AMP-activated protein kinase inhibits Kv 1.5 channel currents of pulmonary arterial myocytes in response to hypoxia and inhibition of mitochondrial oxidative phosphorylation

Abstract

Key points: Progression of hypoxic pulmonary hypertension is thought to be due, in part, to suppression of voltage-gated potassium channels (Kv ) in pulmonary arterial smooth muscle by hypoxia, although the precise molecular mechanisms have been unclear. AMP-activated protein kinase (AMPK) has been proposed to couple inhibition of mitochondrial metabolism by hypoxia to acute hypoxic pulmonary vasoconstriction and progression of pulmonary hypertension. Inhibition of complex I of the mitochondrial electron transport chain activated AMPK and inhibited Kv 1.5 channels in pulmonary arterial myocytes. AMPK activation by 5-aminoimidazole-4-carboxamide riboside, A769662 or C13 attenuated Kv 1.5 currents in pulmonary arterial myocytes, and this effect was non-additive with respect to Kv 1.5 inhibition by hypoxia and mitochondrial poisons. Recombinant AMPK phosphorylated recombinant human Kv 1.5 channels in cell-free assays, and inhibited K(+) currents when introduced into HEK 293 cells stably expressing Kv 1.5. These results suggest that AMPK is the primary mediator of reductions in Kv 1.5 channels following inhibition of mitochondrial oxidative phosphorylation during hypoxia and by mitochondrial poisons.

Abstract: Progression of hypoxic pulmonary hypertension is thought to be due, in part, to suppression of voltage-gated potassium channels (Kv ) in pulmonary arterial smooth muscle cells that is mediated by the inhibition of mitochondrial oxidative phosphorylation. We sought to determine the role in this process of the AMP-activated protein kinase (AMPK), which is intimately coupled to mitochondrial function due to its activation by LKB1-dependent phosphorylation in response to increases in the cellular AMP:ATP and/or ADP:ATP ratios. Inhibition of complex I of the mitochondrial electron transport chain using phenformin activated AMPK and inhibited Kv currents in pulmonary arterial myocytes, consistent with previously reported effects of mitochondrial inhibitors. Myocyte Kv currents were also markedly inhibited upon AMPK activation by A769662, 5-aminoimidazole-4-carboxamide riboside and C13 and by intracellular dialysis from a patch-pipette of activated (thiophosphorylated) recombinant AMPK heterotrimers (α2β2γ1 or α1β1γ1). Hypoxia and inhibitors of mitochondrial oxidative phosphorylation reduced AMPK-sensitive K(+) currents, which were also blocked by the selective Kv 1.5 channel inhibitor diphenyl phosphine oxide-1 but unaffected by the presence of the BKCa channel blocker paxilline. Moreover, recombinant human Kv 1.5 channels were phosphorylated by AMPK in cell-free assays, and K(+) currents carried by Kv 1.5 stably expressed in HEK 293 cells were inhibited by intracellular dialysis of AMPK heterotrimers and by A769662, the effects of which were blocked by compound C. We conclude that AMPK mediates Kv channel inhibition by hypoxia in pulmonary arterial myocytes, at least in part, through phosphorylation of Kv 1.5 and/or an associated protein.

Figure 1. Inhibition of mitochondrial oxidative phosphorylation activates AMPK and reduces K_v1.5 current density in pulmonary arterial myocytes

A, bar chart showing the effect of 10 mm phenformin on the activity of AMPK‐α1 and AMPK‐α2 containing heterotrimers, as determined by immunoprecipitate kinase assay (n = 3; 32 arteries, 8 rats). B, representative records (200 ms pulses from −80 to +40 mV in 10 mV increments, holding potential −80 mV) in pulmonary arterial myocytes before (control) and after extracellular application of 1 μm DPO‐1. C and D, representative records (a) and associated I–V relationships (b; 200 ms depolarization pulses from −80 to +40 mV in 10 mV increments, holding potential −80 mV) from myocytes pre‐incubated with 1 mm phenformin versus time‐matched controls (C, green, n = 9–11), or before and after 5–8 min extracellular application of 1 μm antimycin A (D, n = 6). ^* P < 0.05, ^** P < 0.01 and ^*** P < 0.001.

Figure 2. Time course for reduction in K_v current in response to AMPK activators

A, representative current records for voltage ramp and step protocol at 0 and 10 min after extracellular application of 100 μm A769662 (a), 1 mm AICAR (b) or 30 μm C13 (c). B, time course for reductions in K_v current during 10 min of extracellular application of DMSO (1:1000, vehicle control), 100 μm A769662, 1 mm AICAR and 30 μm C13. C, the effect on K_v currents of pulmonary arterial myocytes of 5 min extracellular application of 30 μm compound C (a), and the time course for inhibition (b). Results are expressed as mean ± SEM, n = 3–8.

Figure 3. AMPK activation inhibits K_v1.5 currents in pulmonary arterial myocytes

A, representative current traces (200 ms depolarization pulses from −80 to +40 mV in 10 mV increments, holding potential −80 mV); and B, I–V relationships for K_v current recorded before (control) and after extracellular application of A769662 (a, 100 μm), AICAR (b, 1 mm) or C13 (c, 30 μm); measurements taken at end of pulse. C, representative current trances (a) and I–V relationships (b) for K_v currents in the absence and presence of 1 μm DPO‐1 and the effect of 100 μm A769662 in the continued presence of DPO‐1. Results are expressed as mean ± SEM, n = 5–7. ^* P < 0.05, ^** P < 0.01 and ^*** P < 0.001.

