Kernel Based Non-Iterative Automatic Fast Capacitance Compensation in Patch-Clamp Experiments

Abstract

Patch-clamp recording is a widely used technique for studying ion channel functions and cellular responses. During patch-clamp experiments, the fast capacitance (CFast) transient is a detrimental artifact and must be eliminated with CFast compensation circuit. When applying the conventional computer-aided CFast compensation procedure, the patch-clamp amplifier faces the risk of saturation due to large square-wave responses, and compensation errors may be caused due to inherent disadvantages of iterative algorithm. Here, we introduce a novel non-iterative automatic CFast compensation method, namely K-method, based on kernel estimation with white noise excitation. The kernel was estimated with cross-correlation technique and captured dynamic properties of the CFast and its related hardware. To achieve the optimal settings of the CFast compensation, the kernel of CFast was fitted to two calibrated kernels indicating two distinct types of compensation effect: the “instantaneous” and “delay” effect. The fitted coefficients were used to adjust compensation circuit. White noise excitation significantly reduced the possibility of saturation, and the K-method suffered from no typical disadvantage of iterative method. We performed compensation experiments on a model circuit and HEK293 cells. The results demonstrated a good accuracy of the K-method and the membrane capacitance measurement could benefit from it.