# Also in the Article

Calculating single-molecule FRET efficiency

Regulation and dynamics of force transmission at individual cell-matrix adhesion bonds

Procedure

Single-molecule data were acquired and analyzed as described previously (14). Briefly, data were acquired with excitation with a 532-nm laser at 5 frames/s for 300 or 600 frames and with direct acceptor excitation at 635 nm for approximately 10 frames at roughly frame 100. The direct excitation helped to distinguish between low-FRET sensors and sensors without an acceptor dye.

Traces were analyzed using a custom MATLAB code, and donor and acceptor channels were aligned using a single-molecule high-resolution colocalization map generated by scanning across a field of beads (50). The positions of individual sensors were then detected using a spot-finding algorithm (T. Ursell, Stanford University) and were determined to be colocalized if within two pixels. Intensities were calculated on the basis of an average of 7 × 7 pixels centered around the detected spot and corrected for spectral bleedthrough.

Intensities for each dye were averaged over manually identified FRETing, non-FRETing, and bleached regions. When the acceptor bleached before the donor, we used the following expression to calculate FRET efficiency$E=(Ia−Ia,back)(Ia−Ia,back)+γ(Id−Id,back)$$γ=Ia−Ia,backId,0−Id$where Ia is the acceptor intensity during FRET, Ia,back is the acceptor background intensity, Id is the donor intensity during FRET, Id,0 is the donor intensity after acceptor photobleaching, Id,back is the donor background intensity, and γ is the correction factor accounting for relative dye quantum yields and instrument detection efficiencies.

When the donor fluorophore bleached first, the FRET efficiency was calculated as$E=(Ia−Ia,back)(Ia−Ia,back)+γ0(Id−Id,back)$

Values for γ0 were 0.40 for MTSlow, 0.52 for MTSFN, and 0.52 for MTShigh.

Events were double-checked by generating a series of z projections for the donor and acceptor molecule during FRETing, non-FRETing, and bleached states. The autoGaussianSurf Matlab function (P. Mineault) was used to fit a two-dimensional Gaussian to the 7 ×7–pixel area to determine whether the spot represented a single emitting fluorophore. Low-FRET events were verified as having a functional acceptor by direct excitation with a 635-nm laser.

Q&A