Mfp 3d controller

AFM force spectroscopy data was acquired on a custom-built AFM operated in closed loop mode by a MFP3D controller (Asylum Research, Santa Barbara, CA, USA). Software for the automated control of AFM head and xy-piezos was programmed in IgorPro6 (Wavemetrics, OR, USA). Strep-Tactin coated Biolever Mini cantilevers were briefly brought in contact with the sample surface and then retracted at 800 nm/s. After each recorded force-distance curve, the surface was horizontally moved in steps of 100 nm distance for iteratively probing a new position. The cantilever spring constant was calibrated using the equipartition theorem method (Hutter and Bechhoefer, 1993 (link)). Typically, datasets contain 30000 force-distance curves and the addition of the substrate was performed in the course of the experiment. If data was not collected within one experiment, but was directly compared via rupture force histograms, the recorded force values were normalized according to the most probable rupture force of the Fn3 domain. For this normalization, only those force-distance curves in a dataset were regarded that detached after unfolding of the Fn3 domain.

SMFS measurements were performed on a custom built AFM40 (link) controlled by an MFP-3D controller from Asylum Research running custom written Igor Pro (Wavemetrics) software. Cantilever spring constants were calibrated using the thermal noise/equipartition method41 . The cantilever was brought into contact with the surface and withdrawn at constant speed ranging from 0.2 to 6.4 μm s⁻¹. An x-y stage was actuated after each force-extension trace to expose the molecules on the cantilever to a new molecule at a different surface location with each trace. Typically 20,000–50,000 force-extension curves were obtained with a single cantilever in an experimental run of 18–24 h. A low molecular density on the surface was used to avoid formation of multiple bonds. While the raw data sets contained a majority of unusable curves due to lack of interactions or nonspecific adhesion of molecules to the cantilever tip, select curves showed single-molecule interactions. We filtered the data using a combination of automated data processing and manual classification by searching for contour length increments that matched the lengths of our specific protein fingerprint domains: Xyn (~89 nm) and CBM (~56 nm). After identifying these specific traces, we measured the loading rate dependency of the final Doc:Coh ruptures based on bond history.