Changes in single K(+) channel behavior induced by a lipid phase transition

Abstract

We show that the activity of an ion channel is correlated with the phase state of the lipid bilayer hosting the channel. By measuring unitary conductance, dwell times, and open probability of the K(+) channel KcsA as a function of temperature in lipid bilayers composed of POPE and POPG in different relative proportions, we obtain that all those properties show a trend inversion when the bilayer is in the transition region between the liquid-disordered and the solid-ordered phase. These data suggest that the physical properties of the lipid bilayer influence ion channel activity likely via a fine-tuning of its conformations. In a more general interpretative framework, we suggest that other parameters such as pH, ionic strength, and the action of amphiphilic drugs can affect the physical behavior of the lipid bilayer in a fashion similar to temperature changes resulting in functional changes of transmembrane proteins.

Figure 1

Thermodynamic characterization of the main phase transition behavior of POPE/POPG 3:1 bilayers. (a) DSC curves acquired starting from high to low temperatures are reported for POPE/POPG 3:1 at varying pH values. (b) Variation of the transition midpoint temperature of the POPE/POPG 3:1 bilayers as a function of pH. Both the temperatures obtained from the upscan and the downscan of the DSC measurements are given highlighting the hysteretic behavior. The markers (○, upscan; □, downscan) display the measured values, while the lines represent linear fits to the experimental data.

Figure 2

Effect of temperature on KcsA conductance. (a) Examples of traces of the current for KcsA in POPE/POPG 3:1 at three different temperatures: 27.0°C, 22.5°C, and 20.0°C. Holding potential: 75 mV. (b) Plot of the conductance of KcsA reconstituted in POPE/POPG 3:1 as a function of temperature for two holding potentials (□, 50 mV and ○, 75 mV). The two lines are the linear fits to the experimental data in the regions of monotonic behavior of the channel conductance (from 22.5°C to 28.5°C: ---, 50 mV; —, 75 mV). Ion conductance changed linearly in a temperature region between 28.5°C and 22.5°C, but upon further cooling, at a temperature of ∼22.0°C a trend inversion appeared. (c) Temperature-dependent ion conductance of KcsA in BLMs of different POPE/POPG compositions at a holding potential of 75 mV. (Solid lines) Respective linear fits in the high temperature region. A nonmonotonic behavior is observed for the two POPE/POPG mixtures (○, 3:1; ▵, 1:1), but a pure linear trend is found for POPG (□). The trend inversion happened at a lower temperature when KcsA was reconstituted in POPE/POPG 1:1 than in 3:1, accordingly to the shift of the phase transition region for the two lipid compositions. In panels c and d, the data are given as mean values together with the standard deviation. (d) Superposition of the downscan DSC traces at pH 3 (shifted by −1.5°C) for the POPE/POPG 3:1 (—) and 1:1 (---) lipid bilayers with channel conductance at a holding potential of 75 mV after subtraction of the linear fits obtained outside the lipid phase transition. The trend was extended into the transition region. Data points were taken at constant temperature (□, 3:1; ■, 1:1) or during the temperature change (▵, 3:1; ▴, 1:1). The changes in ion conductance took place proportionally to the variation of the system's specific heat capacity.

Figure 3

KcsA dwell times as a function of temperature at a holding potential of 50mV. (a) Examples of ion current traces at temperatures of 27.0°C and 22.0°C in a POPE/POPG 3:1 lipid mixture. Channels are open for longer times at the temperature of 20.0°C where the ion conductance presents a maximum. (b and c) Dwell-time histograms for the open (b) and closed (c) states, for KcsA reconstituted into planar bilayers of POPE/POPG 3:1. The open dwell-time and closed dwell-time distributions were fitted with exponential components (^…..) resulting in the overall fit (—). (d) Variation of the characteristic open (longer time, ○; slower time, □) and (e) closed times (○) with temperature compared to the downscan DSC trace at pH 3 (—, shifted by −1.5°C) of the corresponding lipid bilayer mixture. Open times show the same trend as the specific heat capacity. Closed times present a minimum at the temperature corresponding to the maximum of the excess heat capacity and open times.

Figure 4

Normalized open probability for KcsA in lipid bilayers of different composition compared with the respective downscan DSC analysis at pH 3 (shifted by −1.5°C). (a) POPE/POPG 1:1; (b) POPE/POPG 3:1. The normalized open probability NP₀ (○, given as a mean together with its standard deviation) in arbitrary units referring to the right ordinate and the specific heat capacity (—) in arbitrary units referring to the left ordinate are reported. In both cases, the open probability had a trend inversion when the system entered the phase transition region.