By fusing a glutathione S-transferase gene to the JAK2 activation loop, we are able to isolate and directly probe for JAK2 phosphorylation of a bona fide JAK2 substrate. Interestingly, some of the identified mutations in TEL-JAK2 did not translate to resistance in Jak2 V617F. We evaluated the entire panel of mutations in the context of Jak2 V617F with XTT-based BKM120 survival, downstream signaling, and with the GST-J2s kinase assay. We observed only JAK2 V617F G935R to display a striking difference in survival, downstream signaling, and substrate phosphorylation in comparison to the wild-type protein and other mutants. There are at least two possible explanations for this finding. First, the difference may be due to the relative kinase strength of TEL-JAK2 compared to Jak2 V617F. The Jak2 V617F allele is not transforming unless it has a functional FERM domain and is provided with a cytokine scaffold, and even then is relatively indolent without other mutations present. In contrast, TEL-JAK2 is a potent oncogene, thought to be causative in some cases of acute myeloid leukemia. Therefore, even small differences in inhibitor resistance will be evident with TELJAK2, while the homologous mutations may have subtle effects in the context of Jak2 V617F. Second, the mechanisms of activation of TEL-JAK2 and Jak2 V617F are different. The PNT dimerization domain of TEL causes oligimerization of the TELJAK2 protein and constitutive activation. Therefore, the inhibitor resistance observed in some TEL-JAK2 mutations may be due to the oligimerization-specific interaction between the kinase domains. In order to understand how the panel of identified mutations contributes to inhibitor resistance, mutations were modeled using the previously published JAK2 kinase domain crystal structure complexed with JAK Inhibitor-I. The unmutated kinase domain residues isolated in the screen are displayed. G935 lies within the hinge region between the WZ8040 N-lobe and C-lobe. The G935R mutation introduces a spatial clash resulting from the arginine side chain, which prevents inhibitor binding. R975 is located in the catalytic loop region connecting a-helix D with the activation loop. The replacement of arginine by glycine, combined with increased flexibility of the main chain, would influence inter-loop interactions, possibly affecting opening of the pocket. E864K results in a change in side chain charge, and would result in a steric clash with a neighboring lysine. This would result in movement of the b-sheet and occlusion of the pocket. N909K introduces a steric clash that may push neighboring V911 into the binding pocket. The V881A mutation will result in loss of the valine in the hydrophobic core, thereby affecting packing and orientation. A recent publication has identified activating JAK1 mutations selected for by cytokine deprivation. Interestingly, some of these mutations also confer resistance to the JAK inhibitors CMP6 and ruxolitinib.