Figure 1. Rap2 is necessary for... | Rockefeller University Press

Figure 1.

$Rap2 is necessary for MDA-MB-231 migration and invasion. (A) Loss of Rap2 confirmed via Western blot. Two different triple CRISPR KOs were made in MDA-MB-231s (see Materials and methods). All Ctrl cells shown in this figure are dox-inducible Cas9 MDA-MB-231s that were used to generate CRISPR lines. 50 µg of lysate was loaded for each sample. (B) Time-lapse 2D migration. Ctrl and Rap2 KO cells were plated on collagen-coated glass dishes and imaged every 10 min for 16 h using a brightfield 40× objective. Representative still images show the last frame of the time-lapse experiment, where the cell body indicates the location of the cell at the last time point. Cells were manually tracked in Fiji using the Manual Tracking plugin. Colors denote individual cell tracks over 16 h. White dots represent the start and finish of each cell track. Scale bars, 50 µm. (C) 2D migration velocity quantification. From the Fiji manual tracking, velocity data (µm/min) was extracted for each individual cell. Three biological replicates were performed for each cell line. For each experiment, 30 cells (color coded) were randomly chosen for velocity tracking (∼3 cells from ∼10 fields of view). Each cell was treated as its own data point (n = 90). One-way ANOVA with Tukey’s multiple comparisons test. Mean ± SD. Ctrl versus Rap2 KO1, P < 0.0001. Ctrl versus Rap2 KO2, P < 0.0001. (D) Chemotactic cell migration Ctrl versus Rap2 KO1 cells. Cells were imaged every 2 h for 48 h using an IncuCyte S3 instrument. Raw data (the sum area of all migrated cells normalized to the area at time 0) was extracted and averaged for each technical replicate. Data was then adjusted/normalized to the 8-h time point (see Materials and methods). The graph shows relative chemotactic migration over the 48-h time course. Mean ± SD. (E) Chemotactic migration quantification. Three biological replicates were performed, with six technical replicates in each experiment. Statistical analysis (unpaired t test) was performed on the relative chemotactic migration at 48 h. Mean ± SD at 48 h. Ctrl versus Rap2 KO1, P = 0.0273. (F) Boyden chamber invasion assay. Ctrl and Rap2 KO cells were plated in a modified Boyden chamber coated with Matrigel. Cells were allowed 20 h to invade through the Matrigel-coated pores before fixation and crystal violet staining. Inserts were imaged using a brightfield 20× air objective. Representative images are shown. Hollow circles indicate the 8-µm pores. Scale bars, 100 µm. (G) Invasion assay quantification. Three biological replicates were performed, with technical duplicates in each experiment. 5 fields of view were imaged for each Matrigel insert (resulting in 10 fields of view per experiment per condition, color coded). Raw number of cells invaded per field of view were normalized to Ctrl. Rap2 KO1 and Rap2 KO2 cells were analyzed at different times; hence two Ctrl samples. Each field of view was treated as its own data point (n = 30). One-way ANOVA with Tukey’s multiple comparisons test. Mean ± SD. Ctrl versus Rap2 KO1, P < 0.0001. Ctrl versus Rap2 KO2, P < 0.0001. (H) GFP-Rap2a WT localization. MDA-MB-231 cells stably expressing GFP-Rap2a were fixed and stained with phalloidin (magenta) and DAPI (blue). Z-slices are indicated on merged images. In most images, medial z-slices are shown. Here, we also provide a z-slice at the bottom of the cell, to highlight overlap between actin and GFP-Rap2a. Arrows point to example sites of actin and GFP-Rap2a colocalization. The arrowhead indicates the GFP-Rap2a intracellular organelle population. Scale bars, 50 and 5 µm. (I) Colocalization analysis of actin and GFP-Rap2a. Thresholded Mander’s coefficients were calculated using the Fiji Coloc 2 plugin. One biological replicate was performed, with five fields of view and five cells in each field (n = 25). The fraction of actin overlapping with Rap2a (tM1, mean = 0.8678) and the fraction of Rap2a overlapping with actin (tM2, mean = 0.8612) are shown. Mean ± SD. Source data are available for this figure: SourceData F1.$

