Localized Myosin II Activity Regulates Assembly and Plasticity of the Axon Initial Segment

Abstract

The axon initial segment (AIS) is the site of action potential generation and a locus of activity-dependent homeostatic plasticity. A multimeric complex of sodium channels, linked via a cytoskeletal scaffold of ankyrin G and beta IV spectrin to submembranous actin rings, mediates these functions. The mechanisms that specify the AIS complex to the proximal axon and underlie its plasticity remain poorly understood. Here we show phosphorylated myosin light chain (pMLC), an activator of contractile myosin II, is highly enriched in the assembling and mature AIS, where it associates with actin rings. MLC phosphorylation and myosin II contractile activity are required for AIS assembly, and they regulate the distribution of AIS components along the axon. pMLC is rapidly lost during depolarization, destabilizing actin and thereby providing a mechanism for activity-dependent structural plasticity of the AIS. Together, these results identify pMLC/myosin II activity as a common link between AIS assembly and plasticity.

Keywords: actin rings; ankyrin; axon; axon initial segment; axonal plasticity; homeostatic plasticity; myosin II; myosin light chain; node of Ranvier; spectrin.

Figure 1. Phosphorylated myosin light chain is enriched at the AIS and nodes of Ranvier

A. DIV10 cultured rat hippocampal neuron showing colocalization of pMLC (red) with AnkG (green) in the AIS. Soma and dendrites are counterstained with MAP2 (blue). Scale bar, 10μm. B. P30 mouse cortex showing colocalization of pMLC (red) and AnkG (green) in the AIS of projection neurons. Nuclei are stained with Hoechst (blue). Scale bar, 5 μm. C. Adult rat optic nerve (OptN) sections show pMLC (red) is enriched at nodes of Ranvier stained with βIV spectrin (green) and flanked by the paranodal marker Caspr (blue). White arrowhead labels an example node. pMLC is also present in some non-nodal structures. Scale bar, 5 μm. D. Teased adult rat sciatic nerve (ScN) fibers show localization of pMLC (red) at nodes of Ranvier, flanked by the paranodal marker Caspr (blue). White arrowheads label nodes. Scale bar, 5 μm. See also Figures S1 and S2.

Figure 2. pMLC accumulation in axons and the AIS during development

A. At DIV2, pMLC (red) is present throughout the neuron with moderate enrichment in the nascent axon (dotted white line) and growth cone (arrowhead). There is minimal expression of AnkG (green) and MAP2 (blue) in the axon at this stage. Scale bar, 10 μm. B. pMLC (red) is enriched in the developing AIS, where it colocalizes with AnkG (green) at DIV4 and DIV6; pMLC is present at reduced levels in the axon as well. MAP2 (blue) labels the somatodendritic compartment. Scale bar, 10μm. C. Spatial intensity profiles of pMLC (red) and AnkG (green) demonstrate gradual proximal enrichment between DIV3-DIV6. Intensities are normalized in each case to their corresponding DIV6 peaks. Graphs show average intensities ± SEM. D. pMLC and AnkG increase in the AIS and decrease in the axon during maturation. Average AIS intensity ± SEM is compared between DIV4 (black) and DIV10 (gray). Kruskal-Wallis test with Dunn’s correction; *p≤0.05; **p≤0.01; ****p<0.0001.

Figure 3. Myosin II and phosphorylation of MLC are required for AIS assembly

A. Treatment with the myosin ATPase inhibitor 20 M blebbistatin reduces AnkG intensity at the developing AIS. Cultures were treated from DIV3–6. MAP2 (blue) is also shown. Scale bar, 20 μm. B. Spatial intensity profiles of DIV6 axons demonstrate slightly reduced pMLC (red) and substantially reduced AnkG (green) levels after blebbistatin treatment (blue). C. Average intensities ± SEM of pMLC and AnkG after treatment with vehicle (black) or blebbistatin (gray). Mann-Whitney test, ***p≤0.001, ****p<0.0001. D. A 3 hour incubation with the MLCK inhibitor 10μM ML7 decreased pMLC (red) and AnkG (green) levels compared to vehicle (veh) controls in DIV4 neurons; MAP2 (blue). Scale bar, 10 μm. E. Spatial intensity profiles of DIV4 axons show reduced pMLC (red) and AnkG (green) intensities after ML7 treatment (blue). Graphs show average intensities ± SEM and are normalized to the control peak. F. Average intensities ± SEM of pMLC and AnkG after treatment with vehicle (black) or ML7 (gray). Mann-Whitney test; **p≤0.01***p≤0.001. See also Figure S3.

