- Top of page
- Materials and methods
- Supporting Information
Fms-like tyrosine kinase (FLT3) mutations are the most frequent mutations in patients with acute myeloid leukaemia (AML) that confer a poor prognosis. Constitutively active FLT3-ITD (internal tandem duplications) mutations define a promising target for therapeutic approaches using small molecule inhibitors. However, several point mutations of the FLT3 tyrosine kinase domain (FLT3-TKD) have been identified to mediate resistance towards FLT3 tyrosine kinase inhibitors (FLT3-TKI), including secondary mutations of FLT3. We investigated the cellular effects of the recently characterised FLT3-TKI ponatinib (AP24534) on murine myeloid cells transfected with FLT3-ITD with or without additional point mutations of the FLT3-TKD including the (so far) multi-resistant F691I mutation. Ponatinib effectively induced apoptosis not only in the parental FLT3-ITD cell line but also in all stably transfected subclones harbouring additional FLT3-TKD point mutations (N676D, F691I or G697R). These observations correlated with a strong inhibition of FLT3-ITD and its downstream targets STAT5, AKT and ERK1/2 upon ponatinib incubation, as determined by Western blotting. We conclude that ponatinib represents a promising FLT3-TKI that should be further investigated in clinical trials. The targeted therapy of FLT3-ITD-positive AML with ponatinib might be associated with a lower frequency of secondary resistance caused by acquired FLT3-TKD mutations.
During the last decade, an increasing number of distinct molecular aberrations have been identified in acute myeloid leukaemia (AML) (Fröhling et al, 2005; Marcucci et al, 2011). Fms-like tyrosine kinase (FLT3) mutations are the most frequent kind of mutation in AML that is associated with an unfavourable outcome (Nakao et al, 1996; Yokota et al, 1997; Kottaridis et al, 2001; Yamamoto et al, 2001; Schnittger et al, 2002; Thiede et al, 2002). FLT3 internal tandem duplications (FLT3-ITD) and point mutations within the tyrosine kinase domain (TKD) represent the two main subtypes of FLT3 mutations. While FLT3-ITD mutations can be detected in about 25% of all AML patients, FLT3-TKD mutations are found in up to 10% of AML (Kottaridis et al, 2001; Yamamoto et al, 2001). All these FLT3 mutations confer a ligand-independent autophosphorylation of FLT3 that leads to the constitutive activation of STAT5 and important downstream targets, such as ERK1/2 or AKT (Hayakawa et al, 2000; Mizuki et al, 2000). The anti-apoptotic signalling of FLT3 mutations also affects the expression of Bcl-2 family members or survivin at the protein level (Minami et al, 2003; Brandts et al, 2005; Zhou et al, 2009). The up-regulation of Mcl-1 in a presumably STAT5-dependent manner is thought to be a key event in FLT3-induced leukemogenesis (Yoshimoto et al, 2009).
Distinct subclasses of FLT3 tyrosine kinase inhibitors (FLT3-TKI) have been developed, some of which are currently under investigation in prospective clinical trials (Kindler et al, 2010; Fathi & Levis, 2011; Pemmaraju et al, 2011).
Recent phase I/II clinical trials investigating FLT3-TKI monotherapy revealed a primary resistance in approximately 30% of FLT3-mutated AML patients (Stone et al, 2005; Zhang et al, 2008; Weisberg et al, 2009). Primary resistance is known to be either FLT3-independent, through the activation of alternative survival pathways, e.g. STAT5 and MAPK downstream signalling, or FLT3-dependent by mutations of FLT3 that are insensitive to the applied FLT3-TKI. Interestingly, the sensitivity towards FLT3-TKI was shown to diversify between different activating FLT3 mutations (Grundler et al, 2003; Chu & Small, 2009).
Nevertheless, the majority of patients treated with single compound FLT3-TKI experienced only transient response and the development of acquired secondary resistance reflects a special challenge in the treatment with FLT3-TKI (Kancha et al, 2007; Fischer et al, 2010).
One major mechanism of secondary TKI resistance is acquired genetic alterations within the FLT3-TKD. In particular, point mutations near the ATP binding site can cause secondary resistance that impairs drug binding to FLT3. Two in vivo acquired point mutations of FLT3 have been described in clinical investigations with FLT3-TKI. Treatment with midostaurin led to a secondary FLT3 mutation at codon 676 in one AML patient (Heidel et al, 2006). Furthermore, an A848 mutation of FLT3 caused secondary resistance in a patient with chronic myelomonocytic leukaemia (CMML) receiving treatment with sunitinib and sorafenib (von Bubnoff et al, 2010). This is similar to the mechanism of resistance to imatinib in chronic myeloid leukaemia (CML) that is potently overcome by the second line BCR/ABL1 inhibitors nilotinib or dasatinib (Gorre et al, 2001; Weisberg et al, 2007).
