Metformin impairs glucose consumption and survival in Calu-1 cells by direct inhibition of hexokinase-II

Abstract

The anti-hyperglycaemic drug metformin has important anticancer properties as shown by the direct inhibition of cancer cells proliferation. Tumor cells avidly use glucose as a source for energy production and cell building blocks. Critical to this phenotype is the production of glucose-6-phosphate (G6P), catalysed by hexokinases (HK) I and II, whose role in glucose retention and metabolism is highly advantageous for cell survival and proliferation. Here we show that metformin impairs the enzymatic function of HKI and II in Calu-1 cells. This inhibition virtually abolishes cell glucose uptake and phosphorylation as documented by the reduced entrapment of ¹⁸F-fluorodeoxyglucose. In-silico models indicate that this action is due to metformin capability to mimic G6P features by steadily binding its pocket in HKII. The impairment of this energy source results in mitochondrial depolarization and subsequent cell death. These results could represent a starting point to open effective strategies in cancer prevention and treatment.

Figure 1. Effect of metformin on Calu-1 cells glucose consumption and HKs activity.

(A) Cell uptake of FDG was expressed as percentage of total tracer availability according to different metformin concentrations and exposure times. 1 mM metformin did not produce any significant modification, while tracer uptake decreased after 24 hrs exposure to metformin 5 mM. Highest drug concentration (10 mM) caused a significant reduction and virtually abolished glucose consumption at 6 and 24 hr. p values are shown for each comparison that was performed by one way analysis of variance. (B) Calu-1 HKs activity (expressed as percentage of control) is represented as function of metformin concentrations. The reaction was switched on after 10 minutes of metformin pre-incubation with Calu-1 total cell lysate (Lysate) or plus ATP 0.8 mM (Lysate + ATP) or Glucose 100 mM (Lysate + Glu). The reaction was switched on by adding to the solution respectively ATP + Glu (Lysate), Glucose (Lysate + ATP) and ATP (Lysate + Glu). Pre-incubation with metformin and glucose (Lysate + Glu) caused an inhibition of the HK I and II enzymatic activity that was dependent upon metformin concentration. This finding was not observed when the enzymes were pre-exposed to metformin alone (Lysate) nor to metformin and ATP (Lysate + ATP). (C) Enzymatic activity (expressed as percentage of control) of human purified HK I, HK II and HK IV observed after pre-incubation with glucose and different metformin concentrations. The reaction was switched on after 10 minutes by adding to the solution 0.8 mM ATP. Metformin induced a dose-dependent inhibition of catalytic activity of HK I and HK II. By contrast, it did not affect enzymatic activity of HK IV. (D) Dose dependent interference of ATP on human purified HK II inhibition caused by metformin. Phosphorylation rate is expressed as percentage of HKs activity measured after ten minutes pre incubation with glucose (100 mM) and different metformin concentrations and starting the reaction with ATP (0.4–1.2 mM). ATP 0.8 mM, was considered as reference value. Starting the reaction with ATP concentrations ≥1.2 mM fully abolished metformin effect. On the contrary, ATP levels below 1.2 mM reduced the extent of metformin inhibitory action in a dose dependent fashion. * = p < 0.05; ** = p < 0.01:*** = p < 0.001. Error bars indicate standard error.

Figure 2. Molecular mechanism of HK II inhibition by metformin.

(A) Open conformation of HK II in complex with ATP highlighted with the red surface (Table S1 and Figure S2). (B) Closed conformation of HK II in complex with glucose and glucose-6-phosphate highlighted with green and orange surfaces, respectively (Table S2 and Figure S2). (C) Closed conformation of HK II in complex with metformin as suggested by Induced Fit Docking calculation and molecular dynamic simulations. (D) Closed conformation of HK II in complex with glucose, ATP and Mg²⁺ highlighted with green, red and blue surfaces, respectively. (E) Chemical structure of metformin (tautomeric and protonation forms are reported in Figure S3). (F) Chemical structure of glucose-6-phosphate. (G) Interactions map of the most stable binding mode of metformin in the closed HK II conformation. (H) Interactions map glucose-6-phosphate in the closed HK II conformation.

Figure 3. Metformin displaces HK II from Mitochondria.

(A) HK I (green color) and mitochondria (red color) co-localization under control condition and after treatment with 5 mM metformin for 24 hr. Images obtained by immunofluorescence labelling HK I and mitochondria (Mit) show that metformin did not affect HK I interaction with mitochondria. (B) Images display the corresponding finding for HK II staining. HK II binding to mitochondria is impaired by treatment with 5 mM metformin. (C, D) Quantitative measurement of HK I and HK II and mitochondria co-localization under control condition and after metformin treatment. Metformin does not affect the binding of HK I to mitochondria (78% ± 4.7 vs 80% ± 5.2, p = ns) while it decreases the binding of HK II to mitochondria (43% ± 3.8 vs 66% ± 5, p < 0.05). * = p < 0.05. Error bars indicate standard error.

Figure 4. Metformin treatment impairs mitochondrial function inducing cell death.

(A) Calu-1 cells were treated for 1 hr with different doses of metformin; ATP and AMP concentrations (μMol/mg of proteins) were determined by enzymatic assay. Measured AMP/ATP ratio increased as a function of metformin concentration. * = p < 0.05, ** = p < 0.01. Error bars indicate standard error. (B) 10 mM metformin exposure of Calu-1 cells for 24 hr affect mitochondrial potential (ΔΨm). Upper pannels: Bivariate JC-1 analysis of mitochondrial membrane potential by flow cytometry. Botton pannels: Immunofluorescence analysis of Calu-1 cells treated with JC-1.Metformin treatment induced an increase of depolarized regions as indicated by the disappearance of red and increase of green fluorescence. (C) Apoptosis was evaluated after treating Calu-1 cells with metformin and staining with Annexin-V and propidium iodide (PI) at 24 hr. The values indicate the percentage of death (Annexin-V−, PI+), early apoptotic (Annexin-V+, PI−) and necrotic/advanced apoptotic stage (Annexin-V+, PI+) cells.