Cell Cultures

MOLM-13 cells were purchased from AddexBio (2012; San Diego, CA, USA). MV4–11, THP-1, and U937 cell lines were purchased from the American Type Culture Collection (2006, 2014, 2002, respectively; Manassas, VA, USA). Cell lines were cultured using RPMI 1640 media plus 10–20% fetal bovine serum (CLARK Bioscience, Claymont, DE, USA), 2 mM L-glutamine, and 100 U/ml penicillin and 100 μg/ml streptomycin and incubated in a humidified, 5%CO₂/95% air environment at 37 °C. All lines were tested for the presence of mycoplasma using the PCR method described by Uphoff and Drexler (15) on a monthly basis. The cell lines were authenticated in 2017 at the Genomics Core at Karmanos Cancer Institute using the PowerPlex® 16 System from Promega (Madison, WI, USA).

Clinical Samples

Primary patient samples were obtained from the First Hospital of Jilin University, following obtainment of informed consent according to the Declaration of Helsinki. The study itself and the obtainment of patient samples was first approved by the Human Ethics Committee of The First Hospital of Jilin University. All patient samples were screened for FLT3-ITD, NPM1, C-kit, CEBPA, IDH1, IDH2, N-RAS and DNMT3A mutations by PCR amplification and automated DNA sequencing, cytogenetics, and fusion genes utilizing real-time RT-PCR, as described previously (16, 17). Individual patient characteristics are provided in Supplementary Table S1. Patient samples were purified using Ficoll-Hypaque density centrifugation and cultured in RPMI 1640 media plus 20% fetal bovine serum, ITS Solution (Sigma-Aldrich, St Louis, MO, USA), and 20% supernatant of the 5637 bladder cancer cell line (to provide sources of granulocyte–macrophage colony-stimulating factor, granulocyte colony-stimulating factor, interleukin-1 beta, macrophage colony-stimulating factor, and stem cell factor) (16, 18, 19).

Annexin V-fluorescein isothiocyanate (FITC) / Propidium Iodide (PI) Staining and Flow Cytometry Analyses

Cells were treated with either midostaurin, gilteritinib, SCH772984, or venetoclax, alone or in combination, for up to 24 hours. Flow cytometry analysis using the annexin V-FITC/PI Apoptosis Kit (Beckman Coulter; Brea, CA, USA) was performed as previously described (20, 21). Results are displayed as percentage of annexin V positive cells. All experiments were repeated in triplicate independently; displayed cell line data are from one experiment representative of findings of all three, and patient sample experiments were performed once in triplicate due to limited sample. The extent and direction of the antileukemic interaction was identified via calculation of the combination index (CI) using CompuSyn software (Combosyn Inc., Paramus, NJ, USA), where CI<1, CI=1 and CI>1 are indicative of synergistic, additive, and antagonistic effects, respectively (20, 22).

Western Blots

Cell lysis was performed in the presence of protease and phosphatase inhibitors (Roche Diagnostics, Indianapolis, IN, USA). Whole-cell lysates underwent SDS-polyacrylamide gel electrophoresis, electrophoretically transferred onto polyvinylidene difluoride (PVDF) membranes (Thermo Fisher Scientific, Rockford, IL, USA), and subsequently immunoblotted utilizing anti-Mcl-1, -Bcl-2, -Bcl-xL, -Bax, -β-actin, -ERK (Proteintech, Rosemont, IL, USA), -p-AKT (T308), -p-AKT (S473) (Affinity Biosciences, Zhenjiang, Jiangsu, China), -Bim, -cf-Caspase 3, -p-S6 (S240/244), -p-STAT5(Y694), -p-Mcl-1 (T163) (Cell Signaling Technologies, Danvers, MA, USA), -Bak, -p-ERK, -AKT (Abcam, Cambridge, MA, USA) as previously described (23, 24). Visualization of immunoreactive proteins was performed via the Odyssey Infrared Imaging System (Li-Cor, Lincoln, NE, USA), as described by the manufacturer. Western blots were repeated in triplicate (at minimum), with one representative blot displayed. Densitometry measurements were performed using Odyssey V3.0 (Li-Cor) and normalized to β-actin, calculated as fold-change compared to the corresponding control no-drug treatment specimen.

Co-immunoprecipitation

AML cell lines were treated for 4hours with either midostaurin or venetoclax, alone or in combination, then lysed with 1% CHAPS, 5 mM MgCl₂, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 20 mM Tris and 0.05% Tween-20 in the presence of protease inhibitors. Co-immunoprecipitation of Bim was carried out as described previously (25). 2 μg of anti-Bim antibody (Cell Signaling Technology, Danvers, MA, USA), 1 mg protein lysate, and Protein A agarose beads (Roche Diagnostics). Proteins were then eluted with 50 mM glycine at pH 2.0, and then analyzed via Western blotting as described above.

