This month, we highlight the tyrosine kinase KIT and its central role in gastrointestinal stromal tumors (GISTs) and acute myeloid leukemia (AML). Like many of the earliest tyrosine kinases to be characterized, KIT was originally discovered as its viral homolog, the oncogene underlying feline sarcoma, leading to its name (kitten). In healthy blood, KIT regulates the lifecycle of hematopoietic stem cells, from self-renewal through differentiation. Binding of KIT’s ligand, stem cell factor (SCF), to its extracellular domain stabilizes the homodimeric form of the kinase, bringing into proximity the two intracellular kinase domains, leading to their autophosphorylation. The resulting dimeric phospho-KIT undergoes a conformational change that enables it to signal directly through the PI3K pathway and indirectly through SRC and MAP kinase pathways.
Aberrant KIT signaling has been implicated in numerous pathologies, with activating mutations detected in most GIST and AML cases. Although KIT has yet to be discovered as a partner in an activating chromosomal translocation, it can be activated by overexpression, point mutation or internal deletion. KIT is particularly vulnerable to mutations in its conformational switching system: Two separate intracellular domains that serve as activating and inactivating switches compete for binding to an intramolecular switch pocket (SP) binding site adjacent to the catalytic domain. One of the most aggressive KIT mutations studied, D816V in the activation loop, common in systemic mastocytosis, as well as N822H dramatically increase the affinity of the activation loop for the SP while simultaneously perturbing the structure of the ATP pocket, leading to a version of KIT that is both constitutively active and resistant to most ATP-competitive inhibitors that work on wild-type protein. In contrast, mutations in the KIT inactivation switch, including deletions and point mutations like K642E, promote activity without altering the ATP pocket. The autoinhibitory and juxtamembrane regions of KIT overlap, so some inactivation switch mutations may also alter signaling from the extracellular to the intracellular domains.
While primary mutations in the conformational switching system are most likely to initiate a KIT pathology, secondary mutations in the ATP pocket frequently arise as a resistance mechanism in patients treated with KIT inhibitors. Like all RTKs, KIT has a key gatekeeper residue lining the entrance to its ATP pocket. Mutations in the gatekeeper alone are sufficient to abolish inhibitor binding by sterically or chemically occluding entry into the pocket, while still preserving ATP binding. The T670I gatekeeper mutation is most frequently seen in acquired resistance patients, with a companion V654A mutation in a proximal residue also common, creating an even more dramatic change to the shape of the entry region of the ATP pocket.
The earliest therapeutic intervention into KIT pathologies relied on the cross-reactivity of the ABL inhibitor imatinib. Indeed, clinical trials for several indications yielded patients with several years of disease-free survival. The patient populations least like to respond to imatinib posses a mutation in the activation loop such as D816V, mimicking the lack of inhibition at the biochemical level and suggesting the use of second-generation inhibitors like nilotinib and dasatinib. Thus, more recent therapeutic modalities call for the sequencing of KIT from a diseased biopsy before choosing a kinase inhibitor for treatment. Unfortunately, much like imatinib-treated CML patient populations, secondary mutations frequently arise, most often in the gatekeeper residue, that leads to disease recurrence. There is currently a highly-focused effort within the pharmaceutical industry to develop inhibitors that overcome these secondary mutations. A key tool in these efforts is ACD’s panel of matched KIT-dependent cell lines corresponding to wild type protein, activation loop mutations, juxtamamebrane region mutations and gatekeeper mutations. The excellent performance of these lines replicates known literature activities and mimics known patient responses and resistance. The figure to the right shows representative dose-response curves with various inhibitors against the KIT wt and mutant panel.