Figure 4. Intracellular application of active AMPK heterotrimers inhibits K_v1.5 in pulmonary arterial myocytes

A, voltage ramp and step protocol recorded at 0 and 10 min after intracellular dialysis of the indicated recombinant, thiophosphorylated active AMPK heterotrimers (α2β2γ1 or α1β1γ1). B, time course for reduction in K_v current following intracellular dialysis of either active α2β2γ1 (thiophosphorylated, 5 U ml⁻¹), active α1β1γ1 (thiophosphorylated, 5 U ml⁻¹) or inactive α2β2γ1 (D157A mutant) AMPK heterotrimer. Results are expressed as mean ± SEM, n = 5–7.

Figure 5. Hypoxia and mitochondrial inhibitors attenuate K_v1.5 currents and occlude further current inhibition by AMPK activation

A, bar chart showing the reduction in current density at +40 mV from myocytes superfused with a hypoxic solution (4% O₂; n = 4–19). B, time course for reduction in K_v current during extracellular superfusion with hypoxic solution and the effect of subsequent addition of 100 μm A769662; measurements taken at the end of a 100 ms step pulse to +40 mV. Ca, example records for voltage ramp and step protocol under normoxia (1), > 10 min of hypoxia (2), after 8 min of superfusion with hypoxia + 100 μm A769662 (3). Cb‐c, as in a but representative current traces show the effect of 100 μm A769662 on myocytes pre‐incubated with 1 mm phenformin (b) or after 20 min extracellular application of 1 μm antimycin A (c). ^* P < 0.05, ^** P < 0.01 and ^*** P < 0.001 vs. normoxia.

Figure 6. Comparison of the change in current density and voltage–conductance relationships induced by activated AMPK and AMPK activators in pulmonary arterial myocytes

A, bar chart showing mean ± SEM change in current density at the end of each experimental intervention after 6–10 min of DMSO (1:1000), inactive α2β2γ2, hypoxia (∼4% O₂), 5 U active α2β2γ2, 5 U active α1β1γ1, 1 mm AICAR, 100 μm A769662, 30 μm C13, 1 mm phenformin (2–4 h), 1 μm antimycin A, 100 μm A769662 in the presence of hypoxia (∼4% O₂, > 10 min), phenformin and antimycin A (n = 3–16). B, similar to A but showing the net change in V _mid of the voltage–conductance plots for 1 mm AICAR, 100 μm A769662, 30 μm C13, 1 μm antimycin A, 100 μm A769662 in the presence of 1 μm antimycin A and 100 μm A769662 in the presence of hypoxia (∼4% O₂, > 10 min); n = 3–7. C, voltage–conductance plots showing effects of 1 mm AICAR (a), 100 μm A769662 (b), 30 μm C13 (c), 1 μm antimycin A (d), 100 μm A769662 in the presence of 1 μm antimycin A (e) and 100 μm A769662 in the presence of hypoxia (f, ∼4% O₂, > 10 min). Results are expressed as mean ± SEM, n = 3–7. ^* P < 0.05, ^** P < 0.01 and ^*** P < 0.001.

Figure 7. A769662 induces a leftward shift in the activation and inactivation curves of K_v1.5

Plot shows the voltage–conductance relationship for K_v1.5 activation and inactivation in the absence (control, black) and presence of 100 μm A769662 (red). Activation is indicated by filled symbols and continuous lines; inactivation is indicated by open symbols and dashed lines. Data points are mean ± SEM (n = 3–6). Curves were obtained by fitting to the sigmoidal Boltzmann equation.

Figure 8. AMPK activation phosphorylates K_v1.5 and reduces K⁺ currents carried by recombinant K_v1.5 channels stably expressed in HEK 293 cells

A, K⁺ currents carried by K_v1.5 stably expressed in HEK 293 cells before (0 min) and 3.5 min after extracellular application of (a) 100 μm A769662 or 5 min following intracellular dialysis of the active AMPK heterotrimer (b) (α2β2γ1, thiophosphorylated, 2 U ml⁻¹). B, time course for reduction of whole‐cell K⁺ currents during extracellular application of DMSO (1:1000, vehicle control), 100 μm A769662, the combined application of 40 μm compound C and 100 μm A769662; or C, intracellular dialysis of active α2β2γ1 (thiophosphorylated, 2 U ml⁻¹), active α1β1γ1(thiophosphorylated, 2 U ml⁻¹) or inactive α2β2γ1 (D157A mutant). D, bar chart showing mean ± SEM residual current at the end of each experimental intervention (3.5–5 min). Results are means ± SEM, n = 3–7.^* P < 0.05 and ^** P < 0.01 vs. control. E, Coomassie blue stained SDS‐PAGE (a) and autorad (b) of K_v1.5 protein incubated with okodaic acid, 200 μm AMP, ³²P‐ATP in the presence (L1) or absence (L2) of bacterial activated AMPK (α2β2γ1), and (L3) 1 μg of BSA treated as in L1.

Figure 9. Transcripts for all three K_vβ genes are expressed in the HEK_Kv1.5 cell line

Gel showing RT‐PCR amplicons for K_vβ1, K_vβ2, K_vβ3 and the reference gene (RG), GAPDH, from canine brain (CB), canine ventricle (CV) and the HEK_Kv1.5 stable line used in this study. The 500 base pair band of the ladder represents ∼1200 ng of DNA (2‐Log DNA Ladder, New England Biolabs, Ipswich, MA, USA).