Metformin synergizes with gilteritinib in treating FLT3-mutated leukemia via targeting PLK1 signaling

Abstract

Fms-like tyrosine kinase 3 (FLT3) mutations, present in over 30% of acute myeloid leukemia (AML) cases and dominated by FLT3-internal tandem duplication (FLT3-ITD), are associated with poor outcomes in patients with AML. While tyrosine kinase inhibitors (TKIs; e.g., gilteritinib) are effective, they face challenges such as drug resistance, relapse, and high costs. Here, we report that metformin, a cheap, safe, and widely used anti-diabetic agent, exhibits a striking synergistic effect with gilteritinib in treating FLT3-ITD AML. Metformin significantly sensitizes FLT3-ITD AML cells (including TKI-resistant ones) to gilteritinib. Metformin plus gilteritinib (low dose) dramatically suppresses leukemia progression and prolongs survival in FLT3-ITD AML mouse models. Mechanistically, the combinational treatment cooperatively suppresses polo-like kinase 1 (PLK1) expression and phosphorylation of FLT3/STAT5/ERK/mTOR. Clinical analysis also shows improved survival rates in patients with FLT3-ITD AML taking metformin. Thus, the metformin/gilteritinib combination represents a promising and cost-effective treatment for patients with FLT3-mutated AML, particularly for those with low income/affordability.

Keywords: AML; FLT3-ITD; FLT3-mutated; MAPK/mTOR; PLK1; TKI-resistant; combinational therapy; gilteritinib; metformin; synergy.

Graphical abstract

Figure 1

Metformin sensitizes FLT3-ITD AML cell lines to gilteritinib (A and B) IC₅₀ values (48 h) of metformin (A) and gilteritinib (B) across three types of AML cell lines (FLT3-ITD: MOLM13, MV4-11, and MA9.3-ITD; FLT3 TKI resistant: MOLM13-RES; and FLT3-WT: THP1, NOMO-1, and Kasumi-1). Mean value of 3 technical replicates from a representative biological experiment of n = 3 biological replicates. (C–F) The visual diagrams sourced from the CompuSyn report highlight the effects of metformin, gilteritinib, and the combinations (Combo) after 48 h treatment. The concentration-response graph illustrates the correlation between fraction affected (Fa) and concentration in MOLM-13, MV4-11, MOLM13-RES, and THP-1 cells (top of C–F). The combination index (CI) values <1 indicate synergistic interactions. Mean value of 3 technical replicates from a representative biological experiment of n = 3 biological replicates. (G–J) Inhibition rates of different concentrations of gilteritinib before and after combinations with metformin at IC₁₀ value (G–J, left) and the Bliss index values (G–J, right) were shown. Value >1 indicates synergy. n = 3 technical replicates, representative of n = 3 biological replicates. (K) The concentration-response matrix (left) of metformin and gilteritinib at varying concentrations and the Bliss synergy score (determined by SynergyFinder 3.0, middle and right) in MOLM13-RES cells. n = 3 technical replicates, representative data of n = 3 biological replicates. Data are shown as means ± SEM and assessed by 2-tailed Student’s t test (G–J). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ns, not significant. MET, metformin; GIL, gilteritinib; Combo, the combination of metformin and gilteritinib. See also Figure S1.

Figure 2

Gilteritinib and metformin inhibit cell growth/viability and induce cell-cycle arrest and apoptosis in FLT3-ITD cell lines AML cells were treated with gilteritinib and metformin each alone or in combinations (Combo1: gilteritinib 5 nM + metformin 5 mM; Combo2: gilteritinib 10 nM + metformin 10 mM in MOLM13 cells; Combo1: gilteritinib 80 nM + metformin 4 mM; Combo2: gilteritinib 240 nM + metformin 10 mM in MOLM13-RES cells) for 24–48 h and then used in different experiments. (A) MTT assays showing the effects of the treatment of metformin and/or gilteritinib cell growth/viability of MOLM13, MOLM13-RES, and THP-1 cells. n = 3 technical replicates, representative of n = 3 biological replicates. (B and C) Representative flow cytometry plots (left) and the statistical analysis (right) of cell cycle in MOLM13 (B) and MOLM13-RES (C) cells 48 h post-treatment with control (DMSO) or indicated drugs. Numbers represent percentage values. n = 3 biological replicates. (D and E) Representative flow cytometry plots (left) and the statistical analysis (right) of apoptosis in MOLM13 (D) and MOLM13-RES (E) cells 48 h post-treatment with control (DMSO) or indicated drugs. Numbers represent percentage values. n = 3 biological replicates. Data are shown as means ± SD and assessed by two-way ANOVA (A–C) or one-way ANOVA (D and E). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ns, not significant. MET, metformin; GIL, gilteritinib. See also Figure S2.

