Effects of CenpE inhibition on interactions of kinetochores with noncentrosomal microtubules during early prometaphase. (A) Maximum-intensity projection of an early-prometaphase RPE1 cell that entered mitosis in the presence of 20 nM GSK923295. Hec1-tdTomato is shown in red, and Mad2-Venus is shown in green. (B and C) Correlative LM (B) and EM (C) images of two sister kinetochores with intermediate (k01a) or low (k01b) amount of Mad2-Venus. Orange lines in B demark positions of microtubules traced in EM. Yellow arrows in C denote end-on–attached microtubules, and blue arrows point at laterally interacting microtubules. (D) Tukey box plot presenting the number of microtubules (MTs) within the 250-nm area adjacent to the kinetochore plate in cells with inactive CenpE (GSK; 82 kinetochores from three cells; mean = 42) or depleted of CenpE (siRNA; 83 kinetochores from two cells; mean = 34). (E) Distribution of angles between microtubules and the kinetochore plate. The difference between GSK and siRNA distributions is not significant (P = 0.02 in two-sample Kolmogorov-Smirnov test); however, both distributions differ from the distribution of angles in untreated cells (Fig. 4 E; P < 0.001 in two-sample Kolmogorov-Smirnov test). (F) Scatterplot presenting the number of end-on–attached microtubules versus the amount of Mad2 at individual kinetochores. Mad2 intensity is normalized so that the background is 0 and the mean value is 1. The regression line slop coefficient is 0.1863 (R2 = 0.05) for GSK and −1.2453 (R2 = 0.09) for siRNA.
