Abstract
Aberrantly activated tyrosine kinases and their associated signaling pathways are critical to leukemogenesis and primary acute myeloid leukemia (AML) cell viability. While aberrant kinase activation has been confirmed in a significant percentage of AML, constitutive phosphorylation of STAT5, a marker of tyrosine kinase activation, is present in the majority of AML samples indicating that as yet unidentified tyrosine kinases can be aberrantly activated and contribute to leukemogenesis. Efforts to identify activating tyrosine kinase mutations using high-throughput sequencing have identified low frequency mutations of uncertain functional significance. Because these studies failed to detect additional high-frequency kinase mutations, the identity and mechanism of tyrosine kinase activation may be unique in many AMLs.
Methods: To rapidly identify activated kinase pathways in individual, primary AML samples, we have developed a small-molecule inhibitor array which includes 90 small-molecule, cell-permeable inhibitor compounds including a core of 36 tyrosine kinase inhibitors that covers the majority of the tyrosine kinome. Many of the inhibitors are available for clinical use or are in clinical development. In this assay, inhibitors were placed in 96-well plates at four serial dilutions to allow IC50 calculations. Three days after adding primary AML cells to each well, we performed an MTS cell viability assay to evaluate the effect of each inhibitor on cell viability. Because most inhibitors affect multiple kinases, we compared target specificities of compounds that decrease primary AML cell viability with those that have no effect to identify potential targets.
Results: In preliminary proof-of-principal experiments, we tested leukemic cells from five AML patients, including three that were positive for FLT3 internal tandem duplication (ITD), a genetic lesion that is thought to confer a proliferative advantage in approximately 30% of AML patients. Only one sample showed a clear response to small-molecule inhibitors known to target FLT3. The IC50s for the known FLT3 inhibitors MLN518, AST487, CHIR258, Sunitinib, and SU14813 were 10 to 100 fold lower in the leukemic cells from this patient than for the mean and median values for bone marrow samples from normal marrow samples. Interestingly, neither of the remaining FLT3-ITD positive leukemias nor the FLT3 negative leukemias demonstrated decreased viability in the presence of small-molecule inhibitors that target FLT3. A PCR based screen to identify FLT3-ITD alleles showed a near absence of the wild-type FLT3 allele in this sample, while the wild-type allele was present at equal or greater intensity as the ITD allele in the remaining FLT3-ITD positive AML samples. FLT3-ITD positive leukemias with loss of the FLT3 wild-type allele have been shown to have a poorer prognosis than those retaining a wild-type allele indicating aberrantly activated FLT3 may play a crucial role in leukemic cell viability in this setting and are consistent with a previous report showing that efficacy of the FLT3 inhibitor CEP-701 was greatest in pediatric AML with high FLT3-ITD mutant-to-wild-type allelic ratios. Though the remaining AML samples did not show a pattern of inhibitor sensitivity consistent with FLT3 activation, these and many of approximately 15 additional analyzed AML samples showed unique sensitivity patterns implicating other specific kinase targets or kinase families for further investigation while simultaneously providing therapeutic options.
Conclusions: These preliminary data demonstrate that the small-molecule inhibitor functional assays can rapidly identify disease causing genes, provide insights into their mechanism of action, and suggest therapeutic options. The distinct patterns of tyrosine kinase sensitivity in these samples support the hypothesis that tyrosine kinases and related pathways contributing to leukemogenesis in each patient may be different and that targeted therapy will be most effective when administered on an individualized basis.
The MDM2 small-molecule inhibitor RG7388 leads to potent tumor inhibition in p53 wild-type neuroblastoma
