February 25, 2015;
Differential requirement of F-actin and microtubule cytoskeleton in cue-induced local protein synthesis in axonal growth cones.
BACKGROUND: Local protein synthesis (LPS) via receptor-mediated signaling plays a role in the directional responses of axons to extrinsic cues. An intact cytoskeleton is critical to enact these responses, but it is not known whether the two major cytoskeletal elements, F-actin and microtubules, have any roles in regulating axonal protein synthesis.
RESULTS: Here, we show that pharmacological disruption of either microtubules or actin filaments in growth cones blocks netrin-1-induced de novo synthesis of proteins, as measured by metabolic incorporation of labeled amino acids, implicating both elements in axonal synthesis. However, comparative analysis of the activated translation initiation regulator, eIF4E
-BP1, revealed a striking difference in the point of action of the two elements: actin disruption completely inhibited netrin-1-induced eIF4E
-BP1 phosphorylation while microtubule
disruption had no effect. An intact F-actin, but not microtubule
, cytoskeleton was also required for netrin-1-induced activation of the PI3K/Akt/mTOR pathway, upstream of translation initiation. Downstream of translation initiation, microtubules were required for netrin-1-induced activation of eukaryotic elongation factor 2 kinase (eEF2K) and eEF2.
CONCLUSIONS: Taken together, our results show that while actin and microtubules are both crucial for cue-induced axonal protein synthesis, they serve distinct roles with F-actin being required for the initiation of translation and microtubules acting later at the elongation step.
[+] show captions
References [+] :
The effects of pharmacological inhibitors of actin and microtubule polymerization on retinal growth cone morphology. Embryonic eyes at stage 24/25 were cultured for 24 h, treated with either control medium or pharmacological inhibitors for 5 min, and stained for F-actin (Phalloidin-Alexa488) or β-tubulin. Control retinal growth cones (A) contained an extensive network of actin filaments, which was disrupted upon treatment with (B) cytochalasin D (0.1 μM) and (C) latrunculin B (30 nM). Similarly, control retinal growth axons show microtubules in the growth cone central domain and individual microtubules invading the peripheral domain (F). The dynamic microtubules in the growth cone, but not in the axon shaft, were depolymerized upon treatment with (I) colchicine (12.5 μM) or (J) nocodazole (0.1 μM). Importantly, this treatment paradigm of actin or microtubule inhibitors did not affect the ultrastructure of the other cytoskeletal system (G and
E) and also did not increase growth cone collapse at the concentrations used. Scale bar 10 μm.
Intact actin and microtubule cytoskeleton are both required for netrin-1-induced protein synthesis. Stage 24/25 retinal explants were treated with control medium, cytochalasin D (CytoD), latrunculin B (LB), colchicine (Colc), or nocodazole (Noco) for 5 min, followed by stimulation with either control medium or netrin-1 for a further 5 min. Protein synthesis in growth cones was measured by methionine analog L-azidohomoalanine (AHA) incorporation and visualized with fluorescent microscopy (A-I). Protein synthesis in growth cones treated with control medium (A) was markedly increased 5 min following treatment with netrin-1 (B). However, this increase was abolished in growth cones treated with cytochalasin D (C, D) or colchicine (E, F). Protein synthesis within the growth cone was almost completely abolished by pretreatment with cycloheximide (CHX, G) or anisomycin (H). Quantification of fluorescence intensity reveals that netrin-1-induced protein synthesis is inhibited by disruption of either actin or microtubule dynamics (I). The number of growth cones analyzed in each treatment group can be found in the corresponding bar of the graph in panel I. Similarly, protein synthesis, as measured by the incorporation of 3H-leucine in precipitated proteins, was stimulated by netrin-1, but this effect was inhibited by treatment with either cytoclalasin D, latrunculin B, colchicine, and nocodazole (J). ***P < 0.0001 Mann-Whitney test. Scale bar 10 μm.
Netrin-1-induced increase in eIF4E-BP1 phosphorylation requires an intact actin but not microtubule cytoskeleton. A brief schematic of signaling from netrin receptor leading to protein synthesis (A). Stage 24/25 retinal explants were cultured for 24 h, then treated with the actin depolymerizing agents cytochalasin D (CytoD) and latrunculin (B) (LatB), or the microtubule depolymerizing agents colchicine (Colc) and nocodazole (Noco) for 5 min, followed by stimulation with either control medium or netrin-1 for a further 5 min. To assess the activation of translation, retinal cultures were stained for phospho-eIF4E-BP1 (p-eIF4E-BP1; growth cones in the top row of panel (B)) and the intensity of the immunofluorescence signal was measured per unit area (quantified in (C)). Compared to the control (top row of (B), far left panel) netrin-1 induced a significant increase in P-eIF4E-BP1 signal intensity (top row of (B), cytoskeleton intact), which was completely blocked upon pretreatment with actin inhibitors (top row of (B), actin-disrupted), but not with microtubule inhibitors (top row of (B), microtubule-disrupted. Quantification of p-eIF4E-BP1 signal intensity is shown in (C). *P < 0.05 Mann-Whitney test. In contrast, levels of total eIF4E-BP1 signal intensity (bottom row of (B)) were not significantly affected by netrin-1 or the pharmacological inhibitors. The number of growth cones analyzed in each treatment group can be found in the corresponding bar of the graphs in panels (C) and (D). Scale bar 10 μm.
