A series of novel aryl-methanone derivatives as inhibitors of FMS-like tyrosine kinase 3 (FLT3) in FLT3-ITD-positive acute myeloid leukemia

 

Andreas Sellmer, Bernadette Pilsl, Mandy Beyer, Herwig Pongratz, Lukas Wirth, Sigurd Elz, Stefan Dove, Sven Julian Henninger, Karsten Spiekermann, Harald  Polzer,  Susan  Klaeger,  Bernhard  Kuster,  Frank  D.  Bo€hmer, Heinz-Herbert  Fiebig,  Oliver  H.  Kra€mer,  Siavosh  Mahboobi

a Institute of Pharmacy, Faculty of Chemistry and Pharmacy, University of Regensburg, 93040, Regensburg, Germany

b Department of Toxicology, University Medical Center, Mainz, Germany

c Department of Medicine III, University Hospital, LMU Munich, Germany

d Technische  Universita€t  München,  Wissenschaftszentrum  Weihenstephan  für  Erna€hrung,  Landnutzung  und  Umwelt,  Germany

e Universita€tsklinikum  Jena  -  Bachstrasse  18  -  D-07743  Jena,  Germany

f 4HF Biotec GmbH, Am Flughafen 14, 79108, Freiburg, Germany

 

 

A B S T R A C T

Mutants of the FLT3 receptor tyrosine kinase (RTK) with duplications in the juxtamembrane domain (FLT3-ITD) act as drivers of acute myeloid leukemia (AML). Potent tyrosine kinase inhibitors (TKi) of FLT3-ITD entered clinical trials and showed a promising, but transient success due to the occurrence of secondary drug-resistant AML clones. A further caveat of drugs targeting FLT3-ITD is the co-targeting of other RTKs which are required for normal hematopoiesis. This is observed quite frequently. Therefore, novel drugs are necessary to treat AML effectively and safely. Recently bis(1H-indol-2-yl)methanones were found to inhibit FLT3 and PDGFR kinases. In order to optimize these agents we synthesized novel derivatives of these methanones with various substituents. Methanone 16 and its carbamate derivative 17b inhibit FLT3-ITD at least as potently as the TKi AC220 (quizartinib). Models indicate corresponding interactions of 16 and quizartinib with FLT3. The activity of 16 is accompanied by a high selectivity for FLT3-ITD.

 

 

1.            Introduction

Acute myeloid leukemia (AML) is a heterogeneous disorder of hematopoietic progenitor cells, which lose their ability of natural differentiation and to respond to normal regulators of proliferation. Subsequently without treatment this leads to fatal infections, bleeding, or typical organ infiltration within one year after diag- nosis [1]. Usual treatments are intensive chemotherapy and allo- geneic hematopoietic stem cell transplantation [2]. However, 60e80% of relapses are eminently frequent [3] especially for elder patients. Thus, new effective, and precise therapies are urgently required.

Class III receptor tyrosine kinases (RTKs) are required for normal hematopoiesis, and frequently they are dysregulated in AML [4]. Therefore, RTKs are targets for tailored intervention strategies against AML. FMS-like tyrosine kinase 3 (FLT3), macrophage colony stimulating factor receptor (CSF1R, c-FMS), stem cell factor receptor (SCFR, KIT), and platelet-derived growth factor receptor (PDGFRa/ b) belong to this protein family [5]. Each of these receptors consists of five immunoglublinelike domains in the extracellular domain, a transmembrane domain (TMD), a juxtamembrane domain (JM) and an intracellular kinase domain which is split by a kinase insert [6e8].

FLT3 is expressed on hematopoietic stem- and progenitor cells and plays an indispensable role in stem cell development and cell differentiation [9]. FLT3 is activated by the binding of its ligand. After ligand-induced dimerization and autophosphorylation of the receptor, the transduction pathway of the signal is initiated [10]. Mutations in FLT3 are found in about one third of all patients with AML, and aberrations are associated with poor prognosis [9,11,12].

