LSD1 Performs Demethylase-Independent and Context-Specific Roles in Ewing Sarcoma

Lysine specific demethylase 1 (LSD1), encoded by the gene KDM1A, is overexpressed and correlates with poor patient prognosis in Ewing sarcoma. LSD1 colocalizes with the FET-fusion (EWSR::FLI1, EWS::ERG, etc) throughout the genome and has therefore been considered a critical co-regulator driving the progression of Ewing sarcoma. However therapeutic targeting of LSD1 by irreversible and noncompetitive inhibitors has yielded mixed results. Irreversible, enzymatic inhibition seems ineffective, but reversible noncompetitive inhibition has predominant off target mechanisms, leaving open the question of LSD1 function in Ewing sarcoma. Here we take a robust approach through multiple methods of depletion in multiple cell lines to define LSD1 function, both enzymatic and nonenzymatic, in Ewing sarcoma cell lines. We define a core set of 22 genes that are commonly repressed by LSD1 in all cell lines and that repression of these genes downregulates synapse functioning and e-cadherin target genes. De-repression of these genes with LSD1 loss is an early and sustained genotype in all cell lines tested. Furthermore, we define distinct gene sets in each cell line that are regulated by the enzymatic and nonenzymatic functions of LSD1 and find repression of e-cadherin target genes to be nonenzymatically regulated. This finding supports the growing body of evidence that in addition to their canonical catalytic activity, enzymes serve essential noncanonical roles as well. Lastly, we uncovered evidence through use of the irreversible inhibitor OG-L002 that 2D cytotoxicity and proliferation assays may be insufficient to determine how Ewing sarcoma responds to LSD1 inhibition.