Prediction-Based Algorithms for Long-Range High-Speed Force Spectroscopy

We present a high-speed force microscopy platform mounted on a confocal microscope with a z sample stage enabling long-range, high-speed force spectroscopy measurements on biological samples. The control software is built on a field-programmable gate array (FPGA)-based data acquisition and processing system, complemented with a custom graphical user insterface (GUI). We introduce smart algorithms based on probe-engagement prediction algorithms that leverage previously measured probe-sample contact to accelerate the probe engagement to mm/s and decelerate it in proximity of contact to user-defined μm/s velocities. This significantly reduces long-range force curve acquisition time and data density. Using this system, we provide proof-of-concept mechanical maps of cell clusters to extract topography and viscoelastic parameters. To further explore the versatility of our system, we probed the forces required to extract membrane tethers from monocytic cells using ultrashort cantilevers functionalised with adhesion molecules. Our system allowed us to cover retract velocities from ~1 μm/s up to ~6000 μm/s, and to explore the dynamics of tether formation at physiologically relevant velocities. Our results show that coupling extended z displacement with the prediction-based engagement algorithms enables rapid, quantitative mechanical mapping of heterogeneous biological samples with large topography variation and supports measurements requiring long distances at high velocities.