Four AML cell lines, MOLM-13 (AML M5), KASUMI-1 (AML M2), OCI-AML3 (AML M4) and NOMO-1 (AML M5) were obtained from the American Type Culture Collection, and were mycoplasma-tested and authenticated using the LGC Standards Cell Line Authentication service. The cell lines were cultured at 37°C in a 5% CO₂ atmosphere at a density of 0.3×10⁶ cells/ml in complete medium, in T75 flasks. MOLM-13 and KASUMI-1 cells were cultured in RPMI-1640 (Euroclone) supplemented with 20% heat-inactivated FBS (GE Healthcare), 2 mM L-glutamine (GE Healthcare), 100 U/ml penicillin, 100 µg/ml streptomycin (GE Healthcare) and 0.2% Mycozap (Lonza, Inc.). OCI-AML3 cells were cultured in α-MEM (Lonza, Inc.) with 20% FBS, 2 mM L-glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin. NOMO-1 cells were grown in RPMI-1640 with 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin.

Kevetrin powder was kindly provided by Innovation Pharmaceuticals, dissolved in sterile water in a 3.4 mM stock solution, stored at 4°C and used within 1 month. Cells were seeded in 96-well or 6-well plates at 0.5×10⁶/ml in 100 and 3,000 µl of medium, respectively, and treated with increasing drug concentrations (85–340 µM), according to peak plasma concentrations measured in the phase I clinical trial (NCT01664000). For pulsed experiments, cells were exposed to the drug for 6 h and then washed and replated in complete medium [wash-out (wo)]. After 66 h, cells were reseeded in fresh medium containing the drug for 6 h, followed by a 66-h wo. The pulsed treatment was repeated 2–3 times.

Samples were collected at Istituto Scientifico Romagnolo per lo Studio e la Cura dei Tumori (IRST) IRCCS from 4 AML patients at diagnosis (inclusion criteria: Age ≥18 years, confirmed AML diagnosis, available clinical data for review and obtained written informed consent) between December 2018 and October 2019 (Table SI). Bone marrow mononuclear cells (BMMCs) and peripheral blood mononuclear cells (PBMCs) were collected by density gradient centrifugation using Lymphosep (BioWest SAS), then lysed in RLT buffer (Qiagen, Ltd.) supplemented with 1% β-mercaptoethanol, and/or cryopreserved in 90% FBS and 10% DMSO (Sigma-Aldrich; Merck KGaA). After thawing, BMMCs were primed for 24 h with a cytokine cocktail [20 ng/ml Fms-related tyrosine kinase 3 ligand (FLT3-L), interleukin (IL)-3, IL-6, stem cell factor and granulocyte colony-stimulating factor (Miltenyi Biotec GmbH)] and live cells [collected using the Dead Cell Removal Kit (Miltenyi Biotec GmbH)] were then treated with increasing doses of kevetrin (85–340 µM) for 48 h.

Cell viability was determined using the CellTiter 96^® AQueous One Solution Cell Proliferation Assay (Promega Corporation), according to the manufacturer's instructions. The optical density was determined after 3 h at a wavelength of 490 nm by the Thermo Multiskan EX microplate reader (Thermo Fisher Scientific, Inc.). Cell viability in primary samples was evaluated by the trypan blue exclusion assay.

Phosphatidylserine externalization was evaluated using the fluorescein isothiocyanate (FITC) Annexin V Apoptosis Detection kit (eBioscence; Thermo Fisher Scientific, Inc.). After treatment, cells were incubated with 25 µl/ml of Annexin V-FITC for 15 min at 37°C in a humidified atmosphere in the dark. Prior to flow cytometric analysis, propidium iodide (PI) was added to a final concentration of 5 µg/ml. Flow cytometric analysis was performed using a FACSCanto flow cytometer (Becton, Dickinson and Co.) equipped with 488 nm (blue) and 633 nm (red) lasers, and 10,000 events were recorded for each sample. Data acquisition and analysis were performed using FACSDiva software v.6.1.3. (Becton, Dickinson and Co.). In primary samples, Annexin V staining was combined with surface markers using the following antibodies: CD45-APC Vio770 (cat. no. 130-110-635), CD33-APC (cat. no. 130-111-020), CD14-PerCP Vio 700 (cat. no. 130-110-523), CD3-PE (cat. no. 130-113-139) (all from Miltenyi Biotec GmbH, dilution 1:50) and CD19-PE Cy7 (cat. no. 302216, 1:10, BioLegend, Inc.).

