E3 ligase Nedd4 promotes axon branching by downregulating PTEN

Abstract

Regulated protein degradation via the ubiquitin-proteasome system (UPS) plays a central role in building synaptic connections, yet little is known about either which specific UPS components are involved or UPS targets in neurons. We report that inhibiting the UPS in developing Xenopus retinal ganglion cells (RGCs) with a dominant-negative ubiquitin mutant decreases terminal branching in the tectum but does not affect long-range navigation to the tectum. We identify Nedd4 as a prominently expressed E3 ligase in RGC axon growth cones and show that disrupting its function severely inhibits terminal branching. We further demonstrate that PTEN, a negative regulator of the PI3K pathway, is a key downstream target of Nedd4: not only does Nedd4 regulate PTEN levels in RGC growth cones, but also, the decrease of PTEN rescues the branching defect caused by Nedd4 inhibition. Together our data suggest that Nedd4-regulated PTEN is a key regulator of terminal arborization in vivo.

Figure 1

UbK48R Inhibits RGC Axon Branching in the Tectum (A–C) Lateral view of RGC axons in the optic tract and tectum expressing control GFP (A), Myc-UbK48R (B), or Myc-UbWT (C). Axon trajectories are represented in (A′)–(C′). Scale bar, 20 μm. (D) A strategy for quantifying the extent of branching from a population of brains with axons expressing a given construct. Brains that have at least one branched axon are scored as “branched” and brains with no branched axons are scored as “unbranched.” (E) Graph showing the proportions of brains with branched axons that express GFP, UbK48R, or UbWT. Numbers of brains analyzed: GFP (n = 75), UbK48R (n = 11), UbWT (n = 13).

Figure 2

Nedd4 Is Expressed in RGC Axons (A and B) Stage 35/36 retinal explants were cultured for 24 hr and stained with an anti-Nedd4 antibody. Phase contrast image (A) of a GC-expressing Nedd4 (B) is shown. Scale bar, 5 μm. (C) Transverse section through stage 40 retina stained with an anti-Nedd4 antibody. IPL, inner plexiform layer; OPL, outer plexiform layer; OFL, optic fiber layer; ONH, optic nerve head. Scale bar, 50 μm. (D) Western blot of stage 40 head lysate probed with an anti-Nedd4 antibody. The arrowhead points to the full-length Nedd4, and the asterisk denotes a possible proteolytic fragment (Harvey et al., 1998). (E–G) Transverse section through stage 40 brain showing GFP-labeled RGC axons in the optic tract and the tectum (E); the same section stained with anti-Nedd4 is shown in (F), and merged image including DAPI (blue) is shown in (G). (H) Stage 40 brain from an embryo whose right eye was electroporated with GFP and whose left eye was removed before axon outgrowth. The brain was split along the ventral midline and flat-mounted to expose both halves of the tectum. (I) The same brain as in (H) but without the GFP signal. Nedd4 fluorescence intensity was measured along the blue and yellow dashed lines that cut through the region of the tectum where RGC axons arborize (see Experimental Procedures). Scale bar, 100 μm. (J) Graph showing Nedd4 pixel intensity along the dashed lines in (H). Yellow line represents Nedd4 signal in the tectum with no RGC axons, and the blue line represents Nedd4 signal in the tectum innervated by RGC axons.

Figure 3

Nedd4-DN Inhibits Axon Branching in the Tectum (A and B) Lateral view of axons expressing control GFP (A) and Nedd4-DN (B) in the tectum. Corresponding axon trajectories are in (A′) and (B′), where branches of a different order are color coded: white, axon shaft; red, primary; blue, secondary; yellow, tertiary. Scale bars, 20 μm. (C) Graph showing the average number of branches per axon arbor. (D) Formula for Axonal Complexity Index, adapted from Marshak et al. (2007). (E) Graph showing the average ACI value per axon arbor. (F) Pie charts representing the proportions of unbranched axons and branched axons with simple and complex morphologies. Simple arbors, ACI < 1.4; complex arbors, ACI ≥ 1.4. (G) Graph showing the proportion of branches of different order in axon arbors. ^∗∗p < 0.001, ^∗∗∗p < 0.0001, Student's t test.

Figure 4

Nedd4-MO Inhibits Axon Branching in the Tectum (A–C) Nedd4-MO leads to a knockdown in Nedd4 protein levels in GCs. Nedd4 immunofluorescence in a representative GC positive for a Control-MO (A and A′) and Nedd4-MO (B and B′) is shown. Scale bars, 5 μm. (C) A graph showing the average Nedd4 signal intensity per unit area in Nedd4-MO versus Control-MO GCs. Numbers on the bars represent the number of GCs analyzed. ^∗∗∗p < 0.0001, Student's t test. (D and E) Lateral view of Control-MO-positive (D) and Nedd4-MO-positive axons (E). Corresponding axon trajectories are in (D′) and (E′), where branches of a different order are color coded as before. Scale bars, 20 μm. (F) Graph showing the average number of branches per axon arbor. (G) Graph showing the average ACI value per axon arbor. (H) Pie charts representing proportions of Control-MO or Nedd4-MO axons with different morphologies. Simple arbors, ACI < 1.4; complex arbors, ACI ≥ 1.4. (I) Graph showing the proportion of branches of different order in axon arbors. ^∗p < 0.01, ^∗∗p < 0.001, ^∗∗∗p < 0.0001, Student's t test.

