂ at 37°C in the presence of 10% FCS/DMEM for 3 weeks. The spheres were numerated based on different sizes with the aid of an inverted microscope (Olympus, Center Valley, PA) and Image pro-Plus 6.0 software.

Tumorigenesis

Tumor cells were implanted subcutaneously into the flank of 4-week-old (20–22 g) female athymic Ncr-nu/nu mice (National Cancer Institute Animal Production Program, National Cancer Institute-Frederick, Frederick, MD) at 5 × 10⁶ cells per mouse in 100 μl PBS. Tumor size was calculated by the formula lw²/2, where l is the length of the tumor and w is the width. Each testing group contained at least five mice in one experiment. The results were expressed as the mean (±standard error) of tumor sizes in each group.

Statistical analyses

A computer-aided Mann–Whitney U-test program SPSS (Version 11.0) (GraphPad Software, San Diego, CA) was used to determine the statistical significance of the difference between cell responses to testing materials and to controls in experiments of chemotaxis, FPR messenger RNA (mRNA) expression, cell proliferation, tumor microsphere formation and in vivo tumor growth.

Results

Aza promotes U-87 cell differentiation with downregulation of FPR

Aza, a DNA methyltransferase inhibitor, promoted U-87 cell differentiation as shown by increased expression of GFAP, a glial cell differentiation marker, and reduced expression of vimentin, a glial precursor marker (Figure 1A). Aza-treated U-87 cells grew more slowly in vitro (Figure 1B), with significantly reduced capacity to form tumor spheres on soft agar (Figure 1C) and to grow tumors in nude mice (Figure 1D). Aza-treated U-87 cells expressed lower levels of FPR protein (Figure 2A) and mRNA transcripts (Figure 2B), associated with reduced chemotaxis (Figure 2C) and calcium flux (Figure 2D) of U-87 cells induced by the FPR agonist, fMLF.

Fig. 1.

Differentiation of human glioblastoma cell line U-87 after treatment with DNA methyltransferase inhibitor Aza. (A) Analysis of the expression of GFAP and vimentin in U-87 cells treated by Aza. U-87 cells were cultured in 10% FCS/DMEM containing 10 μM Aza in dimethyl sulfoxide (DMSO) for 2 weeks before analyses with monoclonal antibodies against vimentin and GFAP, phycoerythrin-conjugated secondary antibody and a flow cytometry. (B) Growth of Aza-treated U-87 cells. Vehicle-treated (dimethyl sulfoxide) or Aza (10 μM, 14 days)-treated U-87 cells were cultured in 10% FCS/DMEM containing 10% AlamarBlue. After indicated hours, the absorbance at 570 and 600 nm was measured and calculated based on the colorimetric changes according to the assay kit instructions. Cell proliferation was expressed as growth index denoting the fold increase in the cell number at different time points during culture. *Significantly reduced cell growth in the presence of Aza as compared with vehicle control (dimethyl sulfoxide) (P < 0.01, n = 6). (C) Tumor sphere formation on soft agar. Vehicle-treated (dimethyl sulfoxide) or Aza (10 μM, 14 days)-treated U-87 cells suspended in 200 μl 0.3% agar (2500 cells) were layered on solidified bottom agar in the wells of 24-well plates. The cells were grown in 5% CO₂ at 37°C, with medium changes every other day. After 3 weeks, tumor colonies were photographed under light microscope. The results are expressed as the mean (±standard error, n = 6) numbers of colonies. *P < 0.01 and #P < 0.05 indicate significantly reduced sphere numbers as compared with vehicle (dimethyl sulfoxide)-treated cells. (D) Tumorigenesis in nude mice. Vehicle-treated (dimethyl sulfoxide) or Aza (10 μM, 14 days)-treated U-87 cells at 5 × 10⁶ cells in 100 μl of PBS per mouse were injected subcutaneously into the flanks of athymic mice. Mice were examined for tumor formation and size on the indicated days. Tumor size was expressed as the mean volume (in mm³, ±standard error) measured in 10 mice. *Indicates significantly reduced tumor volume as compared with vehicle (dimethyl sulfoxide)-treated tumor cells measured on day 20, 25, 30, 35 and 40 after implantation (P < 0.01).

Fig. 2.