These lesions cause an aberrant activation of FLT3 and its down- stream signaling pathway [13,14]. Of note is that recent evidence shows that mutant FLT3 is an oncogenic driver in AML [15].

Three different activating FLT3 mutations are known. 1) Acti- vating FLT3 internal tandem duplications (FLT3-ITD) are located within the JM domain of FLT3. They were first observed in 1996 by Nakao et al. [11] and are seen in 15e35% of the patients suffering from AML [11,13,16e18]. 2) Point mutations in the activation loop of the kinase domain (FLT3-TKD) occur in 6e8% of patients with AML [12,13,19,20]. 3) Point mutations in the JM are detectable in 2% of AML patients [13,20]. The most common of these mutations, FLT3- ITD, induces the loss of auto-inhibitory functions. Thus, the kinase becomes active constitutively. This leads to altered intracellular signaling [21] and autonomous cell growth [13].

Due to the frequency of FLT3 mutations in AML, the inhibition of FLT3 and its downstream signaling pathways have triggered sig- nificant attention to the discovery of new anti-cancer drugs [22]. Under current development most of the potential therapeutics are direct inhibitors of the FLT3 receptor. These small molecule in- hibitors are investigated in monotherapy or in combination with chemotherapy [22e25]. Inhibitors of FLT3 are classified to the first and the second generation (Fig. 1). The first generation of small molecule inhibitors include midostaurin (1, PKC412)24, approved by the US FDA in April 2017 for treatment of newly diagnosed FLT3- mutant AML [26], lestaurtinib (2, CEP-701) [24,27,28], semaxanib (3, SU 5416) [24,29], and sunitinib (4,SU 11248) [24]. To overcome therapy acquired resistance also covalent irreversible binding in- hibitors were published recently [30,31].

Generally the first generation of small molecule inhibitors are multi-kinase inhibitors, and their efficacies are limited, not durable, and at the expense of undesired side effects [2,25].

Therefore, a second generation of TKi against FLT3-ITD has been developed in order to yield more selective and high potent agents for treating AML [32,33]. Prominent representatives of second generation are tandutinib (5, MLN518) [25,34], KW-2449 (6) [25,35] and quizartinib (7) [24,25,36] (Fig. 1). Nevertheless, sec- ondary FLT3 mutations may substantially confer to in vitro resis- tance of inhibitors, as shown for quizartinib (7) [15].

Further examples are crenolanib (8) [37e39], G-749 (9) [40], TTT-3002 (10) [41], as well as dual kinase inhibitors (e.g., 11, CCT137690 against FLT3 and Aurora kinase) [42]. Crenolanib (8) was originally developed as PDGFR inhibitor [43]. Crenolanib (8) shows activity against FLT3-ITD in vitro and in vivo. In combination with sorafenib (12) crenolanib (8) promisingly decreased leukemic burden and extended the survival of leukemic mice [37]. Ponatinib (13) is an orally active multi-tyrosine kinase inhibitor specifically targeting the BCR-ABL gene mutation, T315I in chronic myeloid leukemia [44]. Moreover, ponatinib (13) exhibits activity against AC220-resistant FLT3-ITD/F691 gatekeeper mutations, but is highly ineffective against FLT3-ITD activation loop mutations, particularly at the D835 residue [45]. Following the midostaurin approval, in November 2018, a second-generation FLT3 inhibitor, gilteritinib, was approved by the US FDA for use as a single agent for adults with relapsed or refractory FLT3-mutated AML [46]. Beside others, further FLT3 inhibitors being active towards therapy associated mutants were described recently. Baska et al. for example found (phenylethenyl)quinazolines as selective inhibitors of the ITD and D835Y mutant FLT3 kinases [47]. Yuan et al. prepared and opti- mized pyrrolo[2,3-d]pyrimidine derivatives for cytotoxic activities against FLT3-ITD mutant cancer cells, exhibiting nanomolar FLT3 inhibitory activities and subnanomolar inhibitory activities against MV4-11 and Molm-13 cells. In part they also showed excellent inhibitory activities in FLT3-ITD-D835V and FLT3-ITD-F691L cells which were resistant to quizartinib (7) [48].

