Smith & Lippiat: Voltage sensing in CLC-5

Supplementary Results

Isolation of gating currents from membrane capacitance

Whole cell capacitance cancellation was used to remove the main component of the membrane

capacitance. Fig. S1 shows recordings from a HEK293 cell expressing EGFP-tagged E268A CLC-5 (Fig.

S1A) before (Fig. S1B) and after (Fig. S1C) whole cell capacitance cancellation. Fig. S1C also shows (in

red) the linear residual capacitance and leak currents calculated using the P/-4 voltage protocol. In this

protocol, voltage pulses one quarter of the amplitude of the test pulses were applied to the opposite polarity,

i.e. in the hyperpolarizing direction for a depolarizing test pulse. This ensured that pulses were applied in the

sub-threshold voltage range. The leak and residual capacitance currents were subtracted from the raw data

to give the subtracted data in Fig. S1D, which were used for data analysis and presentation.

Apparent reversal potential

Measurement of the reversal potential of CLC-5 was attempted by pre-activating CLC-5 with a 10 ms pulse

to 200 mV and stepping down to a range of potentials. This voltage protocol gave rise to transient inward

currents, similar to tail currents observed from voltage-gated ion channels (Fig. S2A). The peak inward

current evoked by repolarization was measured and plotted against voltage (Fig. S2B). A straight line was

fitted to the data yielded the reversal potential, the point where there was no net flow of ions.

Kinetics of wild-type and S520T CLC-5 currents

Whole cell recordings were made from HEK293 cells expressing wild-type (WT) and S520T CLC-5 with the

aim of examining the time-course of current activation upon depolarization. The fitting of curves to the

current at the start of the depolarising step was prevented by the appearance of transient gating currents

superimposed upon the sustained ionic flux (Fig. S3A). Therefore, the limits of the fit were adjusted to

exclude the initial transient component and we assumed that there was no instantaneous component to the

CLC-5 current (offset). We found that both WT and S520T currents were best fit by the sum of two

exponentials (Fig S3A): I(t) = I1.(1-e-t/ 1) + I2.(1-e-t/ 2), where 1 and 2 are the fast and slow time constants,

respectively. The S520T mutation slowed both 1 and 2 at all voltages that gave measurable current. The

S520T mutation did not appreciably alter the steady-state activation kinetics (Fig. S3D), although the

amplitudes of the whole-cell currents appeared to be approximately half those of WT CLC-5 (Fig. S3C).

Effects of altering [Cl-]I on WT CLC-5 currents

We found that lowering intracellular [Cl-] shifted the charge-voltage relationship of the gating currents

recorded from E268A CLC-5 to more-negative potentials and increasing [Cl-]I had the opposite effect (see

main text, Fig. 4). We also found that the activation of WT CLC-5 was similarly shifted when using the same

pipette solutions (Fig. S4). Conductance-voltage relationships were generated and fitted by Boltzmann

functions as detailed in the Materials and Methods section of the main text. Lowering [Cl-]I from 42 to 10 mM

shifted the activation V½ from 106.2 ± 1.2 mV (n=6) to 73.4 ± 3.6 mV (n=5), whilst increasing [Cl-] to 140 mM

shifted the activation V½ to 132.2 ± 1.6 mV (n=5). The mean activation curves (G/GMAX vs voltage) are

shown in Fig. S4C, whilst the mean current-voltage plots show that currents are activated to a similar range

of amplitudes (Fig. S4B).

Effect of lowering pHI on WT CLC-5 currents

Lowering intracellular pHI, equivalent to increasing the concentration of H+ available for coupled exchange

with extracellular Cl-, appeared to shift the V½ to more positive potentials (Fig. 4, main text), although the

data were not statistically different (p=0.13). We also explored if lowering the pH of the pipette had any

effect on the currents of WT CLC-5. Currents recorded from cells expressing WT CLC-5 using a pipette

solution with pH 5.4 were outwardly-rectifying, similar to when a pH 7.4 solution was used (Fig. S5A) and

with comparable current amplitudes (Fig. S5B). By generating activation curves (Fig. S5C) a shift of the G-

V relationship to more positive potentials with pH 5.4 was apparent. The V½ was shifted from 106.2 ± 1.2 mV

(n=6) with pH7.4 to 135.7 ± 4.8 mV (n=7) with pH 5.4 (p<0.001, t-test).

Supplementary Figure Legends

Figure S1: Isolation of gating currents. (A) Images of a HEK293 cell expressing EGFP-tagged E268A CLC-

5 which a sealed patch pipette (top image) and the same subject under epifluorescence (bottom image). (B)

Raw current traces using the voltage clamp protocol detailed in the Materials and Methods prior to whole

cell capacitance cancellation. Raw currents recorded after capacitance cancellation are shown in (C) along

with the calculated residual linear capacitance and leak (red traces), which were subtracted to give the

currents in (D). For clarity, only the current sweeps from voltage pulses to -100, -50, 0, 50, 100, 150, and

200 mV are shown.

Figure S2: (A) Representative current recordings from a HEK293 cell expressing wild-type CLC-5 showing

the transient inward current upon a change in voltage from 200 mV to between 80 and 20 mV, inclusive, in

10 mV intervals. (B) Mean current-voltage relationship of the transient current measured as the peak inward

current immediately after the change in membrane potential. The mean intercept of the x-axis was 53.3 ±

2.3 mV (n=6).

Figure S3: Time course of wild-type and S520T current activation. (A) Representative currents from

HEK293 cells expressing either wild-type (WT) or S520T CLC-5 as indicated evoked by depolarizing steps

from -30 to 200 mV. The presence of transient gating currents at the start of the depolarizing segment is

indicated (zg). The fitted curve, the sum of two exponentials, is indicated in red, and the dotted line is the

zero current level. (B) Mean (± s.e.m.) measurements of 1 (closed symbols) and 2 (open symbols) from

both WT (squares, n=6 cells) and S520T (circles, n=10 cells) CLC-5. Mean (± s.e.m.) current-voltage (C)

and activation (D) curves for both WT (■) and S520T (●) CLC-5. Data were collected from 6 cells expressing

WT and 10 cell expressing S520T CLC-5.

Figure S4: Effects of changing [Cl-]I on WT CLC-5 currents. (A) Representative whole cell currents from

cells expressing WT CLC-5 using pipette solutions with [Cl-] as indicated. (B) mean current-voltage

relationships from cells using pipette solutions containing 42 mM (●), 10 mM (■), and 140 mM (▲) Cl-. (C)

Representation of the same data (symbols as in B) as conductance-voltage plots. Data are mean ± s.e.m. of

at least 5 different cells.

Figure S5: Effect of lowering pHI on WT CLC-5 currents. (A) Representative currents from cells expressing

WT CLC-5 recorded using pipette solutions at either pH 7.4 or 5.4 as indicated. Current-voltage (B) and

conductance-voltage (C) plots of data recorded with internal pH 7.4 (●) and 5.4 (○). Data are mean ± s.e.m.

of at least 5 different cells.

DC

B

1nA1nA5ms5ms

5ms10nA

AFigure S1