A systematic method of noise analysis for channels with single-channel conductance of fA level

电压门控质子通道的单通道电导预计将低于ps。本研究探讨了评估这种小电导通道的噪声分析方法。全细胞记录小胶质细胞的电压门控质子通道。质子通道在去极化电位下的激活过程非常缓慢，需要很长的记录时间才能收集足够数量的电流迹线用于去极化脉冲。对于分析，我们发现最关键的一步是当前轨迹的选择上的趋势下降。在长时间的记录中，电流痕迹会改变它们的形状。目前显示这些系统性变化的痕迹，而不是随机变化，必须从分析中消除。对选择当前踪迹施加了几个标准。从这些选定的轨迹中进行总体平均。提取了不同实验条件下的脉动分量，并进行了功率谱和非平稳噪声分析。功率谱密度采用双洛伦兹拟合法。非平稳噪声分析的截止频率设定为5 kHz。绘制了方差-均值图，发现该图为踪迹的选择提供了有价值的标准。从选择具有不同标准的轨迹到方差均值曲线图的过程重复进行，直到拟合的参数收敛。从单通道电流幅值的方差均值图中估计出活动通道数和最大开放概率。电压门控质子通道的估计单通道电流为12fA。讨论了使用选定的踪迹改进数据分析的基本机制。

Single channel conductance of the voltage-gated proton channel was expected to be below pS. The noise analysis method for evaluating such small conductance channels was examined in this study. Whole-cell recordings of the voltage-gated proton channel from microglial cells were performed. The time course of activation at depolarization potential was very slow for the proton channel and long recording time was necessary to collect enough numbers of current traces for depolarizing pulses. For the analysis we found that the most critical step was detrend upon the selection of current traces. In the long time of recordings current traces change their shape. Current traces showing these systematic changes, rather than stochastic changes, must be eliminated from the analysis. Several criteria for selecting current traces were imposed. Ensemble average was performed from those selected traces. Fluctuating components were extracted and power-spectrum and non-stationary noise analysis were performed at various experimental conditions. Double-Lorentzian fit was done for the power spectral density. The cut-off frequency of 5 kHz was set for the non-stationary noise analysis. The variance-mean plot was drawn and we found that this plot gave valuable criteria for the selection of traces. The process from selecting traces with different criteria to the variance-mean plot was repeated until the fitted parameters were converged. From variance-mean plots amplitudes of single channel current, numbers of active channels and maximal open probability were estimated. The estimated single channel current for the voltage-gated proton channel was 12 fA. Underlying mechanisms for improvements of the data analysis using selected traces were discussed.