We have  previously  reported  on  the  inhibition  of  class  III receptor tyrosine kinase PDGFR and FLT3 by bis(1H-indol-2-yl) methanones 14a-c (Fig. 2) [49,50]. These compounds are biologi- cally active at FLT3 and PDGFR, 14c particularly at FLT3. At least in part the excellent inhibition of FLT3-ITD by AC220 (quizartinib) (7) is based on the bisarylurea motive substituted by a tert-butylisox- azole [51]. Therefore, we generated derivatives of the bis-(1H- indol-2-yl) methanone skeleton by fusion with this substructure (Fig. 2).

Here we report on some of our novel compounds which are potent and selective FLT3 inhibitors. Proliferation assays revealed AML cell growth inhibition at IC50-values in the low nanomolar range. Furthermore, one of our new compounds, 16, inhibits FLT3- ITD as well as FLT3 overexpressing cell lines. Moreover, 16 and especially its carbamate 17b with an enhanced water solubility of 230 mg/L vs < 100 mg/L for 16, shows excellent selectivity for FLT3 and FLT3-ITD. Moreover, compounds 16 and 17b show excellent activity against FLT3-ITD and the clinically relevant FLT3 D835Y mutant in the enzymatic test system, as well as reduced loss off affinity comparing 16 to quizartinib (7) towards FLT3-ITD-F691 L52, an important acquired drug-resistance gatekeeper mutation.

 

 

2.            Results and discussion

2.1.        Chemistry

Based on the promising results from our previous studies, we designed and synthesized a new series of compounds exploring a combination of our bisindolylmethanone lead structure [14,49] with key elements of established tyrosine kinase inhibitors, and evaluated their pharmacological properties. The synthetic route to obtain the desired target compounds is outlined in Schemes 1e5 and in the following: 18a-d (Scheme 1) were  lithiated  using  t- BuLi and the aryl-lithium intermediates reacted with the respective carbonyl chloride 19a, which was prepared in analogy to Mahboobi et al. [49] By removal of the sulfonyl protecting group [49] 20a- d (Scheme 1) were obtained. Acidic cleavage of the Boc-group with TFA yielded 21a-d. On principle by the same reaction sequence 21e- f were obtained by coupling of 18b, 18e and 18f with 19b and 19c, respectively, followed by removals of the sulfonyl protecting group [49] and the tert-butoxycarbonyl-group. From 21e the hydroxyanalogue 23e was obtained by hydrogenolytic cleavage of the benzyloxy-group [49] (way A). Moreover the sequence of steps c) and d) shown in Scheme 2 can be exchanged. By this alternative route, way B. e.g., 23i was obtained (see Scheme 6).

Amino compounds 21a-d and 23e-i were treated with 5-(tert- butyl)-3-isocyanatoisoxazole (24a) to obtain 15a-i (Scheme 3) [36]. As shown by the synthesis of 15f (Scheme 4), the sequence of the steps can be modified. The benzyl-group can also be removed finally from the respective indole.

To obtain the desired aryl-methanone derivative 16, 5- benzyloxyindole 25 was protected by benzenesulfonyl chloride as described [14] in order to get 26. Acetylation led to 27, which was brominated in alpha-position to the carbonyl-group using trime- thylphenylammonium tribromide to give 28 (Scheme 5). Reaction of 28 with 2-hydroxy-5-nitrobenzaldehyde (29) resulted in the formation of (5-(benzyloxy)-1-(phenylsulfonyl)-1H-indol-2-yl)(5- nitrobenzofuran-2-yl)methanone as an intermediate. By alkaline removal of the phenylsulfonyl-protecting group 30 was obtained. By use of ammonium formate and Pd/C the nitro- and Bn-groups were reduced. Finally, reaction of 31 with 5-(tert-butyl)-3- isocyanatoisoxazole (24a) in THF led to the target compound 16.

