The mitochondrial complex V-associated large-conductance inner membrane current is regulated by cyclosporine and dexpramipexole

Abstract

Inefficiency of oxidative phosphorylation can result from futile leak conductance through the inner mitochondrial membrane. Stress or injury may exacerbate this leak conductance, putting cells, and particularly neurons, at risk of dysfunction and even death when energy demand exceeds cellular energy production. Using a novel method, we have recently described an ion conductance consistent with mitochondrial permeability transition pore (mPTP) within the c-subunit of the ATP synthase. Excitotoxicity, reactive oxygen species-producing stimuli, or elevated mitochondrial matrix calcium opens the channel, which is inhibited by cyclosporine A and ATP/ADP. Here we show that ATP and the neuroprotective drug dexpramipexole (DEX) inhibited an ion conductance consistent with this c-subunit channel (mPTP) in brain-derived submitochondrial vesicles (SMVs) enriched for F1FO ATP synthase (complex V). Treatment of SMVs with urea denatured extramembrane components of complex V, eliminated DEX- but not ATP-mediated current inhibition, and reduced binding of [(14)C]DEX. Direct effects of DEX on the synthesis and hydrolysis of ATP by complex V suggest that interaction of the compound with its target results in functional conformational changes in the enzyme complex. [(14)C]DEX bound specifically to purified recombinant b and oligomycin sensitivity-conferring protein subunits of the mitochondrial F1FO ATP synthase. Previous data indicate that DEX increased the efficiency of energy production in cells, including neurons. Taken together, these studies suggest that modulation of a complex V-associated inner mitochondrial membrane current is metabolically important and may represent an avenue for the development of new therapeutics for neurodegenerative disorders.

Fig. 1.

DEX and CsA decreased conductance in SMVs. (A) Example patch-clamp recording from a brain-derived SMV; holding potential, +60 mV, before and after addition of the indicated agents to the bath (Ca²⁺ = 100 μM, DEX = 2 μM, CsA = 1 μM). (B) Histograms represent group data (mean ± S.E.M.) of peak conductance of experiments in (A). Current was measured from 0 pA and presented as peak conductance assuming a linear current-voltage relationship. (Left) Level of conductance before and after switching to high [Ca²⁺], Ca²⁺ plus 2 μM DEX (n = 13 SMVs), followed by Ca²⁺, DEX, and 1.0 μM CsA (n = 5 SMVs). One-way ANOVA, P = 0.0002; preplanned post hoc Bonferroni-corrected t tests, Ca²⁺ versus 2 μM DEX, P = 0.00041; Ca²⁺ versus CsA, P = 0.0048. (Right) Reversal of order of DEX and CsA addition: level of conductance in SMVs in control medium, after switching to high [Ca²⁺], followed by Ca²⁺ plus 1.0 μM CsA (n = 5 SMVs), then Ca²⁺, CsA, and 2 μM DEX (n = 3 SMVs). One-way ANOVA, P = 0.0043; preplanned post hoc Bonferroni-corrected t tests, Ca²⁺ versus CsA, P = 0.0196; Ca²⁺ versus DEX, P = 0.046. (C) Example SMV recording before and after addition of 0.5 mM ATP; holding voltage, +110 mV; discontinuous recording; ATP was added <1 minute prior to the break in the recording. Bar chart indicates group data (mean ± S.E.M.) for the effect of 0.5 mM ATP on peak conductance level (n = 9 SMVs, P = 0.0001, paired t test). (D) Example SMV recording before and after addition of 200 nM DEX; discontinuous recording; holding potential, +100 mV. Histograms indicate group data for the effect of 200 nM DEX on peak conductance level (n = 6 SMVs, P = 0.0003, paired t test). In all cases, recordings are discontinuous; addition of drugs occurred from 30 seconds to 10 minutes prior to recording the steady-state response of the drug at the given concentration. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.

Fig. 2.

