α-Tubulin Tyrosination and CLIP-170 Phosphorylation Regulate the Initiation of Dynein-Driven Transport in Neurons

Abstract

Motor-cargo recruitment to microtubules is often the rate-limiting step of intracellular transport, and defects in this recruitment can cause neurodegenerative disease. Here, we use in vitro reconstitution assays with single-molecule resolution, live-cell transport assays in primary neurons, computational image analysis, and computer simulations to investigate the factors regulating retrograde transport initiation in the distal axon. We find that phosphorylation of the cytoskeletal-organelle linker protein CLIP-170 and post-translational modifications of the microtubule track combine to precisely control the initiation of retrograde transport. Computer simulations of organelle dynamics in the distal axon indicate that while CLIP-170 primarily regulates the time to microtubule encounter, the tyrosination state of the microtubule lattice regulates the likelihood of binding. These mechanisms interact to control transport initiation in the axon in a manner sensitive to the specialized cytoskeletal architecture of the neuron.

Figure 1

CLIP-170 comets serve as docking sites for p150^Glued vesicles in cells. (A) Deconvolved max projection STED image of microtubules in the neuronal growth cone with the cell border outlined in white (left). Schematic of the recruitment of dynein-dynactin organelles in the neuronal growth cone (right). (B) Schematic of p150^Glued (above) showing the CAP-Gly, basic (+), and coiled-coil domains. Schematic of CLIP-170 with relevant CAP-Gly, Ser rich, coiled-coil, and Zinc knuckle (Zn) domains. The phosphorylation sites S-309, -311, -313, -319, -320 and phosphomutant constructs are shown below. (C) Confocal slices of COS-7 cells co-transfected with HT-CLIP-A and either full-length myc-p150^Glued WT or G71R (top row). Upsampled images of individual comets show punctate p150^Glued associated with CLIP-A comets (below). The dashed white lines denote where line profiles were performed in D. (D) Line intensity profiles along upsampled images in C. (E) 3D-STORM super-resolution was used to image microtubules in fixed COS-7 cells (Movie S1). Shown are cropped microtubule plus ends from the two different conditions. CLIP-A (magenta) fully decorates the plus end as comet-like structures and discrete p150^Glued clusters associate with the CLIP-A comet (tubulin is unlabeled in these images). This shows that p150^Glued forms discrete punctate structures (white arrowhead) on CLIP-A comets; the morphology of these punctae are consistent with a vesicular population. In addition, there is a separate population of point-localized individual p150^Glued molecules (gray arrowhead). (F) Histogram of the maximum equatorial diameter for ellipses manually fit around clusters of p150^Glued WT puncta in STORM images (n = 133 puncta total from 2 cells, N= 1).

Figure 2

Reconstitution of CLIP-170 dependent organelle recruitment in vitro. (A) Schematic of in vitro reconstitution assay with p50-GFP-vesicles, EB1, and CLIP-A or CLIP-E in COS-7 cell extract HSS. (B) Western blot of fractions from neuronal vesicle isolation from mouse brain shows p150^Glued and DIC are enriched in the vesicle fraction. Most of the CLIP-170 is retained in the low speed pellet and only a faint band for endogenous CLIP-170 is found in the vesicle fraction. Cytosolic marker GAPDH is not enriched in the vesicle fraction. (C) Representative kymographs of p50-GFP-vesicles on GMPCPP microtubules show that they exhibit processive, bidirectional, and confined motility, as defined by mean squared displacement analysis of particle trajectories. (D) Representative kymographs of p50-GFP-vesicles on single GMPCPP microtubules from each of the three conditions (100nM EB1 + Mock-HSS or CLIP-A or CLIP-E in motility assay buffer). The estimated final concentration of TMR-labeled HT-CLIP in this assay was ~100nM. EB1 + Mock-HSS and EB1 + CLIP-E show a basal landing rate, whereas there is an increase in organelle recruitment with CLIP-A. At least three different preparations of cell lysates were used. (E) Quantification of the landing rates in each of the experimental conditions shows a significant increase in organelle recruitment only with EB + CLIP-A (n >= 22 movies per condition with each movie with > 500 events per movie; N >= 3 independent vesicle isolations; Kruskal-Wallis (KW) one-way ANOVA with Tukey post-tests) (F) Empirical cumulative distribution function (CDF) for microtubule dwell times of all non-fixed particles. The stair plot starts at 0.1s as this is the minimum observable dwell time. Double exponential fit obtained via Maximum Likelihood Estimation (MLE) with 100 bootstrapped fits to obtain 95% CI. The dwell time distributions for Mock-HSS, CLIP-A, CLIP-E were compared in pairs using a two sample Kolmogorov-Smirnov (KS) test with a Bonferroni correction (p > 0.20 for all comparisons; n > 1500 particles per condition; N >= 3).