Rap2 is necessary for MDA-MB-231 migration and invasion. (A) Loss of Rap2 confirmed via Western blot. Two different triple CRISPR KOs were made in MDA-MB-231s (see Materials and methods). All Ctrl cells shown in this figure are dox-inducible Cas9 MDA-MB-231s that were used to generate CRISPR lines. 50 µg of lysate was loaded for each sample. (B) Time-lapse 2D migration. Ctrl and Rap2 KO cells were plated on collagen-coated glass dishes and imaged every 10 min for 16 h using a brightfield 40× objective. Representative still images show the last frame of the time-lapse experiment, where the cell body indicates the location of the cell at the last time point. Cells were manually tracked in Fiji using the Manual Tracking plugin. Colors denote individual cell tracks over 16 h. White dots represent the start and finish of each cell track. Scale bars, 50 µm. (C) 2D migration velocity quantification. From the Fiji manual tracking, velocity data (µm/min) was extracted for each individual cell. Three biological replicates were performed for each cell line. For each experiment, 30 cells (color coded) were randomly chosen for velocity tracking (∼3 cells from ∼10 fields of view). Each cell was treated as its own data point (n = 90). One-way ANOVA with Tukey’s multiple comparisons test. Mean ± SD. Ctrl versus Rap2 KO1, P < 0.0001. Ctrl versus Rap2 KO2, P < 0.0001. (D) Chemotactic cell migration Ctrl versus Rap2 KO1 cells. Cells were imaged every 2 h for 48 h using an IncuCyte S3 instrument. Raw data (the sum area of all migrated cells normalized to the area at time 0) was extracted and averaged for each technical replicate. Data was then adjusted/normalized to the 8-h time point (see Materials and methods). The graph shows relative chemotactic migration over the 48-h time course. Mean ± SD. (E) Chemotactic migration quantification. Three biological replicates were performed, with six technical replicates in each experiment. Statistical analysis (unpaired t test) was performed on the relative chemotactic migration at 48 h. Mean ± SD at 48 h. Ctrl versus Rap2 KO1, P = 0.0273. (F) Boyden chamber invasion assay. Ctrl and Rap2 KO cells were plated in a modified Boyden chamber coated with Matrigel. Cells were allowed 20 h to invade through the Matrigel-coated pores before fixation and crystal violet staining. Inserts were imaged using a brightfield 20× air objective. Representative images are shown. Hollow circles indicate the 8-µm pores. Scale bars, 100 µm. (G) Invasion assay quantification. Three biological replicates were performed, with technical duplicates in each experiment. 5 fields of view were imaged for each Matrigel insert (resulting in 10 fields of view per experiment per condition, color coded). Raw number of cells invaded per field of view were normalized to Ctrl. Rap2 KO1 and Rap2 KO2 cells were analyzed at different times; hence two Ctrl samples. Each field of view was treated as its own data point (n = 30). One-way ANOVA with Tukey’s multiple comparisons test. Mean ± SD. Ctrl versus Rap2 KO1, P < 0.0001. Ctrl versus Rap2 KO2, P < 0.0001. (H) GFP-Rap2a WT localization. MDA-MB-231 cells stably expressing GFP-Rap2a were fixed and stained with phalloidin (magenta) and DAPI (blue). Z-slices are indicated on merged images. In most images, medial z-slices are shown. Here, we also provide a z-slice at the bottom of the cell, to highlight overlap between actin and GFP-Rap2a. Arrows point to example sites of actin and GFP-Rap2a colocalization. The arrowhead indicates the GFP-Rap2a intracellular organelle population. Scale bars, 50 and 5 µm. (I) Colocalization analysis of actin and GFP-Rap2a. Thresholded Mander’s coefficients were calculated using the Fiji Coloc 2 plugin. One biological replicate was performed, with five fields of view and five cells in each field (n = 25). The fraction of actin overlapping with Rap2a (tM1, mean = 0.8678) and the fraction of Rap2a overlapping with actin (tM2, mean = 0.8612) are shown. Mean ± SD. Source data are available for this figure: SourceData F1.