Targeting PFKFB3 alleviates cerebral ischemia-reperfusion injury in mice

Abstract

The glycolytic rate in neurons is low in order to allow glucose to be metabolized through the pentose-phosphate pathway (PPP), which regenerates NADPH to preserve the glutathione redox status and survival. This is controlled by 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase-3 (PFKFB3), the pro-glycolytic enzyme that forms fructose-2,6-bisphosphate, a powerful allosteric activator of 6-phosphofructo-1-kinase. In neurons, PFKFB3 protein is physiologically inactive due to its proteasomal degradation. However, upon an excitotoxic stimuli, PFKFB3 becomes stabilized to activate glycolysis, thus hampering PPP mediated protection of redox status leading to neurodegeneration. Here, we show that selective inhibition of PFKFB3 activity by the small molecule AZ67 prevents the NADPH oxidation, redox stress and apoptotic cell death caused by the activation of glycolysis triggered upon excitotoxic and oxygen-glucose deprivation/reoxygenation models in mouse primary neurons. Furthermore, in vivo administration of AZ67 to mice significantly alleviated the motor discoordination and brain infarct injury in the middle carotid artery occlusion ischemia/reperfusion model. These results show that pharmacological inhibition of PFKFB3 is a suitable neuroprotective therapeutic strategy in excitotoxic-related disorders such as stroke.

Figure 1

In vitro characterization of two PFKFB3 inhibitors. (a) Incubation of A549 cells with AZ67 (1 h) dose-dependently decreases F2,6BP concentration. (b) Incubation of A549 cells with PFK158 (1 h) dose-dependently decreases F2,6BP concentration. (c) Incubation of human recombinant PFKFB3 with AZ67 dose-dependently inhibits PFKFB3 kinase activity. (d) Incubation of human recombinant PFKFB3 with PFK158 does not inhibit PFKFB3 kinase activity. See also Supplementary Data 1.

Figure 2

AZ67 prevents glycolytic activation, redox stress and neuronal death upon excitotoxic stimuli in primary neurons. (a) Western immunoblotting showing that the treatment of neurons with glutamate (100 µM; 10 mins), followed by washout, stabilized PFKFB3 protein levels after 24 h of incubation. Incubation of neurons with AZ67 (10 nM) for 24 h, after glutamate was removed, did not alter PFKFB3 protein levels. (b) Incubation of neurons with AZ67 for 24 h revealed no effect on lactate release. Treatment of neurons with NMDA (100 µM; 10 mins) or glutamate (100 µM; 10 mins), followed by washout, increased lactate released after 24 h of incubation (compare NMDA or glutamate versus none values at 0 nM AZ67). Incubation of neurons with AZ67 for 24 h, after NMDA or glutamate was removed, dose-dependently prevented the increase in lactate release. (c) Incubation of neurons with AZ67 for 24 h revealed no effect on F2,6BP concentrations. Treatment of neurons with NMDA (100 µM; 10 mins) or glutamate (100 µM; 10 mins), followed by washout, increased F2,6BP after 24 h of incubation. Incubation of neurons with AZ67 (10 nM) for 24 h, after NMDA or glutamate was removed, prevented the increase in lactate release. (d) Incubation of neurons with AZ67 for 24 h revealed no significant effect on the NADPH/NADP ratio. Treatment of neurons with NMDA (100 µM; 10 mins) or glutamate (100 µM; 10 mins), followed by washout, decreased the NADPH/NADP ratio after 24 h of incubation. Incubation of neurons with AZ67 (10 nM) for 24 h, after NMDA or glutamate was removed, prevented the decreased NADPH/NADP ratio. (e) Incubation of neurons with AZ67 for 24 h revealed no effect on mitochondrial ROS. Treatment of neurons with NMDA (100 µM; 10 mins) or glutamate (100 µM; 10 mins), followed by washout, increased mitochondrial ROS after 24 h of incubation (compare NMDA or glutamate versus none values at 0 nM AZ67). Incubation of neurons with AZ67 for 24 h, after NMDA or glutamate was removed, dose-dependently prevented the increase in mitochondrial ROS. (f) Treatment of neurons with NMDA (100 µM; 10 mins) or glutamate (100 µM; 10 mins), followed by washout, triggered apoptotic death after 24 h of incubation (compare NMDA or glutamate versus none values at 0 nM AZ67). Incubation of neurons with AZ67 for 24 h, after NMDA or glutamate was removed, dose-dependently prevented apoptotic death up to 10 nM, AZ67; at concentrations of 100 nM to 10 µM, AZ67 showed progressive loss of protection. (g) Transfection of primary neurons with Pfk1-M (4 µg of DNA plasmid) efficiently increased PFK1-M protein levels. (h) Treatment of neurons with glutamate (100 µM; 10 mins), followed by washout, triggered apoptotic death after 24 h of incubation. Incubation of neurons with AZ67 (10 nM) for 24 h, after glutamate was removed, prevented apoptotic death. However, AZ67-mediated protection of apoptotic death was abolished when neurons were previously transfected with the full-length cDNA coding for PFK1-M. Apoptosis was analysed only in the efficiently-transfected, GFP⁺ neurons. (i) Treatment of neurons with NMDA (100 µM; 10 mins), followed by washout, triggered apoptotic death after 24 h of incubation. Incubation of neurons with AZ67 (10 nM) for 24 h, after NMDA was removed, prevented apoptotic death. However, AZ67-mediated protection of apoptotic death was abolished when neurons were previously transfected with the full-length cDNA coding for PFK1-M. Apoptosis was analysed only in the efficiently-transfected, GFP⁺ neurons. In all cases, data are mean ± S.E.M. values for n = 3 independent culture preparations. #p < 0.05 versus none at 0 nM AZ67; *p < 0.05 versus the corresponding treatment at 0 nM AZ67 (ANOVA followed by the least significant difference multiple range test). See also Supplementary Data 1 and Statistics Table 1.