Mitochondrial membrane polarization was evaluated using the Mito-PT JC-1 Assay Kit (ImmunoChemistry Technologies, LLC). Following kevetrin exposure, cells were incubated according to the manufacturer's instructions in JC-1 working solution for 15 min at 37°C and 5% CO₂ in a humidified atmosphere in the dark, suspended in 1X assay buffer and analyzed by flow cytometry. A positive control treated with 50 µM of carbonyl cyanide 3-chlorophenylhydrazone was used for each experiment.

Fragmented DNA was detected by the TUNEL assay (Roche Diagnostics GmbH). After each treatment, samples were fixed in 1% formaldehyde on ice for 15 min, suspended in 70% ice-cold ethanol and stored overnight at −20°C. Cells were incubated for 5 min at 4°C in PBS containing 0.1% Triton X-100 (Bio-Rad Laboratories, Inc.). Thereafter, samples were suspended in 50 µl of solution containing TdT and FITC-conjugated dUTP deoxynucleotides 1:1 (Roche Diagnostics GmbH) and incubated for 90 min at 37°C in a humidified atmosphere in the dark. Samples were counterstained with 2.5 µg/ml of PI (MP Biomedicals, LLC) and 10 kU/ml of RNAse (Sigma-Aldrich; Merck KGaA) for 30 min at 4°C in the dark, then analyzed by flow cytometry. A positive control treated with 80 kU/ml of DNase (Sigma-Aldrich; Merck KGaA) was included for each experiment.

The percentage of active caspase-3 was measured using the FITC Active Caspase-3 Apoptosis Kit (BD Biosciences). After treatment, cells were collected, washed with ice-cold PBS 1X and incubated for 20 min at 4°C in Cytofix/Cytoperm buffer (BD Biosciences). The samples were then incubated with 20 µl of anti-active caspase-3 antibody (cat. no. 550480, 1:5, BD Biosciences) for 30 min at room temperature in the dark, washed in Perm/Wash buffer and analyzed by flow cytometry. A positive control treated with 5 µM camptothecin was used for each experiment.

After treatment, cells were washed in 1X PBS, fixed in 70% ice-cold ethanol, stained with 10 µg/ml PI (MP Biomedicals), 10 kU/ml RNAse (Sigma-Aldrich; Merck KGaA) and 0.01% Nonidet™ P40 (NP40; Sigma-Aldrich; Merck KGaA) overnight at 4°C in the dark and analyzed by flow cytometry. Data analysis was performed using ModFit 4.1 (DNA Modelling System, Verity Software House, Inc.).

DNA was extracted from primary mononuclear cells (BMMCs or PBMCs) using the Maxwell^® RSC Blood DNA Kit (Promega Corporation) according to the manufacturer's instructions. The mutational profile of patients was determined using SOPHiA Myeloid Solution™ (SOPHiA GENETICS), a CE-IVD marked molecular diagnostic application. Libraries were prepared according to the manufacturer's instructions and quantified using the Qubit^® dsDNA High Sensitivity Assay on a Qubit^® Fluorometer (Thermo Fisher Scientific, Inc.). The median average of amplicons was determined by capillary electrophoresis using Agilent High Sensitivity DNA kit (Agilent Technologies, Inc.). The pooled libraries were paired-end (2×301) and sequenced with v3 chemistry on a MiSeq™ instrument (Illumina, Inc.), as described in the manufacturer's protocol. FASTQ sequencing files were uploaded onto the SOPHiA DDM^® platform (version 4), which uses patented advanced technologies for variant calling and annotation. Human Genome Build 19 (Hg19) was used as the reference for sequence alignment. A minimum coverage depth of 1,000× was recommended. A filtering tool was set to exclude known single-nucleotide polymorphisms, variants localized in intronic and UTR regions and synonymous variants. Only exonic and splice site variants with a variant allele frequency (VAF) ≥2.5% were evaluated.

RNA was isolated using TRIzol^® (Invitrogen; Thermo Fisher Scientific, Inc.) from cells untreated or treated for 6 and 48 h with kevetrin at 340 µM. Labeled single-stranded complementary DNA was prepared from 100 ng of RNA and hybridized to Human Transcriptome Array 2.0 (Thermo Fisher Scientific, Inc.), according to the manufacturer's recommendations. Three independent replicates of each condition per cell line were analyzed. Data quality control and normalization were carried out by Expression Console software v1.4.1.46 (Thermo Fisher Scientific, Inc.), while supervised analysis was performed with Transcriptome Analysis Console software v.3.0 (Thermo Fisher Scientific, Inc.). Functional annotation clustering was performed using David Bioinformatics Resources 6.8 (National Institute of Allergy and Infectious Diseases, National Institutes of Health) (21). Gene set enrichment analysis (GSEA) was performed with GSEA software v.3.0 (Broad Institute) (22,23).

