Piceatannol suppresses the metastatic potential of MCF10A human breast epithelial cells harboring mutated H-ras by inhibiting MMP-2 expression

Abstract

Metastasis is one of the most threatening features of the oncogenic process and the main cause of cancer-related mortality. Several studies have demonstrated that matrix metalloproteinases (MMPs) are critical for tumor invasion and metastasis. Resveratrol (3,5,4'-trihydroxystilbene), a phenolic compound of red wine, has been reported to be a natural chemopreventive agent. However, the cancer preventive effects of piceatannol (3,5,3',4'-tetrahydroxystilbene), a metabolite of resveratrol and the underlying molecular mechanisms have not yet been fully elucidated. In this study, we report that piceatannol inhi-bits H-ras-induced MMP-2 activity and the invasive phenotype of MCF10A human breast epithelial cells harboring mutated H-ras (H-ras MCF10A cells) more effectively than resveratrol. Piceatannol attenuated the H-ras-induced phosphorylation of Akt in a time- and dose-dependent manner, whereas resveratrol, at the same concentrations, did not exert an inhibitory effect. In vitro kinase assays demonstrated that piceatannol significantly inhibited phosphatidylinositol 3-kinase (PI3K) activity and suppressed phospha-tidylinositol (3,4,5)-trisphosphate (PIP3) expression in the H-ras MCF10A cells. Ex vivo pull-down assays revealed that piceatannol directly bound to PI3K, inhibiting PI3K activity. Data from molecular docking suggested that piceatannol is a more tight-binding inhibitor than resveratrol due to the additional hydrogen bond between the hydroxyl group and the backbone amide group of Val882 in the ATP-binding pocket of PI3K.

Figure 1

Effects of piceatannol and resveratrol on matrix metalloproteinase-2 (MMP-2) activity in H-ras MCF10A cells. (A) Chemical structures of piceatannol and resveratrol. (B) Effect of piceatannol and resveratrol on MMP-2 activity in H-ras MCF10A cells. Cells were starved in serum-free DMEM/F12 for 24 h and then treated with samples at the indicated concentrations for 24 h. The conditioned medium was analyzed by zymography, as described in Materials and methods. (C) The relative level of MMP-2 activity was quantified densitometrically using Image J software. Data are representative of three independent experiments that produced similar results. (D) Effects of piceatannol and resveratrol on H-ras MCF10A cell viability. Cells were treated for 24 h in the presence or absence of piceatannol or resveratrol. Cell viability was quantified using MTT assay, as described in Materials and methods. Data are the means ± SD of three independent experiments.

Figure 2

Inhibitory effects of piceatannol and resveratrol on the invasion, migration and neoplastic transformation of H-ras MCF10A cells. (A and B) Invaded cells were quantified by counting the cells that migrated to the lower side of the filter in random microscopic fields. Each bar indicates the mean ± SD (n=3). ^*p<0.05, ^**p<0.01 compared with the control. (C) Confluent H-ras MCF10A cells in serum-free medium were treated with resveratrol or piceatannol. Thereafter, the widths of the injury lines were measured at 0, 12, 18 and 24 h. Results are expressed as the widths of the injury lines relative to the untreated controls at 0 h, as determined from three independent experiments. Data are the means ± SD. (D) Comparison of the inhibitory effects of piceatannol and resveratrol on the H-ras-induced neoplastic transformation of MCF10A cells. Cells were treated as described in Materials and methods and the number of colonies was determined 15 days later: (a) untreated control; (b) piceatannol 5 μM; (c) piceatannol 10 μM; (d) piceatannol 20 μM; (e), resveratrol 5 μM; (f) resveratrol 10 μM; (g) resveratrol 20 μM. The colonies were counted under a microscope using Image-Pro Plus software version 4. Data are presented as the mean numbers of colonies ± SD, as determined by three independent experiments. ^**p<0.01 and ^***p<0.001, significant difference between the group treated with piceatannol or resveratrol and the untreated group.

Figure 3

Effects of piceatannol and resveratrol on the phosphorylation of Akt and phosphatidylinositol 3-kinase (PI3K) activity in H-ras MCF10A cells. (A) After 24 h of serum starvation, the cells were treated with 20 μM piceatannol for the indicated periods of time. (B) After 24 h of starvation, the cells were treated with 5, 10, or 20 μM piceatannol or resveratrol for 15 min. The levels of phosphorylated and total Akt protein were determined by western blot analysis, as described in Materials and methods, using specific antibodies against the corresponding phosphorylated and total protein. Each experiment was performed in triplicate. (C) Effects of piceatannol and LY294002 on PI3K activity in vitro. Active PI3K (80 ng) was pre-incubated with piceatannol and LY294002 at 1.25, 2.5, 5, 10, 20, or 40 μM for 10 min at 30°C, then incubated with phosphatidylinositol substrate and [γ-³²P]ATP for an additional 10 min at 30°C. The resulting ³²P-labeled phosphatidylinositol phosphate (PIP) was measured as described in Materials and methods. The relative level of PI3K activity was quantified densitometrically using Image J software. Data are the means ± SD of three independent experiments. (D) An immunofluorescence assay was performed on the cells for PI3K activity using anti-PIP₃ antibody. The immunostained cells (green) were observed under a confocal microscope.

Figure 4

Effects of the phosphatidylinositol 3-kinase PI3K inhibitor (LY294002) and Akt inhibitor (triciribine) on H-ras-induced matrix metalloproteinase-2 (MMP-2) activity, and on the migration and invasion of MCF10A cells. (A and B) Cells were starved in serum-free DMEM/F12 for 24 h and then treated with samples at the indicated concentrations for 24 h. The conditioned medium was analyzed by zymography, as described in Materials and methods. (C) Invaded cells were quantified by counting the cells that had migrated to the lower side of the filter in random microscopic fields. Each bar indicates the mean ± SD (n=3). ^*p<0.05, ^**p<0.01 compared to the control. (D) The confluent H-ras MCF10A cells in serum-free medium were treated with 40 μM LY294002 or 20 μM triciribine. Thereafter, the widths of the injury lines were measured at 0, 12, 18 and 24 h. Results are expressed as the widths of the injury lines relative to untreated controls at 0 h, as determined from three independent experiments. Data are the means ± SD.

Figure 5

Direct binding between piceatannol and phosphatidylinositol 3-kinase (PI3K). (A) Ex vivo binding of PI3K and piceatannol was confirmed by immunoblotting using an antibody against PI3K: lane 1, input control using whole-cell lysate from H-ras MCF10A cells; lane 2, negative control using a precipitated lysate of H-ras MCF10A cells with Sepharose 4B beads; lane 3, whole-cell lysate from H-ras MCF10A cells precipitated by piceatannol-bound Sepharose 4B beads. (B) Hypothetical model of PI3K in complex with piceatannol. The Ras-binding domain, C2 domain and helical domain of PI3K are colored gray. The N-lobe, C-lobe and hinge region of the catalytic domain are colored green, purple and yellow, respectively. Piceatannol (atomic color) binds to the ATP binding site in the catalytic domain of PI3K. ATP (violet) is overlaid on the model structure of the complex for comparison. The hydrogen bonds are depicted as dashed lines.