Cytoplasmic Dynein and LIS1 Are Required for Microtubule Advance during Growth Cone Remodeling and Fast Axonal Outgrowth

Enrichment of dynein within DRG growth cones. A , B , Quantitation of dynein IC immunofluorescence relative to total protein (DTAF) 10 min after treatment with laminin. Stains were ratioed, and line scans were made perpendicular to the leading edge (black line in A ) (see Materials and Methods). Scale bar, 5 μm. Relative fluorescent intensity is greatest close to the leading edge ( B ). Error bars represent the SEM. C–H , TIRF microscopy was performed on chick DRG neurons treated with laminin and stained with antibodies against tubulin and dynein IC. C–E , At 15 min in laminin, enrichment of dynein was observed in the leading portion of the growth cone. F–H , At 25 min in laminin, dynein was observed to be concentrated at the site of a nascent outgrowth associated with bundled microtubules. Dynein staining generally was observed to be punctate in these images. Scale bar: (in H ) C–H , 10 μm.

A , Localization of LIS1 and dynactin to sites of emerging axons. Chick DRG neurons were exposed to laminin for 25 min and then stained with anti-tubulin, anti-LIS1, anti-p150 ^Glued , and anti-Arp1 antibodies. All three of the latter antigens were enriched at sites of nascent axon outgrowth to which dynein was localized and microtubules had converged. Scale bar, 10 μm. B , Colocalization of p150 ^Glued and LIS1 in outgrowths 30 min after laminin. Scale bar, 5 μm. C , Immunoblot of chick brain extract with the antibodies used for immunofluorescence microscopy shows that these antibodies react with chick neural tissue.

Localization of dynein and LIS1 proteins in axons. A , Dynein IC localization in DRG neurons. Undifferentiated neurites grown on a polyamine substrate show little concentration of dynein. Differentiated axons grown on a laminin substrate show a strong concentration of dynein in the growth cone. When each stain is ratioed against a total protein stain (DTAF), the concentration of dynein is confirmed. Scale bar, 10 μm. B , LIS1 localization in hippocampal axons. At differentiation stage 2 undifferentiated neurites show little concentration of LIS1. At stage 3 when an axon has formed, only the axon shows a high concentration of LIS1. This is more apparent when LIS1 is ratioed against a total protein stain. Scale bar, 10 μm. C , Inhibition of LIS1 by RNAi. Shown is tubulin staining of axons from hippocampal 3DIV neurons expressing scrambled control (left) or LIS1 RNAi constructs (right). Scale bar, 10 μm. An extended region of spread and wavy microtubules is observed in the LIS1 knock-down process. D , Average length of axons from hippocampal 3DIV and 5DIV neurons expressing these constructs. These values were significantly different [Student's unpaired two-tailed t test; for p = 0.007, n = 34 control and 57 RNAi (3DIV); for p = 0.00001, n = 34 control and 20 RNAi (5DIV)]. Error bars represent the SEM.

Effect of dynein IC and LIS1 antibody injection on growth cone morphology. A , Morphology of growth cones injected with antibody and then exposed to laminin. The growth cone of a neuron injected with control IgG shows a typical response. At 30 min after exposure to laminin two nascent axons have emerged (arrows). The growth cone of a cell injected with dynein IC antibody shows little change. After laminin treatment the growth cone has enlarged, but there are no outgrowths. The growth cone has a healthy peripheral actin network (arrow). A cell injected with polyclonal LIS1 antibody shows no outgrowths. After 30 min of exposure to laminin the growth cone has advanced a little, and lamellipodial regions are seen along the length of the newly formed neurite shaft (arrow). Scale bar, 10 μm. B , Quantitation of antibody injections. Chick DRG neurons were injected with monoclonal dynein IC antibody (n = 58), polyclonal LIS1 antibody (n = 31), or control mouse IgG (n = 55). Then the neurons were exposed to laminin for 30 min to induce neurite outgrowth. The number of laminin-induced neurites was reduced considerably in cells injected with function-blocking anti-dynein or anti-LIS1 antibody.

