OTS167 blocks FLT3 translation and synergizes with FLT3 inhibitors in FLT3 mutant acute myeloid leukemia

Internal tandem duplication (-ITD) mutations of Fms-like tyrosine kinase 3 (FLT3) provide growth and pro-survival signals in the context of established driver mutations in FLT3 mutant acute myeloid leukemia (AML). Maternal embryonic leucine zipper kinase (MELK) is an aberrantly expressed gene identified as a target in AML. The MELK inhibitor OTS167 induces cell death in AML including cells with FLT3 mutations, yet the role of MELK and mechanisms of OTS167 function are not understood. OTS167 alone or in combination with tyrosine kinase inhibitors (TKIs) were used to investigate the effect of OTS167 on FLT3 signaling and expression in human FLT3 mutant AML cell lines and primary cells. We describe a mechanism whereby OTS167 blocks FLT3 expression by blocking FLT3 translation and inhibiting phosphorylation of eukaryotic initiation factor 4E–binding protein 1 (4E-BP1) and eukaryotic translation initiation factor 4B (eIF4B). OTS167 in combination with TKIs results in synergistic induction of FLT3 mutant cell death in FLT3 mutant cell lines and prolonged survival in a FLT3 mutant AML xenograft mouse model. Our findings suggest signaling through MELK is necessary for the translation and expression of FLT3-ITD, and blocking MELK with OTS167 represents a viable therapeutic strategy for patients with FLT3 mutant AML.


Cell Lysis and Immunoblotting
Cell lysates prepared with CelLytic M (Sigma-Aldrich) with NaV (New England BioLabs) and protease
Briefly, cell lysis was performed with 600μL of urea lysis buffer (8M urea, 50mM Tris.HCl pH8, 150mM NaCl) using a QSonica sonic probe, incubated at room temperature for 1hr with mixing at 1000rpm in an Eppendorf Thermomixer, then clarified by centrifugation at 10,000g for 10min at 25degC. 50ug of each lysate was reduced with 14mM dithiothreitol followed by alkylation with 14mM iodoacetamide, then digested with 20ug sequencing grade trypsin (Promega) at 37degC overnight.
The digest was terminated with formic acid, then centrifuged at 10,000g for 10min.
Each digested sample was processed by solid phase extraction using an Empore C18 (3M) plate under vacuum (5inHg) and lyophilized. For the 10-plex, samples were reconstituted with 200mM HEPES and labeled with TMT labels (ThermoFisher Product # UL292368) by incubating in a Thermomixer for 1.5hr at 25degC and 300rpm. The reactions were quenched with addition of 5% hydroxylamine for 15min at 25degC. Equal amounts of labeled peptide were combined and processed by solid phase extraction using an Empore C18 (3M) 10mg C18 cartridge. 250ug of the labeled pooled sample was separated into 96 discrete fractions using an Agilent 1100 HPLC system equipped with a diode array and fraction collector. The HPLC column was a 2.1 x 150mm Waters XBridge column and peptides were eluted with a basic (pH10) reverse phase buffer system.
Fractions were pooled and each pooled sample was lyophilized and reconstituted in 0.1%TFA to produce 12 samples for LC-MS/MS analysis.
10% of each pooled sample fraction (the equivalent of ~2.5μg of digested sample) was analyzed by nano LC-MS/MS with a Waters M-class LC system interfaced to a ThermoFisher Fusion Lumos.
Peptides were loaded on a trapping column and eluted over a 75μm analytical column at 350nL/min; both columns were packed with Luna C18 resin (Phenomenex). Each fraction was analyzed with a 2hr gradient (24hrs total LC-MS/MS time). The mass spectrometer was operated using a custom MS3 method. MS scans were acquired in the Orbitrap at 120,000 FWHM resolution, MS2 scans were acquired in the ion trap using CID at 35% NCE, product ions were isolated using synchronized precursor selection (SPS) and fragmented using HCD at 65% NCE. MS3 scans were acquired in the Orbitrap at 50,000 FWHM resolution from m/z 100-500. A 2s cycle time was employed for all steps.
Data were processed with MaxQuant version 1.6.5.0 (Max Planck Institute for Biochemistry) which incorporates the Andromeda search engine. The Andromeda settings were as follows: Enzyme, Trypsin/P; Database, SwissProt Human; Fixed modification, Carbamidomethyl (C); Variable modifications, Acetyl (protein N-term), Oxidation (M); Missed cleavages, 2; Reporter ion tolerance, 0.003 Da. The MaxQuant output was further processed using Microsoft Excel. The false discovery rate of proteins and peptides was set to 0.01. The dataset was normalized by subtracting the median, and a fold-change threshold of 1.6 was used to determine differentially expressed proteins.
All datasets were subject to a student's t-test and P values of less than 0.05 were considered statistically significant. For functional enrichment analysis a Welch's t-test (5% permutated FDR) was performed on control vs. 8hr data. A Fisher exact test was performed on the annotations associated with the proteins identified as increased/decreased between conditions. Gene Ontology (GO) (www.geneontology.org), Corum (mips.helmholtz-muenchen.de/corum/#) and Kyoto Encyclopedia of Genes and Genomes (KEGG)(www.genome.jp/kegg/) analysis was applied. Each experimental condition for quantitative proteomic analysis was performed in triplicate.