FPR expression and function in Aza-treated glioblastoma cells. (A) fluorescence activated cell sorting analysis of FPR expression on tumor cells. U-87 cells were cultured in 10% FCS/DMEM in the presence of Aza at 37°C and 5% CO₂. FPR expression was measured by a phycoerythrin-conjugated mouse monoclonal antibody against FPR. %, percentage of FPR-positive cells. MFI, mean fluorescence intensity. (B) The expression of FPR mRNA. U-87 cells were cultured in 10% FCS/DMEM in the presence of 10 μM Aza at 37°C and 5% CO₂ for 3 days. The levels of FPR mRNA were examined by RT²–PCR. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as control. The results are expressed as the mean level of FPR mRNA as compared with glyceraldehyde-3-phosphate dehydrogenase mRNA (%, ±standard error) in three independent experiments. *Indicates significantly reduced FPR mRNA in Aza-treated cells as compared with control (DMSO, dimethyl sulfoxide) cells (P < 0.05). (C) fMLF-induced chemotaxis. Chemotaxis of U-87 cells was assayed by 48-well chemotaxis chambers in response to different concentrations of fMLF. The results are expressed as the mean (±standard error) of three independent experiments. *Indicates significantly reduced migration as compared with vehicle (dimethyl sulfoxide)-treated cells (P < 0.05). (D) Ca²⁺ flux. fMLF-induced Ca²⁺ flux was measured in a fluorescence spectrometer. The FPR agonist peptide fMLF-induced fluorescence intensity in U-87 cells was expressed as the ratio at wavelength 340 to 380 calculated by an FL WinLab program.

Hypermethylation of p53 gene in U-87 cells

We next examined the global changes in gene expression in Aza-treated U-87 cells by using RT² Profiler™ PCR Array (Superarray, Frederick, MD) in which p53 pathway activation was one of the most prominent events (data not shown). Since Aza reduces gene methylation in a number of malignant tumors, we investigated the effects of this compound on the methylation of p53, which is of wild-type (WT), in U-87 cells. Methylation-specific PCR revealed that the CpG island of p53 promoter in U-87 cells was methylated (Figure 3A), and Aza treatment reduced the level of methylation (Figure 3A). Aza treatment also increased the mRNA (Figure 3B) and protein (Figure 3C) levels of p53 in U-87 cells. The increase in p53 protein occurred at 3 days after Aza treatment and remained elevated through the 14 day treatment period by 10 μM Aza (Figure 3D).

Fig. 3.

Methylation of p53 promoter and the expression of p53. (A) Methylation-specific PCR of p53 promoter region. U-87 cells were cultured in 10% FCS/DMEM in the presence of Aza (10 μM) at 37°C 5% CO₂. After 3 days, genomic DNA was modified by treatment with sodium bisulfite and amplified by PCR with specific methylation (M) and unmethylation (U) primers of p53 promoter. (B) The expression of p53 mRNA. The level of p53 mRNA in Aza (0, 5, 10 and 20 μM, 3 days)-treated U-87 cells was examined by reverse transcription–PCR. (C and D) The expression of p53 protein. U-87 cells were treated with different concentrations of Aza for 3 days (C) or 10 μM Aza for different days (D). The protein levels of the cell lysates were measured by immunoblot.

Overexpression of WT p53 promotes U-87 cell differentiation with downregulation of FPR

Since Aza treatment of U-87 cells upregulated p53, but downregulated FPR in association with cell differentiation, we examined the effect of overexpression of WT p53 in U-87 cells. U-87 cells infected with retrovirus containing WT p53-expressing cassettes (pBABE-p53) showed increased p53 protein and a downstream target gene p21 (Figure 4A), suggesting that WT p53 overexpressed in the tumor cells was functional. After expression of the WT p53, U-87 cells contained lower levels of FPR mRNA (Figure 4A) and cell surface protein (Figure 4B). FPR ligand fMLF-induced chemotaxis (Figure 4C) and calcium flux (Figure 4D) of p53-overexpressing U-87 cells were concomitantly reduced. Similar to the effect of Aza treatment, overexpression of p53 increased the level of GFAP, but decreased vimentin level in U-87 cells (Figure 5A). In addition, U-87 cells overexpressing p53 exhibited a slower growth rate (Figure 5B), reduced capacity of tumor sphere formation in soft agar (Figure 5C) and decreased tumorigenesis in nude mice (Figure 5D). It is interesting to note that tumor cells from xenografts formed by pBABE-p53-transfected cells expressed lower levels of p53 as compared with cells maintained in culture (data not shown), indicating that tumor cells with higher pBABE-p53 have lost the capacity to form rapidly growing tumors.

Fig. 4.

