The G-protein-coupled formylpeptide receptor FPR confers a more invasive phenotype on human glioblastoma cells

Abstract

Background: The G-protein-coupled formylpeptide receptor (FPR) that mediates chemotaxis of phagocytic leucocytes induced by bacterial and host-derived chemotactic peptides is selectively expressed by highly malignant human gliomas and contributes to tumour growth and angiogenesis. As invasion of surrounding normal tissues is one of the important features of tumour malignancy, we investigated the function of FPR in the invasive behaviour of human glioblastoma cells.

Methods: Cells (FPR(+) and FPR(-)) were isolated by single-cell cloning from a human glioblastoma cell line U-87MG. The FPR expression was assayed by flow cytometry and reverse transcription PCR. The function of FPR was investigated by chemotaxis and calcium flux induced by FPR agonist fMLF. Tumour cell motility was assayed by a wound-healing model in vitro. The growth and invasive phenotype were observed with subcutaneous implantation of tumour cells in nude mice. Over-expression of FPR in FPR(-) cells was performed by transfection of a plasmid vector-containing human FPR gene.

Results: One of the glioma clones F9 that expressed high level of FPR showed a more 'motile' phenotype in vitro as compared with a clone G3 without FPR expression. Although F9 and G3 clones both formed subcutaneous tumours in nude mice, only F9 tumours invaded surrounding mouse connective tissues. Over-expression of FPR in G3 clone (G3F) increased the cell motility in vitro and the capacity of the cells to form more rapidly growing and invasive tumours in nude mice. We further found that, in addition to supernatant of necrotic tumour cells, foetal calf serum and human serum used in culture media contained FPR agonist activity and increased the motility of FPR-expressing glioblastoma cells.

Conclusion: The expression of FPR is responsible for increased motility of human glioblastoma cells and their formation of highly invasive tumours.

Figure 1

The FPR expression and function in FPR⁺ and FPR⁻ subclones from human U-87MG glioblastoma cells. Clones (FPR⁺ (F9) and FPR⁻ (G3)) were isolated by single-cell clone. (A) FACS analysis of FPR expression on tumour cells. The FPR expression was measured by a PE-conjugated mouse monoclonal antibody against FPR. %=percentage of FPR-positive cells; MFI=mean fluorescence intensity. (B) The expression of FPR mRNA. The levels of FPR mRNA were examined by RT–PCR. GAPDH was used as control. (C) The fMLF-induced chemotaxis and Ca²⁺ flux. Cell chemotaxis was assayed in response to fMLF. The results are expressed as the mean migrated cell number (±s.e.) in three high-powered fields of three independent experiments. ^*Indicates significantly increased cell migration as compared with medium control (0) (P<0.01). The fMLF-induced Ca²⁺ flux was measured in a fluorescence spectrometer. The FPR agonist peptide fMLF-induced fluorescence intensity in F9 and G3 cells was expressed as the ratio at wavelength 340 out of 380 calculated by an FLWINLab program. (D) FPR-mediated activation of signal transduction molecules. Lysates of F9 (FPR⁺) and G3 (FPR⁻) cells stimulated with 100 nM fMLF for 10 min were examined for phosphorylated ERK1/2, p38 and EGFR (Tyr992) by western blotting. Total ERK1/2, p38 and EGFR were used as controls.

Figure 2

Morphology and motility of F9 and G3 cells. (A) Analysis of the expression of GFAP and vimentin in F9 and G3 cells. Anti-vimentin and GFAP antibodies were used to label F9 and G3 cells. The proteins were visualised (green) by an FITC-conjugated secondary antibody under confocal microscope. Nuclei were visualised in blue with DAPI. (B) Motility in wound-healing model. F9 and G3 cells grown to confluence on plastic were scratched to create a wound. Cells in 10% FCS/DMEM were photographed at 0 and 8 h to assess the mean distance (mm) of leading cells moving towards the ‘wound’ area. ^*Significantly slower locomotion of G3 cells (P<0.01) as compared with F9 cells. (C) Invasiveness of xenograft tumours. F9 and G3 cells at 5 × 10⁶ cells in 100 μl PBS per mouse were injected subcutaneously into the flanks of athymic mice. After 30 days, the xenograft tumours were sectioned and stained by HE. Arrows indicate F9 tumour intruding mouse dermis and well-encircled G3 tumour.

