Tip bifurcation involves narrowing of the cleft and can be initiated independent of the mesenchyme. (A) Maximum intensity projections of bifurcating tips, imaged from E17.5–E18.5 mammary glands, with a shallow cleft (left) and deeper clefts (middle, right) visible based on laminin-staining. (B) The angle and depth of the cleft was measured as indicated in A and plotted for 65 bifurcations from 30 glands of 13 embryos. (C) Images of an epithelial surface rendering of a bifurcating tip from a time-lapse imaging video of an epithelial Fucci2a expressing ex vivo cultured mammary gland, at 2-h intervals following the appearance of the cleft with the bifurcation angle indicated. (D) Average bifurcation angles from 10 Fucci2a videos (seven experiments) at incremental hours starting from the appearance of the cleft as indicated by color. (E) Outlines of the epithelial surface rendering of a bifurcating tip at incremental hours from the start of the time-lapse video as indicated by color (ingression of the cleft is indicated by a black arrow and advancement of the daughter tips by white arrows). (F) The depth of the cleft and the length of the daughter tips were measured from time-lapse imaging videos of surface rendered bifurcating tips after the appearance the cleft and their relative velocity calculated based on duration of the video (n = 10, 7 experiments). (G) Maximum intensity projections of mammary epithelial rudiments, derived from R26R-mTmG embryos with (left) or without (right) K14-Cre. (H) Maximum intensity projections of an epithelial rudiment cultured and imaged in 3D Matrigel for a period of 18 h. The boxed region follows a bifurcating tip at 2-h intervals. (I) Quantification of branching events from 26 rudiments of four time-lapse imaging experiments. Data shown in F represents the median (line) with 25th and 75th percentiles (hinges) plus 1.5× interquartile ranges (whiskers). Statistical significance was assessed with the paired two-tailed t test.