Ectopic overexpression of Mcl-1 and shRNA knockdown of Bak and Bax

The pMD-VSV-G and delta 8.2 plasmids were gifts from Dr. Dong at Tulane University. Red fluorescent protein (RFP), Mcl-1 cDNA constructs were purchased from Thermo Fisher Scientific Biosciences (Lafayette, CO, USA). Bak and Bax shRNA constructs were purchased from Sigma Aldrich. Lentivirus production and transduction were carried out as previously described (26).

Quantification of gene expression by real-time RT-PCR

Total RNA was extracted using TRIzol (Thermo Fisher Scientific) and cDNAs were prepared from 2 μg total RNA using random hexamer primers and a RT-PCR Kit (Thermo Fisher Scientific), and then purified using the QIAquick PCR Purification Kit (Qiagen), as described previously (16, 22). Mcl-1 transcripts were quantitated using the TaqMan probe Hs01050896_m1 (Thermo Fisher Scientific) and a LightCycler 480 real-time PCR machine (Roche Diagnostics), based on the manufacturer’s instructions. Real-time PCR results were expressed as means from three independent experiments and were normalized to GAPDH transcripts measured by TaqMan probe (Hs02786624_g1). Fold changes were calculated using the comparative Ct method (27).

Leukemia xenograft model

MV4–11 cells (1 x 10⁶ cells/mouse; 0.2ml/injection) were injected intravenously (day 0) into immunocompromised triple transgenic NSG-SGM3 female mice at eight weeks of age (NSGS, JAX#0103062; non-obese diabetic scid gamma (NOD.Cg-Prkdc^scid Il2rg^tm1Wjl Tg(CMV-IL3, CSF2, KITLG)1Eav/MloySzJ) Jackson Laboratory, Bar Harbor, ME, USA). Mice were randomly placed (5 mice/group; day 4) into vehicle control, 40 mg/kg gilteritinib (administered orally, p.o.), 85 mg/kg ABT-199 (p.o.), or 40 mg/kg gilteritinib + 85 mg/kg ABT-199 cohorts [p.o.; 3% ethanol (200 proof), 1% Tween-80 (polyoxyethylene (20) sorbitan monooleate) and sterile water; all USP grade; v/v]. For combined treatment, gilteritinib was administered first, followed by venetoclax within 1 h. Mice were treated daily for 27 days (treatment was stopped when one mouse in the control group showed leukemic symptoms). Mice were assessed 2x/day minimum for the duration of the study and body weights recorded daily. Mice were humanely euthanized when they presented with: >20% weight loss, decreased mobility limiting food and water access, lymph node metastases, progressive anemia, or lateral recumbency. % increase in lifespan (%ILS) was calculated: % ILS = [T-C/C] x 100 where “T” = treated and “C” = control median day of death. In vivo experiments were approved by the Institutional Animal Care and Use Committee at Wayne State University.

Midostaurin prevents venetoclax-induced upregulation of p-ERK

To begin to examine the mechanism of action in FLT-ITD AML, we assessed protein levels across the MAPK/ERK, PI3K/AKT, and JAK/STAT pathways via western blot in two FLT3-ITD positive AML cell lines and a primary patient sample 24 h post drug treatment. Midostaurin treatment resulted in downregulation of p-S6 and p-STAT5, which was preserved in cells treated with midostaurin in combination with venetoclax in all three samples tested (Figure 2A). Downregulation of p-AKT (both S473 and T308) was detected after venetoclax treatment and was further decreased in the combination treatment; total AKT levels decreased to a similar extent as p-AKT, while the effect of midostaurin was cell-specific (Figure 2A). The level of p-ERK decreased following midostaurin treatment in the primary AML patient sample, however it remained unchanged in MV4–11 cells and increased in MOLM-13 cells. Interestingly, venetoclax treatment also resulted in substantial increase of p-ERK in all three samples, which was abrogated by combination with midostaurin (Figure 2A). To better understand this phenomenon, we conducted a time-course analysis in both MOLM-13 and MV4–11 cell lines. Increased p-ERK was detected within 12–24 hours of venetoclax exposure. Midostaurin treatment downregulated p-ERK as early as 4 hours, unlike at the 24 h endpoint (Figure 2B&C). In the combined treatment, p-ERK levels were nearly undetectable after just 4 hours of treatment and remained below baseline up to 24 h during treatment. Next, we performed a time-course annexin V-FITC/PI flow cytometry experiment and found that the combination induced apoptosis as early as 4 hours (Figure 2D), which coincided with the early enhanced downregulation of p-ERK.

