Abstract Ruxolitinib/Jakafi is a current front line therapy for JAK2-mediated myelofibrosis. Like other chemotherapy agents, prolonged use of the drug confers resistance. However, the mechanisms that allow for the development of Ruxolitinib resistance are poorly understood. Moreover, there is a paucity of alternative treatment options for these drug resistant patients. To address these timely issues, we first took BaF3/JAK2-V617F cells and selected for a Ruxolitinib resistant variant. After selection through increasing concentrations of Ruxolitinib, the cells grow quite rapidly in 500 nM of the drug; normal IC50 is 125 nM. Both the parental and drug resistant cell lines were then subjected to NGS of the whole exome. In the parental cells, JAK2-V617F is the driver mutation. Pathways downstream of JAK2-V617F include the activation of STAT3, STAT5, PI3K/AKT/NFkB and to some extent the SHC1/ERK pathway. Analysis of the variant strain revealed that the cells acquired Ruxolitinib resistance by tuning down all the key pathways downstream of the driver JAK2 mutation. There was low copy number and a consequent knockdown of key genes downstream of JAK2 including STAT3, STAT5, PI3K, SOS1, ELK1, FOS, and MCL1. Additionally there were other newly acquired genomic aberrations including knockdown of dephosphatases DUSP1, PTPRJ, PTPRR; low copy number of epigenetics regulating genes such as TET2 and KAT5; a knockdown of regulators of beta catenin pathway; and also an increase in the copy number of EGFR. Thus, in this model of JAK2 driven tumorigenesis, the cell is dynamically reacting to Ruxolitinib by decreasing flux through the JAK2 signaling pathway and tuning up other pathways that support growth and survival. The signaling network maps for both cell lines were then overlaid onto a list of FDA approved therapeutics that are used for other indications. Predictive simulation modeling of the dys-regulated signaling networks suggested that Roflumilast, a PDE4 inhibitor, and Chloroquine, an autophagy inhibitor, would be highly efficacious in reducing cell viability. When the parental cells were treated individually with Chloroquine (0-40 uM) or Roflumilast (0-40 uM), cell viability was reduced dose-dependently; Chloroquine IC50 = 15 uM and Roflumilast IC50 = 25 uM. Furthermore, when used in combination, the effect of the two agents was synergistic. These drugs also showed efficacy in the Ruxolitinib resistant cell line, but at higher doses. Specifically, the IC50 for Chloroquine doubled to 30 uM in the drug resistant strain while 20 uM Roflumilast only reduced cell viability by ∼25%. Overall, we have created a cell model for Ruxolitinib resistance and identified signaling networks that are dys-regulated in both the drug sensitive and drug resistant cells. Furthermore, using predictive simulation technology, we have begun to identify alternative drug therapies that effectively reduce the viability of both Ruxolitinib sensitive and Ruxolitinib resistant cells. Citation Format: Peter P. Sayeski, Shireen Vali, Ansu Kumar, Neeraj Singh, Susumu Kobayashi, Taher Abbasi. Identification and therapeutic targeting of signaling pathways in Ruxolitinib resistant cells. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4978. doi:10.1158/1538-7445.AM2015-4978
Read full abstract