AAA+ Ring and linker swing mechanism in the dynein motor

Abstract

Dynein ATPases power diverse microtubule-based motilities. Each dynein motor domain comprises a ring-like head containing six AAA+ modules and N- and C-terminal regions, together with a stalk that binds microtubules. How these subdomains are arranged and generate force remains poorly understood. Here, using electron microscopy and image processing of tagged and truncated Dictyostelium cytoplasmic dynein constructs, we show that the heart of the motor is a hexameric ring of AAA+ modules, with the stalk emerging opposite the primary ATPase site (AAA1). The C-terminal region is not an integral part of the ring but spans between AAA6 and near the stalk base. The N-terminal region includes a lever-like linker whose N terminus swings by approximately 17 nm during the ATPase cycle between AAA2 and the stalk base. Together with evidence of stalk tilting, which may communicate changes in microtubule binding affinity, these findings suggest a model for dynein's structure and mechanism.

Figure 1

Cytoplasmic Dynein Motor Domain of D. discoideum (A) Heavy-chain sequence diagram showing six AAA+ modules (numbered), N and C sequences, and the MT-binding domain (MTBD). On the right, the cartoon shows the stalk-head-tail architecture of dynein (tail formerly known as the stem). (B and C) Negative-stain EM and single-particle image processing of the motor domain reveals two characteristic appearances: top and right views. (D) Right view of axonemal dynein-c from Chlamydomonas reinhardtii (modified from Burgess et al., 2003) for comparison. The point of emergence of the coiled-coil stalk (B–D, black arrows) and the tail of dynein-c (D, small arrow) are indicated, as well as a spike (C, white arrow) and stain-filled groove (C, black arrowhead) seen in cytoplasmic dynein (compare with D). The scale bar represents 10 nm.

Figure 2

Stalk Structure in the Unprimed Motor Class averages of (A) top and (B) right views showing the stalk at a range of angles (compare left, middle, and right panels) with the MTBD at the distal end (arrowhead), which is often curved to the right. Diffuse stain-excluding areas (small arrows) are the GFP and BFP of this construct (GN-motor-B2). Coiled-coil lengths of 10.3 ± 0.6 nm (top view, mean ± SD, n = 10 classes) and 10.5 ± 0.7 nm (right view, n = 10 classes) were measured as indicated by double-headed arrows. For further details, see legend to Movie S3.

Figure 3

Structural Impacts of Truncation on the Dynein Motor (A) Sequence diagrams of the motor domain and truncation constructs lacking the C sequence (ΔC), the N sequence (ΔN), and both (ΔNΔC). (B) Class averages showing right view of the motor domain alongside ΔC, ΔN, and ΔNΔC constructs. ΔC molecules resembling the motor domain in right view (upper panel) and with the linker undocked (arrowhead, lower panel) are shown. (C) A stalk emerges from each head at the same position (∼11 o'clock). The scale bars represent 10 nm.

Figure 4

Mapping the Locations of the GFP-Based Tags in Dynein Fusion Proteins in the Unprimed Conformation (A) Motor domain sequence (residues 1383–4725) showing sites of insertion of GFP and BFP in the constructs examined (GN, inserted upstream of V1383; B1, after A2172; B2, after S2471; B5, after K3928; B6, after E4261; B7, after S4450; BC, after I4725). (B) Class averages of top and right views showing each tag (arrowheads) near its mean position. The internal position of the B7 tag is shown by difference mapping (see Figure S2 for details) which, because of superposition with the head, means that the BFP could not be shown by image classification. Difference maps are contoured at 5σ above the mean then at intervals of 2σ, superposed on global averages faded for clarity. No right views were obtained for B6- and BC-tagged constructs. (C) Summary of GFP-based mapping. Mean positions of GN, B1, B2, B5, B6, and BC tags and peak of difference maps for B7 tag are shown (colored circles). The scale bars represent 10 nm.

