DDDR-04. TRANSIENT SUPPRESSION OF XPO1 ACTIVITY IS SUFFICIENT FOR SELINEXOR-INDUCED CYTOTOXICITY IN GLIOBLASTOMA CELLS LINES

Abstract

Abstract Inhibitors of the nuclear export protein exportin 1 (XPO1i) show promise in the treatment of leukemias and solid tumors, including brain tumors. XPO1i such as Selinexor reversibly bind and inhibit XPO1 activity, resulting in a lethal nuclear accumulation of a variety of XPO1 substrates. Although XPO1i activity has been suggested to require prolonged suppression of XPO1 activity, this point has not been directly addressed. To do so we generated glioblastoma and patient-derived glioblastoma xenograft cells that express either endogenous wild-type (WT) XPO1 or endogenous WT XPO1 plus a doxycycline-inducible point-mutant form of XPO1 (C528S) that exhibits full XPO1 activity but cannot bind or be inhibited by Selinexor. These cells were then incubated with doxycycline before, or at various time after, the initiation of continuous Selinexor exposure, after which the ability of C528S XPO1 to rescue cells from Selinexor-induced toxicity could be determined. All four glioblastoma cell lines examined were drug-sensitive (IC50 < 600nM) and exhibited a dose-dependent nuclear retention of the XPO1 substrate RanBP1 following Selinexor exposure. Induced expression of C528S led to a doubling of XPO1 protein levels (WT plus C528S), and variable increases in Selinexor IC50, from 3-fold in GBM43 cells to 12,000-fold in LN229 cells. While induction of C528S expression prior to, or up to 8 h after, Selinexor exposure completely prevented Selinexor-induced cytotoxicity, cells could not be rescued after a 24 hr Selinexor exposure. Furthermore, the resistance provided by prolonged expression of C528S, or by shRNA-mediated suppression of XPO1 levels, was nearly 10-fold less than that provided by complete CRISPR-based replacement of WT XPO1 with C528S. These results therefore show that a transient suppression of XPO1 activity is sufficient for Selinexor-induced cytotoxicity in glioblastoma cells, and suggest that the Selinexor-induced lethality results from a combination of loss of XPO1 activity and Selinexor-XPO1 complex formation.