ANGI-07. 5ALA FLUORESCENCE BASED SORTING IDENTIFIES SERPINE1 AS A NOVEL THERAPEUTIC TARGET ON INVASIVE GBM CELLS

Abstract INTRODUCTION: We previously described separating invasive tumor cells from within normal brain using 5ALA based fluorescent activated cell sorting (FACS). Using RNAseq on these cells we identify new potential therapeutic targets whose expression is elevated in tumor cells invading brain parenchyma. METHOD: 5-aminolevulinic acid (5-ALA) is a clinically used drug for fluorescence guided resection of GBM. Tumors from 11 patients were dissociated and FACS used to separate invasive fluorescent cancer cells from within normal brain. FACS sorted and unsorted mixed samples from tumor core, rim and invasive margin were compared. Gene expression was analyzed by RNA-seq and validated by qPCR, IHC and in vivo xenografts. RESULTS Differential expression analysis identified 2567 genes with differences between core and invasive margin, and 78 genes with differences between 5-ALA FACS positive and FACS negative. Interestingly SERPINE1 expression is reduced in the unsorted invasive margin but expression remains high within sorted invasive tumor cells. Pathway analysis identified a predominance of immune system pathway changes between core and invasive margin. The differential expression of SERPINE1, VEGF, CHI3L1 and RTN1 in qPCR and IHC validated the same changes observed from the RNA-seq data. The invasive and tumorigenic capacity of 5-ALA positive sorted tumor cells was confirmed by enhanced engraftment in a mouse flank model compared to unsorted cells, whereas 5-ALA negative sorted cells failed to engraft. SERPINE1 knockdown had no effect on GBM cell proliferation but significantly reduced the ability of GBM cells to invade in an in vitro assay. CONCLUSION: This study has demonstrated that 5-ALA fluorescent sorting of the invasive region can identify new targets such as SERPINE1, whose high expression in invasive tumor cells would otherwise be overlooked. Our approach gives hope that we can interrogate true residual disease, identifying more relevant therapeutic targets.