Abstract 3765

The inherent preponderance of genetic resistance against tyrosine kinase inhibitor (TKI) therapy poses significant challenge for effective treatments. Recent approval of Janus kinase 2 (JAK2) inhibitor INCB018424 (Jakafi, ruxolitinib) for the treatment of myeloproliferative neoplasms (MPNs) prompted us to identify resistant mutations that may pose clinical challenge. In vitro drug selection screening against two leading JAK2 inhibitors (INCB018424 and TG101348) were performed. Here we show that like other kinase inhibitors, INCB018424 is prone to genetic resistance while TG101348 is recalcitrant to develop in vitro resistance. Sequencing of INCB018424 resistant clones identified 211 amino acid substitutions spanning across FERM, SH2, JH2 and the kinase domain. Biochemical and structural modeling studies of these mutants demonstrate that mutations within the active site confer resistance either by direct steric hindrance or destabilizing the architecture of the active site. And mutations from the allosteric sites destabilize the intermediary state of the active and inactive conformations of JAK2 to which INCB018424 preferentially binds. Furthermore, these resistant variants are cross resistant to other JAK2 inhbitors (Lestaurtinib, CYT-387 and AZD1480). In contrast, these resistant variants are fully sensitive to TG101348 supporting the lack of resistance against this compound as observed during in vitro screening. Structural modeling studies revealed that TG101348 stabilizes the active conformation of the kinase and binds to the substrate-binding pocket. Mutations affecting the substrate-binding pocket may either alter substrate binding/phosphorylation or encode an incompetent kinase that blocks the emergence of resistance. Because JAK2 and BCR/ABL share a common substrate, STAT5, they might have similar architecture of the substrate-binding pocket, which may allow the inhibition of BCR/ABL by TG101348. Indeed, TG101348 can inhibit both native and gatekeeper variants of BCR/ABL and in vitro drug resistant screening failed to develop emergence of resistant clones. These studies provide evidence that the patients developing resistant variants of JAK2 and BCR/ABL can be treated with TG101348 and support for future drug design geared towards targeting the substrate binding sites in other oncogenic kinases for better and sustained therapeutic response.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.