KIT Mutations Define Characteristic Gene Expression Signatures in Core Binding Factor Leukemias.

Core binding factor (CBF) leukemias, characterized by either inv(16)/t(16;16) or t(8;21), constitute acute myeloid leukemia (AML) subgroups with favorable prognosis. However, there exists substantial biological and clinical heterogeneity within these cytogenetic groups, which is not fully reflected by the current classification system. Recently, the identification of novel molecular markers, like the gain-of-function KIT mutations in exons 8 and 17 have been shown to confer higher relapse risk and adversely influence overall survival (Paschka et al, J Clin Onc 2006). To further investigate whether the presence of KIT mutations leads to a characteristic gene expression pattern in CBF AML we profiled gene expression in a large series (n=83) of AML patients with CBF leukemia (KIT WT n=59, KIT mutated n=24). By supervised hierarchical clustering we were able to define a KIT mutation associated gene signature consisting of 168 genes (p=0.005). Gene set enrichment analyses revealed several groups of pathways to be differentially regulated, thereby likely reflecting differences in pathogenic mechanisms which might lead to distinct prognoses of different subgroups of CBF leukemias. For example, in the KIT mutated group, cellular stress response pathways, as for example NFkappaB-signaling and TNF/stress related signaling, were downregulated, whereas the motility-associated Y branching of actin filaments was upregulated. In general, a great overlap of deregulated pathways was found between the overall set and the cytogenetic CBF subgroups inv(16) and t(8;21), respectively. The same was true for KIT mutations in exons 8 and 17 in comparison to all KIT mutations. However, as all these subgroups are characterized by distinct biology and clinics, each subgroup was also associated with a number of deregulated pathways that were not significantly affected in other subgroups. Apart from the expected upregulation of KIT in the KIT mutated group, we also found e.g. a down-regulation of TNF, NFkappaB and PRKCA. Overall, KIT mutations seem to entail a proliferative advantage (higher blast counts in KIT mutated cases; p=0.0015 and p=0.0157 for peripheral blood and bone marrow blast counts, respectively), an increased stem cell self renewal as suggested by the upregulation of LRP6 of the Wnt/LRP6 Signaling pathway and a deregulation of apoptosis. This latter finding is reflected by the NFkappaB-signaling pathway and TNF/stress related signaling, mentioned above, and also, e.g. by the upregulation of the Apoptotic Signaling in Response to DNA Damage pathway in KIT mutated inv(16) cases. Thus, gene expression profiling reveals novel insights into the potential pathomechanism of KIT mutations in CBF leukemias. However, while the leukemogenic relevance of these signatures remains to be validated, their existence nevertheless supports a prognostically relevant biological basis for the heterogeneity observed in CBF leukemia.