Effect of anti-dynein IC injection on growth cone morphology and microtubule distribution. Chick DRG neurons grown on a polyamine substrate were transfected with cDNA encoding EGFP-α-tubulin. Injected and uninjected control cells then were treated with laminin and monitored by DIC and fluorescence microscopy. A , Control uninjected cell. At 30 min after antibody injection and 10 min before laminin addition the microtubules are confined mainly to the central zone some distance from the leading edge (arrows in −10 min laminin). At 15 min after laminin treatment the microtubules have advanced to the leading edge (arrow). At 45 min after laminin treatment foci of microtubules have formed at the leading edge, around which the membrane also has begun to constrict (arrowhead). By 55 min a rapidly advancing neurite with a tightly constricted membrane and tightly bundled microtubules has emerged (arrowhead). B , Growth cone of a cell injected with anti-dynein IC antibody. At 30 min after antibody injection and 10 min before laminin treatment the growth cone displays normal spread morphology with microtubules confined to the central zone. At 15 min after laminin treatment the microtubules fail to reach the leading edge (arrow), and the growth cone advances at its initial slow rate. At 45 min after laminin treatment the microtubules near the leading edge but do not form foci (arrow). At 55 min after laminin treatment the peripheral zone has become reestablished. Lamellipodial protrusions also can be observed along that portion of the neurite that formed during the period of observation (arrowhead). Dotted lines in tubulin images represent the cell boundary. Scale bar, 10 μm. C , Growth rates of the growth cones depicted in A and B . The two growth cones initially advance at a comparable rate before laminin. After laminin treatment the control growth cone speeds up approximately threefold, whereas the rate of advance of the injected growth cone remains unchanged.

Effect of LIS1 antibody injection on growth cone morphology and microtubule distribution. Chick DRG neurons grown on a polyamine substrate were transfected with cDNA encoding EGFP-α-tubulin. Expressing cells were injected with function-blocking LIS1 antibody. These cells were treated with laminin 30 min later and monitored by DIC and fluorescence microscopy. Just before antibody injection the microtubules are contained within a discrete central zone. At 20 min after laminin treatment the microtubules have reached the leading edge but fail to accumulate and form foci. The growth cone has advanced at the slow rate and retains its spread morphology. At 45 min after laminin treatment the peripheral zone has become reestablished (arrowhead). Dotted lines in tubulin images represent the cell boundary. Scale bar, 10 μm.

Effect of dynein antibody injection on neurites with small growth cones. Chick DRG neurons were treated as described in the legend to Figure 4. A , Small growth cone of a control cell. By 20 min after laminin treatment the growth cone is narrower, and the neck is constricted (arrowhead). Constriction of the membrane and microtubules proceeds in subsequent frames as the growth cone advances more rapidly. Dotted lines in tubulin images represent the cell boundary. B , A small growth cone of a cell injected with dynein antibody. At 20 min after laminin treatment the growth cone continues to advance, but there is no constriction of the membrane or microtubules. The growth cone remains at the same width. In contrast to the control cell, lamellipodia remain well spread along the sides of the newly formed process (arrowhead at 70 min). Scale bar, 10 μm. C , Growth rates of the growth cones depicted in A and B . The two growth cones initially advance at a comparable rate before laminin. After laminin treatment the control growth cone speeds up approximately threefold, whereas the rate of advance of the injected growth cone remains unchanged.

Effect of dynein inhibition on microtubule dynamics. A , Appearance of EB3 comet tails (arrows) at a single time point in the growth cones of control IgG and anti-dynein-injected (α74.1) chick DRG neurons. Dotted lines indicate growth cone boundary. B , History maps of comet tail activity under each condition. All frames from a 1 min video sequence (supplemental data 1, 2, available at www.jneurosci.org as supplemental material) were superimposed to visualize the comet tail pathways. Arrow shows a comet tail in control conditions penetrating deep into the peripheral region. Scale bar, 10 μm. C , Analysis of comet tail movements from eight growth cones in each condition. Histograms of the duration of anterograde and retrograde movements of comet tails show dramatic inhibition of anterograde extension in cells injected with dynein function-blocking antibody. Histogram showing numbers of comets exhibiting pauses is superimposed. D , Kymographs of comet tail advance in control cells and rearward movement in anti-dynein-injected cells. A narrow rectangle covering the path of comet tails for 35 frames (2 s apart) was used to create kymographs of each movement. Forward movement associated with the control microtubule at left is rapid at microtubule assembly rates but is marked by a pause, with persistent rapid oscillations during this period. Rearward movement in dynein-inhibited cells shows a slow steady trajectory at rates consistent with retrograde actin flow.