Dynamics of the preprotein translocation channel of the outer membrane of mitochondria

Abstract

The protein translocase of the outer mitochondrial membrane (TOM) serves as the main entry site for virtually all mitochondrial proteins. Like many other protein translocases it also has an ion channel activity that can be used to study the dynamical properties of this supramolecular complex. We have purified TOM core complex and Tom40, the main pore forming subunit, from mitochondria of the filamentous fungus Neurospora crassa and incorporated them into planar lipid bilayers. We then examined their single channel properties to provide a detailed description of the conformational dynamics of this channel in the absence of its protein substrate. For isolated TOM core complex we have found at least six conductance states. Transitions between these states were voltage-dependent with a bell-shaped open probability distribution and distinct kinetics depending on the polarity of the applied voltage. The states with the largest conductance followed an Ohmic I/V characteristic consistent with a large cylindrical pore with very little interaction with the permeating ions. For the lower conductance states, however, we have observed inverted S-shaped nonlinear current-voltage curves reminiscent to those of much narrower pores where the permeating ions have to surmount an electrostatic energy barrier. At low voltages (<+/-70 mV), purified Tom40 protein did not show any transitions between its conductance states. Prolonged exposure to higher voltages induced similar gating behavior to what we observed for TOM core complex. This effect was time-dependent and reversible, indicating that Tom40 forms not only the pore but also contains the "gating machinery" of the complex. However, for proper functioning, additional proteins (Tom22, Tom7, Tom6, and Tom5) are required that act as a modulator of the pore dynamics by significantly reducing the energy barrier between different conformational states.

FIGURE 1

Purification of Tom40. (A) Mitochondria from a N. crassa strain carrying a hexahistidinyl tag on Tom22 were solubilized in 0.3% DDM and passed through a Ni-NTA affinity column. After various washing steps Tom40 was eluted with 3% OG; Tom22 among other proteins was eluted with 300 mM imidazole. Aliquots of the resulting column fractions were analyzed by Coomassie blue stained tricine gels and with Western blots. (Lane 1) Mitochondria; (lane 2) flow through; (lane 3) OG eluate; (lane 4) imidazole eluate. (B) The purity of isolated Tom40 was assessed by Western blot analysis using antibodies against Tom40, Tom22, Tom7, Tom6, and Tom5.

FIGURE 2

Current fluctuations through single TOM core complex channels. (A) Purified TOM core complex was reconstituted into planar lipid bilayers and membrane currents were recorded in symmetrical 1 M KCl, 10 mM Hepes, pH 7.0 using the voltage clamp technique with clamp potentials between −100 and +100 mV. At low voltages between −20 and +20 mV hardly any voltage-dependent gating was observed. At higher voltages, TOM-mediated currents were voltage-sensitive and switched between six nonequidistant conductance states (S0–S5). The conductance states S0–S5 are indicated as dotted lines. (B) Amplitude histograms (bin widths, 4 pA) produced from 10 s of channel activity. Up to six peaks are evident in the histograms (lower panel) and are labeled with the corresponding subconductance level. Gaussian curves were fitted to the data and are superimposed to the histograms. (C) Mean-variance plots of the recordings shown in (A). The window widths used were 15 points, corresponding to 1.5 ms. Transitions between conductance states are visible as parabolic arches in the plots connecting low variance current levels. (D and E) Enlarged view of the shaded areas shown in (A). (D) Currents at −100 mV. (E) Currents at +100 mV. Arrows indicate sojourns in the closed state S0.