Open in a separate window

Figure 2.

Combination of midostaurin and venetoclax mitigates venetoclax-induced upregulation of p-ERK and prevents rebound of p-ERK during midostaurin treatment.

(A-C) FLT3-ITD AML cell lines MOLM-13, MV4–11, and primary patient sample AML#171 were treated with midostaurin alone or in combination with venetoclax for up to 24 hours. Representative Western blots generated using whole cell lysates are shown. Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control. (D) MOLM-13 and MV4–11 cell lines were treated for 4–24 hours with vehicle control, midostaurin, venetoclax, or in combination. Annexin V-FITC/PI staining and flow cytometry results are shown. ***indicates p<0.001. (E) MOLM-13, MV4–11, and primary patient sample AML#180 were treated with variable concentrations of venetoclax. Representative Western blots are shown. Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control. (F-G) AML cells were treated with venetoclax alone or in combination with the ERK- selective inhibitor SCH772984 for 24 hours. Flow cytometry and Western blot analyses are shown. ***indicates p<0.001. Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control. (H-I) AML cells were treated with midostaurin alone or in combination with the ERK selective inhibitor SCH772984 for 24 hours. Flow cytometry and Western blot analyses are shown. ***indicates p<0.001. Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control.

To begin to determine if p-ERK is a mechanism of resistance to venetoclax, we treated AML cells with variable concentrations of venetoclax and found that there was a concentration-dependent increase of p-ERK (Figure 2E). Further, treatment with the ERK-selective inhibitor SCH772984 completely abolished venetoclax-induced expression of p-ERK and significantly enhanced venetoclax-induced apoptosis (Figure 2F&G). Taken together, these results suggest that induction of p-ERK is a mechanism of resistance to venetoclax.

To assess the role of p-ERK in the antileukemic activity of midostaurin, we treated AML cells with midostaurin alone or in combination with SCH772984 and found that SCH772984 significantly enhanced cell death induced by midostaurin (Figure 2H). Western blots confirmed downregulation of p-ERK by SCH772984 alone and reduction of midostaurin-induced upregulation of p-ERK (Figure 2I). These results suggest that upregulation of p-ERK is also a mechanism of resistance to midostaurin.

Midostaurin downregulates Mcl-1 expression

Next, we examined the expression of the Bcl-2 family of proteins across two AML cell lines and a primary AML patient sample since their expression can be modulated by the FLT3 downstream signaling pathways. We identified consistent downregulation of Mcl-1 across the samples when treated with midostaurin alone and in combination with venetoclax for 24 hours (Figure 3A). Bcl-2, Bcl-xL, and Bim protein levels were not notably affected by either agent, alone or in combination. While reduced levels of Bak and Bax were detected, the decrease was not consistent across all three samples. Time-course evaluation showed that Mcl-1 downregulation occurs within 4 hours of exposure to midostaurin in both MOLM-13 and MV4–11 (Figure 3B); these results relate to the previously noted flow cytometric time-point analysis of enhanced cell death at 4 hours (Figure 2D). Co-immunoprecipitation revealed that midostaurin treatment reduced Mcl-1 binding to Bim, but increased Bim binding to Bcl-2, in both MOLM-13 and MV4–11 cells. Consistent with our published work (10, 11, 28), venetoclax treatment reduced Bcl-2 binding to Bim, but increased binding of Mcl-1 to Bim (Figure 3C). In the combined treatment, both Mcl-1 and Bcl-2 binding to Bim were reduced. To further confirm the role Mcl-1 plays in the combined treatment, we overexpressed Mcl-1 in MV4–11 cells. Its overexpression significantly reduced apoptosis induced by venetoclax alone and in combination with midostaurin (Figure 3D), suggesting that Mcl-1 plays an important role in the mechanism of action of combined midostaurin and venetoclax. To determine if the combination induces intrinsic apoptosis, Bak and Bax shRNA knockdown were performed. Knockdown of Bak and Bax significantly reduced annexin V positivity (Figure 3E). While the reduction of annexin V positive cells was only partial, potentially due to the redundancy of Bak and Bax, these results show that the combination induces apoptosis, at least partially through the intrinsic apoptosis pathway.

Open in a separate window

Figure 3.

Midostaurin downregulates Mcl-1, enhancing venetoclax-induced apoptosis.