Figure 3

Combinational treatment of metformin and gilteritinib strikingly suppresses human FLT3-ITD AML progression in vivo (A) Schematic overview of the design and procedures for the FLT3-ITD AML animal models. (B) Bioluminescence imaging of leukemia burden in the vehicle control and treated groups from the ventral side in NSG (MOLM13) mice. Mice were treated with vehicle, metformin (250 mg/kg/day, intraperitoneal injection), gilteritinib (7.5 mg/kg/day, oral gavage), or Combo (metformin 250 mg/kg/day + gilteritinib 7.5 mg/kg/day) once daily after 7 days post-BMT. (C and D) Representative flow cytometry plots (top) and the statistical analysis (bottom) of the engraftment in bone marrow (BM), peripheral blood (PB), and spleen (SP) (C) and statistical analysis of spleen weight (D) in the vehicle control (n = 3) and treated groups (n = 3) on day 21 post-BMT in NSG (MOLM13) mice. (E and F) White blood cell (WBC) counts (E) and body weight (F) from 0 to 18 days post-BMT in NSG (MOLM13) mice. (G) Kaplan-Meier survival curves for NSG mice xenotransplanted with human MOLM13 AML cells. (H) Bioluminescence imaging of leukemia burden in the vehicle control and treated groups from the ventral side in NRGS (MOLM13-RES) mice. Mice were treated with vehicle, metformin (250 mg/kg/day, intraperitoneal injection), gilteritinib (15 mg/kg/day, oral gavage), or Combo (metformin 250 mg/kg/day + gilteritinib 15 mg/kg/day) once daily after 10 days post-BMT. (I and J) Representative flow cytometry plots (top) and the statistical analysis (bottom) of the engraftment in BM, PB, and SP (I) and statistical analysis of spleen weight (J) in vehicle (n = 5) and treated groups (n = 3) on day 26 post-BMT in NRGS (MOLM13-RES) mice. (K and L) WBC counts (K) and body weight (L) from 0 to 21 days post-BMT in NRGS (MOLM13-RES) mice. (M) Kaplan-Meier survival curves for NRGS mice xenotransplanted with human MOLM13-RES AML cells. Data are shown as mean ± SEM and assessed by 2-tailed Student’s t test (C and D, I and J) or two-way ANOVA (E and F, K and L). Log rank test was used for survival analysis (G and M). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. MET, metformin; GIL, gilteritinib. See also Figure S3.