The desired ureas 32b and 32c were formed by coupling the isocyanate  intermediates  24b  and  24c  with  the  5-amino-50-hy- droxy aryl methanone 23e (Scheme 6). The isocyanates 24b and 24c were prepared from commercially available arylamines by reaction with trichloromethyl chloroformate in THF [54].

It turned out that the bisindolylmethanone skeleton with the tert-butyl-isoxazol-urea platform in position 5, adopted from compound 7, was essential for potent biological activity. In position 50  of  the  indole  ring  the  novel  lead  structure  16  bears  a  hydroxyl group (Fig. 2).

In order to modify the phenolic OH group of 16 to the carbamate 17a, 16  was  reacted  with  1,40-bipiperidine-10-carbonyl  chloride  in CH2Cl2/pyridine. The respective hydrochloride 17b with improved water solubility was obtained by treatment of 17a with HCl in 2- propanol (Scheme 7).

 

2.2.        Biological activity of the novel TKi against human AML cells

The pharmacological properties of the novel compounds against FLT3-ITD-positive human MV4-11 cells were evaluated. Based on the data obtained from the proliferation assay in the human MV4- 11 AML cell line, the IC50 values were determined (Table 1).

In a first run we performed a fusion of the bis(1H-indol-2-yl) methanone skeleton of 14b with the bisarylurea-motive taken from AC220 (quizartinib, 7) and investigated the antiproliferative activity of the resulting compounds, fused at different positions of the A- ring system (Fig. 2). It turned out, that an exchange of the phenolic hydroxy group in position 5 by the bisarylurea-motive in position 5 or 6 led to the potent compounds 15b and 15c (Table 1). This finding was applied to prepare compounds based on the substitution pattern of 14c (Fig. 2) by modifying one hydroxy-group by the bisarylurea-motive. The most potent compounds resulting have a bisaroylmethanone-skeleton substituted with the tert-butyl-iso- xazol-urea platform in position 5 of one of the selected aroyl- systems, either indole or benzofurane.

The results obtained in the proliferation assay indicate that the fusion of the basic 5-hydroxy-substituted bisaroylmethanone- structure with the 1-(5-(tert-butyl)isoxazol-3-yl)urea residue is highly active against FLT3-ITD-positive AML cells.

In order to check, whether the antiproliferative activity observed is due to an inhibition of FLT3, the inhibitory activity of all compounds exhibiting IC50 values below 50 nM in the proliferation assay on the enzymatic activity of FLT3-ITD was investigated in a kinase assay using human FLT3-ITD. The therapy-associated FLT3 D835Y mutant was also investigated.

As shown by the data of Table 2, all compounds inhibited FLT3- ITD in the low micromolar until nanomolar range. For 15i, 32b and 16 in particular, the data on antiproliferative activity and FLT3-ITD inhibition correlate very well.

For the FLT3 D835Y mutant, larger differences among the com- pounds were observed, with several compounds being less active than against FLT3-ITD.

Within this group of agents and taking both sets into account, 16 revealed to be the most potent derivative. When given once at only 1 nM it inhibited the proliferation of MV4-11 cells by nearly 80%. In the FLT3-ITD assay 50% inhibition was achieved at 2.3 nM (Table 2).

Moreover 16 and its carbamate 17b are also active towards the D835Y point mutant, whereas for AC220 (quizartinib, 7) a signifi- cant decrease in activity can be observed. Therefore, compound 16 was analyzed in further detail. The carbamate (17b), prepared in order to enhance the solubility, was included into these investigations.

