|
Figure 1. Nonstationary noise analysis of mutant E166D at 120 mV. The figure shows the results from a typical inside-out patch containing many E166D channels. Currents were repeatedly activated at 120 mV (activation time constant ∼6 ms) and the mean was calculated (A). The variance was calculated from the mean square difference of consecutive records (Heinemann and Conti, 1992) (B). In C the variance (binned) is plotted versus the mean current (symbols) together with a fit of Eq. 1 (line) with the particular parameters i = 0.70 pA, N = 103.
|
|
Figure 2. Slow gate of WT, mutants of E166, and heterodimers. Representative recordings are for each construct as indicated in the figure. The left panels show voltage-clamp traces evoked by a standard “slow-gate” protocol (Pusch et al., 1997) that monitors the degree of activation by the variable prepulse at a constant tail pulse of 40 mV. The voltage dependence is shown in the right panels together with fits to a Boltzmann curve with offset. Note that mutants E166A and E166D do not show any voltage dependence of the slow gate, whereas the heterodimers partially recover a voltage-dependent component. Similar results were observed in at least three oocytes for each construct.
|
|
Figure 3. Fast gate of heterodimers. Representative inside-out patch recordings are shown for the EA-WT (A) and the ED-WT (B) dimer. The degree of activation of the fast gate is monitored as the initial current at the constant −140 mV tail pulse, plotted as absolute values in B and D, respectively. The solid lines in B and D represent fits of the equation I(V) = Imax (pmin + (1 − pmin)/(1 + exp((V − V1/2)/k))) with the parameters pmin = 0.42, V1/2 = −89 mV, k = 25 mV for EA-WT; and pmin = 0.16, V1/2 = −75 mV, k = 25 mV for ED-WT. The popen(V) curve of the EA-WT dimer is very similar to that described for WT CLC-0 under identical conditions (Accardi and Pusch, 2003). Similar results were obtained in at least three patches for each kind of tandem dimer.
|
|
Figure 4. Single channel phenotype of EA-WT dimer. In A, a recording at −100 mV of a patch containing a single EA-WT dimer is shown. Three conductance levels can be seen (dashed lines) corresponding to the baseline (both pores closed), one pore open, and two pores open. The behavior can be well described by the superposition of a WT pore with relatively fast gating and a popen around 0.64 and a pore that is mostly open with occasional longer closures (see arrows in A). The fast gating is better seen in B, which shows part of the upper trace at an expanded time scale. C shows the relative areas obtained by fitting the sum of three Gaussian functions to the amplitude histogram obtained over the stretch marked with the bracket with asterisk in A. In this period, no long closure event is seen and the area of the level 2 peak is 64% of the total area whereas the baseline peak is <1%. Similar results were seen in a total of four single-channel patches (one patch WT-EA, three patches EA-WT).
|
|
Figure 5. Phenotype of dimer WT-ED. In A, a current trace recorded from a patch with a single WT-ED dimer is shown. The voltage was stepped from 0 to −100 mV, back to 0 mV for a brief time and then to 80 mV. Currents appear almost like a single WT pore at −100 mV (with a popen ∼ 0.7) and at 80 mV with a popen of almost 1. However, brief openings to a second open level are visible at 80 mV. To visualize the “spikes” at 80 mV, B shows the superposition of 10 consecutive records at an expanded time scale showing only the pulse to 80 mV. To quantify the popen of the ED pore on top of the WT pore at 80 mV, the amplitude histogram of all records at 80 mV was constructed and limited to the part I > 0.7 pA (C, solid line). In this way, interferences from short closures were eliminated. The amplitude histogram was fitted by the sum of two Gaussian functions (dashed line in C). Since the popen of the WT pore is close to one, the relative area of the higher peak directly reflects the popen of the ED pore. In this case it amounts to 0.51%. Similar results were seen in a total of six single-channel patches (two patches WT-ED, four patches ED-WT)
|
|
Figure 6. Dependence of E166D on the extracellular pH. (A and B) Two-electrode voltage clamp recordings from the same oocyte at pHext 7.2 (A) and pHext 5.8 (B). The pulse protocol consisted of a prepulse to 60 mV followed by pulse to a variable voltage (from 80 to −140 mV). (C and D) Recordings from an outside out patch perfused with an extracellular solution at pH 7.2 (C) and 5.8 (D). The pulse protocol consisted of a prepulse to −140 mV followed by pulse to a variable voltage (from 160 to −140 mV) and a final tail pulse to −140 mV.
