Gross morphology of Nedd4 mutant mice and the expression of Nedd4 in mouse embryos. (A) Gross morphology of mice at postnatal day 0 (P0) – the Nedd4^−/− mouse was stillborn and markedly smaller than its wildtype littermate (+/+). (B) Immunoblot analysis of total lysate of the brain, spinal cord, muscle and skin from Nedd4^−/− and wildtype embryos (E18.5), using antibodies against Nedd4, beta-galactosidase (β-gal), and alpha-tubulin (tubulin, as loading control). Nedd4 was detected in the wildtype tissues, but totally absent from the Nedd4 mutants. In contrast, β-gal was detected in Nedd4 mutants, but absent from the wildtypes. Note that β-gal was heavily expressed in the muscle tissues in the Nedd4 mutant. (C–E) Cross-section of an E11.5 Nedd4^+/− embryo at brachial level was processed for X-gal staining to reveal the expression of a reporter gene lacZ inserted into the Nedd4 genomic locus. LacZ expression is detected in the skin as indicated by arrows in C (low magnification) and D (high magnification), and in marginal layers of the spinal cord as indicated by arrowheads in C (low magnification) and D (high magnification). The majority of the other spinal cord regions are devoid of LacZ expression. (E) The ventral region of the spinal cord (E11.5) viewed under high-power magnification: no LacZ is detected in this region. (F–J) Cross-section of Nedd4^+/− spinal cord at brachial level (E18.5): LacZ expression remains highly restricted to the surface of the spinal cord as indicated by arrows in F (low magnification) and I (high magnification). In addition, lacZ-positive cells are present in the dorsal root ganglion (DRG, arrowheads in G), in the cartilage primordium of the vertebra (F, white arrow) and in a small region just underneath the central canal (cc) (arrowheads in F (low magnification) and J (high magnification)). Motoneurons (arrowheads in H) in the ventral spinal cord (high magnification view of the region indicated by the rectangular box in F) are not labeled by LacZ (H). In contrast, the skeletal muscles are intensely stained, as shown in wholemount diaphragm muscle (E18.5) (K) and longitudinal sections of limb muscle (E18.5). Scale bars – C: 200 μm, D: 50 μm, E: 10 μm F: 200 μm, G, H: 30 μm, I, J: 30 μm, K: 400 μm, L: 30 μm.

Reduced muscle fiber size in Nedd4 mutant embryos. (A, B) confocal images of wholemount diaphragm muscle (E18.5) labeled by phalloidin–rhodamine: in the wildtype (A), muscle fibers are aligned in parallel arrays, whereas in Nedd4^−/− embryos (B), muscle fibers appear wavy and disorganized. (C, D) semithin cross-sections of diaphragm muscles (E18.5) stained by toluidine blue. The diaphragm is thinner and muscle fiber sizes are reduced in Nedd4^−/− embryos (D), compared to the wildtype (C). Arrowheads point to peripherally localized muscle nuclei in both mutant (D) and wildtype muscles (C). (E, F) Electron micrograph of the intercostal muscles from the wildtype (E) and Nedd4^−/−g(F) at E18.5: the fine structure of muscles is preserved in Nedd4^−/− embryos; the Z-lines (arrowheads) are clearly visible, suggesting that sarcomeres in Nedd4^−/− embryos are properly assembled. Scale bars: A–B, 20 μm; C–D: 50 μm; E–F, 1 μm.

Aberrant innervation in Nedd4 mutant muscles. Pairs (+/+ and −/−) of diaphragm muscles at E13.5 (A, B), E15.5 (C, D) and E18.5 (E–H) were immunostained with antibodies against neurofilament (NF150) and synaptotagmin-2 (Syt-2). In the wildtype muscles, the secondary (intramuscular) branch (as indicated by “s” in E) of the primary phrenic nerve (p in E) develops as a single bundle perpendicular to the long axis of the muscle fibers. This stereotypic pattern is established as early as E13.5 (A) and maintained throughout embryonic development, as illustrated at E15.5 (C) and E18.5 (E). In addition, numerous tertiary nerves (t) further ramify from the secondary branches and terminate within the central region of the muscle. In Nedd4 mutant, the secondary nerve branches are markedly defasciculated (arrowheads in B, D, F, H), and often develop abnormal loops (arrows in B, D, F). Furthermore, the nerves in the Nedd4^−/− embryo (N=8) fail to innervate the entire diaphragm, leaving an uninnervated gap at the ventral portion of the diaphragm (marked by the bracket in D and H); whereas the ventral portion of the wildtype diaphragm (N=6) is fully innervated (large arrowhead in C, arrows in G). Note the reduction of overall sizes of the diaphragm muscles in the Nedd4^−/− embryos at all stages (E13.5–E18.5). Asterisks indicate the ingrowing sensory nerves from the edge of the diaphragm at E18.5. p: primary phrenic nerve; s: secondary branch; t: tertiary branch. Scale bars: A–F: 1000 μm; G, H: 500 μm.

