Fig. 4.

Effect of blebbing and physical barriers on PGC migration and polarity. (A) Blebbing frequency in PGCs that reside within ectodermal tissues and express control RNA or RNA encoding a dominant-negative version of Rac1. A two-tailed t-test was performed. (B,C) Speed and track straightness of PGCs expressing control RNA or a dominant-negative version of Rac1, migrating among converted ectodermal cells (B) or mesodermal cells (C). A Mann–Whitney test was performed for track straightness comparison in B and both comparisons in C, a two-tailed t-test was performed for speed comparison in B. The graphs show the median; whiskers indicate the interquartile range (IQR) (D) A model illustrating molecular interactions that lead to the protrusion-type shift in different environments. (E) Illustration of the gel injection experiment. Mutant cxcr4b^−/− embryos were injected with a dextran gel of 0.5 kPa or 4.1 kPa stiffness at 3 hpf and were imaged 4 h later. As cells migrate non-directionally, some stochastically come in contact with the gel (left cell right-hand box). Alterations in actin distribution were then monitored in cells that came into contact with the gel and in neighboring PGCs that did not encounter the gel (right cell in magnified box). The comparison between the PGCs was performed to control for the inherent periodic loss of polarity observed in PGCs (tumbling; Reichman-Fried et al., 2004). (F) Examples of the reactions of PGCs (lifeact in magenta) to the gel (green). Behaviors were scored based on the change in the polarity of the actin-rich front. The white arrows indicate the direction of the back-front axis of the cell before the interaction with the gel. (G) Quantification of cell behaviors. For statistical tests, Fisher's exact test was performed (comparing the combined turn and loss of polarity behaviors to no change in polarity). A Bonferroni correction for multiple testing was applied. N and n represent numbers of embryos and cells, respectively.