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

期望电压门控质子通道的单通道电导低于pS。本研究研究了评价这种小电导通道的噪声分析方法。对来自小胶质细胞的电压门控质子通道进行全细胞记录。质子通道在去极化电位下的激活时间过程非常缓慢，需要较长的记录时间来收集足够数量的去极化脉冲电流迹线。在分析中，我们发现电流走线的选择是最关键的一步。在长时间的录音中，电流痕迹会改变它们的形状。必须从分析中消除显示这些系统变化的当前痕迹，而不是随机变化。提出了几种选择电流走线的标准。从这些选定的轨迹中进行总体平均。在不同的实验条件下提取波动分量，进行功率谱分析和非平稳噪声分析。对功率谱密度进行了双洛伦兹拟合。非平稳噪声分析的截止频率设置为5 kHz。绘制了方差-均值图，我们发现该图为选择迹线提供了有价值的标准。从选择具有不同准则的迹线到方差-均值图的过程重复，直到拟合参数收敛。从方差均值图中估计了单通道电流的幅值、有效通道的数量和最大打开概率。电压门控质子通道的单通道电流估计为12fa。讨论了使用选定轨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.