Head of myosin IX binds calmodulin and moves processively toward the plus-end of actin filaments

Abstract

Mammalian myosin IXb (Myo9b) has been shown to exhibit unique motor properties in that it is a single-headed processive motor and the rate-limiting step in its chemical cycle is ATP hydrolysis. Furthermore, it has been reported to move toward the minus- and the plus-end of actin filaments. To analyze the contribution of the light chain-binding domain to the movement, processivity, and directionality of a single-headed processive myosin, we expressed constructs of Caenorhabditis elegans myosin IX (Myo9) containing either the head (Myo9-head) or the head and the light chain-binding domain (Myo9-head-4IQ). Both constructs supported actin filament gliding and moved toward the plus-end of actin filaments. We identified in the head of class IX myosins a calmodulin-binding site at the N terminus of loop 2 that is unique among the myosin superfamily members. Ca(2+)/calmodulin negatively regulated ATPase and motility of the Myo9-head. The Myo9-head demonstrated characteristics of a processive motor in that it supported actin filament gliding and pivoting at low motor densities. Quantum dot-labeled Myo9-head moved along actin filaments with a considerable run length and frequently paused without dissociating even in the presence of obstacles. We conclude that class IX myosins are plus-end-directed motors and that even a single head exhibits characteristics of a processive motor.

FIGURE 1.

Design and purification of C. elegans Myo9 constructs. A, schematic representation of the expressed C. elegans Myo9 constructs Myo9-head and Myo9-head-4IQ. B and C, purified and biotinated Myo9-head and Myo9-head-4IQ proteins. Purified proteins were separated by SDS-PAGE and either stained with Coomassie Blue (lanes 2) or transferred to a membrane and visualized with streptavidin-horseradish peroxidase (lanes 3). Lane 1, molecular mass markers; the positions of the heavy chains Myo9-head, Myo9-head-4IQ, and the calmodulin light chain (CaM) are indicated.

FIGURE 2.

Calmodulin co-purifies with the Myo9-head. A, purified Myo9-head and authentic calmodulin either boiled for 5 min or not and centrifuged. The supernatants of unboiled Myo9-head (HeadUB), boiled Myo9-head (HeadB), and authentic boiled calmodulin (CaMB) were separated by SDS-PAGE in standard sample buffers containing either 5 mm Ca²⁺ or 2 mm EGTA. The positions of Myo9-head heavy chain and calmodulin (CaM) are indicated. B, comparison of the peptide masses obtained from the 17-kDa protein that was co-purified with Myo9-head and peptides from rat calmodulin. Numbers in parentheses indicate the positions of the first displayed amino acid in the calmodulin sequence.

FIGURE 3.

N-terminal region of Myo9 loop 2 binds calmodulin. A, schematic overview of Myo9-head fragments that were fused to GST and expressed in E. coli. B and C, cell homogenates separated on SDS-PAGE, transferred to polyvinylidene difluoride membrane, and either incubated with anti-GST antibody (B) or biotinated calmodulin in the presence of 0.1 mm CaCl₂ (C). Binding was visualized by chemiluminescence using either secondary antibodies or streptavidin that was coupled to horseradish peroxidase. Lane 1, GST; lanes 2–11, fusion proteins indicated in A.

FIGURE 4.

F-actin-activated Mg²⁺-ATPase activity of Myo9 constructs is sensitive to calcium. The steady-state ATPase activities of the Myo9-head (0.1–0.2 μm) (A) and the Myo9-head-4IQ (0.4–0.5 μm) (B) were measured in the absence (filled squares) or presence (open squares) of 50 μm free Ca²⁺ in assay buffer containing 20 mm Hepes, pH 7.4, 50 mm KCl, 2 mm MgCl₂, 1 mm EGTA, 10% glycerol, 1 mm dithiothreitol, 10 μm exogenous calmodulin, and 0–20 μm F-actin at 20 °C using a NADH-coupled assay. Data are the means from three or four independent preparations, and the error bars indicate S.E. The data points were fitted with a hyperbola. Please note the different scales for the ordinates in A and B.

FIGURE 5.

Myo9 moves to the plus-end of actin filaments. Movement of dual-fluorescence-labeled actin filaments was supported by Myo9-head (A), Myo9-head-4IQ (B), and myosin II HMM (C). Individual frames from time-lapse movies are shown on the left. Elapsed time (seconds) is indicated in each panel. Scale bars, 2 μm. The green tip marks the plus-end of an actin filament. Myo9-head, Myo9-head-4IQ, and myosin II HMM move filaments with their green tips trailing, indicating that both Myo9 and myosin II are plus-end-directed motors (see also supplemental movies). Histograms on the right display the number of filaments moving with the plus-end trailing (positive velocities) and leading (negative velocities), respectively. The histograms were fitted with a Gaussian function (gray lines).

FIGURE 6.

Myo9-head exhibits characteristics of a processive motor. A, gliding velocity of actin filaments does not change as a function of Myo9-head density. The surfaces were coated with different concentrations of biotinated Myo9-head. The gliding velocities of actin filaments (n = 5–40) were quantified, and the average velocities (nm s⁻¹) are represented as means ± S.E. on the y axis. B, nodal pivoting of an actin filament at low Myo9-head densities is shown. An actin filament is pivoting around a single contact point (arrow) while it moves unidirectionally. As the end of the actin filament passes the nodal point, the filament diffuses away from the surface (2nd to last panel). The last panel is the overlay of all the frames shown. Elapsed time (seconds) is indicated in each panel. Scale bar, 2 μm.

FIGURE 7.

Movement of qdot-labeled Myo9-head along actin filaments. A, biotinated Myo9-head was mixed with streptavidin-coated qdots at a ratio of 2:1 or 4:1 (qdot/Myo9-head) as indicated. Qdot-labeled Myo9-head was added to immobilized actin filaments as shown in the schematic illustration, and images were recorded every 5 s. Exemplary space-time plots (kymographs) of low traffic (left) and high traffic (right) situations are shown. Arrowhead indicates a qdot passing an obstacle. Scale bars represent 4 μm and 40 s, respectively. B, histograms of run length and velocity are plotted. Velocities fit a Gaussian distribution with a mean of 75 ± 0.9 nm s⁻¹, and the average run length was determined to be 3.1 ± 0.5 μm. Data are derived from six experiments from at least three different preparations.