Polysomal Profiling and Fractionation
Cells were washed twice with Dulbecco's phosphate buffered saline (DPBS) and then lysed in lysis buffer (0.5% Triton X 100, 0.5 % sodium deoxycholate, 5 mM Tris pH 7.5, 2.5 mM MgCl2, 1.5 mM KCl 100μg/ml cycloheximide, 2 mM DTT, protease inhibitor and 1U/μl RNase inhibitor). Lysates were then centrifuged at 20000 x g for 20 minutes at 4C and supernatants were collected and snap frozen in liquid nitrogen. To isolate ribosomal fractions, lysates were layered on a sucrose gradient of 5 to 50%. Samples were centrifuged at 4C for 120 minutes at 35000 rpm in a Beckman SW41-Ti rotor.
Absorbance was measured at 254 nm continuously in an ISCO density gradient fractionator with the following settings: pump speed, 1.5 ml/min; fraction size, 10 drops per fraction; chart speed, 150 cm per hour; sensitivity, 1; peak separator, off; noise filter, 0.5 seconds. Fluorinert TM FC-40 (Sigma-Aldrich #F9755) was used to set the baseline in an UA-6 detector for all experiments. Quantitative measurement (area under the curve) of monosomal and polysomal peaks was performed using ImageJ.
Luminescence was read with a Bio Tek Synergy H4 plate reader using Gen5 software (SCR_017317).

Flow Cytometry
Flow Cytometry analysis was performed using an LSR II or LSR-Fortessa 4-15 (BD Biosciences, Franklin Lakes, NJ). Annexin V staining was performed using eBioscience Annexin V Apoptosis Detection Kit APC (Invitrogen #88-8007, Carlsbad, CA). Splenocytes and bone marrow were harvested and 70um filtered before blocking with anti-mFcR mAb 2.4G2 and then staining with 6 excess primary or control antibodies. Stained cells were washed and run on above flow analyzers in the presence of propidium iodide (PI) or 4′,6-diamidino-2-phenylindole (DAPI) for live/dead cell discrimination. Events displayed were first gated as PIor DAPIthen as singlets (using doublet discrimination). Absolute numbers of CD45+ leukemia cells were calculated using CountBright absolute counting beads (Invitrogen #C36950) added to samples before flow analysis. Data analysis was performed using FCS Express (De Novo Software, Glendale, CA)( SCR_016431).          Table S1. Histology review of MV4:11-xenografted NSG mice treated with a combination of OTS167 and gilteritinib. Comments on hematoxylin and eosin (H&E) tissue slides from pathologist review of slides (slides reviewed from 1 mouse per experimental condition). Slides were generated from a representative MV4:11-engrafted NSG mouse after 14 days of the indicated treatment which started on day 10 after engraftment.