Distinct Biological and Transcriptional Features Of FLT3-ITD Variants In Vitro and In Vivo

Recent publications suggest that mutated FLT3 kinase (FLT3-ITD) is a valid therapeutic target in acute myeloid leukemia (AML). Small molecule FLT3 tyrosine kinase inhibitors (TKIs) of the first and second generation such as midostaurin (PKC412) or quizartinib (AC220) are currently being investigated in advanced clinical trials. Previously, we described that ITD mutations are not always located in the juxtamembrane domain (JMD) of FLT3 but can also be found in the beta1-sheet of tyrosine kinase domain 1 (TKD1) in more than 25% of cases. Moreover, patients harboring ITD mutations in both locations revealed distinct response to chemotherapy: multivariable analyses with regard to achievement of CR, relapse-free survival as well as overall survival revealed consistently similar results, with FLT3-ITD insertion sites in the beta1-sheet of the TKD1 being an unfavorable variable that even outweighed the impact of FLT3-ITD mutant to wildtype allelic ratio.

Here, we investigated an extended spectrum of FLT3-ITD variants subcloned from primary patient material. A total of seven different ITD mutations were investigated: five FLT3 length mutations located in the JM-domain (insertion site (amino acid length): 598/599(12), 598/599(22), 599(7), 601(10), 602(7)) and two ITD-mutations located in the TKD1 domain (611(32) and 613(33)) were subcloned into retroviral (MSCV-based) vectors. Ba/F3 cells were infected with the respective constructs. FLT3-ITDs located in the TKD1 domain displayed significantly less sensitivity to TKIs in vitro: Exposure to clinically relevant doses of midostaurin (PKC412) and quizartinib (AC220) (1nM-100nM) resulted in significantly decreased induction of apoptosis for the TKD1-located ITD-variants (p<0.001).

In order to further elucidate on the biologic impact of FLT3-ITD mutation variants we investigated the respective constructs in vivo using a retroviral bone marrow (BM) transplantation model in Balb/C mice. Stem cell enriched BM cells after 5-FU injection were harvested and infected with FLT3-ITD constructs located either in the JMD (GFP labelled) or the TKD1 (YFP labelled). To assess for transforming capacity of the respective constructs 70,000 or 140,000 GFP positive cells were injected into lethally irradiated recipient mice.

Moreover, cells infected with two different ITD variants (one JMD-ITD-GFP and one TKD1-ITD-YFP) were transplanted competitively into the same recipient animal. Unexpectedly, both assays revealed a competitive advantage for ITD mutations located in the JM-domain. At the lower (limiting dilution) cell dose, TKD1-ITDs showed a reduced penetrance (50% for TKD1 versus 100% for JMD-ITDs) as well as a prolonged onset of leukemia development (20 versus 10 weeks, p<0.001). However, these results are not contradictory to the observed resistance phenotype in vitro, as drug resistance (against chemotherapy or TKI) does not necessarily need to correlate with a proliferative advantage. Drug intervention experiments using midostaurin are currently under way to explore a potential resistance phenotype in vivo.

To determine the downstream targets of the respective ITD variants, we performed gene expression-profiling on murine cell lines (32D cells) and human cell lines (Kasumi-1) transduced with the respective human FLT3-ITD-variants. FLT3-ITD mutations located in the TKD1 domain revealed enrichment of pathways controlling cell signalling (MAPK, adhesion-related molecules, G-protein coupled receptors and small GTPases) and regulation of apoptosis (Granzyme A, JNK). These results are currently being compared to gene expression analyses from primary patients with FLT3-ITD mutation variants treated in large multicenter trials to identify a clinically relevant set of gene expression changes involved in decreased sensitivity to chemotherapy and TKI. Overlapping target genes will be validated in murine and human cell lines using RNAi.

Taken together our data provide evidence that location of ITD mutations within the FLT3 gene may determine differential sensitivity to first- and second-generation inhibitors in patients.