Concentration-response relationship for the inhibition of currents in SMVs by ATP and DEX. (A, Top) Effect of ATP on peak conductance levels; holding potential, +40 mV. (Bottom) Group mean data and logistic fit estimate of the effect of ATP concentration on percentage maximal SMV current inhibition (n = 4 SMVs; EC₅₀ = 224 μM; Hill slope = 1.5). One-way ANOVA, P = 0.0001; preselected post hoc Bonferroni-corrected t tests, control versus 0.4 mM ATP, P = 0.010; 0.6 mM, P = 0.0032; 0.8 mM, P = 0.0023; 1.0 mM, P = 0.0016. (B, Top) Effect of DEX on peak conductance levels in SMVs; holding potential, +50 mV. (Bottom) Group mean data and logistic fit estimate of the mean (±S.E.M.) percentage maximal current inhibition at different concentrations of DEX (n = 5 SMVs; EC₅₀ = 111 nM; Hill slope = 0.31). One-way ANOVA, P < 0.0001; preselected post hoc Bonferroni-corrected t tests, control versus 20 nM DEX, P = 0.0375; 200 nM, P = 0.009; 2 μM, P = 0.0025; 20 μM, P = 0.0005. In all cases, recordings are discontinuous; addition of drugs occurred from 30 seconds to 5 minutes prior to recording the steady-state response. *P < 0.05; **P < 0.01; ***P < 0.001 versus control.

Fig. 3.

DEX modulates complex V activity. (A) F₁F_O ATPase activity (ATP hydrolysis) in the presence of different concentrations of DEX (red squares) or CsA (black squares) shown as a function of the rate of decrease in NADH fluorescence and expressed as percentage of control; DEX, n = 3 determinations/point, one-way ANOVA, P = 0.0014; CsA, n = 3 determinations/point, one-way ANOVA, P = 0.0002. Pre-selected Bonferroni-corrected post hoc comparisons, control versus 200 nM DEX, P = 0.0315; 2 μM, P = 0.0456; 20 μM, P = 0.0438; control versus 2 μM CsA, P = 0.0132; 4 μM, P = 0.010; 6 μM, P = 0.0096; 8 μM, P = 0.0152. (B) ATP synthesis: change in ATP levels over time in liver mitochondria (n = 3 wells/point) following an ADP pulse using a dynamic luciferin-luciferase assay. All drugs added at time 0. *P < 0.05; **P < 0.01 versus control.

Fig. 4.

Denaturation of extramembrane proteins in SMVs eliminated DEX- but not ATP-induced current inhibition. (A) Decrease in oxyluciferin luminescence levels indicating time-dependent decreases in ATP levels in the absence (Blank) and presence (CTL SMV) of SMVs and in the presence of urea-treated SMVs (n = 3 wells for each condition). (B) Immunoblot with an antibody against the F₁ β-subunit in protein from control SMVs or urea-treated SMVs. Identical protein concentrations were loaded in each well. Bottom shows adenine nucleotide translocator (ANT) protein level as a loading control. (C) Peak membrane conductance (percentage of control) of urea-treated SMVs in the presence of the indicated agents (ATP, 1.0 mM; DEX, 200 nM; n = 7 SMVs for ATP; n = 6 SMVs for DEX). Groups represent separate experiments; P = 0.0028 for ATP, unpaired t test. (D) Level of [¹⁴C]DEX binding in SMVs treated with urea, relative to control SMV levels (n = 15 samples/condition; P ≤ 0.0001, unpaired t test). ***P < 0.001 versus control.

Fig. 5.

Evidence for [¹⁴C]DEX binding to isolated complex V subunits. (A) Recombinant human F₁F_O ATP synthase subunits purified from mammalian cell expression system. Myc-Flag–tagged constructs for human F₁F_O ATP synthase subunits (as labeled at bottom of gel) immunoprecipitated with anti-Flag affinity gel and immunoblotted with anti-Myc tag antibody. Control (CTL) lane represents immunoprecipitate with anti-Flag affinity gel of cell lysate from nontransfected cells. (B) Counts per minute of anti-Flag affinity gel immunoprecipitates from cells expressing the indicated constructs exposed to 200 nM ¹⁴C-labeled DEX. One-way ANOVA, Bonferroni-corrected preplanned post hoc comparisons; ***P < 0.001. (C) Counts per minute of anti-Flag affinity gel immunoprecipitates from cells exposed to 200 nM ¹⁴C-labeled DEX and 20 μM unlabeled (“cold”) DEX. One-way ANOVA, Bonferroni-corrected preplanned post hoc comparisons. *P < 0.05; ***P < 0.001; black asterisks are levels of significance for comparisons between control and radiolabeled b and OSCP; red asterisks are levels of significance for comparisons between levels of binding of radiolabeled DEX in b or OSCP columns and levels in the presence of excess unlabeled DEX. ns, not significant.