Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Jun 1;90(5):1000-15.
doi: 10.1016/j.neuron.2016.04.046. Epub 2016 May 19.

The Dynamic Localization of Cytoplasmic Dynein in Neurons Is Driven by Kinesin-1

Affiliations
Free PMC article

The Dynamic Localization of Cytoplasmic Dynein in Neurons Is Driven by Kinesin-1

Alison E Twelvetrees et al. Neuron. .
Free PMC article

Abstract

Cytoplasmic dynein, the major motor driving retrograde axonal transport, must be actively localized to axon terminals. This localization is critical as dynein powers essential retrograde trafficking events required for neuronal survival, such as neurotrophic signaling. Here, we demonstrate that the outward transport of dynein from soma to axon terminal is driven by direct interactions with the anterograde motor kinesin-1. In developing neurons, we find that dynein dynamically cycles between neurites, following kinesin-1 and accumulating in the nascent axon coincident with axon specification. In established axons, dynein is constantly transported down the axon at slow axonal transport speeds; inhibition of the kinesin-1-dynein interaction effectively blocks this process. In vitro and live-imaging assays to investigate the underlying mechanism lead us to propose a new model for the slow axonal transport of cytosolic cargos, based on short-lived direct interactions of cargo with a highly processive anterograde motor. VIDEO ABSTRACT.