After a 48-h treatment, total protein extracts were prepared in ice-cold lysis buffer (0.5% NP40, 250 mM NaCl, 50 mM HEPES, 5 mM EDTA and 0.5 mM EGTA) containing phosphatase inhibitor cocktail 2 (Sigma-Aldrich; Merck KGaA), protease inhibitor (Clontech Laboratories, Inc.), and DTT (Invitrogen; Thermo Fisher Scientific, Inc.). Before use, extract concentrations were normalized using a BCA protein assay kit (Bio-Rad Laboratories, Inc.). Proteins (50–100 µg) were separated by SDS-PAGE using 4–20% polyacrylamide precast gel (Bio-Rad Laboratories, Inc.) and were transferred to PVDF membranes using a TransBlot Turbo system (Bio-Rad Laboratories, Inc.). Blocking and antibody incubations were performed in Tris-buffered saline with 0.1% Tween-20 (TBST) plus 5% dry milk or in TBST plus 5% BSA (Sigma-Aldrich; Merck KGaA). The following antibodies were used: Anti-phosphorylated-p53 (Ser15) (cat. no. 9286, clone 16G8, 1:1,000), anti-p53 (cat. no. 2527, clone 7F5, 1:1,000) and anti-p21 (cat. no. 2946, clone DCS60, 1:2,000), all from Cell Signaling Technologies, Inc. Detection was performed using horseradish peroxidase-conjugated anti-rabbit (cat. no. NA934, 1:5,000) and anti-mouse (cat. no. NA931, 1:10,000) secondary antibodies (GE Healthcare) with WesternBright Sirius (Advansta, Inc.) or SuperSignal West Femto (Thermo Fisher Scientific, Inc.) substrates. Membranes were imaged on a ChemiDoc MP system (Bio-Rad Laboratories, Inc.). Membranes were stripped and reprobed with antibody against β-actin (cat. no. ab49900, clone AC-15, 1:25,000, Abcam) as normalizer. QuantityOne 4.6.8 software (Bio-Rad Laboratories, Inc.) was used for analysis.

Cells were harvested after a 48-h treatment, washed in 1X PBS and fixed in 10% non-buffered formalin in a 37°C water bath for 1 h. Samples were then washed in 1X PBS and resuspended in 1 ml of 70% EtOH. Slides were prepared by cytospin centrifugation at 133 × g for 5 min at room temperature and 5 min incubation at room temperature in ethanol solutions with increasing concentration (50–70–100%). Blocking was performed with 1% BSA and 0.3% Triton X-100 in 1X PBS for 1 h. Slides were incubated overnight at 4°C with anti-p53 antibody (clone 7F5, 1:1,600 Cell Signaling Technologies, Inc.), washed and stained with goat anti-rabbit Alexa Fluor 594 secondary antibody (1:1,000, Invitrogen; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. The samples were washed three times in 1X PBS and mounted using ProLong Antifade DAPI (Invitrogen; Thermo Fisher Scientific, Inc.). Cells were imaged with a N-SIM E laser confocal microscope (Nikon Corporation) at a magnification of ×60 and analyzed with NIS Elements software 5.11 (Nikon Corporation) and ImageJ software 1.52a (National Institutes of Health).

Statistical analysis was carried out with GraphPad Prism 8.0.1 software (GraphPad Software, Inc.). Comparisons between two groups were performed using Student's t-test, whereas multiple comparisons were performed using one-way analysis of variance with Dunnett's post hoc test. Values represent the mean ± standard deviation of three independent experiments. P<0.05 was considered to indicate statistically significant differences.

The response of MOLM-13 (TP53-wt) and KASUMI-1 [mutated (homozygous single-nucleotide variant, p.R248Q)] cells to kevetrin exposure was analyzed, initially using a pulsed treatment to simulate the dosing schedules used in clinical practice. Cell viability was evaluated after exposure to kevetrin doses of 85, 170 and 340 µM for 6 h, or for 6 h followed by a 66-h wo. The latter schedule was repeated consecutively two and three times, respectively. No effect on the viability of MOLM-13 cells was observed (Fig. 1A), whereas a dose- and time-dependent decrease was evident in the KASUMI-1 cell line (Fig. 1B). The role of kevetrin treatment in apoptosis induction was then evaluated by examining phosphatidylserine externalization following exposure to the highest concentration in KASUMI-1 cells, detecting a trend towards increased percentage of Annexin V⁺ cells (Fig. 1C and D and Fig. S1A and B). Kevetrin treatment did not induce cell cycle alterations (Fig. 1E and F and Fig. S1C and D).

