Binding partners regulate unfolding of myosin VI to activate the molecular motor

Figures

• 

  Figure 1.. Backfolding of the myosin VI tail. (A) Cartoon depiction of the key regions of MVI, as discussed in the text. UI Unique Insert; 3HB Three Helix Bundle; SAH Stable Alpha Helix; CC Coiled-coil; CBD Cargo Binding Domain. This highlights position of the large insert (LI), small insert (SI), along with NDP52 and Dab2 binding sites. (B) FRET titration of Alexa555-CBD against 1 µM of FITC-MVITAIL(LI), large insert (LI) containing MVITAIL (residues 814–1091), in the presence of 10 µM NDP52 (dimer) or 10 µM tDab2. (C) FRET titration of Alexa555-CBD against 1 µM of FITC-MVITAIL(NI) non-insert (NI) MVITAIL (residues 814–1060), in the presence of 10 µM NDP52 (dimer) or 10 µM tDab2. All titration data fitting was performed as described in Methods (error bars represent SEM from three independent experiments). (D) Schematic representation of FRET assay to measure backfolding of the MVITAIL. (E) Representative fluorescence spectra of 1 µM GFP-MVITAIL(NI)-RFP ± 5 µM NDP52 (dimer), or 20 µM tDab2. (F) Representative fluorescence spectra of 1 µM GFP-MVITAIL(LI)-RFP (right) ± 5 µM NDP52 (dimer), or tDab2.

• 

  Figure 2.. NDP52 association kinetics with the myosin VI tail. (A) Schematic of the experiments for rapid mixing of the FITC labelled MVITAIL and Alexa-Fluor555 labelled NDP52. Experiments were performed as described in the methods. (B) Representative stopped-flow fluorescence trace depicting both changes in FRET signal, as described in the text for 10 µM NDP52 (dimer). Each process was then analysed separately over the corresponding time scales. (C) Representative stopped-flow fluorescence traces and exponential fitting for the initial increase in fluorescence at the stated concentrations of NDP52 dimers. (D) Representative stopped-flow fluorescence traces and exponential fitting for the decrease in fluorescence at the stated concentrations of NDP52 dimers. (E) The individual traces were fitted to single exponentials and the dependence of the rate constants on concentration was then fitted to a straight line, as shown. The points shown are averages of at least three measurements, where error bars represent SEM. The fit for the first phase gives a slope of 1.65 µM−1 s−1 and an intercept of 3.2 s−1. The second phase is independent of NDP52 concentration and gives an average rate constant of…

• 

  Figure 3.. Dimerization of myosin VI. (A) Cartoon depiction of the key domains of MVI which are used in the ATPase and size-exclusion chromatography measurements. (B) Representative SECMALS traces for 2 mg/ml NDP52 and tDab2 giving molecular weights of 102 kDa and 17 kDa, respectively. A proportion of NDP52 runs outside of the column volume (Supplementary Figure S2). This corresponds to dimeric NDP52 and monomeric tDab2. (C) Michaelis–Menten plot displaying steady-state actin-activated ATPase activity for the MVI constructs. Error bars represent SEM from three-independent experiments. (D) Representative SEC traces for 1 mg/ml of the MVI constructs. (E) Representative SEC traces for 1 mg/ml of the MVITAIL constructs.

• 

  Figure 4.. Binding partner driven dimerization of Myosin VI. (A) Schematic of the experiments for rapid mixing of premixed FITC-MVITAIL and Alexa-Fluor555-MVITAIL with unlabelled NDP52. Experiments were performed as described in the methods. Representative stopped-flow fluorescence traces and exponential fitting to the fluorescence increase, when the two labelled pools of MVITAIL are mixed with the stated concentrations of NDP52 dimers. (B) The individual traces were fitted to single exponentials and the dependence of the rate constants on concentration was then fitted to a straight line, as shown. The points shown are averages of at least three measurements and error bars represent SEM. The rate constants are independent of NDP52 concentration, with an average value of 2.2 s−1. (C) Cartoon depicting binding partner driven dimerization based upon the experiments in Figures 2, 3 and 4. Both pools of labelled MVITAIL are folded. NDP52 binds to either Tail (i) and then triggers their unfolding (ii). This then enables dimerization of the MVITAIL domains (iii).