Figure 5

Structural Changes between Unprimed and Primed Conformations (A) Motor tagged with GFP at the N terminus and BFP at the B1 site in apo/ADP and ADP.Vi states. Small panels: representative class averages showing N-terminal GFP and distal linker (arrows). Larger panels: GFP positions detected automatically (white spots). In ADP.Vi-motors in top view, GFP positions are segregated into primed and unprimed positions (dashed line; see B). (B) Histograms showing angular position of N-terminal GFP measured clockwise relative to an axis passing through the head center and the base of the stalk (cartoon). Mean GFP angle in right view is 10° ± 11° (n = 375) in apo/ADP compared to 145° ± 11° (mean ± SD, n = 557 molecules) in ADP.Vi. Gaussian fits to the top-view data in ADP.Vi give two peaks for GFP angles of 41° ± 18° (mean ± SD) and 124° ± 25°, which intersect at 80° (dashed line). The low-angle peak coincides with that of unprimed motors (also 41° ± 18°). The numbers of molecules in top views are 9,964 (apo/ADP) and 13,527 (ADP.Vi). (C) Average stalk angles show a clockwise tilt between unprimed and primed motors (gray and magenta arrowheads, respectively). Stalk angles: right view: unprimed −16° ± 9° (mean ± SD, n = 3858 molecules), primed 0° ± 8° (n = 526); top view: unprimed −6° ± 6° (n = 1604), primed −4° ± 6° (n = 1667). White spots show positions of distal coiled coil in each of the ten classes used to obtain these values (see also Movie S3). Bifurcation of the stalk is indicated (arrow). Right-view unprimed motor is tagged with BFP at the B2 site rather than at the B1 site. The scale bars represent 5 nm.

Figure 6

Magnitude and Direction of the Linker Swing In this figure, right and top views of the dynein head are related by a rotation about the y axis (see Figure S3 for details): each view has been rotated in the plane of the page relative to previous figures so that tag positions move only in the x direction (dashed lines). (A) Tag positions in the unprimed motor are indicated by colored spots (as in Figure 4C). (B) For the primed motor, the mean position of the N-terminal GFP tag (brighter magenta spots; derived for top view as explained in the main text) also moves only in the x direction. Magnitudes of the displacements from the unprimed positions (faded magenta spots) are indicated (arrows). (C and D) Top and right views are related by a rotation of between 50° and 116° (see Figure S4 for details): illustrated here is a rotation of 90° (see also Movie S4). Tag positions in the unprimed conformation (C) and N-terminal GFP in the unprimed and primed conformations (D). In each case, the bottom surface of the cube shows their projected positions perpendicular to top and right views. The black arrow shows movement of N-terminal GFP during the priming stroke, which has a major component perpendicular to the plane of the top view and parallel to the plane of the right view. The 3D displacement of N-terminal GFP lies between 18.8 and 21.1 nm (depending on the rotation angle between right and top views).

Figure 7

Model for the Structure and Priming Stroke of Dynein (A) Six AAA+ modules (numbered) form a hexameric ring. The C sequence (translucent black) is represented speculatively as an elongated structure interacting with (one or another face of) AAA6, AAA5, and AAA4. The N sequence (magenta) contains the linker which runs from AAA1 across the head to AAA4 (yellow) in the unprimed conformation and switches to a position close to AAA2 in the primed conformation, thereby moving the N terminus of the motor by ∼17 nm (in right view in a plane parallel to the page). The N sequence may also contain nonlinker structures, such as near the junction with AAA1 (magenta ellipse). Tilting of the stalk, shown here to occur entirely in the plane of the page in right view (see Figure S3), displaces the MTBD by ∼5 nm and could shift the registration of the two α helices of the coiled coil (indicated by red and blue spots). Stalk tilting perpendicular to the page is not seen in top-view data (Figure 5C), probably because the stalk flattens down onto the EM grid in this orientation. (B) Model to illustrate how the linker swing and stalk tilt could produce one of dynein's larger displacements along an MT (see Discussion for further details). The attachment geometry proposed here (i and iii) is compatible with Mizuno et al. (2007) and uses the right view, which ensures that movement of the linker N terminus occurs in a plane parallel to the MT axis. With the linker N terminus initially restrained (red asterisk), perhaps by attachment to the second head as observed in dimeric dynein-f (Kotani et al., 2007), ATP binding (ii) causes MT detachment and the priming stroke which displaces the MTBD along the MT, here by 24 nm. Subsequent reattachment and powerstroke (iii) displaces the linker N terminus by 24 nm (green asterisk). α-β tubulin dimers are shown (to scale) as pairs of dark and light gray spheres with red and green dimers showing initial and final binding sites, respectively.