A closed-loop system for millisecond readout and control of membrane tension.

Characterizing the function of force-gated ion channels is essential for understanding their molecular mechanisms and how they are affected by disease-causing mutations, lipids, or small molecules. Pressure-clamp electrophysiology is a method that is established and widely used to characterize the mechanical sensitivity of force-gated ion channels. However, the physical stimulus many force-gated ion channels sense is not pressure but membrane tension. Here, we further develop the approach of combining patch-clamp electrophysiology with differential interference contrast microscopy into a system that controls membrane tension in real time. The system uses machine learning object detection for millisecond analysis of membrane curvature and control of pipette pressure to produce a closed-loop membrane tension clamp. The analysis of membrane tension is fully automated and includes the propagation of experimental errors, thereby increasing throughput and reducing bias. A dynamic control program clamps membrane tension with at least 93% accuracy and 0.3 mN/m precision. Additionally, the absence of tension drift enables averaging open probabilities of ion channels with low expression and/or unitary conductance over long durations. Using this system, we apply a tension step protocol and show that TMEM63A responds to tension with a tension of half-maximal activation of T50 = 5.5 ± 0.1 mN/m. Overall, this system allows for precise and efficient generation of tension-response relationships of force-gated ion channels.

Duke Scholars

Author Jorg Grandl Neurobiology