The FPR expression and function in p53-transfected U-87 cells. (A) The expression of p53 protein and FPR mRNA. U-87 glioblastoma cells were infected by retrovirus produced by pBABE vector containing p53-expressing cassettes (pBABE-p53) or blank vector (Mock). The protein levels of p53 and p21^Waf/Cip1 were measured by immunoblot. The FPR mRNA level was measured by RT²–PCR. *Indicates significantly reduced FPR level in pBABE-p53-infected U-87 cells as compared with Mock cells (P < 0.01, n = 3). (B) The expression of FPR protein. FPR protein levels on the surface of Mock and pBABE-p53 cells were assayed by flow cytometry using a phycoerythrin-conjugated mouse monoclonal anti-FPR antibody. (C and D) Chemotaxis and calcium flux were induced by fMLF. FPR function in Mock and pBABE-p53 cells were assayed by fMLF-induced chemotaxis (C) and calcium flux (D). *Indicates significantly reduced cell migration of pBABE-p53 U-87 cells as compared with Mock cells (P < 0.05, n = 6).

Fig. 5.

Differentiation states of p53-transfected cells. (A) The expression of GFAP and vimentin. pBABE-p53 and Mock cells were fixed, permeabilized and stained by vimentin and GFAP antibodies followed by staining with a phycoerythrin-conjugated secondary antibody. The fluorescence intensity was measured by flow cytometry. (B) Growth of p53-transfected cells. pBABE-p53 or Mock cells were cultured in 10% FCS/DMEM containing 10% AlamarBlue. Cell growth was monitored by measuring the absorbance at 570 and 600 nm and was indicated by mean of growth index (±standard error, n = 6). *Indicates significantly reduced growth of pBABE-p53 U-87 cells as compared with Mock cells (P < 0.01). (C) Tumor sphere formation in soft agar. The Mock and pBABE-p53 U-87 cells in soft agar were cultured on the solidified bottom agar in 24-well plates. After 3 weeks, tumor colonies were analyzed and indicated by the mean numbers (±standard error, n = 6) of the microsphere in different area. *Indicates significantly reduced sphere numbers formed by pBABE-p53 U-87 cells as compared with Mock cells (P < 0.01). (D) Tumorigenesis in nude mice. Mock or pBABE-p53 U-87 cells at 5 × 10⁶ cells in 100 μl of PBS per mouse were injected subcutaneously into the flanks of athymic mice. Mice were examined for tumor formation and size on the indicated days. Tumor size was expressed as the mean volume (in mm³, ±standard error) of the tumors in 10 animals. *Indicates significantly reduced size of tumors formed by pBABE-p53 U-87 cells as compared with Mock cells (P < 0.01).

Increased p53 translocation into the nuclei of U-87 cells after Aza treatment or p53 overexpression

Because p53 acts as a transcription factor that binds to DNA, we examined the nuclear content of p53 in Aza-treated or p53-overexpressing U-87 cells. Immunofluorescence staining shown in Figure 6A reveals that either Aza treatment or overexpressing p53 resulted in the localization of an increased level of p53 in the nuclei. We also found that the nuclear localization of another transcription factor NFκB was reduced in Aza-treated or p53-overexpressing U-87 cells (Figure 6A).

Fig. 6.

p53 and NFκB distribution in the nuclei of U-87 glioblastoma cells and binding of p53 to the FPR promoter region. (A) Visualization of p53 and NFκB translocation. IgG, anti-p53 antibody and anti-NFκB antibody were used to label vehicle-treated (DMSO, dimethyl sulfoxide) or Aza (10 μM, 3 days)-treated U-87 cells, mock cells and pBABE-p53 cells. The proteins were visualized (green) by a fluorescein isothiocyanate-conjugated secondary antibody under confocal microscope. Nuclei were visualized in red with phosphatidylinositol. Arrows indicate nuclear location of p53 or NFκB. Yellow color indicates superimposed p53 or NFκB with phosphatidylinositol in the nuclei. (B) CHIP assay of p53 binding to FPR regulatory region. Aza-treated (10 μM, 3 days) or dimethyl sulfoxide-treated U-87 cells, pBABE-p53 and Mock cells were used for CHIP assays. The cells were cross-linked by 1% formaldehyde. Nuclear DNA–protein complex was extracted and sonicated and then was immunoprecipitated by IgG or antibodies against RNA polymerase II (Rpase), NFκB or p53. One-twentieth of the cross-linked and sonicated DNA were set as ‘Input’ that indicates PCR performed on DNA without any immunoprecipitation. PCR was used to amplify a 200 bp fragment of FPR PRE-1 region. (C) EMSA of p53 binding to the putative PRE in FPR regulatory region. A biotin-labeled 30 bp oligo representing the putative p53-binding site in the FPR regulatory region was incubated with nuclear extracts from pBABE-p53 and Mock cells in the presence or absence of 200-fold excess of unlabeled p53 consensus oligonucleotide. Antibody against p53 or IgG was subsequently added. DNA–protein complexes were fractionated by electrophoresis (non-denaturing polyacrylamide gel) and visualized by horseradish peroxidase-conjugated streptavidin.