Figure 3

CLIP-170 phosphorylation state modulates retrograde transport initiation in neurons. (A) Schematic of live-cell assay for retrograde transport initiation in neurons. (B) Representative kymographs for each condition with a schematic tracing of the analyzed vesicles to the right. The dashed red line indicates 3.5μm from the distal end of the bleaching ROI and is the site for flux analysis. (C) Quantification of retrograde flux for all conditions. CLIP-170 knockdown reduces the efficiency of retrograde flux and this is rescued by CLIP-WT or CLIP-A, but CLIP-E shows incomplete rescue (n= 18–36 neurons total, N > 3 experiments; KW ANOVA with Tukey post-tests). A complete list of the pairwise post-tests is shown in Table S1. (D) Quantification of the number of LAMP-1-RFP vesicles in the growth cone at the start of the movie shows a comparable distribution of vesicles indicating that the reduced retrograde flux in the siRNA or siRNA + CLIP-E is not due to a reduced distal vesicle pool.

Figure 4

The distal axon is rich in tyrosinated α-tubulin. (A) STED super-resolution images of the distal axon shows an enrichment of tyrosinated α tubulin (white arrowhead distal axon; gray arrowhead ~30μm proximal to the distal axon tip) (B) Detyrosinated α tubulin is enriched in the proximal axon, relative to tyrosinated α tubulin. The white arrowhead shows the first axonal branch point after exiting the cell body. The cell body was identified by morphology, but is not shown in these images. (C) Bootstrapped mean and 95% CI for line intensity profiles of Tyr/ Detyr in the distal axon (n = 4 neurons per condition, 10 line profiles per neuron; N = 2). The blue line indicates a ratio of 1. The images for Tyr or Detyr staining represent paired images where the Tyr distal axon in A is from the same neuron as the Tyr cell body staining in B with the same true for Detyr. Representative paired images were chosen from >4 images per condition, per experiment of similar quality.

Figure 5

Tyrosinated α-tubulin promotes robust organelle recruitment in vitro. (A) Schematic of the p50-GFP-vesicle recruitment assay using GMPCPP stabilized Tyr-/ Detyr-microtubules labeled separately with either 5% AF-647 or TRITC. (B) Western of Tyr/ Detyr tubulin purified from HeLa cells shows a homogenous population of fully tyrosinated or fully detyrosinated tubulin, compared to brain tubulin. (C) Representative kymographs of p50-GFP-vesicle landing on Detyr versus Tyr microtubules. (D) Quantification of landing rate shows a robust and significant increase in landing rate on Tyr microtubules (n = 70 movies per condition, each movie with > 500 events; N = 3 independent vesicle isolations; Wilcoxon Rank-Sum test p < 1×10⁻¹⁰). (e) Empirical CDF for microtubule dwell times of all non-fixed particles with double exponential fit and 95% CI for fit parameters. The dwell time distributions for vesicles on Tyr or Detyr microtubules were compared using a two sample KS test (p = 0.07; n > 1×10⁴ particles per condition, N = 3).

Figure 6

CLIP-170 phosphorylation regulates recruitment to Tyr-microtubules. (A) Image of max intensity projection over time shows selective localization of CLIP-WT and CLIP-A to Tyr microtubules. CLIP-E shows reduced localization to Tyr microtubules. (B) Quantification of landing rate on Tyr- or Detyr- microtubules shows a significantly increased landing rate of CLIP-WT and CLIP-A on Tyr-microtubules. CLIP-E retains a significant preference for Tyr-microtubules, but landing is significantly reduced compared to CLIP-WT or CLIP-A (a table of all pairwise post-tests is found in Table S2). Single molecule landing rates for CA-Kinesin-1^[560] in cell extracts show no differences between Tyr-/ Detyr-microtubules (n > 24 movies per condition, each movie with > 500 events; N = 3 independent cell lysate preparations; KW ANOVA with Tukey post-tests). The estimated final concentration of TMR-labeled HT-CLIP in this assay was ~0.5nM for HT-CLIP constructs or 2.5nM for CA-Kinesin-1^[560]. (C) Image panel for ensemble recruitment assay using 100nM purified proteins. Faint GFP-p150^[1-210] signal is present on Detyr microtubules and is likely mediated by the low affinity basic microtubule binding domain. (D) Quantification of the median normalized GFP fluorescent intensity per microtubule. The gray rectangle indicates the 95% CI for the median normalized GFP intensity for areas with no microtubules in each condition (n > 100 microtubules per condition; N = 3; KW ANOVA with Tukey post-tests). No exogenous EB1 was added in the experiments shown in Figure 6A–D.

Figure 7

Simulation of vesicle diffusion and microtubule capture in the distal axon. (A) STED image Tyr-microtubules in the neuronal growth cone with the cell border outlined in white (left) and a schematicgrowth cone (right). (B) Median (with bootstrapped 95% CI of the median) for the effective time to microtubule capture in computer simulations with CLIP-E or CLIP-A, and either Tyr or Detyr binding probabilities (D = 0.006um²/sec). (C) Median time to microtubule capture using Tyr binding probabilities and a range of diffusion coefficients (also see Figure S7). (d) Simulated retrograde flux for four different traced neuronal growth cones used in the simulation. In some neurons, CLIP-A decoration increases the simulated retrograde flux. (e) Working model for the regulation of transport initiation in neurons. A gradient of tyrosinated α-tubulin is enriched in the distal axon provides spatial regulation and promotes efficient binding in a region where many cargos originate. CLIP-170 is highly phosphorylated and not microtubule associated in the growth cone at baseline. Transient or local dephosphorylation may provide temporal regulation and is required for efficient transport initiation in neurons.