Inactive MCAK tip tracks in interphase and mitotic cells. (A) The edge of a cell transfected with GFP-MCAK. (A′) Three-dimensional reconstruction of 30 frames of MCAK tip tracks. Successive moving tips are visible (black arrows). (B) GFP-ML-MCAK tracks at the edge of an interphase cell. (B′) Reconstruction of 30 frames of motorless MCAK show uniform lines (rather than successive tips) because the tips are consistently longer than is seen with GFP-MCAK. (C) Mitotic cell transfected with GFP-hypir-MCAK. Tip tracking is evident in astral MTs (white arrows).

Tip tracking is positively correlated with apparent lattice affinity rather than tubulin affinity. (A) Wild-type MCAK binds MT lattice in the absence of ATP (left, lane 1). Excess free tubulin dimer can compete off 13–15% of lattice-associated wtMCAK (left, lanes 2 and 3). Phosphorylation of MCAK significantly decreases the apparent affinity of wtMCAK for stabilized MTs (right, lane 1). Excess free tubulin competes off a further 2–9% of lattice-bound phosphorylated MCAK (right lanes 2 and 3). (B) MCAK-Q710 binds MT lattice in the absence of ATP (left, lane 1). Excess tubulin dimer can compete off 30–45% of lattice-bound MCAK-Q710 (left, lanes 2 and 3). Phosphorylation of MCAK-Q710 decreases the apparent affinity of MCAK-Q710 for MT lattice to a similar extent as is seen with wtMCAK (right, lane 1). Excess free tubulin competes off a further 9–11% of lattice-bound phosphorylated MCAK-Q710 (right lanes 2 and 3).

MCAK exhibits tip tracking behavior in living cells. (A) GFP-MCAK can be detected at the ends of MTs in an interphase HeLa cell. (B) Kymograph of tips traveling from the centrosomal region to the edge of the cell. (C) RFP-MCAK roughly colocalizes with GFP-EB1 in live HeLa cells cotransfected with both constructs.

Tip tracking of MCAK protein is dependent on phosphorylation. (A) GFP-AAAAA-MCAK binds to MT tips (left), whereas GFP-EEEEE-MCAK is not found on tips (right). See also Videos 4 and 5 (available at http://www.jcb.org/cgi/content/full/jcb.200411089/DC1). (B) Two fields of CHO cells labeled for endogenous MCAK (B) and MTs (B′). The top field of CHO cells are control cells. The bottom field of CHO cells were cultured for 5 h in 10 μM roscovitine and then fixed. Increased association of endogenous MCAK with distal ends of MTs is evident (B, bottom; arrows). (C) MCAK protein tracks on tips (arrows) in a HeLa cell transfected with RFP-MCAK and cultured subsequently in 10 μM roscovitine (Video 6).

Models to explain the presence of MCAK on polymerizing MT tips. (A) EB1 is capable of antagonizing modest levels of MCAK's depolymerizing activity. Cultured cells were transfected with GFP-EB1, RFP-MCAK, or both. Cells expressing low levels of MCAK have a significantly reduced amount of MT polymer relative to control cells (*, T = 0. 0042, P = 0.05). In contrast, there is no significant difference between control levels of MT polymer and low MCAK+EB1 (**, T = 0.4345, P = 0.05). Thus, at low levels of MCAK expression, overexpressed EB1 is capable of restoring bulk MT polymer to control levels. (B) Competition model: EB1 and MCAK compete for high affinity binding sites at the end of the MT. EB1 prevents MCAK from achieving a quorum of occupied protofilaments. (C) The NH₂ and COOH terminus interact with MT ends, relieving the MCAK motor domain from inhibition by the COOH terminus and putting it in a state in which the motor becomes receptive to lattice interactions.