Figure 4. Loss of MYPT subunits increases pMLC and ectopic expression of AIS proteins

A. Mouse neurons were infected with AAV1-EGFP (control; top panels) or AAV1-Cre (knockout; bottom panels) on DIV1 and fixed on DIV12. MYPT1 knockout enhances pMLC (red) and AnkG (magenta) at the AIS. Scale bar, 20μm. B. Spatial intensity profiles demonstrate elevated pMLC and AnkG levels in MYPT1 KO (blue) vs control (green) neurons. Graphs show average intensities ± SEM normalized to the control peak. C. Axons distal to the AIS display significant ectopic accumulation of pMLC (red) and AnkG (magenta) after knockout of MYPT1. Scale bar, 20μm. D. Quantification of pMLC and AnkG in the AIS and axon in control (black) or MYPT1 KO (gray) neurons. Graphs are average intensities ± SEM. Kruskal-Wallis test with Dunn’s correction; **p≤0.01; ****p<0.0001. E. MYPT2 knockdown (KD) modestly increases pMLC (red) or AnkG (magenta) at the AIS. Neurons were infected with scrambled (shScr; top panels) or knockdown (shMYPT2; bottom panels) lentivirus on DIV1 and fixed on DIV12. Scale bar, 20μm. F. Spatial intensity profiles demonstrate slight elevation of pMLC and AnkG in MYPT2 knockdown (blue) versus control (green) infected AIS. Graphs show average intensities ± SEM and are normalized to the control peak. G. Axons distal to the AIS display substantial accumulation of pMLC (red) and AnkG (magenta) after MYPT2 knockdown (KD). Scale bar, 20μm. H. Quantification of pMLC and AnkG in the AIS and axon in control (black) or MYPT2 knockdown (gray) neurons. Graphs are average intensities ± SEM. Kruskal-Wallis test with Dunn’s correction; ****p<0.0001. See also Figure S4.

Figure 5. Depolarization results in rapid loss of pMLC in the AIS

A. Staining for pMLC (red), AnkG (green) and MAP2 (blue) of control (5 mM KCl; top panels) or depolarized (50 mM KCl for 30 minutes; bottom panels) DIV 10 rat Hc neurons. The dotted line demarcates the partially disassembled AIS. Scale bar, 10μm. B. Changes in the AIS staining intensities of pMLC (red) vs. AnkG (green) during 30 minutes of depolarization with 50 mM KCl are shown. Data points are average ± SEM taken at 0, 5, 15, and 30 minutes and are normalized to initial intensity. Kruskal-Wallis test with Dunn’s correction; **p≤0.01; ****p<0.0001. C. Changes in the spatial intensity profiles of pMLC (red) and AnkG (green) after 30 minutes of depolarization with 50 mM KCl (blue) are shown. Graphs are average intensities ± SEM. D. Cumulative probability curves show near complete loss of pMLC (red) and partial loss of AnkG (green) after 30 min depolarization with 50 mM KCl (blue). Curves are calculated from all MAP2+ neurons. E. Burst-like light stimulation of channelrhodopsin 2 (ChR2)-infected rat Hc neurons. Staining for pMLC (red) and AnkG (magenta) of control (dark, top panels) or stimulated (blue light for 15 min, bottom panels). Native fluorescence of the ChR2-Venus fusion protein is shown in green. The dotted line demarcates the partially disassembled AIS. Scale bar, 10 μm. F. Changes in the spatial intensity profiles of pMLC (red) and AnkG (green) after 15 min of depolarization with a burst-like light stimulus (blue) in ChR2-expressing neurons. Graphs are average intensities ± SEM. G. Total MLC intensity at the AIS does not change significantly after 30 minutes of depolarization with 50 mM KCl. Graph shows average intensities ± SEM of control (black) vs. depolarized (gray). Mann-Whitney test; p>0.05. H. Western blots show pMLC but not total MLC levels decrease significantly after 30 min of KCl-mediated depolarization. Graphs show average intensities ± SEM of control (black) vs. depolarized neurons (gray). Values are normalized to tubulin as a loading control. Two-tailed unpaired t-test; ***p<0.001. I. Quantifications of pMLC (red) and AnkG (green) intensities after 60 minutes of depolarization with 50 mM KCl and a subsequent 3 hour recovery in conditioned media (5 mM KCl). Data points show average intensities ± SEM, normalized to the corresponding non-depolarized levels. Mann-Whitney test; **p≤0.01; ****p<0.0001. See also Figure S5.

