(P 456) SECLIDEMSTAT TREATMENT BLOCKS THE TRANSCRIPTIONAL FUNCTION OF MULTIPLE FET-FUSION PROTEINS

Objective: FET-rearranged sarcomas are characterized by chromosomal translocations involving the genes encoding a protein from the FET family (FUS, EWSR1, and TAF15). The fusion partner in these malignancies is most often a transcription factor and the resulting chimeric proteins function as oncogenic transcription factors. These fusion proteins recruit chromatin and transcriptional regulators to change the expression of thousands of genes. The protein domains included in the oncogenic fusions, including both low complexity domains and sequence-specific DNA binding domains, are difficult to pharmacologically target. Instead, efforts have been made to target critical co-regulators of the FET fusion proteins. The histone lysine specific demethylase 1 (LSD1) has been identified as a critical chromatin coregulator and inhibition with the noncompetitive inhibitor, SP-2509, has been shown to block the transcriptional function of EWS::FLI and EWS::ERG. The clinical analog of SP-2509, SP-2577 (seclidemstat) is currently in clinical trials for Ewing sarcoma. Multiple lines of evidence suggest LSD1 interacts with the FET domain of the fusion proteins and we wanted to evaluate whether seclidemstat treatment blocked transcriptional function of other known FET fusion proteins, including EWS::WT1, EWS::ATF, and FUS::DDIT3.

Methods: We used multiple cell lines to evaluate the transcriptional effects of seclidemstat including A673 (EWS::FLI), TTC-466 (EWS::ERG), JN-DSRCT-1 (EWS::WT1), BER (EWS::WT1), SU-CCS-1 (EWS::ATF), DTC1 (EWS::ATF), 402-91 (FUS::DDIT3), 1795-92 (FUS::DDIT3), DL221 (FUS::DDIT3). Dose response curves were generated after 96 hours of treatment for each cell line using Cell Titer Glo. Cells were treated with either the vehicle (0.3% DMSO), the IC50 for the cell line, or the IC90 for the cell line. Samples were collected for bulk RNA-sequencing at 24 hours and 48 hours. Sequencing was performed using standard approaches at the Institute for Genomic Medicine at Nationwide Children’s Hospital, and differential gene expression was determined and analyzed using previously published pipelines.

Results: We found that similar to SP-2509, seclidemstat causes widespread transcriptional changes in Ewing sarcoma cells and reverses both EWS::FLI and EWS::ERG function, in vitro. Having validated these findings, we next defined the fusion-driven transcriptional signature for each of the fusions and compared seclidemstat treatment to the fusion activity. We found that seclidemstat treatment also reverses the transcriptional function of the other FET-rearranged proteins, including both activated and repressed gene signatures. Additionally, we noted a limited change in gene expression at 24 hours, with extensive transcriptional reprogramming at 48 hours. Both the IC90 and IC50 doses induced changes in the same sets of genes. A small set of genes was commonly regulated across all of the cell lines.

Conclusion: These data indicate that seclidemstat blocks the transcriptional activity of FET-rearranged proteins in tumors other than Ewing sarcoma. This suggests LSD1 is critical co-regulator in other FET-rearranged malignancies. Our data also identify the heterogeneity in fusion-driven transcription and the response to seclidemstat, both across tumor types and with different cell lines of the same tumor type. This supports the highly context dependent role of the FET fusion proteins, as well as LSD1. Nonetheless, the commonly regulated genes suggest a key regulatory role for proteotoxic stress in the response to seclidemstat, providing a possible pathway for combination strategies.