Mechanism of efficacy of trabectedin against myxoid liposarcoma entails detachment of the FUS-DDIT3 transcription factor from its DNA binding sites

Abstract

Background: The marine drug trabectedin has shown unusual effectiveness in the treatment of myxoid liposarcoma (MLPS), a liposarcoma characterized by the expression of the FUS-DDIT3 chimera. Trabectedin elicits a significant transcriptional response in MLPS resulting in cellular depletion and reactivation of adipogenesis. However, the role of the chimeric protein in the mechanism of action of the drug is not entirely understood.

Methods: FUS-DDIT3-specific binding sites were assessed through Chromatin Immunoprecipitation Sequencing (ChIP-Seq). Trabectedin-induced effects were studied on pre-established patient-derived xenograft models of MLPS, one sensitive to (ML017) and one resistant against (ML017ET) trabectedin at different time points (24 and 72 h, 15 days). Data were integrated with RNA-Seq from the same models.

Results: Through ChIP-Seq, here we demonstrate that trabectedin inhibits the binding of FUS-DDIT3 to its target genes, restoring adipocyte differentiation in a patient-derived xenograft model of MLPS sensitive to trabectedin. In addition, complementary RNA-Seq data on the same model demonstrates a two-phase effect of trabectedin, characterized by an initial FUS-DDIT3-independent cytotoxicity, followed by a transcriptionally active pro-differentiation phase due to the long-lasting detachment of the chimera from the DNA. Interestingly, in a trabectedin-resistant MLPS model, the effect of trabectedin on FUS-DDIT3 rapidly decreased over time, and prolonged treatment was no longer able to induce any transcription or post-transcriptional modifications.

Conclusions: These findings explain the unusual mechanism underlying trabectedin's effectiveness against MLPS by pinpointing the chimera's role in inducing the differentiation block responsible for MLPS pathogenesis. Additionally, the findings hint at a potential mechanism of resistance acquired in vivo.

Keywords: Adipogenesis; Chromatin immunoprecipitation sequencing; Heterografts; Liposarcoma; Myxoid; Recombinant fusion proteins; Trabectedin.

Fig. 1

FUS-DDIT3 peaks annotation in ML017 controls. A Genomic distribution (%) of the annotated peaks from the consensus of the control (CTRL) condition in ML017. Color regions as reported in the legend. B Enrichment map obtained from the annotated genes of the consensus CTRL of ML017. Pathways are reported as circles, they are connected when sharing at least 50% of the genes. The node size is proportional to the number of genes in the pathway. The node fill color is proportional to the fraction of the significant genes over the total number of genes in the pathway (see Materials and Methods)

Fig. 2

Differential binding in ET-72h versus CTRL in ML017. A Genomic distribution (%) of the annotated peaks from the DBPs of the comparison ET-72h versus CTRL in ML017. Color regions as reported in the legend. B Read enrichment analysis around DBPs in CTRL and ET-72h conditions in ML017. C MAplot of log concentration versus log Fold Change of the identified DBPs. D Enrichment map of the annotated genes associated with pathways. Pathways are reported as circles, they are connected when sharing at least 50% of the genes. The node size is proportional to the number of genes in the pathway. The node fill color is proportional to the fraction of the significant genes over the total number of genes in the pathway (see Materials and Methods)

Fig. 3

Differential binding in ET-15d versus CTRL in ML017. A Genomic distribution (%) of the annotated peaks from the DBPs of the comparison ET-15d versus CTRL in ML017. Color regions as reported in the legend. B Read enrichment analysis around DBPs in CTRL and ET-15d conditions in ML017. C MAplot of log concentration versus log Fold Change of the identified DBPs. D Enrichment map of the annotated genes associated with pathways. Pathways are reported as circles, they are connected when sharing at least 50% of the genes. The node size is proportional to the number of genes in the pathway. The node fill color is proportional to the fraction of the significant genes over the total number of genes in the pathway (see Materials and Methods)

Fig. 4

Comparison between ChIP-Seq and RNA-Seq data in ML017ET. A Figure shows the number (N) of differentially expressed genes (DEGs, in green), differentially bound peaks (DBPs, in pink), and unique genes associated with DBPs (in magenta), under the three trabectedin-treated conditions (ET-24h, ET-72h, ET-15d) in ML017ET model. B Venn diagram comparing the unique genes related to DBPs at ET-24 and the DEGs under the same condition

Fig. 5

Mechanism of action of trabectedin in myxoid liposarcoma. A Venn diagrams showing the common genes between DEGs, divided in up- and down-regulated, and the unique genes annotated to DBPs, in ET-72h and ET-15d, respectively, in the ML017 model. B Enrichment map showing the common pathways between RNA-Seq and ChIP-Seq analysis at ET-15d in ML017. Common pathways are reported as black circles. Common genes are connected by an edge to the pathway they belong to, and the color shows their transcriptional regulation: red for up-regulated genes, blue for down-regulated genes. The darker the color the greater the regulation. D Schematic overview of the possible mechanism of action of trabectedin in myxoid liposarcoma guided by ChIP-Seq and RNA-Seq results