Fig. 3: Localization of regions with high RAC1 activity and high local GTP concertation.

a Schematic representation of the RAC1 FRET (Förster resonance energy transfer) biosensor. CRIB Cdc42/Rac interactive binding motif. b MDA-MB-231 cells co-expressed the RAC1 biosensor and GEVAL30 or GEVALNull. The signal of each individual biosensor was imaged and rendered as described in Methods. c MDA-MB-231 cells co-expressed the RAC1 biosensor and GEVAL30 or GEVALNull as in (b) and assayed for the correlation between biosensor activities. For each cell, biosensor data were collected in a series of images taken at 1-min intervals over the course of 30 min and analyzed as described in Methods. Pixel-wide Pearson correlation between RAC1 activity (measured as FRET (Förster resonance energy transfer) index for the RAC1 biosensor) and GTP index (measured as the activity of the indicated GEVAL variant) was calculated for each image. The correlation values for the image series corresponding to each individual cell were summarized as “bar-and-whiskers” plots, with “whiskers” indicating the first and the fourth quartiles, the horizontal line (median) splitting the bars into the second and the third quartiles, and “X” indicating quartiles, as well as the mean correlation coefficient (r) for each series. The mean correlation coefficients were compared by Mann–Whitney test. d Cells expressing the indicated constructs were probed in RAC1 activity assay as described in Methods. Note that the difference in migration of RAC1^T17N compared to other proteins is likely due to the fact that RAC1^T17N is fused to one Myc-tag, whereas other RAC1 proteins—to two Myc tags. Shown are representative images of at least two independent experiments. e Quantification of (d); n = 2 biologically independent samples (left panel); n = 3 biologically independent samples (right panel) The data represents average ± SEM.