The depletion of F₁ subunit ε in yeast leads to an uncoupled respiratory phenotype that is rescued by mutations in the proton-translocating subunits of F₀

Abstract

The central stalk of the ATP synthase is an elongated hetero-oligomeric structure providing a physical connection between the catalytic sites in F₁ and the proton translocation channel in F₀ for energy transduction between the two subdomains. The shape of the central stalk and relevance to energy coupling are essentially the same in ATP synthases from all forms of life, yet the protein composition of this domain changed during evolution of the mitochondrial enzyme from a two- to a three-subunit structure (γ, δ, ε). Whereas the mitochondrial γ- and δ-subunits are homologues of the bacterial central stalk proteins, the deliberate addition of subunit ε is poorly understood. Here we report that down-regulation of the gene (ATP15) encoding the ε-subunit rapidly leads to lethal F₀-mediated proton leaks through the membrane because of the loss of stability of the ATP synthase. The ε-subunit is thus essential for oxidative phosphorylation. Moreover, mutations in F₀ subunits a and c, which slow the proton translocation rate, are identified that prevent ε-deficient ATP synthases from dissipating the electrochemical potential. Cumulatively our data lead us to propose that the ε-subunit evolved to permit operation of the central stalk under the torque imposed at the normal speed of proton movement through mitochondrial F₀.

FIGURE 1:

A block in ATP synthase subunit ε expression is rapidly followed by F₀-mediated proton leaks that dissipate the mitochondrial membrane potential. The consequences of a block in subunit ε expression were followed using a strain (Tet-ε) in which this protein is under the control of a doxycycline-repressible promoter. (A) Growth curves of Tet-ε in rich glycerol/ethanol medium at 28°C in the presence (+Dox) or absence (–Dox) of doxycycline. Mitochondria were extracted from the two cultures at the time indicated by the arrowhead and used in the experiments shown in the subsequent panels. (B) Mitochondrial proteins were separated via SDS–PAGE and probed with antibodies against the indicated proteins. Setting the amount of each protein in Tet-ε cells grown in the absence of doxycycline at 100%, the relative levels of the proteins in Tet-ε cultured in the presence of the drug (+Dox) are plotted in the bar graph. The results shown are means of three experiments, and the vertical lines denote the SD in the data. (C) Oxygen consumption. The additions were 0.15 mg/ml mitochondrial proteins (mito), 4 mM NADH, 400 μM ADP, 3 μg/ml oligomycin (oligo), and 4 μM CCCP. (D) Energization of the mitochondrial inner membrane. Variations in ΔΨ were monitored by the fluorescence quenching of rhodamine 123. The additions were 0.5 μg/ml rhodamine 123, 0.15 mg/ml proteins (mito), 10 μl of ethanol (EtOH), 4 μg/ml oligomycin (oligo), 2 mM potassium cyanide (KCN), and 4 μM CCCP. Oxygen consumption and fluorescence traces are representative of three to five experimental trials.

FIGURE 2:

Kinetics of ρ⁻/ρ⁰ cell production. The Tet-γ, Tet-δ, and Tet-ε strains and their parental strain SDC22 (WT) were grown in rich galactose in the presence of 10 μM doxycycline. The cultures were refreshed several times with the same medium to produce a total of 40 generations, which was estimated by measuring the turbidity at 650 nm. The contents in ρ⁻/ρ⁰ cells were determined at the indicated number of generations.

FIGURE 3:

Mutations in F₀ a- and c-subunits can bypass the need for subunit ε. (A) Samples of the Tet-ε galactose cultures after 40-generation growth in the presence of doxycycline (Figure 2) produce colonies that grow on rich glycerol/ethanol medium. Arrowheads indicate medium-size (S1) and small (S2) clones that carry mutations in subunit c (c-L57F) and subunit a (a-A128V), respectively. (B, C) Growth of Tet-ε, S1, and S1 retransformed with the Tet-ε gene on glucose plates lacking uracil (B) and in liquid glycerol/ethanol medium (C).

FIGURE 4:

Properties of the c-L57F mutation. All of the experiments shown were performed using mitochondria prepared from cells grown in rich glycerol/ethanol medium. (A) Oxygen consumption. The additions were 0.15 mg/ml mitochondrial proteins (mito), 4 mM NADH, 150 μM ADP, and 4 μM CCCP. (B) Energization of the mitochondrial inner membrane. Variations in ΔΨ were monitored by the fluorescence quenching of rhodamine 123. The additions were 0.5 μg/ml rhodamine 123, 0.15 mg/ml proteins (mito), 10 μl ethanol (EtOH), 4 μg/ml oligomycin (oligo), and 2 mM potassium cyanide (KCN). For comparison, we include the experiment shown in Figure 1, which was performed with mitochondria prepared from Tet-ε cells grown in the presence of doxycycline. Oxygen consumption and fluorescence traces are representative of three to five experimental trials. (C) Steady-state levels of different mitochondrial proteins in S1. Mitochondrial proteins were separated via SDS–PAGE and probed with antibodies against the indicated proteins. The level of each protein in S1 relative to the Tet-ε control is represented by the bars. The results are representative of at least three experiments. The lines indicate standard deviations. (D) Mitochondrial proteins were extracted with digitonin (2 g/g), separated by BN-PAGE, and either assayed for in-gel ATPase activity or transferred to membranes for Western blotting with the indicated antibodies. V₁ and V₂, monomeric and dimeric species of ATP synthase (complex V), respectively.

FIGURE 5:

Contact zones between the different subunits of the central stalk in yeast F₁. Left, surface-rendered model of the yeast F₁ showing only γ (light gray, rendered at 20% transparency), δ (orange), and ε (magenta) subunits. Right, the surface of the δ-subunit was removed and in its place the red and blue dots show the positions of atoms in δ that are within 3.5 Å of the γ- and ε-subunits, respectively. PyMOL was used to calculate the atomic distances and make the Figure with the coordinates from the x-ray structure of yeast F₁C₁₀ ATP synthase (2WPD.pdb).