Fig. 3. Osmolarity changes regulate size and mobility of AQP4 arrays. (a) Representative TIRFM image of DIV9 cortical astrocyte stained for GFAP (cyan) with corresponding dSTORM images of AQP4, rendering localizations with dots of sizes corresponding to the respective localization precision. (b) Cumulative frequencies of the cluster size show an overall increase in cluster sizes in hypertonic condition (Kolmogorov–Smirnov test: D = 0.1832, P < 0.0001). Data were obtained from three independent experiments with 28 FOVs with 17 894 clusters for isotonic medium (246 mOsm) and 27 FOVs with 25 977 clusters for hypertonic medium (370 mOsm). (c) Cumulative distribution of the number of localizations per cluster (Kolmogorov–Smirnov test: D = 0.2835, p < 0.0001). (d) Cluster area per cell (median of distribution of each cell) shows an increase in the cluster size upon hypertonic treatment (black lines show mean value, unpaired t test, p = 0.0057). (e) Cumulative distributions of the diffusion coefficients of AQP4 (mean ± SD) comparing isotonic vs. hypertonic conditions (Kolmogorov–Smirnov test: D = 0.03297, P < 0.0001). Data represent two independent culture replicates (isotonic n = 56 cells, hypertonic n = 56 cells). (f) Percentage of mobile AQP4 arrays per cell showing an decrease in the proportion of mobile vs. immobile arrays with hypertonic conditions (unpaired t test, p = 0.0155).