Fig. 3: The previously unreported JAK2 p.G993A mutation confers resistance to multiple type-I JAK inhibitors.

a Expression of GFP in Ba/F3 cells expressing JAK2 p.G993A-mutant PAX5-JAK2, ETV6-JAK2, or ATF7IP-JAK2 were assessed by flow cytometry. Non-transduced parental Ba/F3 were used as negative controls. Histograms are representative of three biological replicates and the GFP mean fluorescence intensities (MFIs) are shown. b Ba/F3 cells expressing either non-mutant or JAK2 p.G993A-mutant PAX5-JAK2 (blue), ETV6-JAK2 (red), or ATF7IP-JAK2 (purple) were incubated for 72 h with either a DMSO vehicle control or a dose response of the type-I JAK inhibitor, rux. The percentage of cell death was measured following a 20 min incubation with apoptotic markers and analysis by flow cytometry. Linear regression or non-linear regression models were fit to appropriate normalized data. Error bars indicate SEM over the mean of three biological replicates. c–h Ba/F3 cells expressing either non-mutant (gray) or JAK2 p.G993A-mutant (black) JAK2 fusion genes were incubated in 1 µM of type-I JAK inhibitors BMS-911543 (c), AZD-1480 (d), fedratinib (e), rux (f), pacritinib (g), or momelotinib (h) for 1 h. STAT5 p.Y694 phosphorylation was assessed by intracellular flow cytometry. JAK2 Ba/F3 cells starved of IL3 for 5 h were included as a measure of baseline STAT5 phosphorylation. pSTAT5-PE MFIs were plotted. Error bars indicate SEM over the mean of three biological replicates and significance was determined by unpaired t-tests in comparison to respective non-mutant cells (*p < 0.05, **p < 0.01, ***p < 0.001).