Figure 6. Mechanisms that contribute to depolarization-dependent pMLC loss and AIS disassembly

A. Treatment with the PP1/PP2A inhibitor calyculin A (gray) significantly increases pMLC levels (top) in control and neurons depolarized with 50 mM KCl for 30 minutes, relative to vehicle (veh). AnkG is not affected by calyculin treatment (bottom). Graphs show average intensities ± SEM. Kruskal-Wallis test with Dunn’s correction; *p≤0.05, ****p<0.0001. B. Calcium chelation with 5 mM EGTA completely rescues pMLC (red) levels during depolarization with 50 mM KCl; MAP2 (blue) staining is shown. Scale bar, 10μm. C. Quantification of vehicle (control; black) and EGTA (gray) treatments after depolarization with 50 mM KCl. Graphs show pMLC intensity ± SEM. Kruskal-Wallis test with Dunn’s correction; ****p<0.0001. D. Incubation with the calpain inhibitor MDL28170 (gray) partially rescues pMLC (top) and AnkG (bottom) levels during a 30 minute depolarization with 50 mM KCl, relative to vehicle (black). Graphs show average intensities ± SEM. Kruskal-Wallis test with Dunn’s correction; *p≤0.05, ***p≤0.001, ****p<0.0001. E. One hour preincubation with phalloidin-oleate (Phld) to stabilize actin prevents loss of AnkG (green) but not pMLC (red) with a 30 minute depolarization with 50 mM KCl. MAP2 (blue) is shown; scale bar, 10μm. F. Quantification of pMLC and AnkG in the AIS demonstrating rescue of AnkG (right) but not pMLC (left) levels after treatment with phalloidin-oleate (gray) relative to vehicle (black). Graphs show average intensities ± SEM with 5 mM or 50 mM KCl. Kruskal-Wallis test with Dunn’s correction; *p≤0.05, **p≤0.01, ****p<0.0001. G. Spatial intensity profiles of pMLC (red, top) and AnkG (green, bottom) upon depolarization with 50 mM KCl ± phalloidin-oleate (blue); control (non-depolarized) cultures treated with vehicle (black dashed lines) are shown for comparison. Graphs are average intensities ± SEM and are normalized to the control peak. See also Figure S6.

Figure 7. Nanoscale organization of pMLC and actin at the AIS

A. Wide-field epifluorescence image of a control DIV 10 hippocampal neuron, fixed and labeled for actin (top, green on bottom overlay) and pMLC (middle, magenta on bottom overlay). Scale bar, 20 μm. B. STORM image of the AIS from this control neuron (boxed area in A) with actin labeling (top, green on bottom overlay) and pMLC labeling (middle, magenta on bottom overlay). Scale bar, 2 μm. C. Magnified region of interest (boxed area in B) with actin labeling (top, green on bottom overlay) and pMLC labeling (middle, magenta on bottom overlay). Scale bar, 1 μm. D. Intensity profiles along the 1 μm yellow lines highlighted in C showing the actin (green) and pMLC (magenta) labeling intensities. E. Wide-field epifluorescence image of a DIV10 hippocampal neuron after 15 minute depolarization with 50 mM KCl, fixed and labeled for actin (top, green on bottom overlay) and pMLC (middle, magenta on bottom overlay). Scale bar, 20 μm. F. STORM image of the AIS from this depolarized neuron (boxed area in A) with actin labeling (top, green on bottom overlay) and pMLC labeling (middle, magenta on bottom overlay). Scale bar, 2 μm. G. Magnified region of interest (boxed area in F) with actin labeling (top, green on bottom overlay) and pMLC labeling (middle, magenta on bottom overlay). Scale bar, 1 μm. H. Intensity profiles along the 1 μm yellow lines highlighted in C showing the actin (green) and pMLC (magenta) labeling intensities. I. Schematic demonstrating proposed relationship of myosin II (magenta) and thus pMLC with actin rings (green) and spectrin tetramers (purple) in the AIS.