Fluorescent biosensor designs. A. MeroCBD, biosensor of endogenous Cdc42 activation. Here a fragment of Wiskott Aldrich syndrome protein (WASP) that binds only to activated Cdc42 is covalently derivatized with an environmentally sensitive dye. When the WASP fragment encounters and binds to activated Cdc42, the solvation of the dye by water is reduced, leading to a fluorescence change. Advantages of this design include the ability to study endogenous protein, and enhanced sensitivity due to direct excitation of a bright, long wavelength dye (as opposed to indirect excitation via FRET). The disadvantage is the need for microinjection, electroporation or some other means to `load' the covalently tagged protein biosensor into cells. B. RhoA biosensor. Here a fragment of Rhotekin that binds only to activated RhoA is attached to RhoA as part of the same protein chain. Two different fluorescent proteins undergoing FRET are in the chain between RhoA and the Rhotekin fragment, such that binding of the fragment to activated RhoA alters the distance and/or orientation between the fluorescent proteins, affecting FRET. This biosensor is fully genetically encoded, greatly simplifying loading into the cell. Because the Rhotekin fragment is attached to the RhoA, image processing is simplified relative to the dual chain sensor shown in C (see text). C. Rac1 FLAIR, biosensor of Rac1 activation. This design is similar to that of the RhoA biosensor, but here the PAK fragment used to bind activated Rac1 is not part of the same chain as the Rac1. The use of a dual chain, intermolecular FRET design enhances sensitivity because, unlike the single chain design, there is no FRET when the biosensor is in the `off' state. Additional image processing (bleedthrough correction) is required because the biosensor components can distribute differently throughout the cell.

X–Y translational image registration artifacts generate features on ratio images. In panel A, a correctly registered ratio image of MeroCBD is shown. Panel B shows the ratio image from the same cell without registration. In the latter case, the dye image was misaligned by −5 pixels in both x and y directions. The white arrows in both figures point to regions where misalignment produces edge artifacts and other artifactual features. In panel B, a lower ratio rim can be seen on one side of the nucleus, and a higher ratio on the opposite side. Similarly, edge artifacts appear as higher ratio on predominantly one side of the cell, with an artefactually low ratio on the opposite side.

A representative histogram from a shade corrected and background subtracted image, prior to masking. The prominent histogram peak at zero intensity is followed by a continuous distribution of pixels at a range of positive intensity values. Threshold masking to remove pixels of too low an intensity is performed using this histogram. One maximizes selection of pixels from within the cell while removing pixels outside the cell, including noise around the periphery.

Effect of the multiplication factor during ratiometric calculation. In panel A, using a scaling factor of 1000 produced a smooth histogram distribution in a ratio image. Panel B shows the same ratio image with the scaling factor of 25. Though the general distribution pattern is similar, the histogram distribution in panel A retains higher bit-depth resolution.

Transmittance spectra for the Custom dichroic mirror (“Scripps Custom”) (Chroma Technology), Lot# 511111886.

Transmittance spectra for the Custom dichroic mirror (“Quad Custom”) (Chroma Technology), Lot# 511112038.

Image shear between two cameras creates artifacts in ratio images. In the upper panels, calibration grid images from two cameras are overlaid with and without shear correction (red arrow: two channels aligned relative to the corner indicated). In the lower panels, the ratio of RhoA biosensor readouts from the two camera channels are shown with and without shear correction. A peripheral ruffle shows a large artifactual feature when correction is not applied (white arrow).