Deciphering the Transcriptional Modulation Program of Trabectedin in Sensitive and Resistant Patient Derived Xenograft Models of Myxoid Liposarcoma .

Laura Mannarino1 , Ilaria Craparotta , Sara Ballabio , Roberta Frapolli , Enrica Calura , Maurizio Callari , Sergio Marchini , Maurizio D'Incalci.

AUTHOR AFFILIATIONS
1 Mario Negri Institute, Italy.

PICAR SOBRE ESTA IMAGEN.

PRESENTED AT ISMB/ECCB 2019 .

Abstract :

Myxoid liposarcoma (MLP) is a soft tissue sarcoma characterized by the expression of the oncogenic chimeric protein FUS-DDIT3, which prevents adipocytes differentiation. Evidences suggest that trabectedin, a marine alkylating gent, is able to restore adipocytes differentiation in MLP. We used a RNA-Seq approach to investigate trabectedin activity both on tumor cells and the surrounding stroma of two patient-derived xenograft (PDX) models either sensitive (ML017) or resistant (ML017/ET) to the drug. Samples from the two models ML017 and ML017/ET were collected at 24 and 72 hours after the first dose, to study the early effects of the drug, and 15 days after the third dose of trabectedin to study the late effects. RNA-Seq data were processed expecting at least 60 million of reads per sample. Alignment was performed following the ICRG approach (Callari, 2018) using HISAT2 aligner (Kim, 2015) on a human-mouse hybrid-genome. Gene expression was quantified with Salmon (Patro, 2017). Data were processed with DeSeq2 (Love, 2014).

This approach allowed us to study both tumor and stroma cells. Focusing on the first, pathway analysis in ML017 PDX model show that the early effects of the drug were related to the TP53 pathway, while at 15 days after the third dose the main involved pathways were those of extracellular matrix remodelling and collagene production, resulting in the adipose tissue differentiation. None of these pathways was affected by trabectedin exposure in ML017/ET. Moreover, tumor microenvironment analysis revealed that also the expression profiles of mouse stroma genes is modulated by the drug in the ML017 PDX model.

In conclusion, this novel computational approach allowed us to investigate the dual effects of trabectedin on tumor cells and the surrounding microenvironment in PDX models sensitive or resistant to the drug. While the effects on the tumor microenvironment need to be further investigated, transcriptional response to drug clearly show that the resistant tumor does not activate adipocytes differentiation unlike the sensitive model.