Cytoplasmic dynein regulates its attachment to microtubules via nucleotide state-switched mechanosensing at multiple AAA domains

Abstract

Cytoplasmic dynein is a homodimeric microtubule (MT) motor protein responsible for most MT minus-end-directed motility. Dynein contains four AAA+ ATPases (AAA: ATPase associated with various cellular activities) per motor domain (AAA1-4). The main site of ATP hydrolysis, AAA1, is the only site considered by most dynein motility models. However, it remains unclear how ATPase activity and MT binding are coordinated within and between dynein's motor domains. Using optical tweezers, we characterize the MT-binding strength of recombinant dynein monomers as a function of mechanical tension and nucleotide state. Dynein responds anisotropically to tension, binding tighter to MTs when pulled toward the MT plus end. We provide evidence that this behavior results from an asymmetrical bond that acts as a slip bond under forward tension and a slip-ideal bond under backward tension. ATP weakens MT binding and reduces bond strength anisotropy, and unexpectedly, so does ADP. Using nucleotide binding and hydrolysis mutants, we show that, although ATP exerts its effects via binding AAA1, ADP effects are mediated by AAA3. Finally, we demonstrate "gating" of AAA1 function by AAA3. When tension is absent or applied via dynein's C terminus, ATP binding to AAA1 induces MT release only if AAA3 is in the posthydrolysis state. However, when tension is applied to the linker, ATP binding to AAA3 is sufficient to "open" the gate. These results elucidate the mechanisms of dynein-MT interactions, identify regulatory roles for AAA3, and help define the interplay between mechanical tension and nucleotide state in regulating dynein motility.

Keywords: AAA+ ATPases; cytoplasmic dynein; mechanosensing; microtubules; optical tweezers.

Fig. 1.

Dynein–MT bond anisotropy. (A) Model for tension-based regulation of dynein stepping. Splaying of the dynein heads generates intramolecular tension. Under backward tension (front head) MT binding strength is greater, and under forward tension (rear head) it decreases. (B, Left) A polystyrene bead bearing a dynein motor is held in an optical trap as the microscope stage sweeps back and forth parallel to a MT (not to scale). (Right) The motor binds the MT, pulling the bead out of the trap. Force on the motor increases until the dynein–MT bond ruptures at the “unbinding force” (red arrow), here ∼3 pN. (C) Primary and secondary unbinding events. Event 1 is a primary event, beginning from zero force. Secondary events (2 and 3) occur when the motor rebinds the MT before returning to the trap center. These events begin with preload F_start and unbind again at F_end, with force difference ΔF = F_start − F_end. (D) Force (position) vs. time for WT dynein in the apo state. The inserted trace segment corresponds to the data for the period marked by the thick black line. Orange and blue shaded areas show periods of applied backward and forward tension, respectively (loading rate: 5.6 pN/s; k ∼ 0.07 pN/nm, v_stage ∼ 80 nm/s). (E) Normalized histograms of primary forward (n = 575) and backward (n = 512) unbinding forces, with mean values noted above the histograms. Tall vertical bands represent 95% CIs of the means (forward: [1.7, 1.8] pN, backward: [3.1, 3.6] pN) estimated by bootstrapping 4,000 samples. (Inset) ECDFs for the forward vs. backward directions. (F) Mean ΔF vs. F_start for forward (blue) and backward (orange) tension. Events grouped into 1-pN bins for F_start. Shaded regions: 95% CIs for the mean ΔF, estimated by bootstrapping 2,000 samples. For F_start ≳10 pN (gray shaded region), the trap stiffness is not constant (SI Appendix, SI Text and Fig. S4) and 20 or fewer events were recorded. For F_start ≤ 10 pN, each mean ΔF was calculated from 36 to 770 measurements. (G) Unbinding rate vs. loading force derived from the data in E (see SI Appendix, SI Text for details). The shaded areas represent 95% CIs for the mean rates, estimated by bootstrapping 4,000 samples.

Fig. 2.

Effect of 1 mM ATP on dynein’s response to linker-applied tension. (A, Left) Schematic of dynein with GFP fused to the N terminus. (Right) WT dynein forward and backward unbinding forces with mean values noted. Tall vertical bands represent 95% CIs of the means (forward: [1.4, 1.5] pN, backward: [2.6, 2.9] pN) estimated by bootstrapping 4,000 samples. (B) As in A, but for the AAA1 K/A mutant (95% CIs [1.5, 1.9] and [2.5, 3.4] pN). (C) Example of optical trapping data for the AAA1 E/Q mutant. Black trace shows raw data. Green trace shows a fifth-order Savitzky–Golay filter with a 301-sample (0.1-s) window applied to the data to make unbinding events easier to identify. The detail shows the filtered data for the period marked by the thick black line. (D) As in A, but for the AAA1 E/Q mutant (95% CIs [0.7, 0.8] and [0.9, 1.0] pN). (E) As in A, but for the AAA3 E/Q mutant (95% CIs [1.2, 1.4] and [2.4, 2.8] pN). (F) As in A, but for the AAA1 E/Q + AAA3 E/Q mutant (95% CIs [0.8, 0.9] and [0.8, 0.9] pN). (G) As in A, but for the AAA1 E/Q + AAA3 K/A mutant (95% CIs [1.4, 1.7] and [2.9, 3.8] pN). Number of events in the forward, backward directions: (A) (577, 577), (B) (60, 59), (D) (320, 365), (E) (272, 294), (F) (274, 292), and (G) (77, 67).

Fig. 3.

Primary unbinding forces for the WT (A) and the AAA3 E/Q mutant (B) with GFP fused to the C terminus in the presence of 1 mM ATP. Tension is applied via the C terminus (loading rate: 5.6 pN/s). (A, Left) Schematic of dynein with GFP fused to the C terminus. (Right) Histogram of WT dynein forward (blue) and backward (orange) unbinding forces, with the respective mean values noted above each histogram. Tall vertical bands represent 95% CIs of the means (forward: [0.8, 1.1] pN, backward: [0.9, 1.2] pN) estimated by bootstrapping 4,000 samples. (B) As in A, but for the AAA3 E/Q mutant (95% CIs [1.6, 1.8] and [2.7, 3.3] pN). Number of events in the forward, backward directions: (A) (95, 98) and (B) (228, 229).

Fig. 4.

Effect of ADP on dynein’s response to linker-applied tension. (A, Left) Schematic of dynein with GFP fused to the N terminus. (Right) Optical trapping data. The inserted trace segment corresponds to data for the period marked by the thick black line. (B) Histogram of WT dynein forward (blue) and backward (orange) unbinding forces measured in the presence of 2 mM ADP, with the respective mean values noted above each histogram. Tall vertical bands represent 95% CIs of the means (forward: [1.4, 1.5] pN, backward: [2.3, 2.5] pN. (C) As in B, but for the AAA1 K/A mutant (95% CIs [1.1, 1.3] and [1.6, 2.0] pN). (D) As in B, but for the AAA3 K/A mutant (95% CIs [1.6, 1.8] and [3.5, 4.1] pN). Number of events in the forward, backward directions: (B) (996, 869), (C) (325, 387), and (D) (439, 369).

Fig. 5.

Mean forward (blue) and backward (orange) unbinding forces for the various experiments reported. Error bars denote 95% CIs of the mean. The labels on the abscissa denote the experimental condition tested (Top row) and the constructs used (Bottom row).