Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 171 (24), 5708-27

Autocrine Secretion of 15d-PGJ2 Mediates Simvastatin-Induced Apoptotic Burst in Human Metastatic Melanoma Cells

Affiliations

Autocrine Secretion of 15d-PGJ2 Mediates Simvastatin-Induced Apoptotic Burst in Human Metastatic Melanoma Cells

Christine Wasinger et al. Br J Pharmacol.

Abstract

Background and purpose: Despite new therapeutic approaches, metastatic melanomas still have a poor prognosis. Statins reduce low-density lipoprotein cholesterol and exert anti-inflammatory and anti-proliferative actions. We have recently shown that simvastatin triggers an apoptotic burst in human metastatic melanoma cells by the synthesis of an autocrine factor.

Experimental approach: The current in vitro study was performed in human metastatic melanoma cell lines (A375, 518a2) and primary human melanocytes and melanoma cells. The secretome of simvastatin-stressed cells was analysed with two-dimensional difference gel electrophoresis and MS. The signalling pathways involved were analysed at the protein and mRNA level using pharmacological approaches and siRNA technology.

Key results: Simvastatin was shown to activate a stress cascade, leading to the synthesis of 15-deoxy-12,14-PGJ2 (15d-PGJ2 ), in a p38- and COX-2-dependent manner. Significant concentrations of 15d-PGJ2 were reached in the medium of melanoma cells, which were sufficient to activate caspase 8 and the mitochondrial pathway of apoptosis. Inhibition of lipocalin-type PGD synthase, a key enzyme for 15d-PGJ2 synthesis, abolished the apoptotic effect of simvastatin. Moreover, 15d-PGJ2 was shown to bind to the fatty acid-binding protein 5 (FABP5), which was up-regulated and predominantly detected in the secretome of simvastatin-stressed cells. Knockdown of FABP5 abolished simvastatin-induced activation of PPAR-γ and amplified the apoptotic response.

Conclusions and implications: We characterized simvastatin-induced activation of the 15d-PGJ2 /FABP5 signalling cascades, which triggered an apoptotic burst in melanoma cells but did not affect primary human melanocytes. These data support the rationale for the pharmacological targeting of 15d-PGJ2 in metastatic melanoma.