Figures

Figure 1
Figure 1
Dynein-GFP Accumulates in the Axon Terminals of Primary Hippocampal Neurons throughout Development (A) Models for dynein transport in the axon: (i) carrying extra dynein motors on vesicles; (ii) recruitment and surfing of microtubule +TIPs; and (iii) direct transport by kinesin. (B–E) Immunofluorescence and confocal microscopy of dynein-GFP primary hippocampal neurons. (B) Dynein-GFP accumulates in tau-positive axons at 8 DIV (small arrowheads), with particular enrichment in the axon terminals (empty arrowheads). Dynein-GFP accumulation does not colocalize with the presynaptic marker SV2 at 8 DIV (Figure S1A) or 21 DIV (C, empty arrowheads), or with the postsynaptic marker PSD-95 (Figure S1B). (D) Dynein-GFP is colocalized with anti-DIC (colocalization appears white, neuron shown at 12 DIV). (E) Dynein-GFP is colocalized with GAP43 at 3 DIV (filled arrowhead, GAP43-positive terminal; empty arrowhead, GAP43-negative terminal). (F) FRAP analysis of dynein-GFP localization to axon terminals shows recovery of distally accumulated dynein takes >20 min (left). Quantification (right) shows ratio of the mean intensity of the distal 5 μm, d, over the more proximal 5 μm region, p, 50 μm away. (G) Left: first frame, kymograph, and last frame of a dynein-GFP hippocampal axon imaged using near-TIRF. Scale bar indicates anterograde direction. Middle: selected events highlighted for clarity. Right: enlargement of boxed region. Arrowheads point to start (empty) and end (filled) of selected anterograde events. (H) Left: first frame, kymograph and last frame of a dynein-GFP hippocampal axon imaged by near-TIRF after bleaching. Scale bar indicates anterograde direction. Right: selected events highlighted for clarity.
Figure 2
Figure 2
Dynein Accumulates in the Growth Cone Coincident with Axon Specification (A) Still images from Movie S1 showing neurite outgrowth in dynein-GFP stage 2 neuron. Filled arrowhead, neurite with highest dynein-GFP intensity; empty arrowheads, other neurites. Long arrow shows direction of axon exit. Time stamps, hours:minutes; fluorescence intensity scale, bottom left. (B) Maximum projection of Movie S1 with neurite labeling used in (C). (C) Individual neurite integrated density as a percentage of the total in all neurites from Movie S1 through time. Neurites labeled as in (B). (D) Maximum projection of Movie S2, dynein-GFP neuron transfected with K560-Halo, indicating neurite labeling for (E) and (F). (E) Still images of Neurite 2 (see D) through time. Quantification bars show relative integrated density of K560 (orange) and dynein-GFP (green) within Neurite 2 over time. (F) Individual neurite integrated density as a percentage of the total in all neurites from Movie S2 through time. Neurites labeled as in (D). See also Figure S2.
Figure 3
Figure 3
DIC1 Interacts Directly with KLCs through DIC1 Tryptophan Motif Binding to the KLC TPR Domains (A) Schematic of DIC1 showing structural motifs and dynein light chain binding sites relative to the alternatively spliced regions (AS loops 1 and 2) and WD motifs (WD1 and WD2, in gray). See also Figure S3. (B and C) COS cells cotransfected with mCherry-tagged DIC1a and HA-tagged KLC as indicated followed by immunoprecipitation (IP) with anti-mCherry (B, western blotting with anti-DIC and anti-HA). (C) Co-IP efficiency expressed as band intensity relative to KLC1 ± SEM; n = 4 experiments. (D and E) Fluorescence polarization measurements with peptides of the first and second DIC1 tryptophan motifs (WD1pept and WD2pept, respectively) binding to the TPR domain of KLC1 (D) and KLC2 (E). KD values determined at 150 mM NaCl; error bars ± SEM, experiments typically done in triplicate. See also Figure S3. (F–I) COS cells cotransfected with mCherry-tagged DIC1a (wild-type, WT; or point mutations of WD1, WD2, or both WD1 and 2 to alanine, AA) and HA-tagged KLC1 (F) or KLC2 (H) as indicated followed by IP with anti-mCherry (F and H, western blotting with anti-DIC or anti-HA). (G and I) Co-IP efficiency expressed as band intensity relative to WT DIC1a ± SEM; n = 6 and 4 experiments, respectively.
Figure 4
Figure 4
DIC1a Interacts with the Central Stalk Region of KIF5 Heavy Chains (A) Western blot (WB) of COS cells cotransfected with mCherry-tagged DIC1a and myc-tagged KIF5A-C constructs as indicated followed by immunoprecipitation with anti-myc. (B) Schematic of KIF5 showing the constructs used in (C), the sequence isolated by yeast two-hybrid screen and the resulting consensus region for binding DIC relative to key domains: motor domain, coiled-coil, cargo binding domain (CBD), central hinge, and KLC binding region. (C) Western blot (WB) of COS cells cotransfected with mCherry-tagged DIC1a and GFP-tagged KIF5C “head,” “stalk,” and “tail” constructs shown in (B) followed by immunoprecipitation with anti-GFP.
Figure 5
Figure 5
Biochemical Analysis of Endogenous Dynein-GFP and Kinesin Complexes from Brain (A) Experimental procedure for dynein-GFP mouse brain differential centrifugation. (B) Distribution of proteins of interest across centrifugation steps outlined in (A) by SDS-PAGE and western blotting. Antibodies for: vesicular marker GAP-43; slow transport marker synapsin; and motor protein subunits DHC, DIC, p150, KHC, and KLC. (C) Quantification of the relative abundance of each protein in fractions S3 versus P3. n = 3; error bars ± SEM. (D) Distribution of proteins of interest across the 24 fractions of the P3 sucrose density gradient, showing separation of vesicles (V, fractions 10–19) from high-density protein complexes (PC, fractions 22–24). Protein association with vesicles was attenuated by the addition of Triton X-100 to the resuspended P3 fraction (Figure S4). Fractions 11–14 were loaded in duplicate to normalize band intensities across the two gels required to run all fractions. (E) Quantification of the gradient assay in (D) highlighting the vesicular (V) and high-density protein fractions (PC). (F) The ratio of DIC1 to DIC2 in vesicle fractions from the gradient assay compared to high-density protein fractions. n = 5 (gray circles), mean ratio (heavy line) ± SEM (box) is shown. (G) KIF5 and DHC co-immunoprecipitate in an endogenous complex from fractions 23 and 24 (D) with anti-FLAG. (H) Schematic of experimental set up for (I). Rhodamine-labeled microtubules were extended from AMCA-labeled microtubule seeds to indicate polarity, then immobilized onto silanized coverslips by anti-tubulin antibody. Dynein-GFP events were imaged by TIRF microscopy (TIRFM). (I) Two polarity-marked microtubules and example kymographs of dynein-GFP events on those microtubules imaged by TIRFM. Vertical scale is 10 s; horizontal scale is 5 μm.
Figure 6
Figure 6
Discrete Populations of Dynein-GFP Move with an Anterograde Bias in the Axon (A) Example axon before (t = –1) and after (t = 0) bleaching. The relative position of the soma is indicated. (B) Midpoint calculation of dynein-GFP photo-protected region for t = 0. The relative intensity of the axon line scan at t = 0, showing raw data (gray dots) and rolling average (green line). Anterograde and retrograde fronts of the photo-protected population are defined as the positions along the axon where the intensity is half-maximal. The midpoint is defined as halfway between these two intercepts. (C) Line scans and midpoints for all time points of the movie, demonstrating a displacement toward the distal axon with time. Color changes with time indicated by the scale bar, top left. (D) Kymograph of dynein-GFP axon from (A) showing anterograde movement of the photo-protected region, over plotted with the calculated midpoints and intercepts from (C). (E) The mean displacement of the midpoint through time (green line) from n = 11 neurons, ±SEM (green ribbon).
Figure 7
Figure 7
Slow Anterograde Transport of Dynein in the Axon Is Dependent on Microtubules and an Interaction with Kinesin (A) Kymographs of DMSO control and nocodazole-treated axons from dynein-GFP hippocampal neurons showing relative positions of anterograde and retrograde intercepts (blue) and the calculated midpoint displacement (green). Scale bar indicates anterograde direction. (B) The mean relative position of the midpoint with time for DMSO and nocodazole-treated axons: n = 18 and 22 axons, respectively, from three independent primary cultures; solid lines, mean; ribbons, ±SEM. (C) A linear regression was fitted to each axon’s midpoint displacement to find the velocity of displacement. The mean velocity (heavy line) ± SEM (box) is shown. Overlaid spots are the velocities for each measured kymograph with colors indicating overall direction of the kymograph. (D) Kymographs of control and DIC1a peptide-treated axons from dynein-GFP hippocampal neurons showing relative positions of anterograde and retrograde intercepts (blue) and the calculated midpoint displacement (green). (E) The mean relative position of the midpoint with time for control and DIC1a peptide-treated axons: n = 27 and 29 axons respectively from 3 independent primary cultures; solid lines, mean; ribbons, ±SEM. (F) Results of linear regression on each axon’s midpoint displacement to find the velocity of displacement. The mean velocity (heavy line) ± SEM (box) is shown. Overlaid spots are the velocities for each measured kymograph with colors indicating the overall direction of the kymograph.
Figure 8
Figure 8
Different Mechanistic Models of Slow Axonal Transport (A) The “stop and go” model describes the transport of neurofilaments in SCa. Through direct association with motors, neurofilaments switch between “off-track” and “on-track” states and between paused and motile states while “on-track” (Brown et al., 2005, Trivedi et al., 2007). (B) The “dynamic recruitment” model describes the transport of some soluble cytosolic proteins moving in SCb, e.g., synapsin I (Tang et al., 2013). Soluble proteins come together to form larger complexes, which stochastically associate with vesicles (the mobile unit) undergoing transport. (C) The “kinesin-limited” model also describes the transport of soluble cytosolic proteins moving in SCb. SCb cargoes such as dynein can directly associate with kinesin for short bursts of motility. By combining a limited ability to hold kinesin in an active state with a relatively low supply of active kinesin motors, slow transport cargoes would move much more slowly relative to kinesin due to the constant binding and release of cargo producing short bursts of motility.

