Increased Pk2 at anterior AJs enhances anisotropic AJ remodeling in the NE. (A) Distribution of GFP-Pk2 (low dose, 50 pg RNA) in the NE. Scarlet-UtrCH marks AJs. (B) Snapshot images of NEs expressing Scarlet-UtrCH only (purple, left) or Flag-Pk2 (low) with Lifeact-GFP (green right) from time-lapse imaging. Scale bars: 20 μm. (C) Quantification of changes in AJ length. The control and Pk2(low) sides of NEs were compared in the same embryos. Changes in AJ length were categorized as follows: elongation (blue, ≥0.8 μm/3 min), shrinkage (red, less than or equal to −0.8 μm/3 min), or no change (green, −0.8< and <0.8 μm/3 min). AJs were categorized into AP (≤30° from the AP axis), ML (≤30° from the ML axis), and mid (30–45° from either the AP or ML axes). The distributions of AJ length changes in AP and ML junctions were compared between the control and +Pk2(low) NE groups. The Kolmogorov‒Smirnov test was used on original numerical data. Three embryos per group were used. The total counts in the plots were as follows: n = 291 (control #1); n = 280 (Pk2 (low) #1); n = 253 (control #2); n = 183 (Flag-Pk2 #2); n = 360 (control #3); n = 363 (Pk2 (low) #3). (D and E) Rates of AJ elongation (D) and shrinkage (E) in the control and Pk2(low) OE NEs. The same dataset used in C was used for this analysis. The bold and thin lines indicate medians and quartiles, respectively. The Mann‒Whitney U test was used to compare mean ranks. *P < 0.05. n.s., not significant. (F–H) Models of Pk2-mediated tissue fluidity control. Pk2 (red) increases cadherin turnover by modulating the components of AJ complexes or increasing actomyosin dynamics at AJs, promoting AJ remodeling (F). In the absence of directional cues, Pk2-mediated AJ remodeling is directionally random (G). In the presence of directional cues, Pk2 locally increases the susceptibility of AJs to stimulus-induced remodeling (H).