To investigate the binding affinity towards FLT3-ITD-F691L, an important acquired drug-resistance gatekeeper mutation [52], Kd values of compound 16 in comparison to AC220, were determined (KINOMEscan Profiling Service, Eurofins DiscoverX Corporation). In result, 16 (Kd ¼ 115 ± 15 nM) remains an enhanced affinity towards FLT3-ITD-F691L in comparison to AC220 (Kd ¼ 340 ± 60 nM).

 

2.3.        Western-blotting

To prove whether FLT3 kinase inhibition could be recapitulated in vitro a western blotting assay was performed. We could see that 16 was able to dephosphorylate FLT3 strongly, after 3 h of incuba- tion, using doses of 20 nM and 40 nM (Fig. 3).

 

2.4.        Annexin V/PI staining

To analyze the potency of apoptosis induction of compound 16, MV4-11 cells were analyzed with AnnexinV/PI staining using flow cytometry. After 48 h using 20 nM and 40 nM of compound 16, respectively, we could detect 25% of apoptotic MV4-11 cells. With both doses the fraction of cells in late stages of apoptosis and ne- crosis increased up to 40% (Fig. 4).

 

2.5.        Binding modes

Crystal structures of the FLT3 kinase domain in complex with quizartinib 7 (PDB 4XUF [51], PDB 4RT7 [55]) may serve as tem- plates to investigate the binding mode of 16 and its derivatives. Based on previous models of FLT3-bisindolylmethanone in- teractions [49] and on the fit of the (tert-butyl)isoxazole moiety of AC220 (quizartinib, 7) into a hydrophobic specificity pocket present in both crystal structures, 16 was docked into the ATP binding site of FLT3 (PDB 4XUF). Fig. 5 shows the final model. As illustrated in Fig. 5A, the binding modes of AC220 (quizartinib, 7) and 16 closely overlap. Both inhibitors align with the hinge region (F691 e Y696) in the ATP binding pocket. The synperiplanar (sp, cis) conformation of the indole moiety of 16 with respect to the methanone oxygen enables a bidentate hydrogen bond with the backbone oxygen and nitrogen of C694 (Fig. 5B). The sulfur atom of AC220 (quizartinib, 7) may only form a single weak H-bond with this backbone nitrogen [51], but in the second crystal structure (PDB 4RT) the imidazole moiety is involved in water mediated, bidentate H-bonds with backbone atoms of E692 and C694,55 explaining the similar potency of both inhibitors (Table 1). In case of 15h, the nearly equipotent structural isomer of 16 with exchanged benzofuran and indole moieties, both benzofuran and methanone oxygens may form hydrogen bonds with the backbone nitrogen of C694.

For the benzofuran moiety of 16, an antiperiplanar (ap, trans) conformation was assumed which, on the one hand, avoids repul- sion of the oxygen atoms. On the other hand, this conformation enables interactions of the lipophilic benzofuran edge with a hy- drophobic patch consisting of V675, F691 (gatekeeper), L818 and F830 (DFG motif). The phenyl ring of the benzofuran moiety over- laps with the central phenyl linker in quizartinib and contacts like that F691 from one and F830 from the other side. Replacing the benzofuran in 16 or the indole in 15e by a pyrrolopyrimidine moiety, 15f eliminates interactions with F830 and reduces potency by at least two orders of magnitude. However, this large effect may be additionally due to conformational changes in the activation loop behind the DFG motif, induced by the looser fit of 15f between F691 and F830.

The ureido and (tert-butyl)isoxazole moieties of 16 and AC220 (quizartinib, 7) also overlap. The carbonyl and an ureido nitrogen form hydrogen bonds with the backbone NH of D829 (DFG motif) and a carboxyl oxygen of E661, respectively. The side chain conformation of E661 is stabilized by a salt bridge with K644 pre- sent in other tyrosine kinases, too. M665 is directed towards the isoxazole plane. The tert-butyl substituent fits well into a hydro- phobic specificity pocket consisting of M664, I674, L802 and I827 [51]. Obviously, these hydrophobic interactions of the tert-butyl group contribute significantly to the high inhibitory potency of 16 and AC220 (quizartinib, 7) since 15e is about two orders of magnitude more potent than its methyl analogue 32c (Table 1).