• 

  Figure 5.. Cellular clustering of myosin VI and NDP52. (A) Widefield Immunofluorescence imaging against endogenous NDP52 (green) and MVI (red) in the presence and absence of TIP, and JF549 staining against transfected Halo-MVI and Halo-MVI(RRL/AAA) in the cytoplasm of HeLa cells (Scale bar 10 µm). (B) STORM render images of the cells shown in (A) (Scale bar 10 µm). Images were acquired as described in the methods. (C) Depiction of a theoretical example of molecular clustering and random distribution. Molecular clustering is assessed by the linearized plot of Ripley's K function L(r)-r, versus r, where r is the radius. Distribution can be plotted as an L-function, where a randomly distributed set of molecules is a flat line with a value of zero. The green curve corresponds to the organization of NDP52 (peak at r = 270 nm) and the red curve corresponds to the organization of MVI (peak at r = 120 nm). (D) Cluster maps based upon the STORM render in (B). Clusters are shown in Red (NDP52) and Green (MVI). (E) Cluster analysis representing the percentage of molecules in a cluster for the conditions in (B). Individual data points correspond to the average number of molecules per cluster in the…

• 

  Figure 6.. Cluster analysis and colocalization of myosin VI and NDP52. (A) Examples of widefield immunofluorescence imaging against endogenous NDP52 (green) and MVI (red) in the presence and absence of TIP in HeLa cells (Scale bar 10 µm), with their corresponding STORM render. Images were acquired as described in the methods. (B) Cluster colocalization heatmaps corresponding to the STORM renders shown in (A), depicting the colocalization scores for each molecular cluster for MVI and NDP52. Here, values of 1 (dark red) correspond to perfectly colocalized clusters and −1 (dark blue) to clusters separated from each other. (C) Fraction of MVI colocalized with NDP52 in the presence and absence of TIP. (D) Results of cluster analysis of MVI ± TIP and NDP52, displaying the distribution and mean values of the number of clusters per cell ROI, number of molecules per cluster and cluster area. The data are broken down into MVI and NDP52 colocalized and non-colocalized clusters. Individual data points correspond to the corresponding average value per cell ROI. The values represent the mean (±s.d.) from the ROIs (n = 20) for each condition (**** P < 0.0001 by two-tailed t-test compared with…

• 

  Figure 7.. Live cell single-molecule dynamics of myosin VI. (A) Example render of 3D single-molecule trajectories for JF549-labelled Halo-tagged MVI, stably expressed in HeLa cells, under normal conditions. (B) Plot of diffusion constants for wild type (WT) MVI (normal), MVI WWY/WLY (WWY), MVI RRL/AAA (RRL), WT MVI following transient overexpression of GFP-CBD (CBD) and WT MVI following treatment with latrunculin B (LatB). Diffusion constants were derived from fitting trajectories to an anomalous diffusion model, as described in methods. The data represent all trajectories from 100 cells (* P < 0.05, **** P < 0.0001 by two-tailed t-test compared with normal conditions). (C) Diffusion constants (D) were split into three categories corresponding to static/slow moving (D < 0.1 µm2 s−1), mobile (0.1 < D < 2 µm2 s−1) and hyper-mobile fractions (D > 2 µm2 s−1). The percentage of molecules from the experiments in (B) falling into each group were then plotted. (D) Plot of anomalous diffusion alpha values derived from the same experimental set shown in (B), for the indicated conditions. (**** P < 0.0001 by two-tailed t-test compared with normal conditions). (E) Roseplot representing the angular…

• 

  Figure 8.. Binding partners enhance actin binding. (A) Actin pull-down of 5 µM myosin VI wild type (WT) or mutant RRL/AAA in the presence and absence of 15 µM NDP52. P and S represent pellet and supernatant, respectively. (B) Quantification of pulldowns in A. (C) Representative images of transiently transfected eGFP-MVI wild type (WT), RRL/AAA and WWY/WLY MVI in HeLa cells combined with phalloidin staining against actin (magenta). Scale bar 10 µm.

• 

  Figure 9.. Model describing the activation of myosin VI by binding partners. (A) Two routes of binding partner-dependent dimerization of MVI. (i) Dimeric binding partner (e.g. NDP52) binds to backfolded MVI and triggers its unfolding. A second MVI molecule is then recruited through the binding partner, due to its dimeric nature. This results in a stoichiometry of one dimeric partner to two MVI motors. This mechanism does not require MVI internal dimerization. (ii) Each MVI molecule is unfolded by an individual binding partner. Unfolding then exposes internal dimerization sequences within MVI, which can result in a dimer complex with a stoichiometry of two partners and two motors. The data presented here allow option (i) to be excluded. Our data also allows us to propose that the folded state is inactive (non-actin bound) and that the unfolded state (monomer and dimer) becomes active with the ability to bind actin. (B) Recruitment and activation of myosin VI to cellular cargo by binding partners. Monomeric backfolded (inactive) myosin VI (red) can be found freely diffusing in the cytoplasm or bound to actin (dark red). The inactive myosin VI associates with cargo-bound binding partners. This…