(A) Western blot analysis of members of the Bcl-2 protein family following 24 hour treatment with midostaurin alone and in combination with venetoclax in FLT3-ITD AML cell lines MOLM-13 and MV4–11 and primary patient sample AML#171. Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control. (B) Western blot analyses of MOLM-13 and MV4–11 cells treated with midostaurin alone and in combination for up to 24 hours are shown. Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control. (C) MOLM-13 and MV4–11 cells were treated with the indicated drugs for 4 hours. Bim was immunoprecipitated and subjected to Western blot analysis. *indicates light chain of the Bim antibody. (D&E) Lentiviral overexpression (OE) of red fluorescent protein (RFP) control and Mcl-1 (panel D) and shRNA knockdown of Bak and Bax (panel E) were performed as described in the ‘Methods’ section. Whole cell lysates were subjected to Western blot (upper panel). Densitometry measurements, shown below the corresponding blot, were normalized to β-actin and expressed as fold change compared to vehicle control. RFP and Mcl-1 overexpressing and Bak and Bax knockdown cells were treated with midostaurin and venetoclax, alone and in combination, for 4 hours. Flow cytometry analyses results are shown (lower panel). ***indicates p<0.001.

To begin to determine how midostaurin treatment reduces Mcl-1 protein level, we first performed real-time RT-PCR. Midostaurin treatment significantly reduced Mcl-1 transcript levels 4 hours after drug treatment. Reduced levels were maintained in the combined drug-treated cells (Figure 4A), indicating that midostaurin transcriptionally downregulates Mcl-1. To determine if midostaurin treatment affects Mcl-1 protein stability, we treated MOLM-13 and MV4–11 cells with midostaurin and venetoclax, alone or in combination, for 4 hours. Cycloheximide treatment (10 μg/mL) revealed that Mcl-1 levels decreased significantly faster in MV4–11 cells treated with midostaurin alone and in combination with venetoclax (p<0.01; Figure 4B&C), while we did not detect a significant change in MOLM-13 cells. Since phosphorylation of Mcl-1 at T163 has been shown to stabilize Mcl-1 by prolonging its half-life (29), we looked at phosphorylation of Mcl-1 post-midostaurin treatment. Consistent with our cycloheximide results, we detected significantly reduced phosphorylation of Mcl-1 at T163 in MV4–11 cells treated with midostaurin alone and in combination with venetoclax, while we did not detect obvious change for MOLM-13 cells (Figure 4D&E). Taken together, these results show that midostaurin treatment reduces Mcl-1 transcript levels and decreases expression of Mcl-1 protein. While midostaurin can affect Mcl-1 stability, it may be a cell line-specific response. Nonetheless, downregulation of Mcl-1 transcript levels occurred in both cell lines, suggesting that downregulation of Mcl-1 by midostaurin is predominantly due to alteration of transcripts.

Open in a separate window

Figure 4.

Midostaurin downregulates Mcl-1 expression potentially through transcriptional mechanism.

(A) MOLM-13 and MV4–11 cells were treated with midostaurin and venetoclax for 4 hours. Total RNA was isolated and Mcl-1 transcripts were determined by real-time RT-PCR using TaqMan probes. *indicates p<0.05, **indicates p<0.01, and ***indicates p<0.001. (B-C) MOLM-13 and MV4–11 cells were treated with midostaurin and/or venetoclax for 4 h, washed, and then treated with cycloheximide for up to 2 hours. Whole cell lysates were subjected to Western blotting. The fold changes for the Mcl-1 densitometry measurements, normalized to β-actin and then compared with no drug treatment control, are shown. Representative blots are shown on the left side, while densitometry measurements are graphed on the right. **indicates p<0.01 and ***indicates p<0.001 (D-E) MOLM-13 and MV4–11 cells were treated with midostaurin and venetoclax for 4 h. Western blot analyses of p-Mcl-1 and total Mcl-1 are shown. Representative blots are shown in the upper panels and the quantification results are graphed in the lower panels. **indicates p<0.01 and ***indicates p<0.001

Gilteritinib enhances the antileukemic effects of venetoclax in vitro and in vivo

As midostaurin is a first generation FLT3 inhibitor and displays low selectivity, we sought to confirm that the efficacy and mechanism demonstrated were indeed due to FLT3-inhibition and not due to off-target effects. We therefore used the second-generation FLT3-inhibtor gilteritinib. Consistent with midostaurin, gilteritinib synergized with venetoclax to induce annexin V positivity in MOLM-13 and MV4–11 cells (Figure 5A) accompanied by cleavage of caspase 3 (Figure 5B), demonstrating that the two agents synergize in inducing apoptosis in FLT3-ITD AML cells. Synergy was confirmed in 8 primary FLT3-ITD AML patient samples (Figure 5C). Similar to midostaurin, the combination of venetoclax and gilteritinib also demonstrated synergy against FLT3-wt cell lines and patient samples (Figure 5D), though the magnitude of cell death was substantially less than in the FLT3-ITD samples. The combination treatment induced apoptosis in normal PBMCs, though the magnitude was minimal compared to vehicle control treated cells (PBMC#1: 4.4% vs 7.4% and PBMC#2: 2.2% vs 8.8%; Figure 5E).