Contemporaneously, FLT3-TKD mutations at different positions, especially A627, N676, F691, and G697 were identified to confer varying degrees of resistance to several FLT3-TKI. Recently, in vitro analysis demonstrated a large number of FLT3 point mutations within the TKD1 and TKD2 with a differing sensitivity profile towards common FLT3 inhibitors, e.g. midostaurin (PKC412), sunitinib or sorafenib. Importantly, the F691I mutation was resistant to all compounds investigated (Cools et al, 2004; von Bubnoff et al, 2009). Interestingly, the Bis(1H-indol-2-yl)methanone c102 can partially overcome resistance of FLT3-ITD harbouring an additional point mutation at position N676 in Ba/F3 cells (Mahboobi et al, 2006; Heidel et al, 2009).
A recent pre-clinical evaluation (O'Hare et al, 2007) revealed ponatinib (AP24534, Ariad Pharmaceuticals, Cambridge, MA, USA) as a novel multi-targeted TKI. High-level resistance to imatinib, nilotinib, and dasatinib in the treatment of CML due to the BCR/ABL1T315I gatekeeper mutation is caused by an impaired inhibitor binding (O'Hare et al, 2007). Importantly, recent pharmacological research identified ponatinib as an orally active inhibitor of BCR/ABL1 including BCR/ABL1T315I mutation (O'Hare et al, 2009). Furthermore, ponatinib has shown inhibitory activity against tyrosine kinases that are involved in the pathogenesis of other haematological malignancies, including FLT3. A recent evaluation of ponatinib demonstrated a dose-dependent inhibition of FLT3 signalling and the induction of apoptosis in FLT3-ITD-positive AML cells including primary leukaemic blasts (Gozgit et al, 2011). It was also demonstrated that ponatinib effectively inhibited other receptor tyrosine kinases, especially all isoforms of the fibroblast growth factor receptor (Gozgit et al, 2012).
In this study, we sought to investigate the efficacy of ponatinib to inhibit FLT3-ITD harbouring additional point mutations within the FLT3-TKD that have been demonstrated to confer secondary resistance towards different classes of FLT3-TKI. In particular, we hypothesized that the multi-targeted inhibitor ponatinib might overcome resistance of the (so far) multi-refractory point mutation F691I of FLT3.
- Top of page
- Materials and methods
- Supporting Information
We aimed to investigate the efficacy of the recently characterized FLT3-TKI, ponatinib, to induce apoptosis in FLT3-ITD-positive cells dependent on the presence of additional FLT3-TKD mutations that are known to cause resistance to other FLT3-TKI. We were especially interested to determine whether ponatinib could inhibit FLT3-ITD containing the multi-resistant F691I mutation.
The analyses of apoptosis induced by c102, sorafenib, or ponatinib demonstrated a strong resistance of the FLT3 G697R mutation to c102 and a strong resistance of the FLT3 F691I mutation towards c102 and sorafenib up to a concentration of 5 and 1 μmol/l, respectively. These data agree with previously published results (von Bubnoff et al, 2009; Heidel et al, 2009). Nevertheless, exposure of Ba/F3-ITD-691 cells to higher concentrations of sorafenib (up to 10 μmol/l) resulted in apoptosis of Ba/F3-ITD-691 cells with an IC50 value of approximately 3 μmol/l (data not shown).
Furthermore, in Ba/F3 cells expressing FLT3-ITD, apoptosis was induced at subnanomolar concentrations of ponatinib. For the induction of apoptosis in the parental interleukin 3-dependent Ba/F3 cell line, IC50 values of approximately 1800 nmol/l were determined demonstrating that no relevant off-target effect of ponatinib contributes to induction of apoptosis of FLT3-ITD-transfected Ba/F3 cells at low concentrations (O'Hare et al, 2009). Importantly, the FLT3-ITD protein remains highly sensitive to ponatinib in presence of a concomitant F691I point mutation. In addition, ponatinib also induced apoptosis when FLT3-ITD was associated either with the N676D or the G697R mutation. For both of these FLT3-TKD mutations, the IC50 values were, however, one log rank higher as compared with the F691I-mutated FLT3-ITD variant. In consideration of pharmacokinetic data demonstrating a steady state plasma level of more than 100 nmol/l ponatinib in humans, the F691I point mutation of FLT3-ITD should be very effectively inhibited in vivo, and the N676D and the G697R point mutation of FLT3 can also be assumed to be sensitive to ponatinib (Cortes et al, 2010).
For sorafenib, pharmacokinetic data demonstrated steady state plasma levels of about 5 μmol/l when the compound was orally administered at the maximum dosage level. It is therefore possible that sorafenib might also at least inhibit FLT3-ITD-positive leukaemia cells harbouring an additional F691I point mutation when applied at these high doses. Nonetheless, we hypothesize that sorafenib should not be equivalent to ponatinib concerning its efficacy to induce apoptosis of cells harbouring the FLT3-ITD-F691I because the ponatinib IC50 value of 3·8 nmol/l is far below the regular plasma level for this inhibitor, which is more than 10-fold of the IC50.
