Co-incorporation of GDP-tubulin and GTP-tubulin in microtubules favors minus-end survival and hinders growth at the plus-end. (A) Representative kymographs of hybrid microtubules polymerized from 25 µM GTP-tubulin alone or with increasing concentrations of GDP-tubulin as noted. In GDP-/GTP-tubulin mixtures, microtubules are assembled from unlabeled GTP-tubulin and fluorescently labeled GDP-tubulin. S, seed. (B) Growth rate, shrinkage rate, catastrophe frequency, and rescue frequency of microtubule plus and minus ends assembled with different GTP-/GDP-tubulin ratios. Data are represented as the mean ± SD of at least 209 (growth rate), 146 (shrinkage rate), 125 (catastrophe frequency), and 30 (rescue frequency) events from at least three independent experiments (detailed values of parameters are given in Table S1). Triangles represent the mean of each experiment. *P < 0.5, **P < 0.01, ****P < 0.0001 (Kruskal–Wallis ANOVA followed by post-hoc Dunn’s multiple comparison). P values were calculated relative to the condition with no GDP-tubulin (0:1 GDP-tubulin/GTP-tubulin ratio). (C) Microtubule depolymerization length before a rescue is a function of tubulin content. Data are represented as the mean ± SD of 93 (0:1 ratio), 166 (1:5 ratio), 123 (1:2 ratio), 156 (1:1 ratio), 252 (2:1 ratio), and 294 (3:1 ratio) from at least three independent experiments. Triangles represent the mean of microtubule depolymerization length for each experiment. ***P < 0.001, ****P < 0.0001 (Kruskal–Wallis ANOVA followed by post-hoc Dunn’s multiple comparison). P values were calculated relative to the condition with no GDP-tubulin (0:1 GDP-tubulin/GTP-tubulin ratio). (D) Diagram showing how persistent and non-persistent growth were defined. The proportion of microtubules with persistent growth at the minus end increased with the ratio of GDP-tubulin to GTP-tubulin. n, total number of events from at least three independent experiments. Triangles represent the proportion of persistent growth for each experiment. *P < 0.05, **P < 0.01, ***P < 0.001 (two-sided Fisher’s test). P values were calculated relative to the condition with no GDP-tubulin (0:1 GDP-tubulin/GTP-tubulin ratio).