Figure 3

AZ67 prevents the metabolic switch from PPP to glycolysis, redox and mitochondrial stress, and apoptosis in an in vitro model of ischemia/reperfusion in primary neurons. (a) Western immunoblotting showing that the treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h), stabilized PFKFB3 protein levels. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation did not alter PFKFB3 protein levels. (b) Incubation of neurons with AZ67 for 4 h revealed no effect on lactate release. Treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) increased lactate released. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation prevented the increase in lactate release. (c) Incubation of neurons with AZ67 for 4 h revealed no effect on the glycolytic flux, as assessed by the formation of ³H₂O from [3-³H]glucose. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) increased the glycolytic flux. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation prevented the increase in the glycolytic flux. (d) Incubation of neurons with AZ67 for 4 h revealed no effect on the pentose-phosphate pathway (PPP) flux, as assessed by the difference in the formation of ¹⁴CO₂ from [1-¹⁴C]- and from [6-¹⁴C]glucose. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) decreased the PPP flux. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation prevented the decrease in the PPP flux. (e) Incubation of neurons with AZ67 for 4 h revealed no effect on H₂O₂ release, as assessed by the fluorescence of AmplexRed. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) increased H₂O₂ release. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation prevented the increase in H₂O₂ release. (f) Incubation of neurons with AZ67 for 4 h revealed no effect on mitochondrial ROS formation, as assessed by MitoSox fluorescence by flow cytometry. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) increased mitochondrial ROS. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation significantly prevented the increase in mitochondrial ROS formation. (g) Incubation of neurons with AZ67 for 4 h revealed no effect on pyruvate dehydrogenase (PDH) activity, as assessed by the conversion of [1-¹⁴C]pyruvate in ¹⁴CO₂. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) decreased PDH activity that was not altered by incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation. (h) Incubation of neurons with AZ67 for 4 h revealed no effect on mitochondrial membrane potential (∆ψ_m), as assessed by flow cytometry. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) decreased ∆ψ_m. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation prevented the decreased ∆ψ_m. (i) Incubation of neurons with AZ67 for 4 h revealed no effect on caspase-3 activity, a measure of apoptosis. However, treatment of neurons with OGD (3 h) followed by reoxygenation (plus glucose) (4 h) increased apoptosis. Incubation of neurons with AZ67 (10 nM) during the 4 h of reoxygenation prevented the increase in apoptosis. In all cases, data are mean ± S.E.M. values for n = 3 independent culture preparations. #p < 0.05 versus OGD at 0 nM AZ67; *p < 0.05 versus the corresponding normoxic condition (ANOVA followed by the least significant difference multiple range test). See also Supplementary Data 1 and Statistics Table 1.

Figure 4

In vivo AZ67 administration protects mice against neurological impairment, motor discoordination and brain injury caused by a brain ischemia/reperfusion model. (a) Motor coordination, analysed 24 h after a transient MCAO episode in mice, revealed a ~40% performance (rotarod) when compared with the sham-operated animals (100% performance). This effect was significantly prevented by the intravenous administration of AZ67 (60 mg/kg of body weight) immediately after the ischemic episode (~60% performance). (b) Neurological Severity Score (NNS), examined 24 h after a transient MCAO episode in mice, revealed severe neurological deficit according to the Bederson test when compared with the sham-operated animals. This effect was significantly prevented by the intravenous administration of AZ67 (60 mg/kg of body weight) immediately after the ischemic episode. (NNS = 0 both for vehicle or AZ67 in the sham-operated animals). (c) Infarcted brain volume, analysed 24 h after a transient MCAO episode in mice, was ~43% of the brain of the sham-operated animals. This effect was significantly prevented by the intravenous administration of AZ67 (60 mg/kg of body weight) immediately after the ischemic episode (~27% of infarcted volume). Left panel shows pictures of the brain sections of a representative animal for each experimental group. Bar, 1 cm. In all cases, data are mean ± S.E.M. values for 8 male mice. *p < 0.05 (ANOVA followed by the least significant difference multiple range test). See also Supplementary Data 1 and Statistics Table 1.