Actin, but not microtubule, disruption blocks netrin-1 induced PI3K/Akt/mTOR signaling in growth cones. Cultured stage 24/25 retinal explants were treated with a DMSO vehicle control, cytochalasin D (CytD), or colchicine (Colc) for 5 min, followed by 5-min stimulation with either control medium or netrin-1. Quantitative immunofluorescence showed that levels of activated mTOR (p-mTOR; (A)) and Akt (p-Akt; (E)) were elevated by netrin-1 stimulation, a process that was prevented by actin, but not microtubule disruption (quantified in (C) and (G)). Total levels of mTOR (B) and Akt (F) were not significantly altered following stimulation with netrin-1, in either untreated or growth cones treated with cytoskeletal disrupting agents (quantified in (D) and (H)), with the exception of total Akt levels in growth cones treated with colchicine, which showed a small, but significant increase in fluorescence intensity. The number of growth cones analyzed in each treatment group can be found in the corresponding bar of the respective graphs. **P < 0.005; ***P < 0.0001 Mann-Whitney test. (I) Pseudocolored images of a live PI3K biosensor (PHAkt-GFP) before and after netrin-1 stimulation in the presence of cytoskeletal inhibitors. Growth cones expressing low levels of PHAkt-GFP were treated with cytoskeletal inhibitors (0.1 μM cytochalasin D or 12.5 μM colchicine) on stage during time-lapse acquisition on an inverted spinning disk confocal system (60× water immersion). Time-lapse imaging showed an increase in PHAkt-GFP signal after netrin-1 stimulation, which was inhibited by cytochalasin D treatment, but not affected by colchicine treatment (J). Background-subtracted fluorescent signals were normalized to the control medium at time 0. Scale bar 10 μm.
Microtubule depolymerization inhibits netrin-1 induced phosphorylation of EF2K and EF2. Cultured stage 24/25 retinal explants were treated with the mTOR inhibitor rapamycin or a vehicle only control (A, B). Rapamycin induced a significant reduction in the level of phosphorylation of a downstream target of mTOR, eEF2 kinase (A-C). Quantitative immunofluorescence was also used to assess the levels of phosphorylation of both eEF2 kinase (D) and eEF2 (E) in response to netrin-1 stimulation. Netrin-1 induced a significant increase in the fluorescence intensity of both phospho-eEF2 kinase (D) and phospho-eEF2 (E) in growth cones treated with a DMSO vehicle control (ctrl). However, no significant changes in fluorescence intensity following netrin-1 stimulation were observed following inhibition of the microtubule cytoskeleton with colchicine (colc; (D, E)). The number of growth cones analyzed in each treatment group can be found in the corresponding bar of the respective graphs. ns = not significant. ***P < 0.0005 Mann-Whitney test. Scale bar 10 μm.
Inhibition of dynamic microtubules prevents RNP granule movement into the growth cone periphery. Embryos were injected with fluorescently labeled UTP (Cy3-UTP) at the four to eight-cell stage to visualize ribonucleoprotein (RNP) granules. Cultured stage 24/25 retinal explants positive for Cy3-UTP-labeled RNP granules were selected for time-lapse imaging on an inverted spinning disk confocal system (60× water immersion objective, 5-s acquisition interval) and treated on-stage with cytoskeletal disrupting drugs. Prior to colchicine treatment, RNP granules were trafficked to the growth cone periphery (arrowheads, top row in (A)). However, colchicine treatment acutely blocked RNP granule movement into the peripheral domain (bottom row in (A), 8 out of 10 growth cones). In contrast, cytochalasin D treatment did not affect RNP granule trafficking to the periphery (arrowheads in bottom row of (B), 8 out of 11 growth cones). Representative frames per minute from before and 5 min after colchicine (panels in A) or cytochalasin D (panels in B) treatment are shown. Numbers in the lower left of each image represent time in minutes.
Phosphatidylinositol 3-kinase facilitates microtubule-dependent membrane transport for neuronal growth cone guidance.