NuMA’s microtubule-binding region preferentially binds to microtubule minus ends and prevents their growth. (A) Schematic of a TIRF microscopy assay with fluorescent NuMA constructs and tubulin being flowed simultaneously into a channel containing surface-immobilized GMPCPP-microtubule seeds. (B) Representative TIRF microscopy image (left) and kymograph (right) showing minus and plus ends (dim magenta) dynamically elongating from a surface-immobilized Atto647N-labeled GMPCPP-seed (bright magenta) in the presence of Atto647N-tubulin. (C, E, and G) Representative TIRF microscopy images (top) and kymographs (bottom) showing the growth behavior of microtubule ends (dim magenta) elongating from surface-immobilized Atto647N-labeled GMPCPP-seeds (bright magenta) in the presence of Atto647N-tubulin and various concentrations of different mScarlet-tagged NuMA constructs (green). (D, F, and H) Growth velocity distributions of microtubule minus and plus ends related to the experiments shown in C, E, and G (mean of medians ± SEM). Each big circle represents the median velocity of one replicate; each small circle represents the growth velocity of one microtubule end segment; D: n = 109, 60, 69, 71 (left plot) and 109, 59, 68, 70 (right plot); adjusted P values: 0.3126, 0.0030, 0.0027 (left plot) and 0.8469, 0.8469, 0.6588 (right plot). F: n = 109, 72, 66, 67 (left plot) and 109, 72, 68, 65 (right plot); adjusted P values: 0.1334, 0.0188, 0.0041 (left plot) and 0.9642, 0.9642, 0.9642 (right plot). H: n = 109, 74, 70, 66 (left plot) and 109, 74, 70, 75 (right plot); adjusted P values: 0.0785, 0.0042, 0.0031 (left plot) and 0.6572, 0.8526, 0.8526 (right plot). All data are from at least three biological replicates; adjusted P values were calculated by Welch’s ANOVA test with Holm-Sidak’s post-hoc test for multiple comparisons; each condition was compared to the control at 0 nM NuMA. All experiments were performed in a NuMA microscopy buffer. All data are from three biological replicates.