Using Small Molecules To Identify Critical Signaling Pathways Of Mutant N-RAS In Acute Leukemia Cells

Activating point mutations in NRAS are detected in more than 10% of AML patients, making NRAS an important therapeutic target. Using small molecules to directly target NRAS or inhibit post-translational modification, such as farnesylation, have been extensively investigated. The potential of strategies focused on targeting downstream effectors of RAS, such as RAF or MEK, has been limited by the complexity of RAS signaling, including redundancy and feedback loops. Large-scale RNAi screens have been used to identify genes (TBK1, STK33 and GATA2, for example) that are synthetically lethal with RAS mutations and these are being explored as therapeutic targets. Recognizing the complexity of RAS signaling, we tested the notion that small molecule screens designed to simultaneously inhibit multiple signaling pathways might identify combinations of pathways that are critical for NRAS signaling in leukemic cells. Initially, we created an experimental Ba/F3 cell line model that was completely dependent on oncogenic N-RAS-G12D for growth and survival. Knockdown of NRAS suppressed growth >95%, but could be rescued by interleukin-3 (IL-3). A chemical screen using panels of multi-targeted small molecule kinase inhibitors against BaF3-NRAS-G12D cells revealed a lead compound, NRAS1 (N-(4-methyl-3-(1-methyl-7-(6-methylpyridin-3-ylamino)-2-oxo-1,2-dihydropyrimido[4,5-d]pyrimidin-3(4H)-yl)phenyl)-3-(trifluoromethyl)benzamide), with high selectivity and sensitivity toward leukemia cell lines with NRAS mutations in vitro. A number of studies were then performed to investigate the targets of this compound. Transcriptional profiling before and after treatment of two AML cell lines with NRAS mutation (OCI-AML3 and KO52 cells, respectively) showed profiles similar to that obtained by knocking down NRAS, supporting the hypothesis that this compound suppressed NRAS signaling. Biochemical studies demonstrated that NRAS1 did not inhibit several classical targets of RAS signaling, including, RAF, MEK and ERK. In contrast, NRAS1 was found to substantially reduce AKT and RPS6 phosphorylation. Over-expression of a constitutively active allele of AKT, myrAKT, in Ba/F3-NRAS G12D cells conferred strong resistance to NRAS1, confirming that suppression of phospho-AKT may be important for the function of NRAS1. However, direct inhibition of AKT only partially recapitulated the effects of NRAS1. Kinase selectivity profiling of NRAS1 (1μM) in OCI-AML3 cells (EC50: 0.3μM) identified 13 major binding partners with more than 85% efficacy. The targets consisted mainly of SRC family proteins (ie SRC, FGR, and LYN) and MAPK family proteins (ie GCK, KSH, and p38), but not MEK1/2, ERK1/2 or AKT1-3. A series of analogs of NRAS1 was synthesized and structure/function studies were carried out. One compound, (LKB-0304601, 1% EC50 of original compound) lost the ability to bind to the MAP4K family of proteins, especially GCK (MAPK4K2). A combination effect was observed between a known GCK inhibitor, NG25, and a known allosteric AKT inhibitor, MK-2206, against mutant NRAS-expressing cells. This finding supports the hypothesis that simultaneous inhibition of GCK and AKT has suppressive activity against leukemia cells transformed by NRAS. Furthermore, a putative gate-keeper mutation introduced into GCK (GCK G96S) resulted in partial resistance to growth suppression by NG25 or NRAS1. Growth suppression of NRAS-transformed leukemic cells was further induced by knock-down of GCK by shRNAs in cells with mutant NRAS, THP-1 cells and MOLT-3, and this effect could be rescued by over-expression of GCK. Finally, in a xenotransplant model using NRAS-mutant-expressing OCI-AML3 cells and MOLT-3 cells, NRAS1 significantly reduced tumor burden and prolonged survival compared to controls. Overall, by using a chemical screen designed to inhibit multiple signaling pathways simultaneously in oncogene-addicted cells, followed by signaling studies, cell biological studies and kinase selectivity profiling, we found that simultaneous inhibition of AKT and GCK, by either NRAS1 or selective inhibitors, exhibits activity against NRAS-transformed leukemia cells.