FIGURE 3

I/V characteristic of TOM channel conductance states. (A) A representative current trace of a single TOM core complex channel recorded in 1 M KCl, 10 mM Hepes, pH 7.0 (upper panel), in response to a linear voltage ramp from −120 to +120 mV (lower panel). The conductance states S1 to S5 are indicated as dotted lines. (B) I/V characteristic from mean current amplitudes determined by fits to current histograms for substates S1 to S5. Deviation from linear (ohmic) I/V relation with a typical inverted S-shape is particularly evident for the S3 state but also with S2 and S1. The data were fitted with a Nernst-Planck permeation model assuming a single energy barrier in the middle of the pore (solid lines). All fits were based on a channel length of 70 Å (44) an ion diffusion constant of 2 × 10⁻⁹ m²s⁻¹ (36) and a symmetrical salt concentration of 1 M KCl at both boundaries of the channel. (C) Concentration dependence of the conductivities as determined in (D–F). Conductivities follow a Michaelis-Menten relation with an apparent K_S of 2.6 ± 0.8 M and a maximum conductivity γ_max of 7.7 ± 1.7 nS for the main open state S5. (D–F) I/V curves of single TOM core complex channels were recorded in 125 mM, 250 mM, or 1 M KCl, 10 mM Hepes, pH 7.0, as described in (A). Data of 20 voltage ramps were superimposed. The I/V curves for substate S5 are virtually linear in all three salt solutions. I/V relationships of the lower conductance states S3 and S2 display a nonlinear behavior with increasing conductance at high potentials.

FIGURE 4

Voltage-dependent state occupancies of TOM core complex channels. Voltages from −120 to +120 mV were applied in increments of 10 mV to single TOM core complex channels. Membrane currents were measured for 10 s each (n = 8). For each voltage all-point histograms were generated and fitted with Gaussian functions. The occupancy probabilities p of the substates S0–S5 were obtained from the area below each peak at the indicated potentials. The probability of state S5 was fitted to a double Boltzmann distribution (dotted line) with parameters = −33 ± 7 mV, = 68 ± 10 mV, A = 0.1 ± 0.03 mV⁻¹ and A′ = −0.06 ± 0.02 mV⁻¹, respectively. The solid lines represent the subconductance state occupancy probabilities predicted by the kinetic model in (Fig. 5 B).

FIGURE 5

Voltage-dependent rates of TOM core complex channels. (A) Voltage dependence of the transition rate between conductance states S5 → S3. Due to the logarithmic scaling of the ordinate, voltage-dependent rates are represented as linearly correlated clusters of data points. Note the bimodal V-shaped relation that indicates that transitions from conductance states S5 to S3 are governed by two almost symmetric processes with rates and In this particular case both rates were voltage-dependent with = 21 ± 9 s⁻¹ and = 0.013 ± 0.005 V⁻¹ and = 9 ± 10 s⁻¹ and = −0.017 ± 0.012 V⁻¹. (B) Topology for the kinetic scheme for TOM core complex gating. The S_X-notation of the kinetic states refers to the conductance class whereas the S ↔ S′ notation indicates the different polarity of the voltage dependence.

FIGURE 6

Properties of single channels of Tom40. (A) Purified Tom40 was added to one side of a planar lipid membrane and single channel currents were measured in the presence of a membrane potential of 70 mV. Step-like increases in current levels indicate incorporation of Tom40 channels into the bilayer. The aqueous phase contained 1 M KCl, 10 mM Hepes, pH 7.0. (B) Histogram of channel conductances. A total of n = 119 conductance increments as shown in (A) was analyzed. S*, conductance states of Tom40. (C) I/V curves of single Tom40 channels recorded by application of linear voltage ramps between −70 and +70 mV. Four independent recordings with different conductivities were overlaid.

FIGURE 7

Voltage-dependence of single-channels of Tom40. (A) Response of a single Tom40 channel to linear voltage ramps of different amplitude ranging from −70 to +70 mV and from −150 to +150 mV. Initially very little gating between substates could be observed; increasing the amplitude of the ramps induced voltage-dependent gating. Subsequent decrease of the voltage amplitude reversed the gating behavior to the initial gating pattern. (B) Expanded view of the response of a linear voltage ramp ranging from −150 to +150 mV as in (A). (C) Open probability of the main conductance state of Tom40 determined as described in Fig. 4. Application of linear voltage ramps from −70 to +70 mV resulted in very few gating transitions (solid squares). Increasing the voltage range to ±150 mV led to voltage-dependent closures (open squares) approaching the characteristics of TOM core complex (dashed line, Fig. 4). For all records, the cis and trans compartments contained 1 M KCl, 10 mM Hepes, pH 7.0.