Gene expression profile analysis of MOLM-13 and KASUMI-1 cells was performed following treatment with kevetrin or the vehicle for 6 h to identify the transcriptional program regulated by short-term kevetrin exposure. This strategy enabled us to select early kevetrin-responsive genes. The overall transcriptional program was barely affected and we observed a differential expression of a restricted subset of genes (Fig. 2A and B), including the regulator of WNT/β-catenin signaling forkhead box K2 and the transcription factor signal transducer and activator of transcription 5A (STAT5A), which were downregulated 2-fold in KASUMI-1 cells after treatment (Fig. 2B). Moreover, MT 1 and 2, which are involved in the scavenging of oxygen-free radicals, were upregulated by kevetrin exposure in both cell lines (Fig. 2A and B). In addition, KASUMI-1 cells exhibited a significant downregulation of genes in relation to p53 activity, including E2F transcription factor 4 (E2F4), glutamate-rich WD repeat containing 1 (GRWD1), solute carrier family 6 member and elastase, neutrophil expressed (ELANE) (Fig. 2B).

The effects of prolonged kevetrin exposure were investigated by evaluating cell viability after treatment for 24, 48 and 72 h (Fig. S2) using the same concentration range tested in the pulsed experiments. In addition to being tested on MOLM-13 and KASUMI-1 cells, continuous kevetrin treatment was also tested on TP53-wt OCI-AML3 cells and TP53-mutant (heterozygous frameshift deletion, p.C242fs) NOMO-1 models. Cell viability was barely altered after 24 h of treatment (Fig. 3A). At 48 h, a significant decrease in cell viability was observed in MOLM-13 cells at the highest concentration and in OCI-AML3 cells at 170 and 340 µM (Fig. 3B). A dose-dependent inhibition was also detected in KASUMI-1 and NOMO-1 cells (Fig. 3B). To better define the mechanism of action of kevetrin, apoptosis and cell cycle progression were analyzed. In KASUMI-1 cells, a significant apoptosis induction was observed after 24 h of treatment at the highest concentration (Fig. 3C and Fig. S3). After a 48-h exposure to the highest kevetrin concentration (340 µM), a significant increase was observed in Annexin V⁺ cells among MOLM-13 (54.95±5.63% in kevetrin-treated cells vs. 12.53±6.15% in the control) and NOMO-1 (60.93±2.63% in kevetrin-treated cells vs. 22.90±4.63% in the control) cells, and a mild but significant effect in OCI-AML3 cells (10.03±3.79% in kevetrin-treated cells vs. 2.60±0.70% in the control; Fig. 3D and Fig. S4). KASUMI-1 cells exhibited a dose-dependent response at 48 h, with 79.70±4.57% of apoptotic cells at 340 µM (compared with 13.18±0.80% in the control; Fig. 3D and Fig. S4). Apoptotic data in MOLM-13 and KASUMI-1 cells were confirmed in terms of mitochondrial depolarization, DNA fragmentation and caspase-3 activation (Fig. S5A-F). No cell cycle alterations were observed in MOLM-13 and KASUMI-1 cells, while NOMO-1 and OCI-AML3 cells displayed an accumulation of cells in the G0/G1 phase and a decrease of S phase cells after 24 and 48 h of kevetrin treatment at any concentration (Fig. 3E and F and Fig. S6A-D).

BMMCs from 4 patients were treated with increasing kevetrin concentrations to evaluate its effect on primary AML cells. The mutational status of patients was assessed by targeted NGS (Table SII). One of the cases carried a TP53 mutation [no. 01: NM_000546, c.764T>A, p.(Ile255Asn), VAF 92.3%]. A dose-dependent decrease in cell viability (85 µM: 77.8±12.9%, 170 µM: 56.1±10.3% and 340 µM: 33.6.1±13.5%; Fig. 3G) and a significant increase in the proportion of apoptotic cells (170 µM: 31.8±13.3% and 340 µM: 54.3±13.9% vs. control: 11.4±6.5%; Fig. 3H and Fig. S6E) were observed. Kevetrin cytotoxicity was then assessed in BM cell populations and a preferential activity against AML blast cells was noted, indicating a significant dose-dependent decrease in the percentage of alive cells with respect to control, with the TP53-mutant sample exhibiting high sensitivity (Fig. 3I and J). Conversely, the percentage of AnnexinV⁻ monocytes and B and T lymphocytes was only marginally affected by kevetrin treatment (Fig. 3J), suggesting a selective cytotoxic activity of this drug.

