Binding of the inhibitor protein IF(1) to bovine F(1)-ATPase

Abstract

In the structure of bovine F(1)-ATPase inhibited with residues 1-60 of the bovine inhibitor protein IF(1), the α-helical inhibitor interacts with five of the nine subunits of F(1)-ATPase. In order to understand the contributions of individual amino acid residues to this complex binding mode, N-terminal deletions and point mutations have been introduced, and the binding properties of each mutant inhibitor protein have been examined. The N-terminal region of IF(1) destabilizes the interaction of the inhibitor with F(1)-ATPase and may assist in removing the inhibitor from its binding site when F(1)F(o)-ATPase is making ATP. Binding energy is provided by hydrophobic interactions between residues in the long α-helix of IF(1) and the C-terminal domains of the β(DP)-subunit and β(TP)-subunit and a salt bridge between residue E30 in the inhibitor and residue R408 in the C-terminal domain of the β(DP)-subunit. Several conserved charged amino acids in the long α-helix of IF(1) are also required for establishing inhibitory activity, but in the final inhibited state, they are not in contact with F(1)-ATPase and occupy aqueous cavities in F(1)-ATPase. They probably participate in the pathway from the initial interaction of the inhibitor and the enzyme to the final inhibited complex observed in the structure, in which two molecules of ATP are hydrolysed and the rotor of the enzyme turns through two 120° steps. These findings contribute to the fundamental understanding of how the inhibitor functions and to the design of new inhibitors for the systematic analysis of the catalytic cycle of the enzyme.

Graphical abstract

Fig. 1

The structure of bovine F₁-ATPase inhibited with residues 1–60 of the bovine inhibitor protein IF₁. The α-, β-, γ-, δ-, and ɛ-subunits are shown in ribbon representation in red, yellow, dark blue, magenta, and green, respectively. The inhibitor protein is light blue. (a) Side view of the complex towards the catalytic interface between the α_DP- and β_DP-subunits where the inhibitor protein is bound; (b) cross-sectional view of the C-terminal domains of the α- and β-subunits looking up along the axis of the γ-subunit showing interactions of the inhibitor protein with the subunits of F₁-ATPase.

Fig. 2

Alignment of the sequence of residues 1–60 of bovine IF₁ with equivalent portions of F₁-ATPase inhibitor proteins from other species. The purple, yellow, and green stripes denote strictly conserved, highly conserved, and poorly conserved residues, respectively. The black lines above the sequences mark α-helical regions in bovine IF₁. The Expasy accession numbers for inhibitor proteins from Bos taurus, Homo sapiens, Rattus norvegicus, Mus musculus, Saccharomyces cerevisiae, and Pichia jadinii are P01097, Q90112, Q03344, Q35143, P01097, and P09940, respectively.

Fig. 3

Influence of N-terminal deletions on the binding of the inhibitor protein I1-60GFPHis to bovine F₁-ATPase. The data are taken from Supplementary Table S4. For an explanation of the quotient K_imut:K_iwt, see the text.

Fig. 4

Influence of point mutations in the long α-helix of the inhibitor protein I1-60GFPHis on its binding to bovine F₁-ATPase. The mutations and their quantitative effects on binding are given in Supplementary Table S4. K_iwt and K_imut are the values for wild-type and mutant proteins, respectively. Strictly conserved, highly conserved, poorly conserved, and unconserved residues are shown in purple, yellow, green, and white, respectively. The unconserved residues G23, A36, and A38 were not mutated.

Fig. 5

Residues in the long α-helix of the bovine inhibitor protein that contribute to its binding by interacting directly with subunits of bovine F₁-ATPase. The inhibitor protein from residues 8–50 is pale blue, and α-helical regions (residues 14–18 and 21–50) are shown in ribbon representation. The extended N-terminal region (residues 8–13) and the shorter α-helix snake around two antiparallel α-helical regions (residues 1–23 and 232–251) in the γ-subunit, which is part of the central stalk or rotor of the enzyme. Residues with yellow side chains interact with the β_DP-subunit; the residue with a red side chain interacts with the α_DP-subunit; and the residue with a blue side chain interacts with the β_TP-subunit.

Fig. 6

Charged residues in the long α-helix of the bovine inhibitor protein that are required for the binding of the inhibitor to F₁-ATPase but do not interact with F₁-ATPase in the structure of the inhibited complex. The long α-helix of IF₁ (residues 21–50; light blue) occupies an aqueous cleft between the C-terminal domains of the β_DP- and α_DP-subunits (yellow and red, respectively). The side chains of amino acids that do not interact with the β_DP- and α_DP-subunits, but nonetheless are required for the formation of the inhibited complex, are dark blue.

Fig. 7

Possible schemes of the inhibition of F₁-ATPase by the inhibitor protein I1-60. F₁-ATPase is depicted as viewed away from the inner mitochondrial membrane from the foot of the central stalk. For simplicity, only those α- and β-subunits that form the catalytic interface, where I1-60 binds, are shown in ribbon representation. The α- and β-subunits are in red and yellow, respectively, and I1-60 is in light blue. The remaining α- and β-subunits and the γ-subunit are represented schematically, and the corresponding α-subunits have been omitted for simplicity. Panel A: In state I, IF₁ binds first via the longer α-helix to the C-terminal domain of the β_E-subunit in the catalytic interface between the α_E- and β_E-subunits when the enzyme is in the ground state [PDB code 2JDI]. These initial interactions between inhibitor and enzyme involve many of the group 1 residues, E30, Y33, F34, Q41, L42, and L45 (orange), and the formation of this and subsequent intermediates may also include the group 2 charged residues, R25, Q27, R37, K39, and Q41. In state II, rotation of the γ-subunit through 120° closes the α_Eβ_E interface, converting it to the α_TPβ_TP interface [PDB was made by docking I1-60 into the α_TP-β_TP interface of the azide free ground state structure (PDB code 2JDI) at the same angle to which it binds to the α_DPβ_DP interface in the final inhibited state (PDB code 2v7q)⁶]. This α_TP β_TP interface is similar in structure but not identical to the α_DP β_DP interface, and many of the interactions observed in the structure of the final inhibited state probably will be present at this stage. They include the interactions formed by the group 1 residues of IF₁ (E30, Y33, F34, Q41, L42, and L45; orange). Also, it is possible that the group 1 residue A28 (grey) may bind to the α_TP-subunit at this point. However, residue F22 cannot interact with F₁-ATPase at this stage. In state III, further rotation of the γ-subunit through 120° leads to the formation of the final inhibited structure [PDB code 2v7q], either, as depicted in scheme A, a dead end state, where an ATP molecule has been hydrolysed at the same catalytic interface that I1-60 binds, or, as shown in scheme B, a pre-hydrolysis state, where an ATP molecule has not been hydrolysed. Residue F22 (purple) can now interact with the β_TP-subunit, completing the binding interactions between the long α-helix of IF₁ and F₁-ATPase. In this final inhibited state, the N-terminal region (purple) will also interact with the γ- and α_E-subunits.