Fig 2: Absence of RalA in SCs specifically and RalB ubiquitously impairs radial sorting. (A) Electron micrographs of P5 sciatic nerves (top) and 2-mo-old (bottom) control, MpzCre/+:Ralafl/fl, Ralafl/fl:Ralb−/−, and MpzCre/+:Ralafl/fl:Ralb−/− mice. Bundles of unsorted axons are indicated by black arrowheads. White arrowheads point to sorted but not-myelinated fibers. Scale bar of inset: 1 µm. (B–E) Quantification of morphological features of P5 sciatic nerves. The number of myelinated axons (B), all sorted axons (C), and not-myelinated but sorted axons (D) was determined per cross section. The area of endoneurium occupied by bundles of unsorted axons was quantified relative to the total endoneurial area (E). n = 4 mice per genotype; one complete nerve cross section per animal was analyzed. One-way ANOVA with Tukey’s multiple comparisons test. (F–I) Quantification of morphological features of 2-mo-old sciatic nerves. The number of myelinated axons (F; same dataset is shown again in Fig. 3 B), all sorted axons with a diameter >1 µm (G; same dataset is shown again in Fig. 3 C), and not-myelinated but sorted axons with a diameter >1 µm (H; same dataset is shown again in Fig. S2 B) was determined per cross section. The area of the endoneurium occupied by bundles of unsorted axons was quantified relative to the total endoneurial area (I). n = 5 mice per genotype; one complete cross section per animal was analyzed. One-way ANOVA with Tukey’s multiple comparisons test. All data are shown as mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig 3: Interaction of RalA with the exocyst complex is crucial for SC process extension. Exemplary images of a control and a MpzCre/+:Ralafl/fl:Ralb−/− SC derived from P5 sciatic nerves infected with lentivirus (LV) expressing (A) myc-tagged RalA72L (constitutively active RalA [CA-RalA]), (B) myc-tagged RalA28N (dominant-negative RalA [DN-RalA]), (C) myc-tagged RalA72L D49N (CA-RalA uncoupled from the interaction with RalBP1), (D) myc-tagged RalA72L ΔN11 (CA-RalA uncoupled from the interaction with PLD1), (E) myc-tagged RalA72L D49E (CA-RalA uncoupled from the interaction with Exoc2 and Exoc8), and (F) myc-tagged RalA72L A48W (CA-RalA uncoupled from the interaction with Exoc8). SC processes were visualized by immunostaining of α-tubulin (red) and F-actin (green), and infection was controlled by immunostaining for myc (blue). Arrows indicate the far ends of SC processes, and asterisks mark the location of SC nuclei along the length of the cell. Quantifications of average process lengths are depicted on the right. n = cells derived from seven (only in E) or six individual animals; ≥100 cells analyzed per animal. Multiple unpaired two-tailed t tests with Holm–Sidak method for multiple comparisons correction. Scale bar for all images: 50 µm. All data are shown as mean ± SEM. **, P < 0.01; ***, P < 0.001.

Fig 4: Ral-deficient SCs show defective processes and redundant basal laminae in sciatic nerves. (A) Exemplary high-magnification electron micrographs of sciatic nerves of 2-mo-old MpzCre/+:Ralafl/fl:Ralb−/− mice. In the left image, white arrowheads point to redundant loops of basal lamina associated with collagen. In the middle image, black arrowheads mark an area of a bundle of unsorted axons that is not covered by a SC process. In the right image, the arrow indicates an exemplary sorted but not-myelinated axon that is not fully surrounded by a SC. (B) Quantification of incompletely surrounded, not-myelinated sorted axons with a diameter >1 µm per sciatic nerve cross section in 2-mo-old mice. n = 5 mice per genotype, with one complete cross section per animal analyzed. One-sample t test. (C–E) Pull-down assay for active Rac1 from lysates of P5 sciatic nerves. An exemplary Western blot is shown in C. The levels of active Rac1 in D and total Rac1 in E were normalized to controls. n = 4 samples per genotype, each containing sciatic nerves of three animals. Unpaired two-tailed t test. All data are shown as mean ± SEM. **, P < 0.01; ***, P < 0.001.

Fig 5: Cultured Ral-deficient SCs show process extension defects. (A) SCs isolated from P5 sciatic nerves of control and MpzCre/+:Ralafl/fl:Ralb−/− mice, grown in culture for 24 h on laminin 111. Processes and lamellipodia were visualized by immunostaining of α-tubulin (red) and F-actin (green), and nuclei were labeled with DAPI. Arrows point to examples of axial lamellipodia, and arrowheads point to radial lamellipodia. Scale bars of insets: 5 µm. (B and C) Number of radial (B) and axial (C) lamellipodia per SC from P5 sciatic nerves at DIV1, plated on laminin 111. n = cells derived from six individual mice per genotype; ≥100 cells analyzed per animal. Unpaired two-tailed t test. (D) Average process length of SCs from P5 sciatic nerves at DIV5, plated on laminin 111. n = cells derived from five individual mice per genotype; ≥100 cells analyzed per animal. Unpaired two-tailed t test. All data are shown as mean ± SEM. **, P < 0.01; ***, P < 0.001.