To better understand the molecular and biological effects of kevetrin exposure, gene expression profiling of MOLM-13 and KASUMI-1 cells after a 48-h treatment was performed. A total of 1,024 upregulated and 1,563 downregulated genes were identified in MOLM-13 cells, whereas KASUMI-1 cells exhibited increased expression of 325 genes and decreased levels of 1,535 genes. Of note, although a number of genes were uniquely altered, there was also a high degree of overlap in the transcriptional changes between MOLM-13 and KASUMI-1 cells, with a core transcriptional program of 162 upregulated and 812 downregulated genes in both cell lines (Fig. 4A). Enrichment analysis of the core transcripts revealed that upregulated genes were mainly involved in transcription, nucleosome assembly and telomere organization (HIST4H4, HIST1H3H, HIST2H4B and HIST1H4H), apoptosis (CASP10 and CASP8), autophagy (GABARAPL1, ATG14 and WIPI1), nuclear factor-κB pathway regulation (TRIM38, RIPK1, NFKBIA, IL1B, HSPA1A, HSPA1B, S100A4, CASP10, PLK2, CASP8, TLR6, PTGS2 and IL1B) and mitogen-activated protein kinase activity (DUSP1, JUN, MAP3K8, DUSP10, IL1B, HSPA1A, HSPA1B, GADD45B and DUSP22) (Table SIII). Downregulated genes were mostly involved in cell cycle (e.g., CDK4 and CDC7), DNA repair (e.g., EXO1 and FANCI), biosynthetic processes (e.g., PPAT involved in purine nucleotide biosynthesis and FASN in fatty acid synthesis), bioenergetics [e.g., glycolytic enzyme hexokinase 2 (HK2) and mitochondrial electron transport member COX5A), translation (e.g., RPL34 and RPSA), telomere maintenance (e.g., DKC1) and splicing (e.g., HNRNPR and PTBP1) (Table SIV). The common core transcriptional program included critical regulators of myelopoiesis and leukemogenesis, such as MYC, MYB and BCL11A, leukemia-related genes involved in unfolded protein response (UPR; XBP1 and CALR), signaling (RET), glycolysis (HK2) and DNA methylation [isocitrate dehydrogenase 2 (IDH2)], all of which were downregulated by kevetrin treatment (Fig. 4B). Conversely, increased levels of EZH1 and JUN were detected after exposure to kevetrin. Notably, GSEA of microarray data revealed that kevetrin-treated MOLM-13 and KASUMI-1 cells shared the downregulation of glycolysis, DNA repair, UPR and MYC target gene sets (Fig. 4C), in line with the results obtained by pathway analysis.

The effects of kevetrin treatment on the p53 pathway were then analyzed. High kevetrin doses induced a mild reduction in p53 mRNA levels. However, treated cells were enriched with a transcriptional gene signature involved in the p53 pathway (Fig. 4D). A total of 24 bona fide p53 targets (24) exhibited increased expression in MOLM-13 cells, including p21 (CDKN1A), MDM2 and NOXA (PMAIP1) (Table I). The latter gene was also upregulated in KASUMI-1 cells, along with 8 additional p53 targets (Table I). Among these, BLOC1S2, PLK2, RRM2B, TP53INP1, IER5 and LAPTM5 also exhibited increased expression in MOLM-13 cells following kevetrin treatment (Fig. 4E). To better understand the dynamics of p53 expression and its related proteins under kevetrin treatment, western blot and immunofluorescence analyses were performed after a 48-h exposure to increasing kevetrin doses. Western blot analysis did not reveal relevant alterations in the level of total p53 or its active form (phosphorylated on Serine 15), with the exception of KASUMI-1 cells that exhibited reduced p53 expression at the highest kevetrin dose. This may be associated with the high numbers of apoptotic cells lost during the washing steps required for pellet preparation. Immunofluorescence analysis at the highest kevetrin dose revealed an increased p53 expression compared with control cells, which resulted from a fraction of intact cells displaying nuclear p53 staining and apoptotic cells (identified by nuclear fragmentation) with very high p53 levels (Fig. 5B-D). This phenomenon was observed across all cell lines, particularly TP53-mutant ones (Fig. 5C and D). MOLM-13 cells generally displayed lower p53 expression levels across all conditions, thus hampering the detection of its active form by western blotting. Its target protein, p21, exhibited a dose-dependent upregulation following kevetrin treatment in TP53-wt models, in line with transcriptomic data obtained on MOLM-13 cells (Fig. 5A and Table I).