Visualization of single molecules of dynein using HILO microscopy. (a) Left: Confocal microscopy images (with a 20× objective) of anti-m/hDHC (rabbit polyclonal #PA5-68173 primary and donkey anti-rabbit A555 #A32794 secondary) in WT (left) and mDHC-GFP cells (right); the images are represented in the color map indicated on the bottom left and the GFP channel is provided on the bottom right. Right: Plot of mean fluorescence intensity of the anti-m/hDHC in WT and mDHC-GFP cells normalized to the respective no treatment (“−”) values (N = 3 independent experiments, with eight fields imaged per condition; error bars represent SEM). Following hDHC RNAi, WT cells showed normalized anti-m/hDHC intensity of 0.64 whereas DHC-GFP cells had 0.84, indicating that mDHC-GFP constitutes 20% of the dynein population in mDHC-GFP cells. (b) HILO microscopy setup used to visualize dynein molecules in HeLa cells. Depending on the morphology of the cell, the angle β was adjusted such that fluorescent spots of dynein were visible (top, “Front view”). The illumination diameter was kept constant at 30 µm using a field stop (bottom, “Top view”). This resulted in a beam thickness of ∼4.6 µm (calculated according to Tokunaga et al. [2008]). (c) (i) Spinning disk confocal microscopy image of a HeLa cell expressing mDHC-GFP and (ii) HILO microscopy image of the same cell that is partially illuminated in the HILO microscopy setup. The blue lines represent the orientation of the incident laser beam. In the HILO microscopy image, individual fluorescent spots are visible. (d) Box plot for comparison of mean intensities of cells where single-molecule events were visible (“Tracked”), and those where they were not (“Not tracked”). Data are from n >30 cells, n = 3 independent experiments; asterisk indicates P < 0.05, Mann-Whitney Test). (e) HILO microscopy image (left) from a 10-s-long time-lapse video of mDHC-GFP in HeLa cells and the corresponding kymograph (right). Representative stationary, minus end–directed, and plus end–directed events are indicated with the white, teal, and magenta arrowheads, respectively, in the kymograph and in the insets below. (f) HILO microscopy image (left) from a 10-s-long time-lapse of mDHC-GFP in cells treated with 10 µM nocodazole to depolymerize the MTs and the corresponding kymograph (right). The kymograph shows few binding events compared to the cell shown in e, indicating that dynein molecules were stochastically binding to the MTs from the cytosol. (g) HILO microscopy image (left) from a 20-s-long 1 fps time-lapse video of mDHC-GFP in HeLa cells and the corresponding kymograph (right). There are no distinct traces in the kymograph, indicating that the short traces shown in e are not artifacts and dynein molecules do not interact with the MTs for a long duration. (h) Immunofluorescence images of MT (magenta) and EB1 (green) obtained using spinning disk microscopy + SRRF. In these cells with a large aspect ratio, a majority of the MTs are plus end out (pointed by yellow arrowheads). While there might be misoriented MTs or short MTs oriented with their plus ends toward the cell center, these are likely a minority, given the higher intensity of EB1 at the cell periphery compared to the center. In all analyses, movement toward the nucleus was considered as minus end–directed transport and movement away from the nucleus as plus end directed. The plot on the bottom represents quantification of EB1 intensity at the cell periphery (farthest ends of cells covering a quarter of the cell area) and cell center (region close to the nucleus covering a quarter of the cell area; note that in our HILO movies, half a cell spanning the nucleus to the cell tip is typically visible). Data from n = 14 cells, n = 1 independent; asterisk represents P < 0.05, Wilcoxon test for paired data. (i) Histogram of velocities of mDHC-GFP in HeLa cells (data from Fig. 1 e). (j) Temporally color-coded projection of mCherry-tubulin in a 3 min 15 s video obtained using HILO microscopy with the same settings employed for single-molecule dynein imaging. The fact that the MT signal from 00:00 and 03:15 overlap significantly indicates that the MT position and dynamics do not typically vary during a typical single-molecule time-lapse movie, which is ∼10 s. Moreover, a significant length (∼20 µm) of the MT network was visible in HILO microscopy images, indicating that dynein molecules moving on MTs over long distances could be visualized and tracked. (k) HILO microscopy image from a time-lapse video of mDHC-GFP in a cell expressing high levels of mDHC-GFP. Clusters of dynein that are likely at the MT plus end are indicated by the yellow arrowheads. (l) An exponential decay fit to the intensity versus time plot of spots similar to those indicated with yellow arrowheads in k. λ = 0.17 was obtained to give the time constant 1/λ = 5.9 s. This time constant represents the average time required for a dynein molecule to bleach under our imaging conditions. This value is an order of magnitude higher than the average residence time of dynein on MTs (∼0.59 s). Data from n = 119 spots, N = 1 independent experiment with 25 cells. Error bars represent SEM. In c, e–h, and k, “N” marks the location/direction of the nucleus.