Proliferation assay in cytokine-limiting conditions, immunoblot analysis of JAK2 signal transduction, luciferase assay, and colony assay of retrovirally transduced BM cells. (A) MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) test in Ba/F3-EPOR-IL-3–dependent cells stably transfected with different JAK2 vectors. The percentage of proliferating cells was calculated as the percentage of the maximal cell growth observed at EPO concentration of 1.0 U/mL. The transfectants were starved for 12 hours in IL-3–free media and then incubated for 48 hours in the media with different EPO concentration. Results are shown as the mean ± standard deviation (SD) (n = at least 4 tests performed in triplicates). The table below the graph shows statistical evaluation of the growth of individual Ba/F3 transfectants at tested EPO concentrations compared with cell-expressing JAK2 wt. P values were calculated using Origin 6.1 software (OriginLab Corporation). (B) JAK2 downstream signaling in stable Ba/F3-EPOR-IL-3–dependent transfectants. The cells were starved in IL-3–free media for 12 hours and then stimulated with indicated concentrations of EPO for 15 minutes. JAK2_V617F–expressing cells served as a positive control and showed constitutive activation of STAT5 (i-ii). Immunoblot and subsequent densitometry analysis revealed that JAK2_E846D transfectants and also R1063H-expressing cells showed increase in STAT5 activation (i-ii). Each bar represents the ratio of the density of phosphorylated STAT5 (p-STAT5) to the density of normalized total JAK2 and is presented as fold change against the ratio calculated for 1.0 U/mL EPO. For detailed information on JAK2 normalization, see supplemental Materials and methods. Activation of JAK2 and its targets was determined by antibodies recognizing specific phosphorylation sites (as detailed in supplemental Materials and methods); tubulin antibody was used as a loading control. (C) STAT5 transcriptional activity in JAK2-deficient γ-2A cells transfected with various JAK2 complementary DNAs (cDNAs) in the presence of EPOR, MPL, and G-CSF receptor (G-CSFR). The heterozygous configuration was mimicked by cotransfection of JAK2 wt cDNA with JAK2 V617F, JAK2 E846D, or JAK2 R1063H. Four hours after transfection, the cells were stimulated with different concentrations of EPO (i) or stimulated with 10 ng/mL thrombopoietin (ii) or 10 ng/mL G-CSF (iii) and luminescence was detected in cell lysates 24 hours (ii-iii) or 48 hours (i) after transfection using a PerkinElmer Victor X Light analyzer. (i) STAT5 transcriptional activity downstream of JAK2 E846D and R1063H was significantly increased in both EPO-free and EPO-limiting condition in comparison with JAK2 wt in the presence of EPOR; the double mutants expressing both JAK2 E846D and R1063H mutants showed increased STAT5 transcriptional activity over the single mutants in low EPO concentration. (ii-iii) Only JAK2 V617F significantly increased STAT5 transcriptional activity in the presence of MPL and G-CSFR. STAT5 transcriptional activity of JAK2 E846D and JAK2 R1063H cells was comparable to wt cells. The panels show the average of 3 independent experiment ± standard error of the mean (SEM). *P < .05, **P < .01, ***P < .001 using the 2-tailed Student t test. rlu, relative light unit. (D) In vitro colony growth of transduced murine BM. Murine BM cells were infected with retroviruses coding for murine JAK2 wt, V617F, E846D, or R1063H for single mutant configuration and with E846D and R1063H retroviral particles for double mutant configuration. Infected murine BM cells were then plated on methylcellulose media in the absence of EPO (i) or in the presence of 0.1 U/mL EPO (ii). After 2 and 7 days, the plates were evaluated for the formation of CFU-E and BFU-E, respectively. The BM cells infected with the virus containing the empty pMEGIX vector were used as a control. JAK2_E846D and JAK2_E846D/R1063H BM cells have significantly higher numbers of CFU-E and BFU-E colonies in both EPO-free and low EPO conditions when compared with JAK2_wt cells. Cells with JAK2_R1063H mutation had significantly increased CFU-E, but not BFU-E colony formation. The highest number of erythroid colonies was detected in the sample derived from JAK2_V617F positive control. Congruent results were obtained by determination of the replating capacity of primary human cells naturally harboring the JAK2 E846D and R1063H mutations (supplemental Figure 6). The panels show the average of 3 independent experiments. The BFU-E and CFU-E colony numbers are expressed as the number of colonies per transduced cells and are normalized to wt values ± SD. P values were calculated using Origin 6.1 software. *P < .05, **P < .01, ***P < .001. (E) Immunoblot analysis of prolonged activity of E846D-mutant JAK2 kinase. The JAK2 transfectants were after IL-3 starvation stimulated with 20 U/mL EPO for 15 minutes and then incubated in base Iscove modified Dulbecco medium (IMDM) with no additives for indicated periods of time. JAK2_E846D–expressing cells showed prolonged activation of STAT5 and EKR1/2 similarly to JAK2_V617F transfectants as predicted by in silico modeling (supplemental Figure 7).