Computational modeling demonstrates that nucleotide exchange in the microtubule lattice does not recapitulate the characteristics of the microtubule depolymerization observed in experiments. (A) Schematics of the computational model for CLASP-dependent nucleotide exchange leading to depolymerization. See Table S2 for a list of the model parameters. (B) A representative kymograph of an experimental microtubule depolymerizing in the presence of 1 mM GTP and 61.5 nM CLASP. The red dashed lines indicate the traces of the microtubule ends obtained using KymographClear and KymographDirect (Mangeol et al., 2016). (C) A representative kymograph of microtubule depolymerization simulated using the model of nucleotide exchange solely at the microtubule ends. (D) A representative kymograph of microtubule depolymerization simulated using model of nucleotide exchange throughout the entire microtubule lattice in addition to the exchange at the ends. The color map represents the percentage of exchanged sites on the microtubule. The “zoomed in” regions on the kymograph show the tubulin dimers within a 160 nm long segment on the microtubule lattice at three different time points (0, 2.5 and 5 min). Over the course of time, more and more tubulin dimers are exchanged (“green” sites turning into “blue” sites) within the microtubule lattice. (E and F) The time dependence of depolymerization rate at microtubule minus (E) and plus (F) ends as observed in experiments. In E and F, the gray curves denote trajectories obtained from individual microtubules in experiment (N = 20). The blue curve denotes the average curve evaluated from gray trajectories. (G and H) The time dependence of depolymerization rate at microtubule minus (G) and plus (H) ends as obtained from the model of nucleotide exchange solely at the microtubule ends. (I and J) The time dependence of depolymerization rate at microtubule minus (I) and plus (J) ends as obtained from the model of nucleotide exchange throughout the entire microtubule lattice in addition to the exchange at the ends. In G–J, the light green curves denote trajectories obtained from individual microtubules in simulation. The dark green curve denotes the average curve evaluated from individual microtubule trajectories in simulation. The instantaneous depolymerization rates presented in E–J were estimated by averaging over a 1-min time window. The error bars represent SEM (N = 20 in experiment, N = 100 in all simulated cases).