The 5-hydroxyindole moiety of 16 is stacked between amino acids of the hinge region, namely Y693 and the backbone of Y696 and G697. The 5-OH group points out of the ATP site. Larger polar 5substituents    like    the    1,40-bipiperidine-10-carbamate    group    in compound 17a adopt an outward position within the aqueous environment like the morpholinoethoxy group of AC220 (qui- zartinib, 7) and are therefore without significant effect on potency (Table 2) 16 and AC220 (quizartinib, 7) are type II TKis that bind to and stabilize the inactive DFG-out conformation of FLT3. The data in Table 2 show that both compounds are potent inhibitors of acti- vating FLT3-ITD mutations. ITDs release the autoinhibitory in- teractions of the juxtamembrane segment folded onto the kinase domain in the inactive wild type [56]. The crystal structure of FLT3 in complex with AC220 (quizartinib, 7) indicates that binding of this ligand is not compatible with the autoinhibitory juxtamem- brane segment conformation [51]. Therefore, 16 and AC220 (quizartinib, 7) bind to both FLT3 and FLT3-ITD with similar affinity. In case of the wild type, the inhibitors probably displace the jux- tamembrane region from the inner binding pocket.

Quizartinib (7) is not very potent at the D835Y mutant (Table 2). D835 belongs to the activation loop and does not participate in inhibitor binding, but is adjacent to a hydrophobic patch in a model of active FLT3 [51]. Hydrophobic D835 mutations may interact with this patch, stabilizing the activation loop in an extended confor- mation unfavorable for binding of AC220 (quizartinib, 7) and other type II TKis. Thus, the high potency of compounds 16, 17b and 32 b at the D835Y mutant (Table 2) is surprising because of their structural similarity with AC220 (quizartinib, 7) and may even be due to the stabilization of a specific, active-like conformation.

 

2.6.        Determination of the kinase inhibitor profile by proteomics assay

In target based drug discovery of small molecules, protein ki- nases are among the most frequently targeted proteins. Almost a quarter of all kinases are involved in oncogenesis. Therefore, the development of novel antitumor agents that selectively occupy the highly conserved ATP binding pocket is a goal in cancer research. To check the selectivity of our novel compounds, we used a chemical proteomics assay for kinase inhibitor profiling (kinobeads), a helpful tool to elucidate the target spectrum of a compound and to anticipate possible side effects [57]. Briefly, immobilized, broad- spectrum small molecule kinase inhibitors enable purification of endogenous kinases from tissues or cells. When performed as a competition binding assay and measured by quantitative MS, kinobeads allow the label free measurement of the physical inter- action of a compound of interest with thousands of proteins in parallel.

The selectivity of 16 was profiled in a cell mix consisting of a 1: 1: 1: 1 mixture of K562 (CML), Colo205 (human colon adenocar- cinoma), MV4-11 (AML) and SKNBE2 (human neuroblastoma) ly- sates. By use of a cell mix, the repertoire of kinases that can be enriched is much larger than that of a single cell line.

From 249 identified kinases in addition to FLT3 only RET and ZAK/MLTK are significantly affected by 16 (Table 3). RET is responsible for the differentiation of neurons in the autonomic nervous system mainly in embryonic development and has no function in the adult organism [58]. ZAK, on the other hand, be- longs to the family of MAP kinase (mitogen-activated protein ki- nase pathway) and is part of the TGFb/JNK pathway [59], playing a role in the regulation of the cell cycle [60].

Most interestingly, no significant co-targeting of other RTKs required for normal hematopoiesis was observed. The further 237 kinases studied in this experiment showed no reduced binding.

