Fig. 2. Combined strip-FRAP and FLIP-FRAP reveal that a fraction of agonist-liganded GFP-ARs is transiently immobilized. (A) The strip-FRAP method. A strip in the centre of a nucleus is bleached (red rectangle) with high laser power. Subsequently, fluorescence in the strip is measured at regular time intervals. Images are shown in false colour to visualize fluorescence differences more clearly. (B) Combined FLIP and FRAP method (FLIP-FRAP). A strip at one pole of the nucleus was bleached for a relatively long period. The difference between fluorescence signals in the bleached region (FRAP, red rectangle) and a distal region at 10 µm from the bleached region of the nucleus (FLIP, yellow rectangle) was determined at regular time intervals. (C,D) Strip-FRAP and FLIP-FRAP experiments of GFP-AR or the non-DNA-binding mutant GFP-AR(A573D) in the presence of 10^–9 M R1881. (C) Graph showing fluorescence intensities relative to complete redistribution of the non-DNA-binding mutant GFP-AR(A573D) in the presence of R1881 plotted as a function of time. Mean values of at least ten cells of a representative experiment are plotted. All experiments were performed at least three times. (D) Graph showing the difference between fluorescence intensity in the FLIP and FRAP regions (rectangles in B) relative to the difference directly after bleaching, plotted against time. Mean values±two times the s.e.m. of two independent experiments on at least ten cells are plotted. (E,F) Computer simulations (see Materials and Methods) of strip-FRAP and FLIP-FRAP of freely diffusing molecules do not explain the experimental FRAP data obtained with both methods. D is the effective diffusion coefficient. Experimental strip-FRAP data on wild-type GFP-AR lies in between curves representing indicated scenarios of free diffusion (E), whereas experimental FLIP-FRAP data on wild-type GFP-AR lies outside these boundaries (F). (G,H) Computer simulations representing a model where, next to freely diffusing molecules, a fraction is transiently immobilized, fitted to both strip-FRAP and FLIP-FRAP experimental curves on wild-type GFP-AR. Computer simulations correspond to the average of best fits of FRAP and FLIP-FRAP experiments respectively (Table 1), so are not necessarily the best fits of the individual experiments. Absolute value of residuals of the computer simulation fit and the experimental data on each time point are plotted below the x-axis.