Figures

Figure 1
Figure 1
FABP5 expression in metastatic melanoma cells. (A) Analysis of the secretome of 518a2 melanoma cells (by 2D-DIGE) exposed to 10 μM simvastatin (green) or 10 ng·mL−1 vincristine (red) for 48 h. A merged picture revealed three proteins of interest, identified by MS (protein identification number). Fold up-regulation is given as mean ± SD (n = 3). Quantification of the proteins of interest (arrows) was confirmed by three-dimensional illustrations (B). Human metastatic melanoma cells, 518a2 and A375, were treated with simvastatin (Sim) for 48 h in the absence and presence of siRNA targeting FABP5 and were analysed for FABP5 protein (C), FABP5 mRNA (D), or cleaved caspase 3, 8 and 9 (E). Quantitative PCR of FABP5 depicts wild-type cells (grey bars) with FABP5 knockdowns (white bars) (n = 3–23). Asterisks indicate significance versus control (*P < 0.05; **P < 0.005; ***P < 0.0005).
Figure 2
Figure 2
PPAR-γ in simvastatin-treated human melanoma cells. (A) Protein and (B) mRNA levels of PPAR-γ were detected in simvastatin (Sim)-treated metastatic melanoma cells. Analogous to Figure 1D, quantitative PCR for PPAR-γ is illustrated in the absence (grey bars) and presence (white bars) of FABP5 siRNA from simvastatin-treated cells. Similarly, the PPAR-γ targets cyclin D1 and p21 were compared (n = 3–23) (D). (C) PPAR-γ (green) and FABP5 (red) were also visualized in simvastatin-treated cells by confocal microscopy. Nuclei were stained with Hoechst 33342 (grey). Asterisks indicate significance versus control (*P < 0.05; **P < 0.005; ***P < 0.0005; n.s., not significant).
Figure 3
Figure 3
Simvastatin stimulates stress activation via p38. RhoA, Cdc42 and α-tubulin are depicted from cells treated with simvastatin (Sim) for 4 (A) and 24 h (B). The unprocessed forms of the G-proteins are indicated by a red arrow. (C, D) The activation of p38 kinase was monitored by a phospho-specific antibody (p-p38) and compared with total p38 and α-tubulin. (E) Quantification of the p38 phosphorylation is illustrated (n = 3–7). Asterisks indicate significance versus control (*P <0.05; **P < 0.005; ***P < 0.0005).
Figure 4
Figure 4
Simvastatin up-regulates COX-2. (A) Simvastatin (Sim)-exposed melanoma cells were probed for protein and (B) mRNA levels of COX-2 (n = 5–10). (C) Confocal fluorescence microscopy images of simvastatin (10 μM)-treated cells revealing enhanced cytosolic COX-2 (green) staining, compared with controls (CTL); nuclei stained with Hoechst 33342 (blue). Representative images are shown. (D) Melanoma cells were exposed to combinations of simvastatin and the p38 inhibitor SB-203580 (10 μM) and analysed for COX-2. Asterisks indicate significance versus control (*P < 0.05; **P < 0.005; ***P < 0.0005).
Figure 5
Figure 5
Simvastatin-induced caspase 9 activation is prevented by inhibition of p38 or COX-2, but not COX-1. (A) Metastatic melanoma cells (518a2 and A375) were incubated for 48 h in the absence and presence of simvastatin (Sim) or the specific COX-1 inhibitor SC-560 to measure caspase 9 activation. (B) The specific inhibitors of p38 (SB-203580) and COX-2 (NS-398) inhibited simvastatin (Sim)-triggered caspase 9 activity. Data represent means ± SEM (n = 3–6). (C) Cells were stained with JC-1 and analysed by FACS for the ratio of the red/green signal after treatment with simvastatin for 48 h. Data represent the means ± SEM (n = 8–12). Asterisks indicate significance versus control (**P < 0.005; ***P < 0.0005). Hashes indicate significance versus corresponding simvastatin treatment ###P < 0.0005; n.s., not significant).
Figure 6
Figure 6
Simvastatin-induced ROS production is associated with caspase 9 activation. (A) The 518a2 and A375 melanoma cells were exposed to simvastatin (Sim) in the absence and presence of 5 μM N-acetylcysteine (NAC; 5) or 5 μM NAC every 24 h (2 × 5). After 48 h, ROS formation was detected by FACS. (B) Caspase 9 activity was measured after 48 h in cells treated as given in (A). Bars indicate means ± SEM (n = 3–10). Asterisks indicate significance versus control (***P < 0.0005). Hashes indicate significance versus corresponding simvastatin treatment (#P < 0.05; ##P < 0.005; ###P < 0.0005; n.s., not significant).
Figure 7
Figure 7
COX-2 and p38 control simvastatin-induced ROS production. The 518a2 (A, B) and A375 (C, D) melanoma cells were exposed to simvastatin (Sim) in the absence and presence of 10 μM SB-203580 or 50 μM NS-398 for 24 (A, C) or 48 h (B, D). Cells were stained for ROS detection by FACS. Data represent the means ± SEM (n = 8–12). Asterisks indicate significance versus control (*P < 0.05; **P < 0.005; ***P < 0.0005). Hashes indicate significance versus corresponding simvastatin treatment (#P < 0.05; ##P < 0.005; ###P < 0.0005).
Figure 8
Figure 8
Simvastatin mediates 15d-PGJ2 production. (A) elisa of 15d-PGJ2 kinetics in the cytosol of simvastatin (Sim)-treated 518a2 or A375 melanoma cells. (B) The extracellular concentration of 15d-PGJ2 in the medium of simvastatin-treated cells is depicted. Data points represent means ± SEM (n = 3–8). (C) HPLC analysis using a 15d-PGJ2 standard (0.3 μg), or 50 μL of the concentrated medium of untreated (control) and simvastatin (10 μM)-treated A375 cells (48 h). Asterisks indicate significance versus control (*P < 0.05; **P < 0.005).
Figure 9
Figure 9
Simvastatin-induced 15d-PGJ2 secretion is dependent on p38 and COX-2. Human 518a2 (A) and A375 (B) melanoma cells were exposed to simvastatin (Sim), alone or in combination with SB-203580 or NS-398. After 48 h, the extracellular 15d-PGJ2 concentration was measured. Data represent means ± SEM (n = 3–8). Asterisk indicates significance versus control (**P < 0.005). Hash indicates significance versus corresponding simvastatin treatment (#P < 0.05).
None
Simvastatin induces PGDS. (A) The mRNA levels of L-PGDS and H-PGDS were determined by quantitative PCR in 518a2 or A375 cells and compared with simvastatin (3 μM, 48 h) treatment (n = 3–12). (B) L-PGDS protein was detected. (C) The kinetics of L-PGDS up-regulation are depicted. Data represent means ± SEM (n = 6). Asterisks indicate significance versus control (*P < 0.05; ***P < 0.0005).
Figure 11
Figure 11
Simvastatin-induced apoptosis is dependent on 15d-PGJ2. Cells were incubated in the absence and presence of simvastatin (Sim) or the L-PGDS inhibitor AT-56 as indicated. Cell lysates were analysed for caspase 9 (n = 3–8) (A), caspase 3 (n = 3–8) (B) and caspase 8 (n = 4–9) (C) activity. Asterisk indicates significance versus control (***P < 0.0005). Hash indicates significance versus corresponding simvastatin treatment (###P < 0.0005).
Figure 12
Figure 12
Exogenous 15d-PGJ2 triggers ROS formation and apoptosis. The 518a2 and A375 cells were treated with simvastatin (Sim) or 15d-PGJ2 and ROS formation was analysed after 4 (A, B) and 48 h (C, D). Caspase 8 (E) and caspase 9 (F) are activated by 160 nM 15d-PGJ2, comparable to the effects of simvastatin. Data represent means ± SEM (n = 3–4). Asterisks indicate significance versus control (*P < 0.05; **P < 0.005; ***P < 0.0005).
Figure 13
Figure 13
Simvastatin is a trigger for 15d-PGJ2-induced apoptosis in primary human metastatic melanoma cells but not in melanocytes. Primary human melanocytes, ulli (A) and NHEM (B) and primary human metastatic melanoma cells 6F (C) were incubated with simvastatin for 24 or 48 h in order to measure caspase 3 activity and 15d-PGJ2 formation (n = 3). The 6F cells were also treated in the absence (CTL) and presence of 160 nM 15d-PGJ2 for 24 h (D) to measure caspase 8 (n = 4) and 9 (n = 3) activity. Asterisks indicate significance versus control (*P < 0.05; ***P < 0.0005).
Figure 14
Figure 14
Schematic illustration of 15d-PGJ2 signalling in simvastatin-treated metastatic melanoma cells.

Similar articles

See all similar articles

Cited by 4 articles

Publication types

MeSH terms

Feedback