Binding of WT p53 to FPR regulatory region

By bioinformatic analysis, we identified three segments in FPR gene flanking regions that contain putative PRE (PRE-1: 5′-AGACCAGCCTGGCCAACATGTTG-3′, PRE-2: 5′-CAGCCTGTCTCCAGTTGGACTAGCCA-3′ and PRE-3: 5′-AGGCCAGCTCTCAGCAAGAACCTGACATGCAC-3′). We therefore investigated the capacity of p53 to bind to regulatory regions of FPR gene. By CHIP–PCR, we found that p53 binding to the PRE-1 region was increased in Aza-treated and p53-overexpressing U-87 cells (Figure 6B). In contrast, Aza treatment and p53 overexpression (Figure 6B) reduced the binding of RNA polymerase II to this region with reduced FPR mRNA and protein in such cells. We did not observe the binding of p53 to PRE-2 and PRE-3 regions of the FPR gene (data not shown).

To verify the capacity of p53 to bind to the putative PRE-1 in the FPR regulatory region, we used a synthesized oligonucleotide containing putative PRE-1. The oligonucleotide was labeled and incubated with nuclear proteins isolated from U-87 cells. EMSA showed that the nuclear protein from control cells shifted the putative PRE-1 in the regulatory region of FPR gene. Overexpression of p53 increased the intensity of the shifted band, which was supershifted by addition of p53 antibody (Figure 6C). Further, the shifted nucleotide bands were competitively masked by the presence of an oligonucleotide-containing consensus PRE (Figure 6C).

As additional supporting evidence for the regulatory role of p53 in FPR expression, we tested dexamethasone and found that the glucocorticoid increased FPR expression (supplementary Figure 1A and B is available at Carcinogenesis Online) and function (supplementary Figure 1C and D is available at Carcinogenesis Online) in U-87 cells, which was associated with a reduction in p53 expression (supplementary Figure 1B is available at Carcinogenesis Online) and decrease of binding to the PRE-1 region of FPR gene (Figure 6B).

Discussion

In this study, we have shown that demethylation of the tumor suppressor p53 resulted in increased expression of p53 in the highly malignant human glioblastoma cell line U-87, in association with cell differentiation and reduction of the level of a G protein-coupled chemoattractant receptor FPR that has been implicated in promoting tumor cell chemotaxis, survival and production of angiogenic factors (12,14,15).

The differentiation of tumor cells involves a complex program of changes in gene expression patterns that dictate the transformation of cells into functionally specialized phenotypes. DNA methylation is an epigenetic regulator that constitutes an important genetic information coding system controlling eukaryotic gene expression. The hypomethylation of specific sites in some, but not all, genes correlates with increased transcriptional activity (19). Tumor cells are characterized by a paradoxical alteration of DNA methylation pattern: global DNA demethylation together with local hypermethylation of certain genes, in particular tumor suppressor genes (20). The consequent silencing of tumor suppressor genes allow for tumor cells to maintain their uncontrolled growth and invasive capacity. In this context, reversing the methylation status of tumor suppressor genes may promote tumor cell differentiation. Aza is a chemical compound that inhibits the DNA methyltransferase activity in cells and has been shown to reactivate several different tumor suppressor genes that were silenced by epigenetic disregulation (21–26). p53 is one of the most prominent tumor suppressors also known as the ‘guardian of the genome’ owing to its ability to integrate many signals that control life or death of cells (27). Activated by various types of cellular stress elements, including DNA damage, p53 initiates programs that ultimately arrest proliferation and prevent the generation of genetically altered cells. In fact, WT p53 plays a dual role as a direct regulator of cell apoptosis and as a critical inducible regulator of cell responses to DNA damage (28). In the U-87 glioblastoma cells, p53 has been shown to be of the WT (29). However, the constitutive level of p53 activity is not sufficient to attenuate the highly malignant phenotype of the cells, which also ‘hijack’ the chemoattractant receptor FPR, normally expressed in myeloid cells, for tumor cell growth advantage. Nevertheless, enhancement of p53 either by overexpression or by treatment with Aza in U-87 cells is associated with cell differentiation and concomitant reduction in the production of functional FPR as shown by our present study. It should be pointed out that other G protein chemoattractant receptors such as CXCR4 and CXCR3 have also been reported to play a role in promoting tumor progression (11). However, we did not detect functional expression of either CXCR4 or CXCR3 in U-87 cells, suggesting that these receptors may not be involved in the growth and tumorigenesis of this cell line.