Figure 5

Nedd4 Regulates PTEN Levels in GCs (A) Transverse section through stage 40 retina stained with anti-Nedd4 (red) and anti-PTEN (green) antibodies. Blue signal indicates nuclear staining (DAPI). IPL, inner plexiform layer; OPL, outer plexiform layer; OFL, optic fiber layer; ONH, optic nerve head. Scale bars, 50 μm for (A″); 5 μm for (A″) inset. (B and C) Nedd4-DN (B) or Nedd4-MO (C) leads to an increase in PTEN signal intensity in GCs. (Left) Representative GCs containing the control constructs (Myc-GFP or Control-MO) or the experimental constructs (Myc-Nedd4-DN or Nedd4-MO) as shown in the upper panels, and the corresponding PTEN fluorescence in the lower panels. Scale bars, 5 μm. (Right) Graph showing average PTEN pixel intensity per unit. Numbers on the bars represent the number of GCs analyzed. ^∗∗p < 0.001, ^∗∗∗p < 0.0001, Student's t test. (D and E) Inhibiting proteasomes increases colocalization of Nedd4 and PTEN in GCs. Nedd4 (red) / PTEN (green) immunoflurescence (D′ and E′) and the resulting pixel-by-pixel intensity histogram (D″ and E″) in representative GCs (D and E) incubated for 16 hr without (D) or with (E) a proteasome inhibitor (50 μM LnLL) is shown. Scale bars, 5 μm. (F) Nedd4 and PTEN form complexes. Xenopus Nedd4 tagged with a Myc tag (1M-Nedd4) and Xenopus PTEN tagged with EGFP (GFP-PTEN) were transfected into HEK293T cells and their interactions were examined by coimmunoprecipitation using a control IgG (control) or an IP antibody (IP: anti-Myc for the upper two panels and anti-GFP for the lower two panels) followed by western blot (IB) using an anti-Myc or anti-GFP antibody.

Figure 6

Overexpressed PTEN Inhibits RGC Axon Branching in the Tectum (A and B) Lateral view of control (GFP) (A) and GFP-PTEN (B) axons in the tectum with corresponding axon trajectories in (A′) and (B′). Axon trajectories are color-coded as before. Scale bars, 20 μm. (C) Graph showing the average number of branches per axon arbor. (D) Graph showing the average ACI value per axon arbor. (E) Pie charts representing proportions of axons with different morphologies. Simple arbors, ACI < 1.4; complex arbors, ACI ≥ 1.4. (F) Graph showing the proportion of branches of a different order in axon arbors. ^∗∗∗p < 0.0001, Student's t test.

Figure 7

PTEN-MO Rescues the Branching Defect Caused by Nedd4-MO (A) PTEN-MO leads to a specific knockdown in PTEN protein levels in retina. Retinae isolated from embryonic stage 40 embryos injected with Control-MO or PTEN-MO were subjected to western blot using an anti-PTEN antibody (upper panel). (B–G) Nedd4-MO was cointroduced with Control-MO (B) or with PTEN-MO (C) to examine the branching phenotype. (B and C) Lateral view of Nedd4-MO+Control-MO-positive (B) and Nedd4-MO+PTEN-MO-positive axons (C). Corresponding axon trajectories are in (B′) and (C′), where branches of a different order are color coded as before. Scale bars, 20 μm. (D) Graph showing the average number of branches per axon arbor. (E) Graph showing the average ACI value per axon arbor. (F) Pie charts representing proportions of axons with different morphologies. Unbranched, ACI < 1; simple arbors, 1 ≤ ACI < 1.4; complex arbors, ACI ≥ 1.4 (G) Graph showing the proportion of branches of different orders in axon arbors. ^∗∗p < 0.001; ^∗∗∗p < 0.0001, Student's t test.

Figure 8

Netrin-1 Induces UPS-Dependent PTEN Degradation and GC Collapse Stage 35/36 retinal explants were cultured for 24 hr, after which they were stimulated for 5 min with 3 μg/ml recombinant human Netrin-1, after which they were fixed and stained with an anti-PTEN antibody. (A–D) Images of representative GCs unstimulated (A and C) or stimulated with Netrin-1 (B and D). Lactacystin (10 μM) was added to cultures immediately prior to stimulation in (C) and (D). Scale bars, 5 μm. (E) Graph showing quantification of average PTEN pixel intensity per unit area normalized to the unstimulated control. The numbers in the bars represent numbers of analyzed GCs. ^∗∗p < 0.001; ^∗∗∗p < 0.0001, Student's t test. (F) Stage 32 retinal explants were cultured for 24 hr on a high laminin (20 μg/ml) substrate. Axons were stimulated with 1.2 μg/ml Netrin-1, or control buffer in which Netrin-1 was dissolved. After 10 min they were fixed and the number of collapsed GCs was counted. The numbers in the bars represent numbers of analyzed GCs. ^∗∗∗p < 0.0001, Dunn's Multiple Comparisons Test followed by Kruskal-Wallis test.

Figure 9

A Proposed Model for How Nedd4-Mediated Downregulation of PTEN Promotes PI3K Signaling in an Axon and Its Branching In this model, external signals promote or reduce axon branching by modulating Nedd4-dependent PTEN downregulation, and consequently, the levels of downstream PI3K signals. An increase in Nedd4 activity results in reduced PTEN levels and therefore promotes the PI3K pathway and the downstream cytoskeletal rearrangements that favor branch growth. Conversely, an elevation in PTEN levels due to compromised Nedd4 activity inhibits PI3K signaling and reduces arbor growth.