14-3-3sigma gene silencing during melanoma progression and its role in cell cycle control and cellular senescence

Abstract

Background: The family of 14-3-3 proteins plays an important role in cancer biology by interfering with intracellular signalling pathways and cell cycle checkpoints. The 14-3-3sigma isoform acts as a tumor suppressor and is often inactivated during tumor development.

Results: Here, we demonstrate enhanced CpG methylation of the 14-3-3sigma gene in lymph node and cutaneous melanoma metastases compared with primary tumors, associated with dramatically reduced mRNA expression. In line with this, treatment of different metastatic melanoma cell lines with 5-aza-2'-deoxycytidine (5-Aza-CdR), a potent inhibitor of cytosine methylation, significantly induces 14-3-3sigma protein expression. Additional treatment with histone deacetylase inhibitor 4-phenylbutyric acid (Pba) further enhances 14-3-3sigma expression. Induction of 14-3-3sigma expression by 5-Aza-CdR/Pba treatment leads to almost complete inhibition of cell proliferation, with cells predominantly arrested in G2-M. The antiproliferative effect of 5-Aza-CdR/Pba was reversed in 14-3-3sigma knockdown cells. Similarly, melanoma cell lines stably overexpressing 14-3-3sigma show dramatically reduced cell proliferation rates. Moreover, synchronous 14-3-3sigma stably overexpressing cells do not progress through cell cycle, but display a permanent increase in the population of 4n DNA containing cells. Interestingly, overexpression of 14-3-3sigma induces senescence of melanoma cells and is involved in melanoma cell senescence under genotoxic stress. Finally, 14-3-3sigma knockdown supports migratory capacity of melanoma cells in vitro, while 14-3-3sigma overexpression has opposing effects.

Conclusion: Taken together, the present report indicates that epigenetic silencing of 14-3-3sigma might contribute to tumor progression in malignant melanoma via loss of cell cycle control, impaired cellular senescence program and support of migratory capacity.

Figure 1

14-3-3σ mRNA is downregulated in melanoma metastases compared with primary melanomas. Total mRNA was extracted from melanoma cells excised by laser-capture microdissection from primary melanomas (n = 11), lymph node (n = 10) and cutaneous metastases (n = 12), respectively. Quantitative real-time PCR was performed and relative mRNA expression levels are given as means +/- S.E.M. MM, primary melanomas; Meta (ln), lymph node metastases; Meta (s), cutaneous metastases.

Figure 2

The 14-3-3σ gene shows enhanced methylation in melanoma metastases compared with primary melanomas. Methylation of the 14-3-3σ gene was analyzed by methylation-specific PCR. For this purpose genomic DNA extracted from laser-microdissected melanoma tissues was bisulfite-treated and subjected to PCR amplification of the 14-3-3σ gene with primers specific for either unmethylated (U) or methylated (M) DNA alleles. (a) The percentage of methylated fraction is given for a representative set of tumors in each group after densitometric quantification of the PCR products separated by agarose gel electrophoresis. (b) Summary of the percentages of methylated alleles of all samples tested (15 primary melanomas, 20 lymph node metastases and 19 cutaneous metastases). Data are given as means +/- SD, P values represent statistical significance after applying Student's t-test. MM, primary melanomas; Meta (ln), lymph node metastases; Meta (s), cutaneous metastases.

Figure 3

Treatment of melanoma cells with 5-Aza-CdR, Pba, and 5-Aza-CdR/Pba, respectively, induces 14-3-3σ expression and leads to inhibition of cell cycle progression. (a) 14-3-3σ immunoblot analyses of melanoma cell lines SK-Mel-19, SK-Mel-29 and SK-Mel-147, treated with 5-Aza-CdR (3 μM), Pba (3 mM), and a combination of both, respectively, for indicated times points or left untreated. (b) Cell cycle analysis of melanoma cell line SK-Mel-147 by flow cytometry after treatment as described in (a). Data are given as mean values +/- SD of four independent experiments.

Figure 4

14-3-3σ significantly contributes to 5-Aza-CdR/Pba-induced cell cycle inhibition of melanoma cells. (a) Cell cycle analyses were performed for SK-Mel-147 melanoma cells transduced with empty vector, a functional (si14-3-3B4) and non-functional (si14-3-3B7) shRNA, respectively, treated with a combination of 5-Aza-CdR (3 μM) and Pba (3 mM) and (upper panel). 14-3-3σ stably ovexpressing cells after lentiviral transduction served as positive control. Data are given as mean values +/- SD of four independent experiments. Asterisks indicate statistical significance of differences in the number of cells in S or G2-M between vector and si14-3-3σ transduced cells (P < 0.05; Student's t-test). Lower panel shows respective immunoblots. (b) Time course of cell cycle progression of synchronized SK-Mel-147 melanoma cells after release of double thymidine block. Data are given as mean values +/- SD of four independent experiments. (c)Cell numbers of (b) at different time points after release of double thymidine block, together with cell numbers from an untreated control culture of SK-Mel-147 melanoma cells.

Figure 5

14-3-3σ induces cellular senescence in melanoma cells. (a) Expression of senescence marker SA-β-Gal in untransduced, vector control or 14-3-3σ cDNA transduced SK-Mel-147 melanoma cells. (b) SK-Mel-19 melanoma cells were exposed to 3.0 μg/ml adriamycin for indicated time points to induce cellular senescence. Induction of p53, 14-3-3σ and SA-β-Gal was analyzed by immunoblotting or immunohistochemistry in untransduced, vector control or 14-3-3σ shRNA (si14-3-3B4) transduced cells.

Figure 6

14-3-3σ negatively interferes with melanoma cell migration. Stably transfected SK-Mel-147 melanoma cells with respective empty control vector, 14-3-3σ cDNA and 14-3-3σ shRNA (si14-3-3B4), respectively, were used to test the influence of 14-3-3σ on melanoma cell migration. A wounding assay was performed and wound closure was analyzed after 24 h. For inhibition of cell proliferation 10 μg/ml mitomycin C was added 3 h before wounding of the cell layer.