A novel Netrin-1-sensitive mechanism promotes local SNARE-mediated exocytosis during axon branching

Abstract

Developmental axon branching dramatically increases synaptic capacity and neuronal surface area. Netrin-1 promotes branching and synaptogenesis, but the mechanism by which Netrin-1 stimulates plasma membrane expansion is unknown. We demonstrate that SNARE-mediated exocytosis is a prerequisite for axon branching and identify the E3 ubiquitin ligase TRIM9 as a critical catalytic link between Netrin-1 and exocytic SNARE machinery in murine cortical neurons. TRIM9 ligase activity promotes SNARE-mediated vesicle fusion and axon branching in a Netrin-dependent manner. We identified a direct interaction between TRIM9 and the Netrin-1 receptor DCC as well as a Netrin-1-sensitive interaction between TRIM9 and the SNARE component SNAP25. The interaction with SNAP25 negatively regulates SNARE-mediated exocytosis and axon branching in the absence of Netrin-1. Deletion of TRIM9 elevated exocytosis in vitro and increased axon branching in vitro and in vivo. Our data provide a novel model for the spatial regulation of axon branching by Netrin-1, in which localized plasma membrane expansion occurs via TRIM9-dependent regulation of SNARE-mediated vesicle fusion.

Figure 1.

TRIM9 directly binds the Netrin-1 receptor DCC and colocalizes with DCC in cortical neurons. (A) Bacterially expressed GST-SPRY domain interacts with DCC in embryonic mouse cortical lysate. Protein purity is shown by Coomassie. IB, immunoblot. (B) Sequential overlapping peptides within the cytoplasmic tail of DCC were arrayed on nitrocellulose and probed with GST-SPRY, GST antibodies, and HRP secondary antibodies. The SPRY domain binds two sequences within the cytoplasmic tail of DCC. (C) E15.5 cortical neuron transfected with MycTRIM9 and HA-DCC and cultured for 48 h. Boxes denotes the ROIs shown in the enlarged color-combined image. (D) Neuron transfected with GFP-TRIM9 and mCherry (mCh)-DCC imaged by TIRF. Arrowheads denote colocalization, and time is given in seconds (Video 1).

Figure 2.

TRIM9 is expressed in the developing murine cortex. (A) Murine TRIM9 isoforms. COS, C-terminal subgroup one signature. (B–D) Immunoblots (IB) using a TRIM9 polyclonal antibody demonstrate TRIM9 expression in embryonic whole brain lysate (B), in embryonic cortex (C), and in cortical neurons (D) cultured for the indicated times. βIII-Tubulin is a loading control. The data shown in C are shown again in Fig. S2 E alongside TRIM9^−/− samples. (E and E′) Immunohistochemistry (IHC) of E15.5 brain using TRIM9 monoclonal antibody (black) reveals TRIM9 expression in the developing cortex. The dotted square is the ROI shown in E′. (F and F′) EGFP-TRIM9 localizes to punctae at the periphery of the cell body, along the axon shaft (arrow) and at the tips of F-actin–positive filopodia (arrowheads; F′); phalloidin is shown in red. Boxes show the ROI represented in the enlarged color-combined image in F’.

Figure 3.

TRIM9 ubiquitin ligase function is required for Netrin-1–dependent axon branching. E15.5 cortical neurons transfected with Myc or MycTRIM9ΔRING (red) stained with phalloidin to mark F-actin (green) in Netrin-1 and FGF2 branching assays. Arrowheads denote axon branching points. Mean axon branch density ± SEM; n = 593 neurons. P-values were obtained from ANOVA with Tukey’s post-hoc analysis.

Figure 4.

TRIM9 is required for Netrin-1–dependent branching. (A) E15.5 TRIM9^+/+ and TRIM9^−/− cortical neurons were untreated or Netrin-1 stimulated and stained with fluorescent phalloidin. Arrowheads denote branch points. (B) Mean axon branch density ± SEM. n > 80 neurons/condition. Expression of Myc-TRIM9 (T9) in TRIM9^−/− neurons reduced constitutive branching and rescued Netrin-1 sensitivity. MycTRIM9ΔRING (T9ΔRING) or MycTRIM9LD (T9LD) introduction failed to rescue Netrin-1–dependent branching. (C) E15.5 TRIM9^−/− cortical neurons expressing either Myc or MycTRIM9ΔRING were untreated or FGF2 stimulated and stained with fluorescent phalloidin and antibodies to Myc and βIII-tubulin (not depicted). Arrowheads denote branch points. (D) Mean axon branch density ± SEM. Stimulation of TRIM9^−/− neurons with FGF2 fails to increase already exaggerated axon branching except in cells expressing MycTRIM9ΔRING. (E) Axon length is not affected by Netrin-1 or genetic loss of TRIM9. Means ± SEM. All graphs are n > 80 neurons/condition from more than three independent experiments. P-values were obtained from ANOVA with Tukey’s post-hoc analysis.

Figure 5.