Comment in

Similar articles

See all similar articles

Cited by 25 articles

See all "Cited by" articles

References

    1. Ayloo S., Lazarus J.E., Dodda A., Tokito M., Ostap E.M., Holzbaur E.L.F. Dynactin functions as both a dynamic tether and brake during dynein-driven motility. Nat. Commun. 2014;5:4807. - PMC - PubMed
    1. Baas P.W., Black M.M., Banker G.A. Changes in microtubule polarity orientation during the development of hippocampal neurons in culture. J. Cell Biol. 1989;109:3085–3094. - PMC - PubMed
    1. Brown A., Wang L., Jung P. Stochastic simulation of neurofilament transport in axons: the “stop-and-go” hypothesis. Mol. Biol. Cell. 2005;16:4243–4255. - PMC - PubMed
    1. Dillman J.F., 3rd, Dabney L.P., Karki S., Paschal B.M., Holzbaur E.L.F., Pfister K.K. Functional analysis of dynactin and cytoplasmic dynein in slow axonal transport. J. Neurosci. 1996;16:6742–6752. - PMC - PubMed
    1. Dillman J.F., 3rd, Dabney L.P., Pfister K.K. Cytoplasmic dynein is associated with slow axonal transport. Proc. Natl. Acad. Sci. USA. 1996;93:141–144. - PMC - PubMed

Publication types

Feedback