High-Throughput Screening of Myxoid Liposarcoma Cell Lines: Survivin Is Essential for Tumor Growth

Abstract

Myxoid liposarcoma (MLS) is a soft tissue sarcoma characterized by a recurrent t(12;16) translocation. Although tumors are initially radio- and chemosensitive, the management of inoperable or metastatic MLS can be challenging. Therefore, our aim was to identify novel targets for systemic therapy. We performed an in vitro high-throughput drug screen using three MLS cell lines (402091, 1765092, DL-221), which were treated with 273 different drugs at four different concentrations. Cell lines and tissue microarrays were used for validation. As expected, all cell lines revealed a strong growth inhibition to conventional chemotherapeutic agents, such as anthracyclines and taxanes. A good response was observed to compounds interfering with Src and the mTOR pathway, which are known to be affected in these tumors. Moreover, BIRC5 was important for MLS survival because a strong inhibitory effect was seen at low concentration using the survivin inhibitor YM155, and siRNA for BIRC5 decreased cell viability. Immunohistochemistry revealed abundant expression of survivin restricted to the nucleus in all 32 tested primary tumor specimens. Inhibition of survivin in 402-91 and 1765-92 by YM155 increased the percentage S-phase but did not induce apoptosis, which warrants further investigation before application in the treatment of metastatic MLS. Thus, using a 273-compound drug screen, we confirmed previously identified targets (mTOR, Src) in MLS and demonstrate survivin as essential for MLS survival.

Figure 1

Hits of high throughput in vitro drug screen of three MLS cell lines. List of 27 drugs with a reduction in cell viability of >50% in two or all three cell lines at a drug concentration of 100 nM. Strong inhibitory effect of the survivin inhibitor YM155 is observed in two out of three MLS cell lines. Also, a good response is observed to several conventional chemotherapeutics, like doxorubicin, gemcitabine, and paclitaxel. Per drug, four concentrations (1, 10, 100, and 1000 nM) are tested; green boxes correspond to a high cell viability (~100%) and red boxes to a loss of cell viability (~0%). For comparison, at the bottom, five compounds are randomly shown that did not meet the criteria.

Figure 2

Role of survivin in MLS cell lines. Dose-response curves for YM155 (72 hours) in MLS cell lines. Error bars represents three experiments performed in triplicates (A). Normalized RNA expression of three survivin isoforms in cell lines showing relative abundance of the Δex3 isoform (B). Survivin immunohistochemistry of FFPE cell pellets revealed a strong nuclear survivin expression in the cell lines 402-91 (C), 1765-92 (D), and DL-221 (E) (20× magnification).

Figure 3

High nuclear expression of survivin in MLS. Immunohistochemical analysis of nuclear survivin expression in 32 MLS patients showing high expression in tumor components with myxoid, intermediate, and round cell morphology (A). High nuclear survivin expression in tumor with myxoid morphology (B). High nuclear survivin expression in tumor with round cell morphology (C) (20× magnification).

Figure 4

Effect of YM155 on apoptosis and cell cycle. YM155 does not cause PARP-dependent apoptosis but increases S-phase in two of the MLS cell lines. Western blot analysis for PARP and cleaved PARP expression in MLS cell lines (A). FACS cell cycle analysis of MLS cell lines treated with YM155 for 48 hours, measured in two independent experiments. Two cell lines, 402-91 and 1765-92, show a decrease in G₁ and an increase in S-phase after treatment. DL-221 does not show a difference in cell cycle distribution (B).

Supplementary Figure 1

Dose-response curves of MLS cell lines after monotherapy with everolimus (A) and panobinostat (B) after 72 hours of treatment.

Supplementary Figure 2

Knockdown BIRC5. BIRC5 siRNA resulted in a partial decrease of cell viability (402-91, 69%; 1765-92, 55%; and DL-221, 69%). The transfection was successful as all three MLS cell lines and the control siRNA GAPDH showed a minor reduction of cell viability, whereas transfection with the positive control siPLK1 resulted in all three cell lines with a pronounced decrease in cell viability. The transfection reagent alone hardly affected cell viability (A). A reduction in GAPDH and survivin protein expression was found after transfection with, respectively, siGAPDH and siBIRC5 in all three cell lines, although knockdown efficiency was variable (B).

Supplementary Figure 3

Combination treatment of cell line 402-91 and DL-221 with panobinostat and trabectedin (A and B) showed no synergism or antagonism. Combination therapy of three MLS cell lines with trabectedin and YM155 also demonstrated no synergism or antagonism (C-E).