TRIM9 regulates SNARE complex formation and exocytosis. (A) Immunoblot showing SDS-resistant SNARE complexes in TRIM9^+/+ and TRIM9^−/− cortical neurons probed for VAMP2, Syntaxin-1a (Synt-1), and SNAP25. βIII-Tubulin (Tub) is a loading control. Syntaxin-1 and SNAP25 monomer are shown in boiled samples; n = 3 independent experiments. These samples are shown again in Fig. S4 A alongside boiled samples. (B) Quantitation of SNARE complexes (>40 kD) was normalized to βIII-tubulin. Significance is from ANOVA with Tukey’s post-hoc analysis; *, P < 0.5. (C) Inverted image montage of VAMP2-pHluorin exocytosis events in TIRF images. Arrowheads denote exocytic events, and time is noted in seconds. (D) Maximal projections of VAMP2-phluorin time lapse. Ovals denote location of exocytic events within 90 s (Videos 2 and 3). (E) Mean frequency of VAMP2-pHluorin exocytosis events ± SEM; n ≥ 10 cells/condition. P-value was obtained from Kruskal–Wallis nonparametric ANOVA within the entire image (top), soma (middle), or neurites (bottom). (F) DIC image montage showing the initial plasma membrane protrusion and subsequent axon branch formation of a TRIM9^+/+ cortical neuron after Netrin-1 stimulation (time is in hours and minutes). White arrowheads denote active plasma membrane protrusions; black arrowheads denote axon branches ≥20 µm long (Video 4). Scatter plot shows the time of initial axon protrusions at future branch sites in hours after Netrin-1 stimulation plotted against the time a stable 20-µm-long branch appears after Netrin-1 stimulation; n = 13 cells. Error bars show SEMs.

Figure 6.

SNARE proteins are required for cortical axon branching. (A) E15.5 TRIM9^+/+ cortical neurons were untreated or Netrin-1 stimulated, with the addition of 10 nM BoNTA, 10 nM TeNT, and/or longin expression (blue), and stained with fluorescent phalloidin (red) and for βIII-tubulin (green). Arrowheads denote branch points. (B) Mean axon branch density ± SEM. (C) E15.5 TRIM9^−/− cortical neurons were untreated or Netrin-1 stimulated, with the addition of 10 nM BoNTA, and stained with fluorescent phalloidin (red) and for βIII-tubulin (green). Arrowheads denote branch points. (D) Mean axon branch density ± SEM. Red asterisks in B and D denote statistical significance from TRIM9^+/+ neurons in the absence of Netrin-1 (P < 0.5). Black asterisks denote statistically significant differences between untreated and Netrin-1–stimulated conditions (P < 0.5). All graphs are at least n > 54 neurons/condition from more than three independent experiments (n = 1,353 neurons). P-values are from ANOVA with Tukey’s post-hoc analysis.

Figure 7.

The SNAP25 binding domain of TRIM9 is required to minimize exocytosis and axon branching. (A) Domain diagram of TRIM9 and TRIM9ΔCC. COS, C-terminal subgroup one signature. (B) HEK293 cells expressing HA-DCC, GFP-SNAP25, and either Myc, MycTRIM9, or MycTRIM9ΔCC were precipitated with Myc antibody-conjugated beads (n = 5 independent experiments; p-values are from Kruskal–Wallis nonparametric ANOVA). The immunoblot for HA-DCC contains an empty lane between samples 3 and 4. IB, immunoblot. (C) Maximal projection of a VAMP2-pHluorin–transfected TRIM9^−/− neuron expressing mCherryTRIM9ΔCC (not depicted; Video 5). Ovals denote exocytic events. (D) mCherryTRIM9ΔCC expression significantly increased VAMP2-mediated exocytic events as measured by VAMP2-pHluorin experiments in TRIM9^+/+, but not TRIM9^−/−, cortical neurons; n ≥ 10 cells/condition. P-values are from Kruskal–Wallis ANOVA. (E) Neurons transfected with MycTRIM9ΔCC (blue) stained with phalloidin (green) and for βIII-tubulin (red) in a Netrin-1 branching assay. Arrowheads denote branches. (F) Mean axon branching density ± SEM; n > 75 neurons/condition from three independent experiments. P-values are from Kruskal–Wallis nonparametric ANOVA. Error bars show SEMs.

Figure 8.

Genetic loss of TRIM9 leads to aberrant axon branching within and thickening of the corpus callosum. (A) Coronal sections through corpus callosum from adult TRIM9^+/+ and TRIM9^−/− mice stained for nissl substance. Lines denote callosal thickness measurements. Graph shows quantitation of callosal thickness; n = 5 mice/genotype; **, P < 0.05 obtained from t test. (B) Inverted maximal projections of Thy1-GFP/TRIM9^+/+ and TRIM9^−/− corpus callosum sections. Arrowheads denote branch points. (B′) Inset shows an example of a branch. Video 6. (C) Quantitation of axonal branch points in 3-wk-old littermates; n = 3 mice/genotype. P-values obtained from t test. (D) Example images used in quantitation of the number of GFP-expressing axons crossing the midline in Thy1-GFP/TRIM9^+/+ and Thy1-GFP/TRIM9^−/− brains. Error bars show SEMs.

Figure 9.

Molecular mechanism of Netrin-1–dependent axon branching. (top left) In the absence of Netrin-1, TRIM9 (orange) is bound to SNAP25 (green). The TRIM9 interaction blocks SNAP25 binding to VAMP2 or VAMP7 (blue), which inhibits SNARE complex formation and exocytic vesicle fusion. Prevention of plasma membrane expansion constrains basal levels of axon branching. (top right) After Netrin-1 stimulation, TRIM9 releases SNAP25, which promotes SNARE complex formation, vesicle fusion, local plasma membrane expansion, and axon branching. (bottom) Genetic deletion of TRIM9 renders cells nonresponsive to Netrin-1. As the interaction between SNAP25 and VAMP2 or VAMP7 is no longer inhibited, constitutive levels of SNARE complex formation, exocytosis, and axon branching increase. To reduce and rescue the exocytosis and axon branching phenotypes associated with loss of TRIM9, introduction of TRIM9 containing the SNAP25 binding domain is required. To rescue Netrin-1 sensitivity, TRIM9 containing ubiquitin ligase activity must be introduced.