Role of the DELSEED loop in torque transmission of F1-ATPase

Abstract

F(1)-ATPase is an ATP-driven rotary motor that generates torque at the interface between the catalytic β-subunits and the rotor γ-subunit. The β-subunit inwardly rotates the C-terminal domain upon nucleotide binding/dissociation; hence, the region of the C-terminal domain that is in direct contact with γ-termed the DELSEED loop-is thought to play a critical role in torque transmission. We substituted all the DELSEED loop residues with alanine to diminish specific DELSEED loop-γ interactions and with glycine to disrupt the loop structure. All the mutants rotated unidirectionally with kinetic parameters comparable to those of the wild-type F(1), suggesting that the specific interactions between DELSEED loop and γ is not involved in cooperative interplays between the catalytic β-subunits. Glycine substitution mutants generated half the torque of the wild-type F(1), whereas the alanine mutant generated comparable torque. Fluctuation analyses of the glycine/alanine mutants revealed that the γ-subunit was less tightly held in the α(3)β(3)-stator ring of the glycine mutant than in the wild-type F(1) and the alanine mutant. Molecular dynamics simulation showed that the DELSEED loop was disordered by the glycine substitution, whereas it formed an α-helix in the alanine mutant. Our results emphasize the importance of loop rigidity for efficient torque transmissions.

Figure 1

Sequence and structure of DELSEED loop. (A) Amino-acid sequences of the DELSEED loop and the mutants prepared in this study. Residues are numbered according to the TF₁ sequence. Residues Ile³⁸⁶ to Asp³⁹⁴ are defined as the DELSEED loop. (B) Crystal structure of the β-subunit (green) in its AMPPNP-bound form, i.e., β_TP form, and the γ-subunit (cyan) (PDB:2JDI). The positions of the both ends of glycine or alanine substitutions are shown (sticks) in the magnified images.

Figure 2

Time course of rotation and kinetic analyses of Ala and Gly mutants. (A) Time courses of rotation at 1 mM ATP. The probe used for the rotation was a 60-nm gold colloidal bead. The rotational velocity was determined from 30 continuous revolutions. (B) Rotational velocity and [ATP]. The average rotational rates and the estimated rotational rate from the ATPase activity (1/3 ATPase) are shown (solid circles and open squares, respectively). (Error bars) Standard deviations from at least four different molecules (rotation assay) or three independent measurements (ATPase activity measurement). (Data points) Fits using the Michaelis-Menten equation, V = V_max × [ATP]/(K_m + [ATP]). The values of V_max, K_m, and the calculated k_on^ATP are summarized in Table 1. (C) (Histogram) Catalytic dwell of the Gly^loop+13 mutant at 1 mM ATP probed by a 60-nm gold colloid (2 molecules, 339 events). The histogram was fitted using the consecutive reaction model of two reactions. The time constants were determined to be 1.77 and 0.48 ms.

Figure 3

Torque determination of Ala and Gly mutants. The fluctuation theorem was employed for the measurement. The ln[P(Δθ)/P(−Δθ)] value is plotted against Δθ for Δt = 3 ms (solid circles), 4 ms (solid squares), 5 ms (open circles), 6 ms (open squares), and 7 ms (cross). The slope of this plot represents the torque/k_BT generated by F₁. The average torque was determined from a linear approximation of all the data points (shaded line) using a fitting width of ±0.22 (rad). The results are summarized in Table 2.

Figure 4

Fluctuation analysis of pausing mutant F₁. (A) Typical x-y trajectory of the rotation of wild-type F₁. The rotation was observed at 300 nM ATP. (B) Analysis of the fluctuations in circumferential directions. The probability distributions of the wild-type F₁ (red) and the Gly^loop+13 (black) and Ala^loop (orange) mutants are shown.

Figure 5

Molecular dynamics simulations of Gly^loop+13 and Ala^loop. (A) The average structures of β_empty during the last 15 ns of the simulations are shown (ribbon representation). The structures of β_empty in the wild-type F₁, Gly^loop+13, and Ala^loop are shown (red, black, and orange, respectively). The γ-subunit is also shown (light gray). (Lower images) Expanded views of the structures of the DELSEED loop regions (enclosed with blue rectangles in the upper images). (B) Results of DSSP analyses of β_empty (empty), β_DP (DP), and β_TP (TP). The probabilities of α-helix formation at individual residues of the DELSEED loop and adjacent regions in the wild-type F₁ (red), Gly^loop+13 (black), and Ala^loop (orange) are plotted. Residue numbers 384–405 in MF₁ correspond to 380–401 in TF₁. (C) (Histograms) Retention times of the specific contacts are shown. Bin-width is 5% of 15 ns (0.75 ns). The wild-type and Gly^loop+13 and Ala^loop mutants are represented (red, black, and orange plots, respectively). (D) ΔRMSF analysis of Gly^loop+13 (black) and Ala^loop (orange). The RMSF values of wild-type F₁ were subtracted from the RMSF values of Gly^loop+13 and Ala^loop to obtain the ΔRMSF values of β_empty (empty), β_DP (DP), and β_TP (TP). A positive ΔRMSF represents a larger fluctuation than that for the wild-type F₁, whereas negative values represent smaller fluctuations. (Horizontal black and orange bars in the figures represent the glycine- or alanine-substituted regions.)