Figure 4

Metformin synergizes with gilteritinib in treating FLT3-ITD AML by co-targeting PLK1 (A) Venn diagram showing the overlapping of the hallmark pathways (left and middle) and genes (right) down-regulated after 24 h of treatment with gilteritinib (240 nM), metformin (10 mM), or Combo1 (gilteritinib: 80 nM + metformin: 4 mM), relative to the control (DMSO) group in MOLM13-RES cells. (B) Volcano plot displaying genes upregulated or down-regulated in untreated MOLM13-RES cells compared to MOLM13 cells, based on the analysis of datasets: GSE180180, GSE180181 (left), and Venn diagram showing the overlap of genes upregulated in untreated MOLM13-RES cells relative to MOLM13 cells and those that are consistently down-regulated in all three groups of drug-treated samples (i.e., gilteritinib: 240 nM; metformin: 10 mM; and Combo1: gilteritinib 80 nM + metformin 4 mM) and shared by the three commonly suppressed pathways (right). (C) Quantitative reverse-transcription PCR (qRT-PCR) validation of PLK1 mRNA abundance in MOLM13 cells, MOLM13-RES cells with and without gilteritinib, metformin, or Combo treatment. n = 3 technical replicates, representative of n = 2 biological replicates. (D) qRT-PCR validation (2ˆ-ΔCT) of PLK1 mRNA abundance in PB samples from MOLM13-RES xenotransplanted NRGS mice. 5 samples from each group, n = 3 technical replicates. (E) Western blot analysis of PLK1 expression level changes in MOLM13-RES cells 24 h post-treatment with different drug concentrations. Representative data of n = 2 biological replicates. (F) Western blot showing the protein abundance of PLK1 in multiple subtypes of human AML cells compared to healthy BM controls. Representative data of n = 2 biological replicates. (G–I) IC₅₀ value changes of gilteritinib upon PLK1 knockdown in MOLM13 and MOLM13-RES cells (G) or after the overexpression of PLK1 in MOLM13 (H) or MOLM13-RES cells (I), compared to the control cells. Mean value of 3 technical replicates from a representative biological experiment of n = 3 biological replicates. (J) Representative western blot analysis of PLK1 expression levels (left) and cell growth/viability assays (right) at 24 h post-treatment with indicated drug concentrations in MOLM13-RES cells overexpressing PLK1 or empty vector (EV). n = 3 biological replicates. (K) Representative flow cytometry plots (left) and the statistical analysis (right) of apoptosis at 48 h post-treatment with indicated drug concentrations in MOLM13-RES cells overexpressing PLK1 or EV. Numbers represent percentage values. n = 3 biological replicates. Data are shown as mean ± SD and assessed by 2-tailed Student’s t test (C, D, G, H, I, and K) or two-way ANOVA (J). ∗p < 0.05, ∗∗∗p < 0.001, and ns, not significant. MET, metformin; GIL, gilteritinib; EV, empty vector; OE, overexpression. See also Figure S4.

Figure 5

Metformin and gilteritinib co-target the PLK1 signaling in FLT3-ITD AML cells (A) Western blot analysis of PLK1 and downstream signaling protein phosphorylation levels post 24-h treatment with gilteritinib (5 nM), metformin (5 mM), gilteritinib (10 nM), metformin (10 mM), and combinations (Combo1: GIL 5 nM + MET 5 mM; Combo2: GIL 10 nM + MET 10 mM) in MOLM13 cells. Representative data of n = 2 biological replicates. (B) Western blot analysis of PLK1 and downstream signaling protein phosphorylation levels after 24-h treatment with single gilteritinib (80 nM), metformin (4 mM), gilteritinib (240 nM), metformin (10 mM), or combinations (Combo1: GIL 80 nM + MET 4 mM; Combo2: GIL 240 nM + MET 10 mM) in MOLM13-RES cells. Representative data of n = 2 biological replicates. (C) Western blot showing PLK1 and downstream signaling protein phosphorylation level changes upon PLK1 knockdown or overexpression in MOLM13-RES cells. Representative data of n = 2 biological replicates. (D) Western blotting of PLK1 and the downstream signaling protein phosphorylation level changes after 24-h treatment with specified drug concentrations in MOLM13-RES cells overexpressing PLK1 or EV. Representative data of n = 2 biological replicates. (E) Scheme of the PLK1 signaling-mediated mechanisms underlying gilteritinib and metformin treatment in FLT3-ITD AML. MET, metformin; GIL, gilteritinib; EV, empty vector; OE, overexpression.

Figure 6

Metformin improves the OS of patients with FLT3-ITD AML (A) Survival analysis of patients with concurrent FLT3-ITD AML and diabetes. (B and C) Summary of treatment outcomes (B) and characteristics (C) for samples presented in A. (D) qRT-PCR validation of PLK1 expression in clinical samples. Data are shown as mean ± SD and assessed by 2-tailed Student’s t test (D). Log rank test was used for the survival analysis (A). ∗∗p < 0.01, ∗∗∗p < 0.001. See also Figure S5.