Figure 3

The GBM cell motility and adhesion in vitro. (A, B) Cell motility. The motility of CMPTX-labelled G3M (red) and CMFDA-labelled G3F (green) in wound-healing model were assayed in the presence of 100 nM fMLF or 10% HS by time-lapse photographing under confocal microscope (A). The cell number in a rectangle gate was counted after 4, 8 and 16 h (B). ^*Indicates significant increased cell number of G3F cells as compared with G3M cells and media (P<0.05). (C) Adhesion. Serum-starved F9, G3, GM and G3F cells were cultured on 0.5% gelatin and 30 μg ml^–1 laminin-coated culture plates at 37°C for 30 min followed by washing with ice-cold PBS. The adhesion rate was the percentage of adhered cells in total input cells. ^*Indicates significantly increased adhesion of FPR-expressing cells (F9 and G3F) as compared with G3M cells (P<0.05). Significantly increased cell adhesion as compared with cells treated with medium alone (Media) (P<0.05).

Figure 4

The MMP expression and cell growth in vitro. (A) Expression of MMP2 and MMP9. G3F cells were cultured in the presence or absence of fMLF or 10% HS for 24 h at 37°C, followed by RT–PCR detection of mRNA for MMP2 and MMP9. (B) Zymogram. G3M and G3F cells were cultured in the presence or absence of 100 nM fMLF or 10% HS for 8, 24 and 48 h at 37°C. The supernatants were collected to measure the proteolytic activity by gelatin zymography. (C) Growth of FPR-transfected cells in culture. The FPR-transfected G3F or mock G3M cells were cultured in 0.5% FCS/DMEM containing 10% AlamarBlue with or without 1 μM fMLF. Cell growth was monitored by measuring the absorbance at 570 and 600 nm. The results are presented as the mean of ‘growth index’ (±s.e., n=6). ^#Indicates significantly increased growth of G3F cells as compared with G3M cells (P<0.05) and ^*indicates significantly increased growth of G3F cells in the presence of fMLF as compared with growth in the absence of fMLF (P<0.05). (D) Colony formation in soft agar. Tumour 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 DMEM containing 10% FCS at 37°C. After 3 weeks, tumour colonies were photographed under light microscopy. The results are expressed as the mean (±s.e., n=6) numbers of colonies. ^*Indicates significantly increased sphere numbers formed by G3F cells as compared with G3M cells (P<0.05).

Figure 5

Tumourigenesis, invasiveness and angiogenesis. (A) Tumourigenesis in nude mice. Wild-type G3 (G3WT), G3M and G3F cells at 5 × 10⁶ cells in 100 μl of PBS per mouse were injected subcutaneously into the flanks of athymic mice. Tumour size was expressed as the mean volume (in mm³±s.e.) of the tumours in 10 animals. ^*Indicates significantly increased size of tumours formed by G3F cells as compared with G3M and G3WT cells (P<0.05). (B) Invasiveness in vivo. Sections of the xenograft tumours of G3M and G3F cells were stained by HE revealing fibrous capsule surrounding xenograft tumours. Numbers in the tumour section shown in centre panels represent areas amplified in surrounding panels. Arrows indicate tumour margin with fibrosis. (C) VEGF in tumour cell supernatant. The GBM cells were incubated in 24-well plates (10⁴ cells in 200 μl 1% FCS DMED in each well) in the presence or absence of 100 nM fMLF for 72 h. Culture media were measured for VEGF by ELISA. ^*Indicates significantly increased VEGF level in culture media of fMLF-containing cultures as compared with control medium (P<0.01). (D) VEGF and vessels in xenograft tumours. Frozen tumour sections having VEGF were detected by immunofluorescence staining in red. Tumour cell nuclei were revealed by DAPI in blue. The microvessels in tumours were visualised by staining of CD105. Arrows indicate VEGF fluorescence and CD105-positive microvessels. The levels of VEGF were quantified by the average optical density of the fluorescence. The microvessel densities were expressed as the mean microvessel number (±s.e.) in six high-powered fields of three independent experiments. ^*Indicates significantly increased VEGF level and microvessel density as compared with G3M (P<0.05).

Figure 6

The FPR agonist activity in FCS and HS. (A) FCS-induced migration of F9 and G3 cells. F9 or G3 cells were measured for migratory response to 10% FCS. ^*Indicates significantly increased migration of F9 cells in response to FCS as compared with G3 cells (P<0.05). (B, C) Inhibition of HS-induced tumour cell migration by FPR antagonist cyclosporin H (CsH). F9 (B), G3F and G3M (C) cells were preincubated with 10 μM CsH for 30 min, then were measured for migration in response to EGF, fMLF, HS and FCS. ^*Indicates reduced migration of CsH-treated cells as compared with medium-treated cells (P<0.05). ^#Indicates increased migration of G3F cells as compared with G3M cells (P<0.05). (D) Ca²⁺ flux in G3F cells. G3F cells loaded with Fura-2 were stimulated with different concentrations of HS, followed 100 s later by 1000 nM fMLF (left panel), or the cells were stimulated with different concentration of fMLF, followed by 2% HS. The intensity of the fluorescence was recorded.