|
|
Figure 7. Dependence of E166D on the intracellular pH. (A) Families of current traces recorded from the same inside-out patch at the indicated pHint values. The pulse protocol consisted of a prepulse to −140 mV followed by pulse to a variable voltage (from 220 to −140 mV) and a final tail pulse to −140 mV that served to estimate the open probability as explained in the text. The boxed insets show on a magnified time scale the initial phase of the deactivation at the final tail pulse to −140 mV (horizontal scale bar in inset, 1 ms; vertical scale bar, 30 pA; all insets are drawn at the same scale). Experiments from a different patch are shown in B at the indicated pHint values (scale bars in inset, 1 ms and 10 pA, respectively). Voltage pulses in B were from 180 to −140 mV.
|
|
Figure 8. Analysis of the pHint dependence of E166D for pH > 5. (A) Averaged relative popen obtained from the tail currents as explained in the text are plotted versus voltage for each pHint with symbols as indicated. Error bars indicate SD. The solid lines represent fits of Eq. 2. The “slope” parameter, k, was ∼33 mV for pHint 5.3, 5.6, and 6.3, whereas it was manually fixed to 34 mV for pHint 6.8 and 7.2. (B) Average V1/2 values obtained from fits of the individual experiments plotted versus the respective pHint value (error bars are SD). The solid line is a linear fit with a slope of 81 mV per pH unit. The dashed line is a fit of Eq. 5 resulting in K1(0) = 4.3*10−8, pKO = 12.2, pKC = 5.3, z = 0.59.
|
|
Figure 9. Proof of specificity of the persistent inward current. (A) Currents recorded from an inside-out patch containing many E166D channels in control, in the presence of 5 and 20 mM CPA in the intracellular solution and after washout. The pulse protocol consisted of a prepulse to −140 mV, a variable pulse to voltages from −140 to 160 mV and a constant tail pulse to 100 mV. Note the block of inward currents by CPA. The slight increase of outward currents by CPA was observed consistently, but was not analyzed in detail. (B) Current traces from an inside-out patch with 104 mM Cl− inside (left) and 14 mM Cl− inside (right). Intracellular and extracellular pH had standard values in these experiments.
|
|
Figure 10. Small unitary conductance underlying the persistent current. In each box (A–C) are shown the results of a nonstationary noise analysis conducted with an inside-out patch with the indicated pHext values. In each case, subpanel a shows the mean current response from >60 stimulations evoked by a step from 0 to 160 mV and then to −140 mV. Subpanel b shows the corresponding baseline-subtracted variance. Subpanel c shows the plot of the binned variance versus the absolute value of the mean for the segment of the records at −140 mV. Solid lines are best fits of Eq. 1 with i = 0.68 pA (A), i = 0.7 pA (B), i = 0.13 pA (C). The number of channels was very large in A and B, and N = 583 in C. The dashed lines in A and B are linear fits with slopes of 0.79 pA (A) and 1 pA (B) and a current offset of 1.1 pA (A) and 4.2 pA (B). The residual variance at the end of the −140 mV pulse was 0.46 pA2 in B.
|
|
Figure 11. Persistent inward current is not associated with H+ inward movement. The extracellular pH was measured with a pH-sensitive microelectrode close to the oocyte surface as described in MATERIALS AND METHODS. Currents recorded from a CLC-5–expressing oocyte are shown in A (pulses are from −120 to 80 mV with a constant “tail” pulse to 60 mV). In B, the pH response, shown as a function of time, was evoked by a train of 300-ms pulses to 60 mV (current at 60 mV is ∼1.8 μA) with a 300-ms holding period between the pulses at −20 mV. The arrow indicates the switch off of the voltage clamp, nulling the current and leading to a recovery of pHext. (C and D) Results from a similar experiment from an oocyte from the same batch expressing mutant E166D. The train of 300-ms pulses was delivered to −100 mV (Vhold = −30 mV), where a current of ∼−2 μA was measured. No significant pH change could be detected in >10 oocytes with large expression of E166D, whereas a pH change was robustly detected in all CLC-5–expressing oocytes.
|
|
(MODEL 1).
|
|
(MODEL 2).
|