Open in a separate window

Figure 5.

Gilteritinib enhances venetoclax activity both in vitro and in vivo.

(A) MOLM-13 and MV4–11 cells were treated with gilteritinib and venetoclax, alone or in combination, for 24 hours. Annexin V-FITC/PI staining and flow cytometry analyses results are shown. ***indicates p<0.001. (B) MOLM-13 and MV4–11 cells were treated as indicated for 24 hours. Whole cell lysates were subjected to Western blotting. (C-E) Primary AML patient samples (panels C&D) and normal peripheral blood mononuclear cells (PBMC; panel E) were treated with gilteritinib and venetoclax, alone or in combination, for 24 hours. Annexin V-FITC/PI staining and flow cytometry analyses results are shown. ***indicates p<0.001. (F-H) MV4–11 cells (1 x 10⁶ cells/mouse) were injected through the tail vein of immunocompromised NSGS mice. Four days post cell injection the mice were randomized (5 mice/group) and treated daily with vehicle control (3% 200 proof ethanol, 1% polyoxyethylene (20) sorbitan monooleate, and USP water), 40 mg/kg gilteritinib (p.o.), 85 mg/kg venetoclax (p.o.), or 40 mg/kg gilteritinib + 85 mg/kg venetoclax (p.o.) for 27 consecutive days (treatments were stopped when one mouse in the vehicle control group showed leukemic symptoms; panel F). Body weights were measured on a daily basis and are graphed as mean ± SEM (panel G). Overall survival probability, estimated with the Kaplan-Meier method, is shown (panel H).

In vivo efficacy potential of gilteritinib in combination with venetoclax was evaluated in an early stage MV4–11-derived xenograft mouse model. We first conducted a toxicity/preliminary efficacy study with 2 mice/group. The mice were treated with vehicle control, 40 mg/kg gilteritinib, 85 mg/kg venetoclax, or gilteritinib + venetoclax on a daily basis for 13 days (Figure S1A). Moderate body weight loss (3% or less) was observed during treatment, though this was completely reversible (Figure S1B). Median survival for vehicle control and venetoclax treated mice was 41.5 days and 43.5 days, respectively, while gilteritinib alone was 61.5 days and the combination was 95.5 days, an ILS of 130% was achieved (Figure S1C). Based on these encouraging results, we conducted a full trial using 5 mice/group and extending treatment to 27 consecutive days (Figure 5F). Moderate body weight loss (6% or less) was observed during treatment, though it was completely reversible (Figure 5G). One mouse from both the venetoclax group and the combination group was excluded due to technical issues. Median survival for vehicle control, venetoclax, and gilteritinib treated mice was 37 days, 41 days, and 91 days, respectively (Figure 5H). On day 190, the one remaining gilteritinib treated mouse and the four combination treated mice were asymptomatic. At this point, we rechallenged the mice with tumor to rule-out a leaky system. Two weeks after implanting tumors, the tumors were significantly larger, demonstrating the survival of these mice was not due to regain of immunity.

This study was supported by Jilin University, Changchun, China, the Barbara Ann Karmanos Cancer Institute, Wayne State University School of Medicine, and by grants from the National Natural Science Foundation of China, NSFC 31671438 and NSFC 31471295, Graduate Innovation Fund of Jilin University, Hyundai Hope on Wheels, LaFontaine Family/U Can-Cer Vive Foundation, Kids Without Cancer, Children’s Hospital of Michigan Foundation, Decerchio/Guisewite Family, Justin’s Gift, Elana Fund, Ginopolis/Karmanos Endowment and the Ring Screw Textron Endowed Chair for Pediatric Cancer Research. The Animal Model and Therapeutics Evaluation Core and Genomics Core are supported, in part, by NIH Center Grant P30 CA022453 to the Karmanos Cancer Institute at Wayne State University. The funders had no role in study design, data collection, analysis and interpretation of data, decision to publish, or preparation of the manuscript.