It has been reported that methylation of the promoter region of tumor suppressor genes maybe associated with transcriptional silencing and tumor progression. Although the 5′ region of the p53 gene does not contain a CpG island, there are 16 CpG dinucleotides in a segment essential for full promoter activity (30). It has been reported that methylation in the promoter region of the p53 gene reduces reporter gene activity and p53 was downregulated in cultured cells transfected with a plasmid containing p53 promoter with methylated CpG dinucleotides (31). In our study, by using methylation-specific PCR, we observed methylation of the p53 promoter in an human U-87 glioblastoma cell line, and treatment with Aza upregulated the expression of p53 mRNA and protein. These results suggest that promoter methylation is associated with reduced expression of p53, probably an important determinant of the highly malignant phenotype of U-87 glioblastoma cells. This notion was supported by our observation that overexpression of WT p53 in U-87 cells reproduced the effect of Aza treatment to promote tumor cell differentiation and to reduce the level of the chemoattractant G protein-coupled receptor, FPR, a promoter of the invasive phenotype of the tumor cells. Similar results have been reported in experiments with other cell types (32–35) such as K562 leukemic cells that underwent terminal differentiation with overexpression of exogenous WT p53 (33).

It remains to be elucidated whether WT p53 regulates the transcription of FPR in glioblastoma cells directly, despite our observation that the regulatory region of FPR contains putative PREs and both CHIPs and EMSA revealed physical binding of p53 to the PRE-1 region. It has been reported that in tumor cells, WT p53 represses the expression of prometastatic and antiapoptotic genes by direct binding to promoters, thus interfering with the transcriptional activity of tumor promoters such as SP-1, AP-1 and Ets-1 or alternatively by recruitment of histone deacetylase complexes containing mSin3a (36–38). Interestingly, WT p53 has been shown to suppress the transcription and expression of the chemokine receptor CXCR4 in breast cancer cells by interacting with the receptor promoter (39). These findings, together with our results showing the capacity of dexamethasone to reduce p53 expression and its binding to FPR regulatory region with concomitant upregulation of FPR, support the potential for WT p53 to directly repress FPR transcription.

It is intriguing that demethylation of p53 gene, while increased the nuclear localization of p53 protein, reduced the presence of NFκB in the nuclei. Although in CHIP–PCR we detected that DNA products of FPR PRE-1 region complexed with NFκB, we failed to obtain evidence of direct binding of NFκB to the regulatory region of FPR gene, suggesting that the results of CHIP–PCR with NFκB might be complicated by the presence of other DNA segments containing NFκB-binding sites that are not present in FPR promoter or other transcription cofactors that mediate the ‘indirect’ transcriptional activity of NFκB (40,41). Nevertheless, since NFκB is linked to tumor growth and is constitutively active in a number of malignant tumors (42–48), its downregulation by Aza or p53 overexpression maybe considered as additional evidence for glioblastoma cell differentiation.

In summary, our study provides novel evidence for the mechanisms potentially involved in the selective hijacking of the G protein-coupled chemoattractant receptor FPR by human glioblastoma cells to their advantage. Further investigation into the genetic and epigenetic regulation of glioma cell transformation will be important for the identification of pivotal therapeutic targets.

Supplementary materials

Supplementary Figure 1 can be found at http://carcin.oxfordjournals.org/

Funding

Federal funds from the National Cancer Institute, National Institutes of Health (Contract No. NO1-CO-12400); Intramural Research Program of the National Cancer Institute, National Institutes of Health.

Glossary

Abbreviations

Aza

5-Aza-2′-deoxycytidine

CHIP

chromatin immunoprecipitation

DMEM

Dulbecco’s modified Eagle’s medium

EMSA

electrophoretic mobility shift assay

FCS

fetal calf serum

fMLF

N-formyl-methionyl-leucyl-phenylalanine

FPR

formylpeptide receptor

GFAP

glial fibrillary acidic protein

mRNA

messenger RNA

NF

nuclear factor

PBS

phosphate-buffered saline

PCR

polymerase chain reaction

PRE

p53-responsive element

RT²

real-time reverse transcription

WT

wild-type