Decrease of motoneuron and axon number in Nedd4 mutants. (A, B) Cross-sections of embryonic spinal cord (E10.5) immunolabeled by anti-HB9 antibodies: HB9-possitive motoneurons are markedly reduced in the Nedd4^−/− embryos (B), compared to the wildtype (+/+, A). (C, D) Thionin-stained cross-section of E18.5 lumbar spinal cord (4th segment, L4) of a wildtype (+/+, C) and Nedd4^−/− (D) embryo. The spinal cord volume and the dorsal root ganglion (DRG) volume are markedly reduced in Nedd4^−/− embryos (for detailed quantification, see Table 2). (E, F) Higher magnification of the L4 ventral horn: number of motoneurons (arrows in E–F) is markedly reduced in the Nedd4^−/− embryos (F), compared to the wildtype (E). The dotted line indicates the medial border of the ventral horn. (G, H) Cross-section of the phrenic nerve (E18.5) of the wildtype (G) and Nedd4^−/− embryos (H). Schwann cells (black arrows in G–H) are present in both wildtypes and Nedd4 mutants. The number of axons (black arrowhead) in the Nedd4^−/− embryo is markedly reduced, compared to the wildtype. Note that the nerves are not yet myelinated at E18.5. Scale bars: A, B: 50 μm; C, D: 250 μm; E, F: 50 μm. G, H: 10 μm.

Electrophysiological analysis of neuromuscular synaptic transmission in Nedd4 mutants. (A) Sample traces (superimposed 60×1-second traces) of spontaneous miniature endplate potentials (mEPPs) in normal Ringers recorded from E18.5 wildtype (+/+, WT) and Nedd4 mutant (−/−, mut). (B) Quantification of mEPP frequency and amplitude. The frequency of mEPPs is significantly increased in Nedd4 mutant (Student’s t-test, **p<0.001). (C) Depolarization (40 mM K⁺)-induced mEPP frequency changes in responding to various Ca²⁺ concentration (0.5, 1, 2 and 4 mM): 73.8±12.8 min⁻¹ (n=20, WT), 37.1±6.1 min⁻¹ (n=14, mut) (0.5 mM Ca²⁺); 80.4±14.7 min⁻¹ (n=20, WT), 42.6±9.9 min⁻¹ (n=10, mut) (1 mM Ca²⁺); 105.9±9.6 min⁻¹ (n=22, WT), 68.8±8.9 min⁻¹ (n=12, mut) (2 mM Ca²⁺); 156±13.6 min⁻¹ (n=22, WT), 102.1± 16.0 min⁻¹ (n=12, mut) (4 mM Ca²⁺). Evoked mEPP frequencies are significantly lower in Nedd4 mutant, compared to the wildtype (*p<0.05). However, the Ca²⁺-dependency of mEPP frequency is similar between the two genotypes. (D) Action potentials evoked by nerve stimulation in both wildtype and Nedd4 mutants.

Ultrastructural analysis of neuromuscular synapses in Nedd4 mutants. (A, B) Electron micrograph of neuromuscular synapses in the wildtype (A) and Nedd4^−/− (B) muscles (E18.5). In both wildtype and Nedd4^−/−muscles, pre-synaptic terminals are capped by terminal Schwann cells (SC). Basal lamina (arrowheads in A–B) is clearly visible, and multiple pre-synaptic terminals are present at an individual synaptic site in both wildtype (A) and Nedd4^−/− (B) muscles. Within the terminals, docked vesicles (small arrows in A–B) are present in both wildtypes and mutants, however, nerve terminals in the mutants (B) are smaller and more numerous, compared to the wildtypes (A). (C, E) Quantification of nerve terminal size (C) and the distribution of pre-synaptic terminal number per synapse in the wildtype (D) and Nedd4^−/− (E) embryos. Scale bar: A–B, 1 βm. ***p<0.005.

Formation of the neuromuscular junction in Nedd4 mutants. (A) Wholemount diaphragm muscles (E18.5) were stained with Texas-Red conjugated α-bgt to label post-synaptic AChRs (endplates). AChRs (arrowheads) are clustered along the central region (the endplate band) in both wildtype and Nedd4^−/−muscles. (B) Quantitative comparison of individual endplates (AChR clusters) in wildtype and Nedd4^−/− muscles. The mean gray value of the fluorescence intensity (α-bgt labeling intensity), the average area, perimeter and Feret’s diameter (the length of the greatest axis) of AChR clusters are similar between the wildtype and Nedd4^−/− muscles. (C) Comparison of AChE staining pattern between the wildtype and the Nedd4^−/−diaphragm muscles (E18.5): AChE (dark brown staining) as indicated by black arrowheads in C) is detected at the central region of the muscle in both the wildtype and Nedd4^−/−embryos. (D) E18.5 diaphragm muscles were labeled by Texas-Red conjugated α-bgt (red) and antibodies against synaptotagmin 2 and neurofilament (green). Nerve terminals in both wildtype and Nedd4^−/−embryos (green, arrowheads in D) are nicely aligned with post-synaptic AChR clusters (red), as indicated in the merged images. Scale bars: A: 50 μm; C: 100 μm; D: 20 μm.

Quantitative analysis of endplate numbers in dissociated single muscle fiber. Dissociated single muscle fibers were double-labeled with phalloidin (purple) and α-bgt (pink). In both wildtype (A) and Nedd4^−/− muscles (B, C), the majority of muscle fibers bear only one endplate (1 patch, arrowheads); only a small fraction of the muscle fibers contained two endplates (2 patches, arrows). (D) Quantitative analyses of endplate numbers in the wildtype (n=562 fibers, N=3 embryos) and Nedd4^−/−muscles (n=368 fibers, N=3 embryos). The distribution of AChR clusters in individual fibers is similar between the wildtype and Nedd4^−/− muscles. Scale: 20 μm.