In addition, we performed cell cycle analyses of all cell lines investigated in this study dependent on the incubation with ponatinib. For Ba/F3-ITD-598 cells, an IC50 of between 3 and 10 nmol/l was demonstrated for the inhibition of proliferation based on the S+G2/M fraction. A similar efficiency of cell cycle inhibition could be detected in Ba/F3-ITD-691 cells. These data support the notion that ponatinib can effectively inhibit proliferation of both cell lines, as well as that of MV4-11 cells (Fig S3), at concentrations below 10 nmol/l and induce apoptosis in a similar concentration range as demonstrated by the Annexin V/PI assay.
Next, we investigated the inhibition of FLT3-ITD and its downstream signalling pathways (STAT5, AKT and ERK1/2) by the FLT3-TKIs sorafenib and ponatinib in Ba/F3-ITD-598 and Ba/F3-ITD-691 cells, respectively. The inhibition of FLT3 phosphorylation, as determined by phospho-specific immunoblot analyses of tyrosine 589, showed a good correlation with the dose-dependent inhibition of STAT5 activation for both FLT3-TKI and in both Ba/F3 cell lines. Thus, the IC50 value of phospho-STAT5 can be estimated at 3 nmol/l for ponatinib in both Ba/F3-ITD-598 and Ba/F3-ITD-691 cells. Furthermore, the IC50 value of sorafenib was significantly lower when STAT5 phosphorylation was analysed in Ba/F3-ITD-598 cells as compared with Ba/F3-ITD-691 cells (3 vs. 30 nmol/l). These data support the observation that the F691I mutation is resistant to the FLT3-TKI sorafenib. Importantly, a significant reduction of STAT5 phosphorylation was observed when Ba/F3-ITD-691 cells were incubated with sorafenib at a concentration of 30 nmol/l for 1 h. This seems to be contradictory to the results of the Annexin V/PI assay, wherein we confirmed the strong resistance of F691I-mutated FLT3-ITD to sorafenib in Ba/F3-ITD-691 cells (IC50 > 1000 nmol/l) (von Bubnoff et al, 2009). One possible explanation is based on pharmacokinetic aspects, i.e., that a short half-life of the inhibitor may cause a reduction of sorafenib concentration during the 48 h incubation.
We also investigated the impact of sorafenib and ponatinib on downstream pathways, e.g. the phosphorylation of AKT and ERK1/2, in Ba/F3-ITD-598 and Ba/F3-ITD-691 cells, respectively. Incubation of both murine cell lines with different concentrations of ponatinib resulted in a strong reduction of phospho-AKT in Ba/F3-ITD-598 cells and a significant decrease of phospho-ERK1/2 in Ba/F3-ITD-691 cells. In contrast, the basal AKT phosphorylation in Ba/F3-ITD-691 cells and the basal phosphorylation of ERK1/2 in Ba/F3-ITD-598 cells were both too low and therefore not appropriate to determine a concentration-dependent effect of the FLT3-TKI. Surprisingly, in both Ba/F3 cell lines, incubation with sorafenib was associated with an increase of AKT phosphorylation at one distinct concentration (30 nmol/l in Ba/F3-ITD-598 cells versus 10 nmol/l in Ba/F3-ITD-691 cells). Furthermore, the phosphorylation of ERK1/2 also increased when Ba/F3-ITD-691 cells were treated with 30 nmol/l sorafenib. Mechanisms of such a paradoxical downstream activation are a matter of speculation. Similar observations have been described for other multi-targeted inhibitors like the BCR/ABL1 inhibitor imatinib (Chu et al, 2004; König et al, 2008).
We further determined the expression of the important anti-apoptotic protein Mcl-1 by intracellular staining and subsequent analysis by flow cytometry. We demonstrated a strong reduction of Mcl-1 expression in Ba/F3-ITD-598 cells with increasing concentrations of ponatinib. Surprisingly, not only the inhibition but also the expression of Mcl-1 itself was much lower in Ba/F3-ITD-691 cells. Importantly, ponatinib at 10 nmol/l, which effectively induced apoptosis in F691I-mutated Ba/F3-ITD cells, demonstrated only a slight reduction of Mcl-1 expression. Because immunoblotting of phospho-STAT5 does not necessarily represent its activation state to induce downstream pathways, a potential explanation for the low Mcl-1 expression observed in Ba/F3-ITD-691 cells might be reflected by a discrepancy between STAT5 phosphorylation and its activation. In detail, F691I-specific induction of signalling pathways that can directly or indirectly inhibit STAT5-dependent transcriptional activity might modulate FLT3-ITD-induced Mcl-1 expression. Further experiments will be necessary to address these open questions.
Taken together, the development of resistance towards FLT3 inhibitors remains a serious challenge in the treatment of AML. The knowledge about the underlying molecular mechanisms in the pathogenesis of resistance is necessary to prevent them and further improve the therapeutic options for FLT3-ITD-positive AML patients. Ponatinib, however, is a new promising compound for overcoming secondary resistance towards available FLT3-TKI due to the evolution of leukaemic subclones acquiring an additional point mutation in the TKD of FLT3. Furthermore, ponatinib might also be associated with a lower mutation rate when applied as first line therapy with FLT3-TKI in AML. Therefore, clinical trials are warranted to elucidate the efficacy and in vivo resistance profile of ponatinib in FLT3-ITD-positive AML.