 

2.7.        Selectivity of 16 and 17b in cell viability assays

Additional experiments to assess the specificity of 16 were performed. Other human blood cell lines that overexpress FLT3 or carry FLT3-ITD as well as further hematological tumor cell lines were treated with 16 and analyzed by using the CellTiter-Blue® Cell Viability Assay. The cell line panel comprises 12 cell lines derived from acute myeloid leukemia (AML) and 7 cell lines derived from multiple myeloma (MM).

16 showed activity across the 19 hematological cell lines tested with a geometric absolute mean IC50 value of 0.34 mM and a sig- nificant selectivity for the three highly sensitive cell lines EOL-1, MOLM-13 and MV4-11. At least for MV4-11 and MOLM-13 an ITD (internal tandem duplicate) of FLT3 was described [61] causing strong sensitivity towards inhibitors of FLT3. These findings match well the results presented in Table 1. The EOL-1 cell line represents an AML cell line overexpressing wild-type FLT3 [62] and bearing a PDGFRA mutation in addition [63], sensitizing it towards 7. Regarding cell lines derived from multiple myeloma, the sensitivity was clearly lower.

 

2.8.        In vitro assessment of AC220 (quizartinib, 7), midostaurin 1, and 17b in 27 selected hematological, human cancer cell lines

Compound 16 and its carbamate 17b, respectively, show strong activity across the hematological cell lines tested. With a geometric absolute mean IC50 value of 0.131 mM Midostaurin 1 was the most potent compound followed by 16 (0.337 mM), 17b (0.502 mM) and AC220 (quizartinib, 7) (0.846 mM).

All five highly sensitive cell lines EOL-1, MOLM-13, MV4-11, MONO-MAC-1, MONO-MAC-6 exhibit FLT3 overexpression. MONO-MAC-1, and MONO-MAC-6 have an activating point muta- tion of FLT3 (V592A) in the juxtamembrane domain [64].

16 and its carbamate 17b achieved IC50 values in the sub nanomolar till low nanomolar range in these cell lines and were extremely specific. For the remaining AMLs, the compounds were completely ineffective up to the highest dose of 30 mM. The IC50 for 17b on the sensitive cell lines was 0.09e11 nM. In contrast it showed IC50 values between 1 and 5 mM in the other cell lines, indicating 100 to 1000-fold selectivity.

 

2.9.        In vivo tolerability of 17b

To investigate possible toxicity effects in vivo, we started a preliminary tolerability study using NMRI nu/nu mice. The mice were given 75 mg/kg/day of 17b orally for 3 days. Daily observation of the treated mice (n   6) showed no lethality and no impairment of general condition till euthanasia on day 4. Moreover, a body weight gain of 3% during the therapy from day 0e3 was observed. In summary, no signs of toxicity were observed. Comparative data were obtained for 17b and AC220 in the pharmacologic compari- son, extended experiments which will be published in a forth- coming paper by some of the authors.

 

 

3.            Conclusion

In enzymatic assays and in cell lines the new FLT3 inhibitor 16 and its carbamate derivative 17b show high efficiency, compa- rable to AC220 (quizartinib, 7) in FLT3 activating mutations, namely FLT3-ITD and activating point mutations. This activity is accompanied by a high specificity for FLT3. No significant co- targeting of other RTKs required  for  normal  hematopoiesis  can be observed.

Moreover, 16 and its carbamate 17b exhibit selectivity of 100 until 1000-fold for sensitive versus resistant cell lines. No signs of toxicity were observed in a preliminary in vivo study.

Our data suggest that 16 or its carbamate derivative 17b, respectively, might become a novel and well-tolerated TKi against FLT3-ITD-positive AML.

Cells are cultured in Roswell Park Memorial Institute (RPMI) 1640 medium plus 10% fetal bovine serum and 100 U/mL penicillin and 100 mg/mL streptomycin.